Download the Full Ebook and Access More Features - ebooknice.

com

(Ebook) Advanced Soft Computing Techniques in Data

Science, IoT and Cloud Computing by Sujata Dash
(editor), Subhendu Kumar Pani (editor), Ajith
Abraham (editor), Yulan Liang (editor) ISBN
9783030756567, 3030756564
https://ebooknice.com/product/advanced-soft-computing-
techniques-in-data-science-iot-and-cloud-computing-36306376

OR CLICK HERE

DOWLOAD EBOOK

Download more ebook instantly today at https://ebooknice.com

 Instant digital products (PDF, ePub, MOBI) ready for you
Download now and discover formats that fit your needs...

Start reading on any device today!

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason;

Viles, James ISBN 9781459699816, 9781743365571,
9781925268492, 1459699815, 1743365578, 1925268497
https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

ebooknice.com

(Ebook) Biomedical Data Mining for Information Retrieval:

Methodologies, Techniques, and Applications by Subhendu
Kumar Pani; Sujata Dash; S. Balamurugan; Ajith Abraham
ISBN 9781119711247, 111971124X
https://ebooknice.com/product/biomedical-data-mining-for-information-
retrieval-methodologies-techniques-and-applications-34242538

ebooknice.com

(Ebook) Matematik 5000+ Kurs 2c Lärobok by Lena

Alfredsson, Hans Heikne, Sanna Bodemyr ISBN 9789127456600,
9127456609
https://ebooknice.com/product/matematik-5000-kurs-2c-larobok-23848312

ebooknice.com

(Ebook) SAT II Success MATH 1C and 2C 2002 (Peterson's SAT

II Success) by Peterson's ISBN 9780768906677, 0768906679

https://ebooknice.com/product/sat-ii-success-
math-1c-and-2c-2002-peterson-s-sat-ii-success-1722018

ebooknice.com
(Ebook) Vagabond, Vol. 29 (29) by Inoue, Takehiko ISBN
9781421531489, 1421531488

https://ebooknice.com/product/vagabond-vol-29-29-37511002

ebooknice.com

(Ebook) AI, Edge and IoT-based Smart Agriculture

(Intelligent Data-Centric Systems) by Ajith Abraham
(editor), Sujata Dash (editor), Joel J.P.C. Rodrigues
(editor), Biswaranjan Acharya (editor), Subhendu Kumar
https://ebooknice.com/product/ai-edge-and-iot-based-smart-agriculture-
Pani (editor) ISBN 9780128236949, 0128236949
intelligent-data-centric-systems-37244888

ebooknice.com

(Ebook) Master SAT II Math 1c and 2c 4th ed (Arco Master

the SAT Subject Test: Math Levels 1 & 2) by Arco ISBN
9780768923049, 0768923042
https://ebooknice.com/product/master-sat-ii-math-1c-and-2c-4th-ed-
arco-master-the-sat-subject-test-math-levels-1-2-2326094

ebooknice.com

(Ebook) Soft Computing in Data Science: 5th International

Conference, SCDS 2019, Iizuka, Japan, August 28–29, 2019,
Proceedings by Michael W. Berry, Bee Wah Yap, Azlinah
Mohamed, Mario Köppen ISBN 9789811503986, 9789811503993,
https://ebooknice.com/product/soft-computing-in-data-science-5th-
9811503982, 9811503990
international-conference-scds-2019-iizuka-japan-
august-2829-2019-proceedings-10805960
ebooknice.com

(Ebook) Cambridge IGCSE and O Level History Workbook 2C -

Depth Study: the United States, 1919-41 2nd Edition by
Benjamin Harrison ISBN 9781398375147, 9781398375048,
1398375144, 1398375047
https://ebooknice.com/product/cambridge-igcse-and-o-level-history-
workbook-2c-depth-study-the-united-states-1919-41-2nd-edition-53538044

ebooknice.com
Studies in Big Data 89

Sujata Dash
Subhendu Kumar Pani
Ajith Abraham
Yulan Liang Editors

Advanced Soft
Computing
Techniques in Data
Science, IoT and
Cloud Computing
Studies in Big Data

Volume 89

Series Editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
The series “Studies in Big Data” (SBD) publishes new developments and advances
in the various areas of Big Data- quickly and with a high quality. The intent is to
cover the theory, research, development, and applications of Big Data, as embedded
in the fields of engineering, computer science, physics, economics and life sciences.
The books of the series refer to the analysis and understanding of large, complex,
and/or distributed data sets generated from recent digital sources coming from
sensors or other physical instruments as well as simulations, crowd sourcing, social
networks or other internet transactions, such as emails or video click streams and
other. The series contains monographs, lecture notes and edited volumes in Big
Data spanning the areas of computational intelligence including neural networks,
evolutionary computation, soft computing, fuzzy systems, as well as artificial
intelligence, data mining, modern statistics and Operations research, as well as
self-organizing systems. Of particular value to both the contributors and the
readership are the short publication timeframe and the world-wide distribution,
which enable both wide and rapid dissemination of research output.
The books of this series are reviewed in a single blind peer review process.
Indexed by SCOPUS, EI Compendex, SCIMAGO and zbMATH.
All books published in the series are submitted for consideration in Web of Science.

More information about this series at http://www.springer.com/series/11970

Sujata Dash · Subhendu Kumar Pani ·
Ajith Abraham · Yulan Liang
Editors

Advanced Soft Computing

Techniques in Data Science,
IoT and Cloud Computing
Editors
Sujata Dash Subhendu Kumar Pani
Department of Computer Science School of Computer Science
and Application and Engineering
North Orissa University Orissa Engineering College
Bhubaneswar, India Bhubaneswar, Odisha, India

Ajith Abraham Yulan Liang

Faculty of Electrical Engineering Department of Family and Community
and Computer science Health
Machine Intelligence Research Labs (MIR University of Maryland
Labs) Baltimore, MD, USA
Auburn, WA, USA

ISSN 2197-6503 ISSN 2197-6511 (electronic)

Studies in Big Data
ISBN 978-3-030-75656-7 ISBN 978-3-030-75657-4 (eBook)
https://doi.org/10.1007/978-3-030-75657-4

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Overview

The new applications of soft computing can be regarded as an emerging field

in computer science, automatic control engineering, medicine, biology applica-
tion, natural environmental engineering, and pattern recognition. Now the exemplar
model for soft computing is the human brain. The use of various techniques of soft
computing is nowadays successfully implemented in many domestic, commercial,
and industrial applications due to the low-cost and very high-performance digital
processors and also the declining price of the memory chips. This is the main reason
behind the wider expansion of soft computing techniques and their application areas.
These computing methods also play a significant role in the design and optimiza-
tion of diverse engineering disciplines. With the influence and the development of the
Internet of Things (IoT) concept, the need for using soft computing techniques has
become more significant than ever. In general, soft computing methods are similar to
biological processes very closely than traditional techniques, which are mostly based
on formal logical systems, such as sentential logic and predicate logic, or rely heavily
on computer-aided numerical analysis. Soft computing techniques are anticipated to
complement each other. The aim of these techniques is to accept imprecision, uncer-
tainties, and approximations to get a rapid solution. However, recent advancements in
the representation of soft computing algorithms (Fuzzy Logic, Evolutionary Compu-
tation, Machine Learning, and Probabilistic Reasoning) generate a more intelligent
and robust system providing a human interpretable, low-cost, approximate solution.
Soft computing-based algorithms have demonstrated great performance in a
variety of areas including Multimedia retrieval, Fault tolerance, System modeling,
Network architecture, Web semantics, big data analytics, time series, biomedical
and health informatics, etc. Soft computing approaches such as Genetic Program-
ming (GP), Support Vector Machine–Firefly Algorithm (SVM-FFA), Artificial
Neural Network (ANN), and Support Vector Machine–Wavelet (SVM-Wavelet)

v
vi Preface

have emerged as powerful computational models. These have also shown significant
success in dealing with massive data analysis for a large number of applications.
This book, Advanced Soft Computing Techniques in Data Sciences, IoT and
Cloud Computing, aims to play a significant role in improvising human life to a
great extent. All the researchers and practitioners who are working in the field
of computer engineering, medicine, biology application, signal processing, and
mechanical engineering will be highly benefited.
This book would be a good collection of state-of-the-art approaches for soft
computing-based applications to various engineering fields. It will be very bene-
ficial for the new researchers and practitioners working in the field to quickly know
the best-performing methods. They would be able to compare different approaches
and can carry forward their research in the most important area of research which has
a direct impact on the betterment of the human life and health. This book would be
very useful because there is no book in the market that provides a good collection of
state-of-the-art methods of soft computing-based models for Multimedia retrieval,
Fault tolerance, System modeling, Network architecture, Web semantics, big data
analytics, time series, and biomedical and health informatics.

Objective

The purpose of this book is to report the latest advances and developments in the field
of Multimedia retrieval, Fault tolerance, System modeling, Network architecture,
Web semantics, big data analytics, time series, and biomedical and health informatics.
The book comprises the following three parts:
• Soft Computing Techniques for Internet of Things (IoT) devices
• Soft Computing Techniques in Cloud Computing and Computer Networking
• Soft Computing Techniques in Data Science

Organization

The book, “Advanced Soft Computing Techniques in Data Sciences, IoT and Cloud
Computing”, consists of 19 edited chapters and the whole contents of the book are
organized into the following three sections:
• Part 1: Soft Computing Techniques for IoT Devices. This section has focused on
Soft Computing Techniques for Internet of Things paradigms and their applica-
tion in wearable assistive devices for visually impaired people, security of IoT
devices, medical data security in healthcare systems, content-based video retrieval
systems, and also applications in intelligent farming. There are six chapters in this
Preface vii

section. The first chapter looks into the application of soft computing techniques
for designing IoT-based wearable assistive devices for visually impaired people.
The second, third, fourth, and fifth contributions focus on the security of IoT
devices and their application in medical data in healthcare systems and content-
based video retrieval systems. The sixth chapter discusses various existing soft
computing techniques and their applications for decision support in intelligent
farming.
• Part II: Soft Computing Techniques in Cloud Computing and Computer
Networking. The second part comprises five chapters. The first contribution
discusses about the hybrid cloud data protection using the machine learning
approach. The second chapter audits the analysis of LSTM networks for COVID-
19 impact prediction. The third chapter focuses on the progress and systematic
review of soft computing techniques for energy consumption and resource allo-
cation on the cloud. The fourth chapter discusses automatic segmentation and
classification of Brain Tumor from MR Images using DWT-RBFNN.
• The fifth chapter discusses on the application of unsupervised learning on
automatic localization of optic disc in Retinal Fundus Image.
• Part III: Soft Computing Techniques in Data Science. There are eight chapters in
this section. The first chapter discusses the performance evaluation of Hybrid
Machine Learning Algorithms for Medical Image Classification. The second
chapter discusses the issues and challenges of soft computing techniques for
Healthcare Decision Support Systems and the third chapter focuses on the truth
values of Modus Ponens and Modus Tollens Rule for Linguistic Truth-Valued
Propositions and Its Application in Taking Decisions in Health Care. The fourth
chapter focuses on a case study of Amazon based on the analysis of customers’
reviews using Soft Computing Classification Algorithms. The fifth chapter throws
light on Pattern Mining—FTISPAM applying Hybrid Genetic Algorithm and
the sixth chapter highlights soft computing techniques for Medical Diagnosis,
Prognosis, and Treatment. The concluding chapter explains Artificial Intelligence
Applications for COVID-19 Pandemic and the eighth chapter deliberates on the
prediction of transmittable diseases in a location using ARIMA.

Target Audiences

The current volume is a reference text aimed to support a number of potential

audiences, including the following:
• Researchers in this field who wish to have the up to date knowledge of the current
practice, mechanisms, and research developments.
• Students and academicians of the biomedical and informatics field who have an
interest in further enhancing the knowledge of the current developments.
viii Preface

• Industry and people from Technical Institutes, R & D Organizations and working
in the field of machine learning, IoT, Cloud Computing, biomedical engineering,
health informatics, and related fields.

Baripada, India Sujata Dash

Bhubaneswar, India Subhendu Kumar Pani
Auburn, USA Ajith Abraham
Baltimore, USA Yulan Liang
Acknowledgements

The editors would like to acknowledge the help of all the people involved in this
project and, more specifically, the reviewers who took part in the review process.
Without their support, this book would not have become a reality.
First, the editors would like to thank each one of the authors for their time,
contribution, and understanding during the preparation of the book.
Second, the editors wish to acknowledge the valuable contributions of the
reviewers regarding the improvement of quality, coherence, and content presentation
of chapters.
Last but not the least, the editors wish to acknowledge the love, understanding,
and support of their family members during the preparation of the book.

