P.D.E.A.

’S
PROF. RAMKRISHNA MORE ARTS, COMMERCE & SCIENCE
COLLEGE AKURDI, PUNE – 411 044.

CERTIFICATE

This is to certify that the project entitled

Character recognition
(Using CNN)

Has been successfully completed by

Anubhav Sharma
Arun Pillay
Mayur Koli

Towards the fulfillment of M.Sc. (Computer Science) Sem-II

For the academic year 2019-2020.

Ms. R. J. Katkar Mr. S. G. Lakhdive

Project Guide Head of Department

Internal Examiner External Examiner

 INDEX
1. Introduction
1.1 Existing System
1.2 Need for System
1.3 Hardware and Software Specification
2. Proposed System
2.1 Proposed System
2.2 Objectives of the System
3. Analysis and Design
3.1 Activity Diagram
3.2 Sequence Diagram
4. Implementation
4.1 Input and Output Screen shots
5. Testing
6. Future Enhancements and Conclusion
6.1 Future Enhancements
6.2 Conclusion
7. Bibliography
 ACKNOWLEDGEMENT

I sincerely wish to express deep sense of gratitude to my project

guide Ms. R. J . Katkar & Head of Department Mr. Lakhdive S.G for
supporting & guiding me in this project.
I would also like to extend my sincere thanks to all our teachers of
department for their excellent guidance & helpful suggestions. I would
also like to thank them for their valuable inspiration & for their co-
operation during the project work & also thank them for providing
us all required facilities.
And finally I thank our colleagues for their immense encouragement
for the success completion of my project.
With thanks from.

MAYUR KOLI
ANUBHAV SHARMA
ARUN PILLAY
 1. Introduction

• Character recognition is neural network model which is used for recognizing

handwritten Hindi characters.
• Multilayer feedforward neural network algorithm is applied.
• To train neural network, dataset with size of 70mb is used.
• ReLU (Rectified Linear Unit) activation function is applied.

1.1 Existing System:

• There is no any model that can recognize Handwritten Hindi Character,

which is cheaper in cost.
• Existing system is expensive so no can easily accommodate it.

1.2 NEED OF THE SYSTEM:

• The purpose to develop handwritten characters is to recognize different

handwritings that matches with original character.
• But the handwritten characters should meet maximum parameter of
actual character.
 1.3 HARDWARE AND SOHTWARE
SPECIFICATIONS

HARDWARE: -
Hard Disk : 2GB or above.
Ram : 20 MB
GPU : 2GB or above

SOFTWARE: -
Operating System: Windows 10, MAC OS, Linux, Raspberry pi
Front End : Kotlin, Js
Back End : Python
 2. Proposed System

2.1 Proposed System

➢ Handwritten Hindi character recognizer with the help neural network

➢ 70K+ data is used to train the neural network
➢ 10K+ data is used to test the model .
➢ Different handwritten character is passed as an input to the input layer
of the model
➢ Learning technique used by model is Supervised learning technique.

2.2 Objective of the System

A) Objectives of the System:

To showcase machine learning concepts that can be used in day to

day life and computers can work as human brain.

Specific Objective of the site:

Objective of doing this project is to enhance our knowledge in the field of E-

com technology. Some of the client requirement and objectives of this site is as
under:

➢ To increase the business of Client.

➢ To make it Global.

B) Benefits of the Proposed System:

➢ Save the time and the efforts
➢ Maintain the data in very easy way
➢ Easy access to the information as when required
➢ Optimum utilization of the resource
 3.Analysis and Design
3.1 Sequence Diagram
3.2 Activity Diagram
4.0 Implementation
 5.Testing

Definition:

Testing is the set of activities that can be planned in Advanced and conducted systematically to ensure
the proper functioning of the software.
There are various types of testing as listed below: -

1) Unit testing
2) Black box Testing
3) White box Testing

Unit Testing: -

Unit testing focuses verification effort on the smallest unit of software design –software component or
module. It is white box oriented and the steps can be conducted in parallel for multiple components.

Black –box Testing: -

In this, structure of the program is not considered. The software is considered as black box to which
defined inputs are given. From this the defined outputs are obtained. In this the tester only knows what
the software is supposed to do but he cannot look in the box to see how it operates.
This testing has to be done in alpha testing & at user site. During planning stage the time for alpha &
beta are fixed.
The black box testing finds the errors such as interface errors, incorrect & missing function,
initialization & termination errors, performance errors, errors in data structure and external database
accessed.

White-box Testing: -

In this the structure of a program is taken into considerations. The objective of this testing is to ensure
each and every line of the code is tested. The tester derives test data from an examination of program
logic and structure. White box testing is more involved then black box testing. The software utilities
called “Run Time Profilers” are specifically used which give the statistics and on which statement
executed how many times. Using this testing method we can develop highly reliable software.
 6. Future Enhancements and Conclusion

6.1 Future Enhancements

The project has a very vast scope in future. The project can be implemented on intranet
in future. Project can be updated in near future as and when requirement for the same
arises, as it is very flexible in terms of expansion. With the proposed software of
database Space Manager ready and fully functional the client is now able to manage and
hence run the entire work in a much better, accurate and error free manner. The
following are the future scope for the project.

➢ Multi language hand written character will able to recognize.

➢ Image recognition with advance technology that gives higher accuracy.

➢ Solving mathematical equations.

6.2 Conclusion:

It was a wonderful learning experience for us while working on this project. This
project took us through the various phases of project development and gave us real
insight into the world of software engineering. The joy of working and the thrill
involved while tackling the various problems and challenges gave us a feel of the
developers’ industry.
 7.Bibliography

Reference sites:

www.google.com

www.Tensorflow.org
YouTube channels: Sentdex

Andrew NG videos

Courses:

https://www.coursera.org

https://www.udacity.com