Odisha, India Sujata Dash

Bhubaneswar, India Subhendu Kumar Pani
Auburn, USA Ajith Abraham
Baltimore, USA Yulan Liang

ix
Contents

Soft Computing Techniques for IoT Devices

A Wearable Assistive Device for Safe Travel Using Transfer
Learning and IoT for Visually Impaired People . . . . . . . . . . . . . . . . . . . . . . 3
R. Priyatharshini, R. Senthil Kumar, M. Sanjay Sivakumar,
A. Mathumathi, and Nancy Sharon Johnson
Soft Computing Techniques for Physical Layer Security of IoT
Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
C. Ismayil Siyad and S. Tamilselvan
Linear Congruence Generator and Chaos Based Encryption Key
Generation for Medical Data Security in IoT Based Health Care
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Anirban Bhowmik, Sunil Karforma, Joydeep Dey, and Arindam Sarkar
Content Based Video Retrieval—Methods, Techniques
and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Reddy Mounika Bommisetty, P. Palanisamy, and Ashish Khare
Building the World of Internet of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Seema Sahai, Richa Goel, and Gurinder Singh
Applicability of Machine Learning Algorithms for Intelligent
Farming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Bharti Verma, Nikhil Sharma, Ila Kaushik, and Bharat Bhushan

Soft Computing Techniques in Cloud Computing and Computer

Networking
Hybrid Cloud Data Protection Using Machine Learning Approach . . . . . 151
D. Praveena, S. Thanga Ramya, V. P. Gladis Pushparathi, Pratap Bethi,
and S. Poopandian

xi
xii Contents

Analysis of Long Short Term Memory (LSTM) Networks

in the Stateful and Stateless Mode for COVID-19 Impact Prediction . . . . 167
Vinayak Ashok Bharadi and Sujata S. Alegavi
Soft Computing Techniques for Energy Consumption and Resource
Aware Allocation on Cloud: A Progress and Systematic Review . . . . . . . . 191
Sukhpreet Kaur, Yogesh Kumar, and Sushil Kumar
Automatic Segmentation and Classification of Brain Tumor
from MR Images Using DWT-RBFNN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Hari Mohan Rai, Kalyan Chatterjee, and Anand Nayyar
Automatic Localization of Optic Disc in Retinal Fundus Image
Based on Unsupervised Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
J. Prakash and B. Vinoth Kumar

Soft Computing Techniques in Data Science

Performance Evaluation of Hybrid Machine Learning Algorithms
for Medical Image Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
N. T. Renukadevi
Computing Truth Values of Modus Ponens and Modus Tollens
Rule for Linguistic Truth-Valued Propositions and Its Application
in Taking Decisions in Health Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Sumita Basu and Bithi Chattaraj
Analysis of Customers’ Reviews Using Soft Computing
Classification Algorithms: A Case Study of Amazon . . . . . . . . . . . . . . . . . . 331
Koushal Kumar and Bhagwati Prasad Pande
Pattern Mining—FTISPAM Using Hybrid Genetic Algorithm . . . . . . . . . 353
L. Mary Gladence, S. Shanmuga Priya, A. Shane Sam,
Gladis Pushparathi, and E. Brumancia
Soft Computing Techniques for Medical Diagnosis, Prognosis
and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
Surabhi Adhikari, Surendrabikram Thapa, and Awishkar Ghimire
Role of Artificial Intelligence in COVID-19 Pandemic . . . . . . . . . . . . . . . . . 401
Abhishek Mehta and Trupti Rathod
Prediction of Transmittable Diseases Rate in a Location Using
ARIMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Varun Totakura and E. Madhusudhana Reddy
Editors and Contributors

About the Editors

Sujata Dash is currently working as an Associate Professor of Computer Science

at North Orissa University in the Department of Computer Science, Baripada, India.
She has been consistently delivering her services in teaching and guiding students
for more than two and a half decades now. She is a recipient of Titular Fellow-
ship from the Association of Commonwealth Universities, UK. She has worked as a
visiting professor of the Computer Science Department of the University of Mani-
toba, Canada. She has published more than 170 technical papers in international
journals/ proceedings of international conferences/ edited book chapters of reputed
publishers like Springer, Elsevier, IEEE, IGI Global USA, 10 national and interna-
tional patents, and has published many textbooks, monographs, and edited books.
She is a member of international professional associations like ACM, IRSS, CSI,
IMS, OITS, OMS, IACSIT, IST, IEEE and is a reviewer of around 15 international
journals which include World Scientific, Bioinformatics, Springer, IEEE ACCESS,
Inderscience, and Science Direct publications. In addition, she is a member of the
editorial board of around 10 international journals. She has visited many countries
and delivered keynotes, invited speech, and chaired many special sessions in Interna-
tional conferences in India and abroad. Her current research interest includes Machine
Learning, Data Mining, Big Data Analytics, Bioinformatics, Soft Computing, and
Intelligent Agents.

Subhendu Kumar Pani received his Ph.D. from Utkal University Odisha, India in
the year 2013. He is working as a Professor in the Department of Computer Science
and Engineering and also a Research coordinator at Orissa Engineering College
(OEC) Bhubaneswar. He has more than 16 years of teaching and research experience.
His research interests include Data mining, Big Data Analysis, web data analytics,
Fuzzy Decision Making, and Computational Intelligence. He is the recipient of 5
researcher awards. In addition to research, he has guided two Ph.D. students and 31
M.Tech students. He has published 7 patents and 51 International Journal papers (25

xiii
xiv Editors and Contributors

Scopus index). His professional activities include roles as Associate Editor, Edito-
rial board member, and/or reviewer of various International Journals. He serves as
Associate Editor/Editorial Board Member for more than 50+ International Jour-
nals/Conferences of high repute and impact. He is an Associate with no. of confer-
ence societies. He has more than 100 international publications, 5-authored books, 4
edited books, and 10 book chapters into his account. He is a fellow in SSARSC and
the life member in IE, ISTE, ISCA, OBA, OMS, SMIACSIT, SMUACEE, and CSI.

Ajith Abraham is current working as the Director of Machine Intelligence Research

Labs (MIR Labs), which has members from more than 100 countries. Dr. Abraham’s
research and development experience includes more than 27 years in the industry and
academia. He received the M.S. degree from the Nanyang Technological University
(NTU), Singapore, and the Ph.D. degree in Computer Science from Monash Univer-
sity, Melbourne, Australia. He worked in a multidisciplinary environment involving
machine (network) intelligence, cybersecurity, sensor networks, Web intelligence,
scheduling, and data mining and applied to various real-world problems. He has given
more than 100+ conference plenary lectures/tutorials and invited seminars/lectures
in over 100 Universities around the globe.

Yulan Liang holds a Ph.D. in Statistics and has been a tenured faculty at the Univer-
sity of Maryland, Baltimore (UMB) since 2008. Before that, she worked as a tenure-
track assistant professor at the School of Public Health and School of Medicine at
the University at Buffalo. Dr. Liang’s expertise is in the cutting-edge multidisci-
plinary fields including big data science, statistics, data mining, neural networks,
machine learning, Bayesian inferences, artificial intelligence, and bioinformatics
with applications to large biomedical, health science, omics, and healthcare data
for risk assessments, predictions, and medical decision-making. She has published
more than a hundred peer-reviewed journal and conference proceeding articles in
the data science fields. She was funded as PI by NSF and served as a statistics
core director of NIH P30, U01, and P20 grants and co-Investigator on multiple
NIH-funded grants. She also served as co-Investigator and biostatistician for private
organizations in health associations and other funding agencies, including the Robert
Wood Johnson Foundation, Alzheimer’s Association, and National Council of State
Boards of Nursing.

Contributors

Surabhi Adhikari Department of Computer Science and Engineering, Delhi Tech-

nological University, New Delhi, India
Sujata S. Alegavi Department of Electronics Engineering, Thakur College of
Engineering and Technology, Mumbai, India
Editors and Contributors xv

Sumita Basu Department of Mathematics, Bethune College, University of Calcutta,

Kolkata, West Bengal, India;
Guest Faculty, Department of Mathematics, The Heritage College, University of
Calcutta, Kolkata, West Bengal, India
Pratap Bethi OPG Power Generation Pvt. Ltd., Chennai, India
Vinayak Ashok Bharadi Department of Information Technology, Finolex
Academy of Management and Technology, Ratnagiri, Maharashtra, India
Anirban Bhowmik Department of Computer Science, M.U.C. Women’s College,
Burdwan, WB, India
Bharat Bhushan School of Engineering and Technology, Sharda University,
Greater Noida, India
Reddy Mounika Bommisetty Department of Electronics and Communication
Engineering, National Institute of Technology Tiruchirappalli, Tiruchirappalli, India
E. Brumancia Sathyabama Institute of Science and Technology, Chennai, India
Bithi Chattaraj Department of Mathematics, Bethune College, University of
Calcutta, Kolkata, West Bengal, India
Kalyan Chatterjee Department of Electrical Engineering, Indian Institute of Tech-
nology (ISM), Dhanbad, Jharkhand, India
Joydeep Dey Department of Computer Science, M.U.C. Women’s College,
Burdwan, WB, India
Awishkar Ghimire Department of Computer Science and Engineering, Delhi
Technological University, New Delhi, India
V. P. Gladis Pushparathi Department of Computer Science and Engineering,
Velammal Institute of Technology, Chennai, India
Richa Goel Amity International Business School, Amity University Uttar Pradesh,
Noida, India
C. Ismayil Siyad Pondicherry Engineering College, Puducherry, India
Nancy Sharon Johnson Easwari Engineering College, Chennai, India
Sunil Karforma Department of Computer Science, The University of Burdwan,
Burdwan, WB, India
Sukhpreet Kaur Department of Computer Science & Engineering, Chandigarh
Engineering College Landran, Mohali, India
Ila Kaushik Krishna Institute of Engineering & Technology, Ghaziabad, India
Ashish Khare Department of Electronics and Communication Engineering,
University of Allahabad, Prayagraj, India
xvi Editors and Contributors

Koushal Kumar Sikh National College, Qadian, Guru Nanak Dev University,
Amritsar, Punjab, India
Sushil Kumar Chandigarh Engineering College, Landran, India
Yogesh Kumar Department of Computer Science & Engineering, Chandigarh
Group of Colleges, Mohali, India
E. Madhusudhana Reddy Department of Computer Science and Engineering,
Guru Nanak Institutions Technical Campus, Hyderabad, India
L. Mary Gladence Sathyabama Institute of Science and Technology, Chennai,
India
A. Mathumathi Easwari Engineering College, Chennai, India
Abhishek Mehta Faculty of IT and Computer Science, Parul University, Vadodara,
Gujarat, India
Anand Nayyar Graduate School, Duy Tan University, Da Nang, Vietnam;
Faculty of Information Technology, Duy Tan University, Da Nang, Vietnam
P. Palanisamy Department of Electronics and Communication Engineering,
National Institute of Technology Tiruchirappalli, Tiruchirappalli, India
Bhagwati Prasad Pande Department of Computer Applications, LSM Govern-
ment PG College, Pithoragarh, Uttarakhand, India
S. Poopandian Infosys Pvt., Ltd., Chennai, India
J. Prakash Department of CSE, PSG College of Technology, Coimbatore, India
D. Praveena Department of Information Technology, R.M.D Engineering College,
Kavaraipettai, India
R. Priyatharshini Easwari Engineering College, Chennai, India
Gladis Pushparathi Velammal Institute of Technology, Chennai, India
Hari Mohan Rai Department of Electronics and Communication Engineering,
Krishna Engineering College, Ghaziabad, UP, India
Trupti Rathod Vidyabharti Trust College of MCA, Gujarat Technological Univer-
sity, Bardoli, Gujarat, India
N. T. Renukadevi Department of CT-UG, Kongu Engineering College, Erode,
India
Seema Sahai Amity International Business School, Amity University Uttar
Pradesh, Noida, India
M. Sanjay Sivakumar Easwari Engineering College, Chennai, India
Arindam Sarkar Department of Computer Science and Electronics, R.K.M.
Vidyamandira, Belur Math, WB, India
Editors and Contributors xvii

R. Senthil Kumar Easwari Engineering College, Chennai, India

A. Shane Sam St. Joseph’s College of Engineering, Chennai, India
S. Shanmuga Priya M.I.E.T Engineering College, Trichy, India
Nikhil Sharma Department of Computer Science & Engineering, Delhi Techno-
logical University, Delhi, India
Gurinder Singh Amity International Business School, Amity University Uttar
Pradesh, Noida, India
S. Tamilselvan Pondicherry Engineering College, Puducherry, India
S. Thanga Ramya Department of Information Technology, R.M.D Engineering
College, Kavaraipettai, India
Surendrabikram Thapa Department of Computer Science and Engineering, Delhi
Technological University, New Delhi, India
Varun Totakura Tata Consultancy Services, Hyderabad, Telangana, India
Bharti Verma HMR Institute of Technology & Management, Delhi, India
B. Vinoth Kumar Department of IT, PSG College of Technology, Coimbatore,
India
Abbreviations

2D/3D Two Dimension/Three Dimension

3D Three-dimensional
AANN Adaptive Artificial Neural Network
ABC Artificial Bee Colony
ACO Ant Colony Optimization
AD Alzheimer’s Disease
AdaBoost Adaptive Boost
ADNI Alzheimer’s Disease Neuroimaging Initiative
AES Advanced Encryption Standard
AFBNN Adaptive Firefly Backpropagation Neural Network
AFCNN Advanced Fuzzy Cellular Neural Network
AI Artificial Intelligence
AIC Akaike’s Information Criterion
ANFIS Adaptive Neuro Fuzzy Inference System
ANFIS Adaptive Neuro-fuzzy Inference System
ANN Artificial Neural Network
ANN Artificial Neural Networks
API Application Programming Interface
AR Auto Regressive
ARIMA Auto Regressive Integrated Moving Averages
ARMA Auto Regressive Moving Averages
ARTISANA Artificial Intelligence in Sleep Analysis Algorithm
ASIN Amazon Standard Identification Number
ATD Artificial Training Data
BA Bat Algorithm
BFO Bacterial Foraging Optimization
BLE Bluetooth Low Energy
BMRI Brain Magnetic Resonance Imaging
BP Back Propagation
BPNN Back Propagation Neural Network
BRIEF Binary Robust Independent Elementary Features

xix
xx Abbreviations

BRIoT Behavior Rule specification-based misbehavior detection for IoT

BRISK Binary Robust Invariant Scalable Key point
C Cases
CAD Computer-Aided Diagnosis
CANFIS Coactive Neuro-Fuzzy Inference System
CBDHDSE Cross-Border Digital Healthcare Decision Support Ecosystem
CBIR Content-Based Image Retrieval
CBSE Component-Based Software Engineering
CBVR Content-Based Video Retrieval
CDRC Consumer Data Research Centre
CE Convergent Encryption
CFT Clustering Fusion Technique
CGS Chronic-disease Guideline Support
CG-SVM Coarse Gaussian SVM
CIR Channel Impulse Response
CK Component Kosh
CLARA Clustering Large Applications
CloudSim Cloud Simulator
CM Classification Methods
CMS Content Management System
CNN Convolution Neural Network
CNN Convolutional Neural Networks
CNTK Cognitive Toolkit
COVID-19 Novel Coronavirus Disease
CPNN Convolutional Pre-processing Neural Network
CPU Central Processing Unit
CR Coverage Ratio
CRNN Convolutional Recurrent Neural Network
CSI Channel State Information
CSP Cloud Service Provider
CT Computed Tomography
CW Continuous Wavelet
DAM Data-Adaptive Matrix
DAS Distance between Adjacent Signals
DBN Deep Belief Network
DBSCAN Density-Based Spatial Clustering of Applications with Noise
DDoS Distributed Denial of Attack
DDPA Deduplication Processing Algorithm
DFM Data Fusion of MRI
DICOM Digital Imaging and Communications in Medicine
DL Deep Learning
DOI Document Identifier
DS Dataset Size
DSA Digital Signature Algorithm
DSRBACA Dynamic Spatial Role-Based Access Control Algorithm
Abbreviations xxi

DT Decision Tree
DT Decision Trees
DW-LSTM Dynamic Watermarking LSTM
DWT-RBFNN Discrete Wavelet Transform–Radial Basis Function Neural
Network
ECDH Elliptic Curve Diffie–Hellman
ECG Electrocardiogram
ECG Electrocardiography
EHR Electronic Health Record
EL Ensemble Learning
ELM-IPSO Extreme learning machine-improved particle swarm optimization
EMG Electromyography
ERDIP Electronic Record Development and Implementation Program
FA False Alarm
FAST Features from Accelerated Segment Test
FCFS First Come First Serve
FCM Fuzzy c-means
FCM Fuzzy C Means
FCM Fuzzy C-Mean
FCMC Fuzzy C-Means Clustering, MLS
FDR False Discovery Rate
FFNN Feed Forward Neural Networks
FIVR Fine-grained Incident Video Retrieval
FKM Fuzzy K Means
FN False Negative
FP False Positive
FS Feature Selection
FS Fuzzy Systems
FT Fourier Transform
FTISPAM Fuzzy Time Interval Sequential Pattern Mining
FWT Fast Wavelet Transform
GA Genetic Algorithm
GA-ANN Genetic Algorithm–Artificial Neural Network
GBHSP Gene-Based Health Screening Personalization
GLCM Gray Level Co-occurrence Matrix
GLOGTH Global and Local Oriented Gabor Texture Histogram
GOA Grasshopper Optimization
GPU Graphical Processing Units
HBR Harvard Brain Repository
HCAPN Hierarchical Context-Aware Aspect-Oriented Petri Net
HGA Hybrid Genetic Algorithm
HIPAA Health Insurance Portability and Accountability Act
HOG Histogram of Oriented Gradients
HOG Histogram of the oriented gradient
HRF High-Resolution Fundus
xxii Abbreviations

HSO Hybrid Swarm Optimization Algorithm

HSP Health Screening Personalization
HSV Hue, Saturation, Value
I Images
I/O Input/Output
IaaS Infrastructure as a Service
ICA Independent Component Analysis
ICT Information Communication Technology
IDF Inverse Document Frequency
IIoT Industrial IoT
IoT Internet of Things
IoTD IoT Devices
KM Kernel Machine
KMC K-Means Clustering
KNN K-Nearest Neighbor
LBP Local Binary Pattern
LBPV LBP Variance
LC Linear Classifiers
LDPC Low Density Parity Check
LIA Lattice Implication Algebra
LOG Laplacian of Gaussian
LOS Line of Sight
LPLA Lightweight Physical Layer Authentication
LSTM Long Short Term Memory
LSTM Long Short Term Memory Networks
LSTM Long Short-Term Memory
LT Luby Transform
LTVP Linguistic Truth-Valued Propositions
MA Moving Average
MAD Mean Absolute Deviation
MAE Mean Absolute Error
MAFD Modify Adaptive Frame Differencing
MAP Mean Arterial Pressure
mAP Mean Average Precision
MATLAB Matrix Laboratory
MCI Most Common Imputation
MD Multipath Delay
MDPSO Modified Discrete Particle Swarm Optimization
MEI Mean Imputation
MI Moment Invariant
ML Machine Learning
MLP Multilayer Perceptron
MLP Multi-layer Perceptron
MOGA Multi-Objective Genetic Algorithm
MR Magnetic Resonance
Abbreviations xxiii

MRI Magnetic Resonance Imaging

MRS Magnetic Resonance Spectroscopy
MSE Mean Square Error
MSE Mean Squared Error
MSER Maximally Stable Extremal Regions
MSER Maximum Stable Extremal Regions
MSNE Mixed Strategy Nash Equilibrium
NBC Naïve Bayes Classifiers
NBTS National Brain Tumor Society
NER Named-Entity Recognition
NFC Near Field Communication
NLOS Non-Line of Sight
NLP Natural Language Processing
NR Novelty Ratio
NTM New Thresholding Method
OASIS Open Access Series of Imaging Studies
OCC One Class Classification
OCR Optical Character Recognition
OS Operating System
OSCR Ontology of Software Component for Retrieval
OWL Web Ontology Language
P Patients
PA Physician’s Assistant
PACS Picture Archiving and Communication Systems
PCA Principal Component Analysis
PCC Pearson Correlation Coefficient
PD Proton Density
PEAM Performance Evaluation Assessment Metrics
PET Positron Emission Tomography
PID Pima Indian Diabetes
PLS Physical Layer Security
PM Practice Management
PN Pseudo Noise
POS Part of Speech
PPA Peripapillary Atrophy
PPV Positive Predicted Value
PSA Particle Swarm Algorithm
PSD Power Spectral Density
PSO Particle Swarm Optimization
QLIA Quasi-Lattice Implication Algebra
QLTVP Quasi-Linguistic Truth-Valued Propositions
QoS Quality of Service
RACS Role-Based Access Control Scheme
RAI RAdiometric Identification
RBAC Role-Based Access Control
xxiv Abbreviations

RBF Radial Basis Function

RBM Restricted Boltzmann Machines
RCM Revenue Cycle Management
RCT Randomized Control Trials
RCUO Retrieval of the Component Using Ontology
ReLU Rectified Linear Unit
RF Random Forest
RFID Radio Frequency Identification
RGB Red, Green, and Blue
RGM Region Growing Method
RIM Reference Information Model
RMS Root Mean Square
RMSE Root Mean Squared Error
RNN Recurrent Neural Networks
ROI Region of Interest
RP-CNN CNN and Recurrence Plot
RR Round Robin Algorithm
RS Reed Solomon
RSA Rivest Shamir Adleman
RSS Received Signal Strength
SC Sentiment Classification
SC Soft Computing
SCI Similarity Computation Index
SD Standard Deviation
SGLD Spatial Gray Level Dependence Matrix
SHA Secure Hash Algorithm
SI Swarm Intelligence
SIFT Scale Invariant Feature Transform
SLA Service Level Agreement
SM Segmentation Methods
SOA Service-Oriented Architecture
SOAP Simple Object Access Protocol
SOM Self-Organizing Map (SOM)
SPAM Sequential Pattern Mining
SPARQL SPARQL Protocol and RDF Query Language
SPDTH Similarity-Preserving Deep Temporal Hashing
SPECT Single Photon Emission Computed Tomography
SPECT Single Proton Emission Computed Tomography
SR Sparse Representation
STFT Short-Time Fourier Transform
SURF Speeded Up Robust Feature
SVM Support Vector Machine
TF Term Frequency
TF-IDF Term Frequency–Inverse Document Frequency
TLBO Teaching Learning-Based Optimization
Abbreviations xxv

TN True Negative
TP True Positive
TPU Tensor Processing Units
U.S United States
UI User Interface
UK United Kingdom
ULBP Uniform Local Binary Pattern
URL Uniform Resource Locator
US Ultrasound
USRP Universal Software Radio Peripheral
VOC Pascal Visual Object Classes
WHO World Health Organization
WLBP Weber Local Binary Pattern
YOLO You Only Look Once
Soft Computing Techniques for IoT
Devices
A Wearable Assistive Device for Safe
Travel Using Transfer Learning and IoT
for Visually Impaired People

R. Priyatharshini, R. Senthil Kumar, M. Sanjay Sivakumar, A. Mathumathi,

and Nancy Sharon Johnson

Abstract There are a number of discomforts faced by visually impaired people

every day in both indoor and outdoor surroundings. Assistive Technology for people
with visual disabilities plays a vital role in their Independent living. Various systems
have been developed to help them to live a better life even with the low or no
vision. Visual Mobility plays a vital role in their Independent living such as arrival
of buses, recognizing the Route Number from number plates, finding the doorstep
in a train etc. This paper proposes an efficient approach for recognizing the Route
details from a bus that helps in easy commuting in bus stations for the visually
impaired. The You Only Look Once (YOLO v3) model can be used to detect the
real time object bus and also segment the bus name board i.e. the Region of Interest
(ROI) by transfer learning it with a custom dataset. To further improve the pattern
recognition accuracy the numbers of anchoring boxes were increased from 3 to 5 in
all the strides which provided more precise results. The route details present in the
segmented ROI is converted into text using Tesseract tool that uses LSTM (Long
Short Term Memory) engine for producing text from recognized characters in image
and then regular expression is used to filter out only the bus number (alphanumeric
or numeric values). The obtained bus number is converted as voice output (using text
to speech library e-speak) along with bus details based on the bus number extracted
from the number plate. An approach to detect the door step at railway compartments
has also been proposed for visual mobility. During the travel, the device can be further
used to know the current location and the distance remaining to reach the destination.

Keywords Pattern recognition · Deep learning · Image processing · Transfer

learning

1 Introduction

Globally the number of humans of all ages who are visually impaired is estimated to
be around 285 million. Partial or complete loss of vision is common among most of

R. Priyatharshini (B) · R. Senthil Kumar · M. Sanjay Sivakumar · A. Mathumathi · N. S. Johnson

Easwari Engineering College, Chennai, India

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 3

S. Dash et al. (eds.), Advanced Soft Computing Techniques in Data Science, IoT and Cloud
Computing, Studies in Big Data 89, https://doi.org/10.1007/978-3-030-75657-4_1
4 R. Priyatharshini et al.

the old age people and it is difficult for a person to deal with such kind of impairment
after experiencing niche lifestyle with perfect vision. Visually impaired people has
less ability to perform everyday task like other people so their quality of life gets
affected and lose the ability to experience the necessary things in the world. Visually
impaired people are comfortable only with their familiar places because they know
the layout of the place and they would have memorized it, but when it comes to new
environment they may find several difficulties. There are a number of technological
advancements and everyday developments which are focused on providing a better
lifestyle for handicapped and old age peoples. Although there are bright ideas and
thoughts, there are practical difficulties in implementing them as a product that can
be used every day by this section of the population.
One of the leading difficulties for visually impaired people in day to day scenarios
is to travel alone without any dependencies from others. Some of them can have the
blurry scene of a bus arriving at the bus terminal, still recognizing the bus route
remains a challenging and an impossible task for them. People also feel embarrassed
to request others to help them to identify bus numbers. Another major problem
they face is in the railway stations. It is difficult for visually impaired person to
choose which exit side of the train to use to get out from the train. Sometimes they
accidentally trip over the wrong exit side and fall down to the track.
These types of incidents has occurred several times. These all leads to rise in
demand to provide a simple and powerful solution to address these problems. Most
of the time they are present alone in the station and they find it difficult to receive
any assistance from other people. Numbers of bus recognition system ideas are
introduced, like identifying the bus route with help of smartphone using Maximum
Stable Extremal Regions (MSER) applications, or in another system histogram of
the oriented gradient (HOG) identifies the codes of texts from detected text regions
and produces output as audio notification. There are also systems introduced for
bus detection using wireless networks and satellite signals. It requires installation of
sensors and other modules in bus and bus stations along with periodic maintenance.
Most of these current systems lack performance required for real time detection
and usage along with few other technical snags. There are a number of breakthrough
researches and technologies coming out throughout the year. They highly involve
attributes like quick performance, efficient methodologies, less implementation cost
and many others. Therefore, coining enhanced solutions to all kinds of problems is
not difficult and this includes for the above stated problems of visually impaired too.

2 Review of Literature

In the research field of computer vision and in other areas a number of bus recognition
systems are proposed; however, most of them use active devices and sensors such as
Global Positioning System tracking system, RFID (Radio Frequency Identification),
Beacons etc. For vision based approaches, Wongta [1] introduced a system that
uses MSER (Maximally Stable Extremal Regions) to recognize bus numbers. Their
A Wearable Assistive Device for Safe Travel … 5

solution finds all the texts in an image rather than the required ones i.e. the bus
number. Guida et al. [2] proposed a system for bus route number which uses a number
of interlinked classifiers along with adaboost in order to identify the route number
or other elements present in the front of the bus and some corrections are done to the
extracted features to obtain rectified numeric values. The object is then converted
to Hue, Saturation, Value (HSV) colorspace and then partitioning of each numeric
or digit value. Finally Optical character Recognition (OCR) is used to identify the
digits and output is produced as voice to the user. Viola and Jones [3] proposed a
face detection framework that involved three key contributions which were integral
images, ada boost and cascading classifiers. It was a simple and powerful technique
to identify faces from binary images. Pan et al. [4] proposed a bus detection system
to help the visually impaired where Histogram of the oriented gradients is used to
extract features from bus images and a cascading Support Vector Machine model
is implemented for a bus classifier to detect the bus facade in frames of windows.
Tsai and Yeh [5] also introduced a system for bus detection to support the blind.
The system included the functionalities of detection of the moving bus, bus panel
detection and text detection from the text region of the panel. The system made use of
MAFD(Modify Adaptive Frame Differencing) and was found to have great accuracy
when features were extracted from different frames of the video. In [6] Bouhmed used
an ultrasonic sensor and camera in a walking cane to identify obstacles in the path
and inform it to the blind. The system produces output through voice. Zahir et al. [7]
developed a prototype of wearable head mountable device by customizing the Virtual
Reality glass with ultrasonic sensors and HC-SR04 because it take minimum amount
of time for detection and can also find obstacle in the longer range. The prototype
is developed using Arduino. Ani et al. [8] introduced a voice assistive text reading
system that is integrated with eyeglasses. A Camera is inbuilt with the eyeglass to
capture a image and with Tesseract-OCR text is extracted from the captured image.
Open Software E-speak is used for TTS. A method for end-to-end real-time scene
text localization and recognition is presented in [9] and the robustness of the proposed
method against noise and low contrast of characters is well demonstrated using “false
positives”. In [12] a Convolutional Neural Networks (CNN) model is developed to
detect English and Thai text from natural scene images with improved accuracy
compared to the other existing methods A method for Multi-script Text Extraction
from Natural Scenes is presented in [10] which exploits collaboration of proximity
and similarity laws to create text-group hypotheses. The problem of establishing
correspondences between a pair of images taken from different viewpoints is analysed
in [11]. There are a number of techniques and various innovative ideas proposed to
help the blind. Most of these systems are dynamic and they have less difficulty when
it comes to real time usage. Our proposed system is for real time usage and uses video
recognition rather than capturing only images. The bus arrival and waiting time are
always varying therefore every frame must be checked instead of single image snaps
which would be less efficient for the detection of bus and bus board. The system
proposed can be integrated with any of the existing ones, where the components
are almost similar or it can be developed as a separate one in which any other new
features can be extended.
6 R. Priyatharshini et al.

3 Overview of Proposed System

A visually impaired person after loss of vision limits them from indulging in any
activities that involves travelling in public transport facilities. They become hesitant
to ask help from other people. For them to able to live their life independently we
have developed Assistive Device for visually impaired to people that helps them to
travel.
A single integrated system is developed to assist the visually impaired. User can
choose the required operation by commanding through voice. Raspberry Pi serves
the processing unit for all the operations and it is connected to an external portable
power supply. An approach to accurately detect the real time bus object and to
segment the ROI for extracting and recognizing the route details has been proposed
to assist visually impaired people. An enhanced version of the You Only Look Once
algorithm is proposed in this paper which detects the objects in an image at a single
instance. It is an algorithm based on regression and it was pre trained on a number
of classes.
The algorithm initially splits the image into cells of 19 × 19 grid and each cell
is responsible for predicting 5 bounding boxes. Each of the bounding boxes can be
described using 5 descriptors as shown in (Fig. 1). They are,
1. Width (bw ): Width of the bounding box.
2. Height (bh ): Height of the bounding box.
3. Center of a bounding box (bx by ): Center coordinates of the bounding box.
4. Value c: Corresponding to a class of an object (i.e. car, traffic lights)
5. Value pc : Probability that there is an object in the bounding box.
For areas under each of the bounding box, the Convolution Neural Network (CNN)
processes to identify to which class the object under the area belongs. The boxes
with low object probability (pc ) and unknown classes (c) are removed. Areas under
high overlapping boundary regions across cells along with greater pc are the detected
objects in the image. By transfer learning the YOLO v3 model with a custom dataset,
the same above procedure is involved to detect the bus board in our system. The
detected bus board image is then used to obtain the text with help of tesseract optical
character recognition tool. It gathers the outlines into blobs by nesting. Blobs are
further ordered into lines of text and each of these text lines are drilled down into
probable terms and further into characters. Multiple passes are involved before finally
predicting the character, words and the sentences. Latest versions of tesseract use
deep learning frameworks which are accurate and fast. Output from the OCR is
filtered with regular expressions which are constrained to find at maximum three
characters length data that are either numeric or alphanumeric. The obtained data is
the bus number as bus route number usually falls under these constraints in modern
bus route naming systems. The identified bus number is informed to blind through
text to speech method (Fig. 2).
Speaker/headphones are present to give output notifications. Microphone is used
to obtain the voice input from the user. If the voice input is found to be difficult for
A Wearable Assistive Device for Safe Travel … 7

Fig. 1 A 19 × 19 grid with descriptors (Source https://pylessons.com)

the user or if there is a lot of noise, button inputs can be added to the system. GPS can
be added as an extended feature which can inform the user where they are currently
located and to know the remaining time it would take to reach the destination.
For railway platform detection, an Ultrasonic Sensor is present at a certain angle
facing ground every time. It is also connected to the raspberry pi and it gives the
frequent depth measures from ground to the pi in which the depth identification
algorithm is executed. Speaker/headphones are present to give output notifications.
Microphone is used to obtain the voice input from the user. If the voice input is found
to be difficult for the user or if there is lot of noise, button inputs can be added to the
8 R. Priyatharshini et al.

Fig. 2 Proposed system

components architecture

system. GPS can be added as an extended feature which can inform the user where
they are currently located and to know the remaining time it would take to reach the
destination.

4 Proposed Methodology

The Proposed system mainly consists of two sub modules that are chosen for different
scenarios by the user. A methodology to detect the real time object (bus) and segment
the number plate (region of interest) has been proposed for recognizing the route
number in a bus. In this approach, there are three sub modules involved. First the bus
object is recognized and the bus board is segmented for which the core functionality
is based on Convolution Neural Network and transfer learning. For the next module,
tesseract optical character recognition engine is used to recognize the text and finally
with regular expressions, the bus route number is recognized (Fig. 3).

4.1 Bus Board Detection for Route Information Extraction

This subsystem deals with identifying the bus board in real time. A high definition
wide angle camera is mounted to the head wear and is connected to the pi. The camera
is used to capture the bus arrival scene and from the video frames, ROI detection and
extraction takes place.
More than 400 images were used to create a customized model. Sample images
from training dataset is shown in Fig. 4. The images obtained were in a daylight envi-
ronment. Preprocessing is done to remove the blurry and dark images before initiating
the training process. The region of interests were marked and their labeling was given.
Pascal visual object classes (VOC) is a format for providing object detection data, i.e.
A Wearable Assistive Device for Safe Travel … 9

Fig. 3 Overview of proposed system

images with bounding boxes. Using the annotation tool, the bus boards are marked
with bounding boxes in each image manually and these values are saved in a xml file.
The bus boards are the Region of Interests for our model. The process involved in
building the model for bus board detection module using transfer learning is shown
in Fig. 5.
In order to ensure that the trained custom models have better detection accuracy,
transfer learning from a pretrained YOLO v3 model was involved in the training.
Transfer learning helps us to construct precise models in a timesaving way. Using
transfer learning instead of starting afresh, we start from patterns that have been
learned while addressing a different problem. Pre-trained model are used to initiate
the transfer learning process. A pre-trained model is the one that is used to address a
problem which is similar to a problem that we want to work out. A lot of computa-
tional costs are involved in training such models therefore models from well estab-
lished literatures are used (e.g. MobileNet, VGG, Inception, YOLO) as shown in
Fig. 6.
10 R. Priyatharshini et al.

Fig. 4 Sample training images

Fig. 5 System Architecture of proposed bus board detection module

A Wearable Assistive Device for Safe Travel … 11

Fig. 6 Traditional ML versus transfer learning (Source https://towardsdatascience.com)

You Only Look Once (YOLO) is a fully convolutional network model and its
outputs are generated by applying a 1 × 1 kernel on a feature map. In YOLO v3, the
detection is done by applying 1 × 1 detection kernels on feature maps of three
different sizes at three different places in the model network. These places are known
as strides. Each of these strides are used for processing large, medium and small
images respectively which is shown in Fig. 7.

Fig. 7 YOLO v3 network architecture (Source https://towardsdatascience.com)

12 R. Priyatharshini et al.

Transfer learning with annotated images was completed along with a pretrained
YOLOv3 model. The obtained customized model is saved for later usage in bus
recognition and bus board identification. After training is done, the model is evaluated
for accuracy. They are identified based on the decrease in the validation loss. In most
cases, the lower or less is the loss, the more accurate the model will be in detecting
required objects. However, some models may experience overfitting and have lower
losses. Therefore, to ensure that the best model is picked for our custom detection, we
evaluate with mAP (mean Average Precision) of all the trained models saved in the
detection folder. The detection accuracy of the model is improved with better mAP.

4.1.1 Bus Object Detection

The detection mechanism in a deep learning module mainly has two phases. In the
first phase, an image is taken as an input and a number of blocks or boxes are formed
around the possible entities with statistical features to label objects. Then this output
is predicted to detect the object with its class name.
The feature to identify the bus object comes from the pretrained YOLO v3 model
which has a Convolutional Neural Network as underlying layer and there is no
requirement to train the bus images separately for its identification process. The
initial phase consists of a number of convolution layers and pooling layers through
which the image pixel values move in an array format. Activation functions are

Fig. 8 Detecting ROI from

random image (Source
https://heartbeat.friz.ai)
A Wearable Assistive Device for Safe Travel … 13

Fig. 9 a Case 1—No buses recognized. b Case 2—Two buses recognized

present in the middle and through all of these, feature learning takes place. The
next phase has flattened and fully connected layers with weights and before the final
output classification is given soft maxing is done as shown in Fig. 8. The system
moves to next processing only when the bus object is present in the video frame.
Results of bus detection module is shown in Fig. 9.

4.1.2 Bus Board Segmentation

Our customized model is used for detection of bus boards from the video frames. For
every frame, the detection algorithm is performed. The algorithm initially recognizes
14 R. Priyatharshini et al.

Fig. 10 Sample output of detected ROI (bus name board)

the bus. Then in the image, segmentation as 13 × 13 grids is done and each cell has
a maximum of five bounding boxes. Prediction for each of the bounding box areas is
made and scores are provided. If it is identified to be below the threshold value from
the trained model, they are neglected. For the areas with overlapping bounding box
boundaries, features are collected together and prediction scores are given. Along
with them, confidence score is also present which gives distinction between that
class (ROI) and rest of the background image. The bus board (Region Of Interest)
is finally detected if it is present in the image. If there are multiple bus boards in
the frame, the same technique is applied and all of them are extracted and they are
saved as separate grayscale images which are used for the next operation i.e., Optical
Character Recognition (OCR). It is necessary that the detected bus board image
must not be blurry or dark as it makes it difficult for OCR operation. The results of
segmented bus board is shown in Fig. 10.

4.2 Route Recognition

For obtaining text from an image. Tesseract OCR is used in our system. The latest
recognition engine of tesseract is based upon Long Short Term Memory (LSTM). For
the system pytesseract module is chosen as it acts as a replacement for the command
line tesseract with specified configuration arguments. As all of the other modules
are python based, we also make use of Tesseract OCR in python as pytesseract.
A Wearable Assistive Device for Safe Travel … 15

Tesseract makes a two step approach for text detection. It makes use of adaptive
recognition where in the initial step it identifies the pattern in letters, pixels, words
and sentences, hence recognizing the characters. If it is not sure or successful in
finding some characters, in the next step it tries to fill in with the character or word
that matches the word or sentence context. In character recognition step the outlines
are arranged into blobs by the nesting process. These blobs are then structured into
lines of text and each of these text lines are broken into possible words and further into
characters as shown in Fig. 11. Multiple passes are involved before finally predicting
the character, words and the sentences. The process involves a number of iterations
before final text is produced. In each iteration, the accuracy of recognition improves
and finally finely converted text data is obtained.
In our system all the information present in the bus board image is obtained by
operating the tesseract tool on the image. This obtained text data is stored as string

Fig. 11 Working of Tesseract OCR (Source https://blog.cedric.ws)

16 R. Priyatharshini et al.

type data for further processing. For obtaining text with good accuracy, quality images
are required from the previous module.

4.3 Text to Speech Conversion for Route Identification

Regular expressions are defined to obtain the numeric or alphanumeric bus number
from the text data stored in a text data format. A regular expression is a special text
string for describing a search pattern. Strings of text are compared to the pattern
in order to identify the string that matches the logical pattern defined by regular
expression. On the basis of these comparisons the regular expressions can be used to
identify strings of text that meet specific requirements or to validate that strings meet
a required pattern. It is very important to define regular expressions properly because
even if a single letter or number is missing, the route number becomes ambiguous.
The process involved in extracting the bus route and converting it in to speech output
is shown in Fig. 12.
The constraint laid for our regular expression is to work in a way such that it
only extracts maximum 3 characters that are either numeric or alphanumeric and are
continuous characters present one after another. The constraint is based in such a
way because the bus numbers in the modern transport system are in numeric or in
alphanumeric format and their characters length are usually two or three characters
long. This constraint can be adjusted based on the locality requirements.
A Wearable Assistive Device for Safe Travel … 17

Fig. 12 Process involved in bus route identification

18 R. Priyatharshini et al.
A Wearable Assistive Device for Safe Travel … 19

An additional module, a database can be added to the system so that with the
extracted bus number, all other information regarding the bus can be provided to the
user. The bus number serves as the primary key with which the database is queried
to provide details like destination location, stops present in the middle, ticket cost
and other necessary information.

4.4 An IOT Based Railway Platform Detection

This subsystem involves detection of railway platform when train stops at the stations.
An ultrasonic sensor is mounted to the headwear like cap/hat at a certain angle facing
the ground. Initially the user’s height data is loaded in pi memory. This height along
with a small variable length is added together and the value is stored as a prefixed
threshold value. The variable length is the difference in height from platform to the
train door. An ultrasonic sensor measures the distance based on the working given
in Fig. 13.
The transmitter sends out waves which travels at the speed of sound waves
(0.034 cm/µs) and the receiver obtains them when they are reflected back by an
object or surface that is present in their path. The time difference (t) is noted and
along with the known speed of the sound, distance value can be calculated (v*t/2)
in the pi. The algorithm for the platform detection mainly identifies whether this
measured distance exceeds the prefixed threshold value. If so, the user is notified
quickly through speakers to get down on the other side of the train. Ultrasonic sensor
is very much responsive therefore this operation takes place quickly. It is a light
weight and simple process to assist the blind for safe exit from trains. The same
methodology can be extended in future also for other transports. Some ambiguity
may arise if there are obstacles present between the sensor and the ground/surface.

Fig. 13 Working of an ultrasonic sensor (Source https://howtomechatronics.com)

20 R. Priyatharshini et al.

This can be overcome by adding multiple ultrasonic sensors to the system as shown
in Fig. 14.

Fig. 14 Flow diagram for platform detection module

A Wearable Assistive Device for Safe Travel … 21
22 R. Priyatharshini et al.

4.5 Assistive Device for Visual Mobility and Safe Travel

A single integrated system is developed and the whole system is attached to a cap to
assist the visually impaired. User can choose the required operation by commanding
through voice or using a button. Raspberry Pi serves the processing unit for all the
operations and it is connected to an external portable power supply.
For bus identification, a High Definition wide angle camera is used and the real
time feed starts once bus number identification operation is chosen to obtain the bus
number and all other details.
For railway platform detection, an Ultrasonic Sensor is present at a certain angle
facing ground every time. It is also connected to the raspberry pi and it gives the
frequent depth measures from ground to the pi in which the depth identification
algorithm is executed. Speaker/headphones are present to give output notifications.
Microphone is used to obtain the voice input from the user. If the voice input is found
to be difficult for the user or if there is lot of noise, button inputs can be added to the
system. GPS can be added as an extended feature which can inform the user where
they are currently located and to know the remaining time it would take to reach the
destination as shown in Fig. 15.

Fig. 15 Results of IOT based railway platform detection

Another Random Scribd Document
with Unrelated Content
and the Athenian people,—an exasperation heightened by
circumstances immediately preceding. For the resolution to send
auxiliaries into Laconia, when the Lacedæmonians first applied for
them, had not been taken without considerable debate at Athens:
the party of Periklês and Ephialtês, habitually in opposition to Kimon,
and partisans of the forward democratical movement, had strongly
discountenanced it, and conjured their countrymen not to assist in
renovating and strengthening their most formidable rival. Perhaps
the previous engagement of the Lacedæmonians to invade Attica on
behalf of the Thasians may have become known to them, though
not so formally as to exclude denial; and even supposing this
engagement to have remained unknown at that time to every one,
there were not wanting other grounds to render the policy of refusal
plausible. But Kimon, with an earnestness which even the philo-
Laconian Kritias afterwards characterized as a sacrifice of the
grandeur of Athens to the advantage of Lacedæmon,[615] employed
all his credit and influence in seconding the application. The
maintenance of alliance with Sparta on equal footing,—peace among
the great powers of Greece, and common war against Persia,—
together with the prevention of all farther democratical changes in
Athens,—were the leading points of his political creed. As yet, both
his personal and political ascendency was predominant over his
opponents: as yet, there was no manifest conflict, which had only
just begun to show itself in the case of Thasos, between the
maritime power of Athens, and the union of land-force under Sparta:
and Kimon could still treat both of these phenomena as coexisting
necessities of Hellenic well-being. Though no way distinguished as a
speaker, he carried with him the Athenian assembly by appealing to
a large and generous patriotism, which forbade them to permit the
humiliation of Sparta. “Consent not to see Hellas lamed of one leg,
and Athens drawing without her yoke-fellow;”[616] such was his
language, as we learn from his friend and companion, the Chian
poet Ion: and in the lips of Kimon it proved effective. It is a speech
of almost melancholy interest, since ninety years passed over before
such an appeal was ever again addressed to an Athenian assembly.
[617] The despatch of the auxiliaries was thus dictated by a generous
sentiment, to the disregard of what might seem political prudence:
and we may imagine the violent reaction which took place in
Athenian feeling, when the Lacedæmonians repaid them by singling
out their troops from all the other allies as objects of insulting
suspicion,—we may imagine the triumph of Periklês and Ephialtês,
who had opposed the mission,—and the vast loss of influence to
Kimon, who had brought it about,—when Athens received again into
her public assemblies the hoplites sent back from Ithômê.
Both in the internal constitution, indeed,—of which more
presently,—and in the external policy of Athens, the dismissal of
these soldiers was pregnant with results. The Athenians immediately
passed a formal resolution to renounce the alliance between
themselves and Lacedæmon against the Persians. They did more:
they looked out for land enemies of Lacedæmon, with whom to ally
themselves. Of these by far the first, both in Hellenic rank and in
real power, was Argos. That city, neutral during the Persian invasion,
had now recovered from the effects of the destructive defeat
suffered about thirty years before from the Spartan king Kleomenês:
the sons of the ancient citizens had grown to manhood, and the
temporary predominance of the Periœki, acquired in consequence of
the ruinous loss of citizens in that defeat, had been again put down.
In the neighborhood of Argos, and dependent upon it, were situated
Mykenæ, Tiryns, and Midea,—small in power and importance, but
rich in mythical renown. Disdaining the inglorious example of Argos,
at the period of danger, these towns had furnished contingents both
to Thermopylæ and Platæa, which their powerful neighbor had been
unable either to prevent at the time, or to avenge afterwards, from
fear of the intervention of Lacedæmon. But so soon as the latter was
seen to be endangered and occupied at home, with a formidable
Messenian revolt, the Argeians availed themselves of the opportunity
to attack not only Mykenæ and Tiryns, but also Orneæ, Midea, and
other semi-dependent towns around them. Several of these were
reduced; and the inhabitants robbed of their autonomy, were
incorporated with the domain of Argos: but the Mykenians, partly
from the superior gallantry of their resistance, partly from jealousy
of their mythical renown, were either sold as slaves or driven into
banishment.[618] Through these victories Argos was now more
powerful than ever, and the propositions of alliance made to her by
Athens, while strengthening both the two against Lacedæmon,
opened to her a new chance of recovering her lost headship in
Peloponnesus. The Thessalians became members of this new
alliance, which was a defensive alliance against Lacedæmon: and
hopes were doubtless entertained of drawing in some of the habitual
allies of the latter.
The new character which Athens had thus assumed, as a
competitor for landed alliances, not less than for maritime
ascendency, came opportunely for the protection of the neighboring
town of Megara. It appears that Corinth, perhaps instigated, like
Argos, by the helplessness of the Lacedæmonians, had been making
border encroachments on the one side upon Kleônæ, on the other
side upon Megara:[619] on which ground the latter, probably
despairing of protection from Lacedæmon, renounced the
Lacedæmonian connection, and obtained permission to enrol herself
as an ally of Athens.[620] This was an acquisition of signal value to
the Athenians, since it both opened to them the whole range of
territory across the outer isthmus of Corinth to the interior of the
Krissæan gulf, on which the Megarian port of Pegæ was situated,
and placed them in possession of the passes of Mount Geraneia, so
that they could arrest the march of a Peloponnesian army over the
isthmus, and protect Attica from invasion. It was, moreover, of great
importance in its effects on Grecian politics: for it was counted as a
wrong by Lacedæmon, gave deadly offence to the Corinthians, and
lighted up the flames of war between them and Athens; their allies,
the Epidaurians and Æginetans, taking their part. Though Athens
had not yet been guilty of unjust encroachment against any
Peloponnesian state, her ambition and energy had inspired universal
awe; while the maritime states in the neighborhood, such as Corinth,
Epidaurus, and Ægina, saw these terror-striking qualities threatening
them at their own doors, through her alliance with Argos and
Megara. Moreover, it is probable that the ancient feud between the
Athenians and Æginetans, though dormant since a little before the
Persian invasion, had never been appeased or forgotten: so that the
Æginetans, dwelling within sight of Peiræus, were at once best able
to appreciate, and most likely to dread, the enormous maritime
power now possessed by Athens. Periklês was wont to call Ægina
the eyesore of Peiræus:[621] but we may be very sure that Peiræus,
grown into a vast fortified port, within the existing generation, was
in a much stronger degree the eyesore of Ægina.
The Athenians were at this time actively engaged in prosecuting
the war against Persia, having a fleet of no less than two hundred
sail, equipped by or from the confederacy collectively, now serving in
Cyprus and on the Phenician coast. Moreover, the revolt of the
Egyptians under Inaros, about 460 b. c., opened to them new means
of action against the Great King; and their fleet, by invitation of the
revolters, sailed up the Nile to Memphis, where there seemed at first
a good prospect of throwing off the Persian dominion. Yet in spite of
so great an abstraction from their disposable force, their military
operations near home were conducted with unabated vigor: and the
inscription which remains,—a commemoration of their citizens of the
Erechtheid tribe, who were slain in one and the same year, in
Cyprus, Egypt, Phenicia, the Halieis, Ægina, and Megara,—brings
forcibly before us that energy which astonished and even alarmed
their contemporaries. Their first proceedings at Megara were of a
nature altogether novel, in the existing condition of Greece. It was
necessary for the Athenians to protect their new ally against the
superiority of Peloponnesian land-force, and to insure a constant
communication with it by sea; but the city, like most of the ancient
Hellenic towns, was situated on a hill at some distance from the sea,
separated from its port Nisæa by a space of nearly one mile. One of
the earliest proceedings of the Athenians was to build two lines of
wall, near and parallel to each other, connecting the city with Nisæa,
so that the two thus formed one continuous fortress, wherein a
standing Athenian garrison was maintained, with the constant means
of succor from Athens in case of need. These “long walls,” though
afterwards copied in other places, and on a larger scale, were at that
juncture an ingenious invention, for the purpose of extending the
maritime arm of Athens to an inland city.
The first operations of Corinth, however, were not directed
against Megara. The Athenians having undertaken a landing in the
territory of the Halieis, the population of the southern Argolic
peninsula, bordering on Trœzen and Hermionê, were defeated on
land by the Corinthian and Epidaurian forces: possibly it may have
been in this expedition that they acquired possession of Trœzen,
which we find afterwards in their dependence, without knowing
when it became so. But in a sea-fight which took place off the island
of Kekryphaleia, between Ægina and the Argolic peninsula, the
Athenians gained the victory. After this victory and defeat,—neither
of them apparently very decisive,—the Æginetans began to take a
more energetic part in the war, and brought out their full naval force,
together with that of their allies,—Corinthians, Epidaurians, and
other Peloponnesians: while Athens equipped a fleet of
corresponding magnitude, summoning her allies also; though we do
not know the actual numbers on either side. In the great naval
battle which ensued off the island of Ægina, the superiority of the
new nautical tactics, acquired by twenty years’ practice of the
Athenians since the Persian war,—over the old Hellenic ships and
seamen, as shown in those states where, at the time of the battle of
Marathon, the maritime strength of Greece had resided,—was
demonstrated by a victory most complete and decisive. The
Peloponnesian and Dorian seamen had as yet had no experience of
the improved seacraft of Athens, and when we find how much they
were disconcerted with it, even twenty-eight years afterwards, at the
beginning of the Peloponnesian war, we shall not wonder at its
destructive effect upon them in this early battle. The maritime power
of Ægina was irrecoverably ruined: the Athenians captured seventy
ships of war, landed a large force upon the island, and commenced
the siege of the city by land as well as by sea.[622]
If the Lacedæmonians had not been occupied at home by the
blockade of Ithômê, they would have been probably induced to
invade Attica as a diversion to the Æginetans; especially as the
Persian Megabazus came to Sparta at this time on the part of
Artaxerxes to prevail upon them to do so, in order that the Athenians
might be constrained to retire from Egypt: this Persian brought with
him a large sum of money, but was nevertheless obliged to return
without effecting his mission.[623] The Corinthians and Epidaurians,
however, while they carried to Ægina a reinforcement of three
hundred hoplites, did their best to aid her farther by an attack upon
Megara; which place, it was supposed, the Athenians could not
possibly relieve without withdrawing their forces from Ægina,
inasmuch as so many of their men were at the same time serving in
Egypt. But the Athenians showed themselves equal to all these three
exigencies at one and the same time,—to the great disappointment
of their enemies. Myrônidês marched from Athens to Megara at the
head of the citizens in the two extremes of military age, old and
young; these being the only troops at home. He fought the
Corinthians near the town, gaining a slight, but debatable
advantage, which he commemorated by a trophy, as soon as the
Corinthians had returned home. But the latter when they arrived at
home, were so much reproached by their own old citizens, for not
having vanquished the refuse of the Athenian military force,[624] that
they returned back at the end of twelve days and erected a trophy
on their side, laying claim to a victory in the past battle. The
Athenians, marching out of Megara, attacked them a second time,
and gained on this occasion a decisive victory. The defeated
Corinthians were still more unfortunate in their retreat; for a body of
them, missing their road, became entangled in a space of private
ground, inclosed on every side by a deep ditch, and having only one
narrow entrance. Myrônidês, detecting this fatal mistake, planted his
hoplites at the entrance to prevent their escape, and then
surrounded the enclosure with his light-armed troops, who, with
their missile weapons, slew all the Corinthian hoplites, without
possibility either of flight or resistance. The bulk of the Corinthian
army effected their retreat, but the destruction of this detachment
was a sad blow to the city.[625]
 Splendid as the success of the Athenians had been during this
year, both on land and at sea, it was easy for them to foresee that
the power of their enemies would presently be augmented by the
Lacedæmonians taking the field. Partly on this account,—partly also
from the more energetic phase of democracy, and the long-sighted
views of Periklês, which were now becoming ascendent in the city,—
the Athenians began the stupendous undertaking of connecting
Athens with the sea by means of long walls. The idea of this
measure had doubtless been first suggested by the recent erection
of long walls, though for so much smaller a distance, between
Megara and Nisæa: for without such an intermediate stepping-stone,
the idea of a wall forty stadia long (equal to four and a half miles) to
join Athens with Peiræus, and another wall of thirty-five stadia
(equal to about four miles) to join it with Phalêrum, would have
appeared extravagant even to the sanguine temper of Athenians,—
as it certainly would have seemed a few years earlier to
Themistoklês himself. Coming as an immediate sequel of great
recent victories, and while Ægina, the great Dorian naval power, was
prostrate and under blockade, it excited the utmost alarm among
the Peloponnesians,—being regarded as the second great stride,[626]
at once conspicuous and of lasting effect, in Athenian ambition, next
to the fortification of Peiræus. But besides this feeling in the bosom
of enemies, the measure was also interwoven with the formidable
contention of political parties then going on at Athens. Kimon had
been recently ostracized; and the democratical movement pressed
by Periklês and Ephialtês—of which more presently—was in its full
tide of success, yet not without a violent and unprincipled opposition
on the part of those who supported the existing constitution. Now,
the long walls formed a part of the foreign policy of Periklês,
continuing on a gigantic scale the plans of Themistoklês when he
first schemed the Peiræus. They were framed to render Athens
capable of carrying on war against any superiority of landed attack,
and of bidding defiance to the united force of Peloponnesus. But
though thus calculated for contingencies which a long-sighted man
might see gathering in the distance, the new walls were, almost on
the same grounds, obnoxious to a considerable number of
Athenians: to the party recently headed by Kimon, who were
attached to the Lacedæmonian connection, and desired above all
things to maintain peace at home, reserving the energies of the
state for anti-Persian enterprise: to many landed proprietors in
Attica, whom they seemed to threaten with approaching invasion
and destruction of their territorial possessions: to the rich men and
aristocrats of Athens, averse to a still closer contact and
amalgamation with the maritime multitude in Peiræus: lastly,
perhaps, to a certain vein of old Attic feeling, which might look upon
the junction of Athens with the separate demes of Peiræus and
Phalêrum as effacing the special associations connected with the
holy rock of Athênê. When, to all these grounds of opposition, we
add, the expense and trouble of the undertaking itself, the
interference with private property, the peculiar violence of party
which happened then to be raging, and the absence of a large
proportion of military citizens in Egypt,—we shall hardly be surprised
to find that the projected long walls brought on a risk of the most
serious character both for Athens and her democracy. If any farther
proof were wanting of the vast importance of these long walls, in the
eyes both of friends and of enemies, we might find it in the fact,
that their destruction was the prominent mark of Athenian
humiliation after the battle of Ægos Potamos, and their restoration
the immediate boon of Pharnabazus and Konon after the victory of
Knidus.
Under the influence of the alarm now spread by the proceedings
of Athens, the Lacedæmonians were prevailed upon to undertake an
expedition out of Peloponnesus, although the Helots in Ithômê were
not yet reduced to surrender. Their force consisted of fifteen
hundred troops of their own, and ten thousand of their various allies,
under the regent Nikomêdês. The ostensible motive, or the
pretence, for this march, was the protection of the little territory of
Doris against the Phocians, who had recently invaded it and taken
one of its three towns. The mere approach of so large a force
immediately compelled the Phocians to relinquish their conquest, but
it was soon seen that this was only a small part of the objects of
Sparta, and that her main purposes, under instigation of the
Corinthians, were directed against the aggrandizement of Athens. It
could not escape the penetration of Corinth, that the Athenians
might presently either enlist or constrain the towns of Bœotia into
their alliance, as they had recently acquired Megara, in addition to
their previous ally, Platæa: for the Bœotian federation was at this
time much disorganized, and Thebes, its chief, had never recovered
her ascendency since the discredit of her support lent to the Persian
invasion. To strengthen Thebes, and to render her ascendency
effective over the Bœotian cities, was the best way of providing a
neighbor at once powerful and hostile to the Athenians, so as to
prevent their farther aggrandizement by land: it was the same policy
as Epaminondas pursued eighty years afterwards in organizing
Arcadia and Messênê against Sparta. Accordingly, the Peloponnesian
force was now employed partly in enlarging and strengthening the
fortifications of Thebes herself, partly in constraining the other
Bœotian cities into effective obedience to her supremacy: probably
by placing their governments in the hands of citizens of known
oligarchical politics,[627] and perhaps banishing suspected
opponents. To this scheme the Thebans lent themselves with
earnestness; promising to keep down for the future their border
neighbors, so as to spare the necessity of armies coming from
Sparta.[628]
But there was also a farther design, yet more important, in
contemplation by the Spartans and Corinthians. The oligarchical
opposition at Athens were so bitterly hostile to the Long Walls, to
Periklês, and to the democratical movement, that several of them
opened a secret negotiation with the Peloponnesian leaders, inviting
them into Attica, and entreating their aid in an internal rising for the
purpose not only of putting a stop to the Long Walls, but also of
subverting the democracy. And the Peloponnesian army, while
prosecuting its operations in Bœotia, waited in hopes of seeing the
Athenian malcontents in arms, encamping at Tanagra, on the very
borders of Attica, for the purpose of immediate coöperation with
them. The juncture was undoubtedly one of much hazard for
Athens, especially as the ostracized Kimon and his remaining friends
in the city were suspected of being implicated in the conspiracy. But
the Athenian leaders, aware of the Lacedæmonian operations in
Bœotia, knew also what was meant by the presence of the army on
their immediate borders, and took decisive measures to avert the
danger. Having obtained a reinforcement of one thousand Argeians
and some Thessalian horse, they marched out to Tanagra, with the
full Athenian force then at home; which must, of course, have
consisted chiefly of the old and the young, the same who had fought
under Myrônidês at Megara; for the blockade of Ægina was still
going on. Nor was it possible for the Lacedæmonian army to return
into Peloponnesus without fighting; for the Athenians, masters of the
Megarid, were in possession of the difficult highlands of Geraneia,
the road of march along the isthmus; while the Athenian fleet, by
means of the harbor of Pegæ, was prepared to intercept them, if
they tried to come by sea across the Krissæan gulf, by which way it
would appear that they had come out. Near Tanagra, a bloody battle
took place between the two armies, wherein the Lacedæmonians
were victorious, chiefly from the desertion of the Thessalian horse,
who passed over to them in the very heat of the engagement.[629]
But though the advantage was on their side, it was not sufficiently
decisive to favor the contemplated rising in Attica: nor did the
Peloponnesians gain anything by it, except an undisturbed retreat
over the highlands of Geraneia, after having partially ravaged the
Megarid.
Though the battle of Tanagra was a defeat, yet there were
circumstances connected with it which rendered its effects highly
beneficial to Athens. The ostracized Kimon presented himself on the
field as soon as the army had passed over the boundaries of Attica,
requesting to be allowed to occupy his station as an hoplite and to
fight in the ranks of his tribe,—the Œnêis. But such was the belief,
entertained by the members of the senate and by his political
enemies present, that he was an accomplice in the conspiracy
known to be on foot, that permission was refused and he was forced
to retire. In departing, he conjured his personal friends, Euthippus,
of the deme Anaphlystus, and others, to behave in such a manner as
might wipe away the stain resting upon his fidelity, and in part also
upon theirs. His friends retained his panoply, and assigned to it the
station in the ranks which he would himself have occupied: they
then entered the engagement with desperate resolution, and one
hundred of them fell side by side in their ranks. Periklês, on his part,
who was present among the hoplites of his own tribe, the
Akamantis, aware of this application and repulse of Kimon, thought
it incumbent upon him to display not merely his ordinary personal
courage, but an unusual recklessness of life and safety, though it
happened that he escaped unwounded. All these incidents brought
about a generous sympathy and spirit of compromise among the
contending parties at Athens, while the unshaken patriotism of
Kimon and his friends discountenanced and disarmed those
conspirators who had entered into correspondence with the enemy,
at the same time that it roused a repentant admiration towards the
ostracized leader himself. Such was the happy working of this new
sentiment that a decree was shortly proposed and carried,—
proposed too, by Periklês himself,—to abridge the ten years of
Kimon’s ostracism, and permit his immediate return.[630] We may
recollect that, under circumstances partly analogous, Themistoklês
had himself proposed the restoration of his rival Aristeidês from
ostracism, a little before the battle of Salamis:[631] and in both cases,
the suspension of enmity between the two leaders was partly the
sign, partly also the auxiliary cause, of reconciliation and renewed
fraternity among the general body of citizens. It was a moment
analogous to that salutary impulse of compromise, and harmony of
parties, which followed the extinction of the oligarchy of Four
Hundred, forty-six years afterwards, and on which Thucydidês dwells
emphatically as the salvation of Athens in her distress,—a moment
rare in free communities generally, not less than among the jealous
competitors for political ascendency at Athens.[632]
So powerful was this burst of fresh patriotism and unanimity
after the battle of Tanagra, which produced the recall of Kimon, and
appears to have overlaid the preëxisting conspiracy, that the
Athenians were quickly in a condition to wipe off the stain of their
defeat. It was on the sixty-second day after the battle that they
undertook an aggressive march under Myrônidês into Bœotia: the
extreme precision of this date,—being the single case throughout
the summary of events between the Persian and Peloponnesian
wars, wherein Thucydidês is thus precise, marks how strong an
impression it made upon the memory of the Athenians. At the battle
of Œnophyta, engaged against the aggregate Theban and Bœotian
forces,—or, if Diodorus is to be trusted, in two battles, of which that
of Œnophyta was the last, Myrônidês was completely victorious. The
Athenians became masters of Thebes as well as of the remaining
Bœotian towns; reversing all the arrangements recently made by
Sparta,— establishing democratical governments,—and forcing the
aristocratical leaders, favorable to Theban ascendency and
Lacedæmonian connection, to become exiles. Nor was it only Bœotia
which the Athenians thus acquired: Phocis and Lokris were both
successively added to the list of their dependent allies,—the former
being in the main friendly to Athens and not disinclined to the
change, while the latter were so decidedly hostile that one hundred
of their chiefs were detained and sent to Athens as hostages. The
Athenians thus extended their influence,—maintained through
internal party-management, backed by the dread of interference
from without in case of need,—from the borders of the Corinthian
territory, including both Megara and Pêgæ, to the strait of
Thermopylæ.[633]
These important acquisitions were soon crowned by the
completion of the Long Walls and the conquest of Ægina. That
island, doubtless starved out by its protracted blockade, was forced
to capitulate on condition of destroying its fortifications, surrendering
all its ships of war, and submitting to annual tribute as a dependent
ally of Athens. The reduction of this once powerful maritime city,
marked Athens as mistress of the sea on the Peloponnesian coast
not less than on the Ægean. Her admiral Tolmidês displayed her
strength by sailing round Peloponnesus, and even by the insult of
burning the Lacedæmonian ports of Methônê and of Gythium. He
took Chalkis, a possession of the Corinthians, and Naupaktus
belonging to the Ozolian Lokrians, near the mouth of the Corinthian
gulf,—disembarked troops near Sikyon with some advantage in a
battle against opponents from that town,—and either gained or
forced into the Athenian alliance not only Zakynthus and
Kephallênia, but also some of the towns of Achaia; for we afterwards
find these latter attached to Athens without knowing when the
connection began.[634]
During the ensuing year the Athenians renewed their attack upon
Sikyon, with a force of one thousand hoplites under Periklês himself,
sailing from the Megarian harbor of Pêgæ in the Krissæan gulf. This
eminent man, however, gained no greater advantage than Tolmidês,
—defeating the Sikyonian forces in the field and driving them within
their walls: he afterwards made an expedition into Akarnania, taking
the Achæan allies in addition to his own forces, but miscarried in his
attack on Œniadæ and accomplished nothing. Nor were the
Athenians more successful in a march undertaken this same year
against Thessaly, for the purpose of restoring Orestes, one of the
exiled princes or nobles of Pharsalus. Though they took with them
an imposing force, including their Bœotian and Phocian allies, the
powerful Thessalian cavalry forced them to keep in a compact body
and confined them to the ground actually occupied by their hoplites;
while all their attempts against the city failed, and their hopes of
internal rising were disappointed.[635]
Had the Athenians succeeded in Thessaly, they would have
acquired to their alliance nearly the whole of extra-Peloponnesian
Greece: but even without Thessaly their power was prodigious, and
had now attained a maximum height, from which it never varied
except to decline. As a counterbalancing loss against so many
successes, we have to reckon their ruinous defeat in Egypt, after a
war of six years against the Persians (b. c. 460-455). At first, they
had gained brilliant advantages, in conjunction with the insurgent
prince Inarôs; expelling the Persians from all Memphis except the
strongest part, called the White Fortress: and such was the alarm of
the Persian king, Artaxerxes, at the presence of the Athenians in
Egypt, that he sent Megabazus with a large sum of money to Sparta,
in order to induce the Lacedæmonians to invade Attica. This envoy,
however, failed, and an augmented Persian force being sent to Egypt
under Megabyzus, son of Zopyrus,[636] drove the Athenians and their
allies, after an obstinate struggle, out of Memphis into the island of
the Nile called Prosôpîtis. Here they were blocked up for eighteen
months, until at length Megabyzus turned the arm of the river, laid
the channel dry, and stormed the island by land. A very few
Athenians escaped by land to Kyrênê: the rest were either slain or
made captive, and Inarôs himself was crucified. And the calamity of
Athens was farther aggravated by the arrival of fifty fresh Athenian
ships, which, coming after the defeat, but without being aware of it,
sailed into the Mendesian branch of the Nile, and thus fell unawares
into the power of the Persians and Phenicians; very few either of the
ships or men escaping. The whole of Egypt became again subject to
the Persians, except Amyrtæus, who contrived, by retiring into the
inaccessible fens, still to maintain his independence. One of the
largest armaments ever sent forth by Athens and her confederacy
was thus utterly ruined.[637]
It was about the time of the destruction of the Athenian army in
Egypt, and of the circumnavigation of Peloponnesus by Tolmidês,
that the internal war, carried on by the Lacedæmonians, against the
Helots or Messenians at Ithômê, ended. These besieged men, no
longer able to stand out against a protracted blockade, were forced
to abandon this last fortress of ancient Messenian independence,
stipulating for a safe retreat from Peloponnesus with their wives and
families, with the proviso, that if any one of them ever returned to
Peloponnesus, he should become the slave of the first person who
seized him. They were established by Tolmidês at Naupaktus, which
had recently been taken by the Athenians from the Ozolian Lokrians,
[638]—where they will be found rendering good service to Athens in
the following wars.
 After the victory of Tanagra, the Lacedæmonians made no farther
expeditions out of Peloponnesus for several succeeding years, not
even to prevent Bœotia and Phocis from being absorbed into the
Athenian alliance. The reason of this remissness lay, partly, in their
general character; partly, in the continuance of the siege of Ithômê,
which occupied them at home; but still more, perhaps, in the fact
that the Athenians, masters of the Megarid, were in occupation of
the road over the highlands of Geraneia, and could therefore
obstruct the march of any army out from Peloponnesus. Even after
the surrender of Ithômê, the Lacedæmonians remained inactive for
three years, after which time a formal truce was concluded with
Athens by the Peloponnesians generally, for five years longer.[639]
This truce was concluded in a great degree through the influence of
Kimon,[640] who was eager to resume effective operations against
the Persians; while it was not less suitable to the political interests of
Periklês that his most distinguished rival should be absent on foreign
service,[641] so as not to interfere with his influence at home.
Accordingly, Kimon equipped a fleet of two hundred triremes, from
Athens and her confederates, and set sail for Cyprus, from whence
he despatched sixty ships to Egypt, at the request of the insurgent
prince Amyrtæus, who was still maintaining himself against the
Persians amidst the fens,—while with the remaining armament he
laid siege to Kitium. In the prosecution of this siege, he died, either
of disease or of a wound. The armament, under his successor,
Anaxikrates, became so embarrassed for want of provisions that
they abandoned the undertaking altogether, and went to fight the
Phenician and Kilikian fleet near Salamis, in Cyprus. They were here
victorious, first on sea, and afterwards on land, though probably not
on the same day, as at the Eurymedon; after which they returned
home, followed by the sixty ships which had gone to Egypt for the
purpose of aiding Amyrtæus.[642]
From this time forward no farther operations were undertaken by
Athens and her confederacy against the Persians. And it appears
that a convention was concluded between them, whereby the Great
King on his part promised two things: To leave free, undisturbed,
and untaxed, the Asiatic maritime Greeks, not sending troops within
a given distance of the coast: to refrain from sending any ships of
war either westward of Phasêlis (others place the boundary at the
Chelidonean islands, rather more to the westward) or within the
Kyanean rocks at the confluence of the Thracian Bosphorus with the
Euxine. On their side, the Athenians agreed to leave him in
undisturbed possession of Cyprus and Egypt. Kallias, an Athenian of
distinguished family, with some others of his countrymen, went up to
Susa to negotiate this convention: and certain envoys from Argos,
then in alliance with Athens, took the opportunity of going thither at
the same time, to renew the friendly understanding which their city
had established with Xerxes at the period of his invasion of Greece.
[643]

As is generally the case with treaties after hostility,—this

convention did little more than recognize the existing state of things,
without introducing any new advantage or disadvantage on either
side, or calling for any measures to be taken in consequence of it.
We may hence assign a reasonable ground for the silence of
Thucydidês, who does not even notice the convention as having
been made: we are to recollect always that in the interval between
the Persian and Peloponnesian wars, he does not profess to do more
than glance briefly at the main events. But the boastful and
inaccurate authors of the ensuing century, orators, rhetors, and
historians, indulged in so much exaggeration and untruth respecting
this convention, both as to date and as to details,—and extolled as
something so glorious the fact of having imposed such hard
conditions on the Great King,—that they have raised a suspicion
against themselves. Especially, they have occasioned critics to ask
the very natural question, how this splendid achievement of Athens
came to be left unnoticed by Thucydidês? Now the answer to such
question is, that the treaty itself was really of no great moment: it is
the state of facts and relations implied in the treaty, and existing
substantially before it was concluded, which constitutes the real
glory of Athens. But to the later writers, the treaty stood forth as the
legible evidence of facts which in their time were passed and gone;
while Thucydidês and his contemporaries, living in the actual fulness
of the Athenian empire, would certainly not appeal to the treaty as
an evidence, and might well pass it over, even as an event, when
studying to condense the narrative. Though Thucydidês has not
mentioned the treaty, he says nothing which disproves its reality,
and much which is in full harmony with it. For we may show, even
from him: 1. That all open and direct hostilities between Athens and
Persia ceased, after the last-mentioned victories of the Athenians
near Cyprus: that this island is renounced by Athens, not being
included by Thucydidês in his catalogue of Athenian allies prior to
the Peloponnesian war;[644] and that no farther aid is given by
Athens to the revolted Amyrtæus in Egypt. 2. That down to the time
when the Athenian power was prostrated by the ruinous failure at
Syracuse, no tribute was collected by the Persian satraps in Asia
Minor from the Greek cities on the coast, nor were Persian ships of
war allowed to appear in the waters of the Ægean,[645] nor was the
Persian king admitted to be sovereign of the country down to the
coast. Granting, therefore, that we were even bound, from the
silence of Thucydidês, to infer that no treaty was concluded, we
should still be obliged also to infer, from his positive averments, that
a state of historical fact, such as the treaty acknowledged and
prescribed, became actually realized. But when we reflect farther,
that Herodotus[646] certifies the visit of Kallias and other Athenian
envoys to the court of Susa, we can assign no other explanation of
such visit so probable as the reality of this treaty: certainly, no
envoys would have gone thither during a state of recognized war;
and though it may be advanced as possible that they may have gone
with the view to conclude a treaty, and yet not have succeeded,—
this would be straining the limits of possibility beyond what is
reasonable.[647]
We may therefore believe in the reality of this treaty between
Athens and Persia, improperly called the Kimonian treaty:
improperly, since not only was it concluded after the death of Kimon,
but the Athenian victories by which it was immediately brought on
were gained after his death. Nay, more,—the probability is, that if
Kimon had lived, it would not have been concluded at all; for his
interest as well as his glory led him to prosecute the war against
Persia, since he was no match for his rival Periklês, either as a
statesman or as an orator, and could only maintain his popularity by
the same means whereby he had earned it,—victories and plunder at
the cost of the Persians. His death insured more complete
ascendency to Periklês, whose policy and character were of a cast
altogether opposite:[648] while even Thucydidês, son of Melêsias,
who succeeded Kimon, his relation, as leader of the anti-Periklean
party, was also a man of the senate and public assembly rather than
of campaigns and conquests. Averse to distant enterprises and
precarious acquisitions, Periklês was only anxious to maintain
unimpaired the Hellenic ascendency of Athens, now at its very
maximum: he was well aware that the undivided force and vigilance
of Athens would not be too much for this object,—nor did they in
fact prove sufficient, as we shall presently see. With such
dispositions he was naturally glad to conclude a peace, which
excluded the Persians from all the coasts of Asia Minor, westward of
the Chelidoneans, as well as from all the waters of the Ægean,
under the simple condition of renouncing on the part of Athens
farther aggressions against Cyprus, Phenicia, Kilikia, and Egypt. The
Great King on his side had had sufficient experience of Athenian
energy to fear the consequences of such aggressions, if prosecuted;
nor did he lose much by relinquishing formally a tribute which at the
time he could have little hope of realizing, and which of course he
intended to resume on the first favorable opportunity. Weighing all
these circumstances, we shall find that the peace, improperly called
Kimonian, results naturally from the position and feelings of the
contracting parties.
Athens was now at peace both abroad and at home, under the
administration of Periklês, with a great empire, a great fleet, and a
great accumulated treasure. The common fund collected from the
contributions of the confederates, and originally deposited at Delos,
had before this time been transferred to the acropolis at Athens. At
what precise time this transfer took place, we cannot state: nor are
we enabled to assign the successive stages whereby the
confederacy, chiefly with the freewill of its own members, became
transformed from a body of armed and active warriors under the
guidance of Athens, into disarmed and passive tribute-payers,
defended by the military force of Athens,—from allies free, meeting
at Delos, and self-determining, into subjects isolated, sending their
annual tribute, and awaiting Athenian orders. But it would appear
that the change had been made before this time: some of the more
resolute of the allies had tried to secede, but Athens had coerced
them by force, and reduced them to the condition of tribute-payers,
without ships or defence; and Chios, Lesbos, and Samos were now
the only allies free and armed on the original footing. Every
successive change of an armed ally into a tributary,—every
subjugation of a seceder,—tended of course to cut down the
numbers, and enfeeble the authority, of the Delian synod; and, what
was still worse, it altered the reciprocal relation and feelings both of
Athens and her allies,—exalting the former into something like a
despot, and degrading the latter into mere passive subjects.
Of course, the palpable manifestation of the change must have
been the transfer of the confederate fund from Delos to Athens. The
only circumstance which we know respecting this transfer is, that it
was proposed by the Samians,[649]—the second power in the
confederacy, inferior only to Athens, and least of all likely to favor
any job or sinister purpose of the Athenians. It is farther said that,
when the Samians proposed it, Aristeidês characterized it as a
motion unjust, but useful: we may well doubt, however, whether it
was made during his lifetime. When the synod at Delos ceased to be
so fully attended as to command respect,—when war was lighted up,
not only with Persia, but with Ægina and Peloponnesus,—the
Samians might not unnaturally feel that the large accumulated fund,
with its constant annual accessions, would be safer at Athens than
at Delos, which latter island would require a permanent garrison and
squadron to insure it against attack. But whatever may have been
the grounds on which the Samians proceeded, when we find them
coming forward to propose the transfer, we may reasonably infer
that it was not displeasing, and did not appear unjust, to the larger
members of the confederacy,—and that it was no high-handed and
arbitrary exercise of power, as it is often called, on the part of
Athens.
After the conclusion of the war with Ægina, and the
consequences of the battle of Œnophyta, the position of Athens
became altered more and more. She acquired a large catalogue of
new allies, partly tributary, like Ægina,—partly in the same relation
as Chios, Lesbos, and Samos; that is, obliged only to a conformity of
foreign policy and to military service. In this last category were
Megara, the Bœotian cities, the Phocians, Lokrians, etc. All these,
though allies of Athens, were strangers to Delos and the confederacy
against Persia; and accordingly, that confederacy passed insensibly
into a matter of history, giving place to the new conception of
imperial Athens, with her extensive list of allies, partly free, partly
subject. Such transition, arising spontaneously out of the character
and circumstances of the confederates themselves, was thus
materially forwarded by the acquisitions of Athens extraneous to the
confederacy. She was now not merely the first maritime state of
Greece, but perhaps equal to Sparta even in land-power,—
possessing in her alliance Megara, Bœotia, Phocis, Lokris, together
with Achæa and Trœzen, in Peloponnesus. Large as this aggregate
already was, both at sea and on land, yet the magnitude of the
annual tribute, and still more the character of the Athenians
themselves, superior to all Greeks in that combination of energy and
discipline which is the grand cause of progress, threatened still
farther increase. Occupying the Megarian harbor of Pêgæ, the
Athenians had full means of naval action on both sides of the
Corinthian isthmus: but, what was of still greater importance to
them, by their possession of the Megarid, and of the highlands of
Geraneia, they could restrain any land-force from marching out of
Peloponnesus, and were thus, considering besides their mastery at
sea, completely unassailable in Attica. Ever since the repulse of
Xerxes, Athens had been advancing in an uninterrupted course of
power and prosperity at home, as well as of victory and ascendency
abroad,—to which there was no exception, except the ruinous
enterprise in Egypt. Looking at the position of Greece, therefore,
about 448 b. c.,—after the conclusion of the five years’ truce
between the Peloponnesians and Athens, and of the so-called
Kimonian peace between Persia and Athens,—a discerning Greek
might well calculate upon farther aggrandizement of this imperial
state as the tendency of the age; and accustomed as every Greek
was to the conception of separate town-autonomy as essential to a
freeman and a citizen, such prospect could not but inspire terror and
aversion. The sympathy of the Peloponnesians for the islanders and
ultra-maritime states, who constituted the original confederacy of
Athens, was not considerable; but when the Dorian island of Ægina
was subjugated also, and passed into the condition of a defenceless
tributary, they felt the blow sorely on every ground. The ancient
celebrity and eminent service rendered at the battle of Salamis, of
this memorable island, had not been able to protect it; while those
great Æginetan families, whose victories at the sacred festival-
games Pindar celebrates in a large proportion of his odes, would
spread the language of complaint and indignation throughout their
numerous “guests” in every Hellenic city. Of course, the same anti-
Athenian feeling would pervade those Peloponnesian states who had
been engaged in actual hostility with Athens,—Corinth, Sikyon,
Epidaurus, etc., as well as Sparta, the once-recognized head of
Hellas, but now tacitly degraded from her preëminence, baffled in
her projects respecting Bœotia, and exposed to the burning of her
port at Gythium, without being able even to retaliate upon Attica.
Putting all those circumstances together, we may comprehend the
powerful feeling of dislike and apprehension now diffused so widely
over Greece against the upstart despot city; whose ascendency,
newly acquired, maintained by superior force, and not recognized as
legitimate,—threatened, nevertheless, still farther increase. Sixteen
years hence, this same sentiment will be found exploding into the
Peloponnesian war; but it became rooted in the Greek mind during
the period which we have now reached, when Athens was much
more formidable than she had come to be at the commencement of
that war: nor shall we thoroughly appreciate the ideas of that later
period, unless we take them as handed down from the earlier date
of the five years’ truce, about 451-446 b. c.
Formidable as the Athenian empire both really was and appeared
to be, however, this wide-spread feeling of antipathy proved still
stronger, so that, instead of the threatened increase, the empire
underwent a most material diminution. This did not arise from the
attack of open enemies; for during the five years’ truce, Sparta
undertook only one movement, and that not against Attica: she sent
troops to Delphi, in an expedition dignified with the name of the
Sacred War,—expelled the Phocians, who had assumed to
themselves the management of the temple,—and restored it to the
native Delphians. To this the Athenians made no direct opposition:
but as soon as the Lacedæmonians were gone, they themselves
marched thither and placed the temple again in the hands of the
Phocians, who were then their allies.[650] The Delphians were
members of the Phocian league, and there was a dispute of old
standing as to the administration of the temple,—whether it
belonged to them separately or to the Phocians collectively. The
favor of those who administered it counted as an element of
considerable moment in Grecian politics; the sympathies of the
leading Delphians led them to embrace the side of Sparta, but the
Athenians now hoped to counteract this tendency by means of their
preponderance in Phocis. We are not told that the Lacedæmonians
took any ulterior step in consequence of their views being frustrated
by Athens,—a significant evidence of the politics of that day.
The blow which brought down the Athenian empire from this its
greatest exaltation, was struck by the subjects themselves. The
Athenian ascendency over Bœotia, Phocis, Lokris, and Eubœa, was
maintained, not by means of garrisons, but through domestic parties
favorable to Athens, and a suitable form of government; just in the
same way as Sparta maintained her influence over her
Peloponnesian allies.[651] After the victory of Œnophyta, the
Athenians had broken up the governments in the Bœotian cities
established by Sparta before the battle of Tanagra, and converted
them into democracies at Thebes and elsewhere. Many of the
previous leading men had thus been sent into exile: and as the same
process had taken place in Phocis and Lokris, there was at this time
a considerable aggregate body of exiles, Bœotian, Phocian, Lokrian,
Eubœan, Æginetan, etc., all bitterly hostile to Athens, and ready to
join in any attack upon her power. We learn farther that the
democracy,[652] established at Thebes after the battle of Œnophyta,
was ill-conducted and disorderly: which circumstances laid open
Bœotia still farther to the schemes of assailants on the watch for
every weak point. These various exiles, all joining their forces and
concerting measures with their partisans in the interior, succeeded in
mastering Orchomenus, Chæroneia, and some other less important
places in Bœotia. The Athenian general, Tolmidês, marched to expel
them, with one thousand Athenian hoplites and an auxiliary body of
allies. It appears that this march was undertaken in haste and
rashness: the hoplites of Tolmidês, principally youthful volunteers,
and belonging to the best families of Athens, disdained the enemy
too much to await a larger and more commanding force: nor would
the people listen even to Periklês, when he admonished them that
the march would be full of hazard, and adjured them not to attempt
it without greater numbers as well as greater caution.[653] Fatally,
indeed, were his predictions justified. Though Tolmidês was
successful in his first enterprise,—the recapture of Chæroneia,
wherein he placed a garrison,—yet in his march, probably incautious
and disorderly, when departing from that place, he was surprised
and attacked unawares, near Korôneia, by the united body of exiles
and their partisans. No defeat in Grecian history was ever more
complete or ruinous. Tolmidês himself was slain, together with many
of the Athenian hoplites, while a large number of them were taken
prisoners. In order to recover these prisoners, who belonged to the
best families in the city, the Athenians submitted to a convention
whereby they agreed to evacuate Bœotia altogether: in all the cities
of that country, the exiles were restored, the democratical
government overthrown, and Bœotia was transformed from an ally
of Athens into her bitter enemy.[654] Long, indeed, did the fatal issue
of this action dwell in the memory of the Athenians,[655] and inspire
them with an apprehension of Bœotian superiority in heavy armor
on land: but if the hoplites under Tolmidês had been all slain on the
field, their death would probably have been avenged and Bœotia
would not have been lost,—whereas, in the case of living citizens,
the Athenians deemed no sacrifice too great to redeem them. We
shall discover hereafter in the Lacedæmonians a feeling very similar,
respecting their brethren captured at Sphakteria.
The calamitous consequences of this defeat came upon Athens in
thick and rapid succession. The united exiles, having carried their
point in Bœotia, proceeded to expel the philo-Athenian government
both from Phocis and Lokris, and to carry the flame of revolt into
Eubœa. To this important island Periklês himself proceeded
forthwith, at the head of a powerful force; but before he had time to
complete the reconquest, he was summoned home by news of a still
more formidable character. The Megarians had revolted from Athens:
by a conspiracy previously planned, a division of hoplites from
Corinth, Sikyon, and Epidaurus, was already admitted as garrison
into their city: the Athenian soldiers who kept watch over the Long
Walls had been overpowered and slain, except a few who escaped
into the fortified port of Nisæa. As if to make the Athenians at once
sensible how seriously this disaster affected them, by throwing open
the road over Geraneia,—Pleistoanax, king of Sparta, was
announced as already on his march for an invasion of Attica. He did,
in truth, conduct an army, of mixed Lacedæmonians and
Peloponnesian allies, into Attica, as far as the neighborhood of
Eleusis and the Thriasian plain. He was a very young man, so that a
Spartan of mature years, Kleandridês, had been attached to him by
the ephors as adjutant and counsellor. Periklês, it is said, persuaded
both the one and the other, by means of large bribes, to evacuate
Attica without advancing to Athens. We may well doubt whether
they had force enough to adventure so far into the interior, and we
shall hereafter observe the great precautions with which Archidamus
thought it necessary to conduct his invasion, during the first year of
the Peloponnesian war, though at the head of a more commanding
force. Nevertheless, on their return, the Lacedæmonians, believing
that they might have achieved it, found both of them guilty of
corruption. Both were banished: Kleandridês never came back, and
Pleistoanax himself lived for a long time in sanctuary near the
temple of Athênê, at Tegea, until at length he procured his
restoration by tampering with the Pythian priestess, and by bringing
her bought admonitions to act upon the authorities at Sparta.[656]
So soon as the Lacedæmonians had retired from Attica, Periklês
returned with his forces to Eubœa, and reconquered the island
completely. With that caution which always distinguished him as a
military man, so opposite to the fatal rashness of Tolmidês, he took
with him an overwhelming force of fifty triremes and five thousand
hoplites. He admitted most of the Eubœan towns to surrender,
altering the government of Chalkis by the expulsion of the wealthy
oligarchy called the Hippobotæ; but the inhabitants of Histiæa, at
the north of the island, who had taken an Athenian merchantman
and massacred all the crew, were more severely dealt with,—the
free population being all or in great part expelled, and the land
distributed among Athenian kleruchs, or out-settled citizens.[657]
But the reconquest of Eubœa was far from restoring Athens to
the position which she had occupied before the fatal engagement of
Korôneia. Her land empire was irretrievably gone, together with her
recently acquired influence over the Delphian oracle; and she
reverted to her former condition of an exclusively maritime
potentate. For though she still continued to hold Nisæa and Pegæ,
yet her communication with the latter harbor was now out off by the
loss of Megara and its appertaining territory, so that she thus lost
her means of acting in the Corinthian gulf, and of protecting as well
as of constraining her allies in Achaia. Nor was the port of Nisæa of
much value to her, disconnected from the city to which it belonged,
except as a post for annoying that city. Moreover, the precarious
hold which she possessed over unwilling allies had been
demonstrated in a manner likely to encourage similar attempts
among her maritime subjects,—attempts which would now be
seconded by Peloponnesian armies invading Attica. The fear of such
a combination of embarrassments, and especially of an irresistible
enemy carrying ruin over the flourishing territory round Eleusis and
Athens, was at this moment predominant in the Athenian mind. We
shall find Periklês, at the beginning of the Peloponnesian war,
fourteen years afterwards, exhausting all his persuasive force, and
not succeeding without great difficulty, in prevailing upon his
countrymen to endure the hardship of invasion,—even in defence of
their maritime empire, and when events had been gradually so
ripening as to render the prospect of war familiar, if not inevitable.
But the late series of misfortunes had burst upon them so rapidly
and unexpectedly, as to discourage even Athenian confidence, and
to render the prospect of continued war full of gloom and danger.
The prudence of Periklês would doubtless counsel the surrender of
their remaining landed possessions or alliances, which had now
become unprofitable, in order to purchase peace; but we may be
sure that nothing short of extreme temporary despondency could
have induced the Athenian assembly to listen to such advice, and to
accept the inglorious peace which followed. A truce for thirty years
was concluded with Sparta and her allies, in the beginning of 445
b. c., whereby Athens surrendered Nisæa, Pegæ, Achaia, and
Trœzen,—thus abandoning Peloponnesus altogether,[658] and leaving
the Megarians—with their full territory and their two ports—to be
included among the Peloponnesian allies of Sparta.
It was to the Megarians, especially, that the altered position of
Athens after this truce was owing: it was their secession from Attica
and junction with the Peloponnesians, which laid open Attica to
invasion. Hence, arose the deadly hatred on the part of the
Athenians towards Megara, manifested during the ensuing years,—a
sentiment the more natural, as Megara had spontaneously sought
the alliance of Athens a few years before as a protection against the
Corinthians, and had then afterwards, without any known ill-usage
on the part of Athens, broken off from the alliance and become her
enemy, with the fatal consequence of rendering her vulnerable on
the land-side. Under such circumstances we shall not be surprised to
find the antipathy of the Athenians against Megara strongly
pronounced, insomuch that the system of exclusion which they
adopted against her was among the most prominent causes of the
Peloponnesian war.
Having traced what we may call the foreign relations of Athens
down to this thirty years’ truce, we must notice the important
internal and constitutional changes which she had experienced
during the same interval.
 CHAPTER XLVI.
CONSTITUTIONAL AND JUDICIAL CHANGES AT
ATHENS UNDER PERIKLES.

The period which we have now passed over appears to have been
that in which the democratical cast of Athenian public life was first
brought into its fullest play and development, as to judicature,
legislation, and administration.
The great judicial change was made by the methodical
distribution of a large proportion of the citizens into distinct judicial
divisions, by the great extension of their direct agency in that
department, and by the assignment of a constant pay to every
citizen so engaged. It has been already mentioned that even under
the democracy of Kleisthenês, and until the time succeeding the
battle of Platæa, large powers still remained vested both in the
individual archons and in the senate of Areopagus: which latter was
composed exclusively of the past archons after their year of office,
sitting in it for life,—though the check exercised by the general body
of citizens, assembled for law-making in the ekklesia, and for judging
in the heliæa, was at the same time materially increased. We must
farther recollect, that the distinction between powers administrative
and judicial, so highly valued among the more elaborate
governments of modern Europe, since the political speculations of
the last century, was in the early history of Athens almost unknown.
Like the Roman kings,[659] and the Roman consuls before the
appointment of the prætor, the Athenian archons not only
administered, but also exercised jurisdiction, voluntary as well as
contentious,—decided disputes, inquired into crimes, and inflicted
punishment. Of the same mixed nature were the functions of the
senate of Areopagus, and even of the annual senate of Five
Hundred, the creation of Kleisthenês. The stratêgi, too, as well as
the archons, had doubtless the double competence—in reference to
military, naval, and foreign affairs—of issuing orders and of
punishing by their own authority, disobedient parties: the imperium
of the magistrates, generally, enabled them to enforce their own
mandates as well as to decide in cases of doubt whether any private
citizen had or had not been guilty of infringement. Nor was there
any appeal from these magisterial judgments; though the
magistrates were subject, under the Kleisthenean constitution, to
personal responsibility for their general behavior, before the people
judicially assembled, at the expiration of their year of office,—and to
the farther animadversion of the ekklesia, or public deliberative
assembly, meeting periodically during the course of that year: in
some of which ekklesiæ, the question might formally be raised for
deposing any magistrate, even before his year was expired.[660] Still,
in spite of such partial checks, the accumulation, in the same hand,
of powers to administer, judge, punish, and decide civil disputes,
without any other canon than the few laws then existing, and
without any appeal,—must have been painfully felt, and must have
often led to corrupt, arbitrary, and oppressive dealing: and if this be
true of individual magistrates, exposed to annual accountability, it is
not likely to have been less true of the senate of Areopagus, which,
acting collectively, could hardly be rendered accountable, and in
which the members sat for life.[661]
I have already mentioned that shortly after the return of the
expatriated Athenians from Salamis, Aristeidês had been impelled,
by the strong democratical sentiment which he found among his
countrymen, to propose the abolition of all pecuniary qualification for
magistracies, so as to render every citizen legally eligible. This
innovation, however, was chiefly valuable as a victory and as an
index of the predominant sentiment: notwithstanding the enlarged
promise of eligibility, little change probably took place in the fact,
and rich men were still most commonly chosen. Hence the
magistrates, possessing the large powers administrative and judicial
above described,—and still more the senate of Areopagus, which sat
for life,—still belonging almost entirely to the wealthier class,
remained animated more or less with the same oligarchical interest
and sympathies, which manifested themselves in the abuse of
authority. At the same time the democratical sentiment among the
mass of Athenians went on steadily increasing from the time of
Aristeidês to that of Periklês: Athens became more and more
maritime, the population of Peiræus augmented in number as well
as in importance, and the spirit even of the poorest citizen was
stimulated by that collective aggrandizement of his city to which he
himself individually contributed. Before twenty years had elapsed,
reckoning from the battle of Platæa, this new fervor of democratical
sentiment made itself felt in the political contests of Athens, and
found able champions in Periklês and Ephialtês, rivals of what may
be called the conservative party, headed by Kimon.
We have no positive information that it was Periklês who
introduced the lot, in place of election, for the choice of archons and
various other magistrates, but the change must have been
introduced nearly at this time, and with a view of equalizing the
chances of office to every candidate, poor as well as rich, who chose
to give in his name, and who fulfilled certain personal and family
conditions ascertained in the dokimasy, or preliminary examination.
But it was certainly to Periklês and Ephialtês that Athens owed the
elaborate constitution of her popular dikasteries, or jury courts
regularly paid, which exercised so important an influence upon the
character of the citizens. These two eminent men deprived both the
magistrates and the senate of Areopagus of all the judicial and penal
competence which they had hitherto possessed, save and except the
power of imposing a small fine. This judicial power, civil as well as
criminal, was transferred to numerous dikasts, or panels of jurors
selected from the citizens; six thousand of whom were annually
drawn by lot and sworn, and then distributed into ten panels of five
hundred each, the remainder forming a supplement in case of
vacancies. The magistrate, instead of deciding causes, or inflicting
punishment by his own authority, was now constrained to impanel a
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com