Department : - Electronics and Telecommunication .

Project group no. : -

Name of the Project : - Analysis of WBC , RBC and Platelets using Deep Learning

Subject area : - Deep Learning

Nature of the Project : - Software

Internal guide : - Dr. Prachi Mukherji ( Allocated on ____________ )

( Date on which the Guide Name was
communicated through Notice )

Member information : -

Roll no. Student name

4249 Jui Oak

4267 Kajal Shrawagi
4265 Srushti Shinde

Sponsor details : - College / Company / Self

( Strike out whichever is not applicable . )

P .T .O.
Sr. Task details Sign. of the Int. Remarks ( Sign. of the
no. Guide for Completing by Int. Guide ) Co-ordinator
the said Task ( with Date )
1 Search for a Project ----------------X------------------ -----------------X----------------
Idea [ From __________ (dt.)
to ____________ (dt.) ]
2. Rough Synopsis ----------------X------------------ Approved / NOT Approved
submitted for FINAL [Suggestions : -
Approval ________________________
(Date : - _________ ) ________________________
___________________________
___________________________
_________________________ ]
3. NEXT Rough Synopsis ----------------X------------------
of submitted to Int. Approved / NOT Approved
Guide.( In case the
previous idea is
Rejected.) (Date : -
_________ )
4 Reporting to the
Int.Guide &
Registration of
Personal Information
in the Int.Guide’s File
5. Fair Synopsis
submitted to Int.
Guide. (with WORK
PLAN enclosed .)
6 Review ----------------X------------------ ----------------X------------------
7 Review ----------------X------------------ ----------------X------------------
8 Internal Seminar & ----------------X------------------ ----------------X------------------
Sem. I Assessment
9 Seminar Report
DRAFT copy APPROVAL Approved / NOT Approved

10 Seminar Report FAIR ----------------X------------------

copy APPROVAL
11 Project Report ( Draft of
Initial Chapters )
12 Review ----------------X------------------ ----------------X------------------

13 Review ----------------X------------------ ----------------X------------------

14 Internal Project ----------------X------------------ ----------------X------------------
Working
Demo.(Competition) &
Sem. II Assessment
15 Project Report DRAFT Approved / NOT Approved
copy APPROVAL
16 Project Report FAIR ----------------X------------------ ----------------X------------------
copy APPROVAL

[ Note : - a) Put a DASH in those Columns , which are NOT APPLICABLE .

b) The matter can be continued on a NEW page if NOT fitting in this single page . ]
Sr. Details of the References searched Name of the Name of the Sign.
no. Internet IEEE Previous About Faculty Company of the
Sites Research Academic extending member(s), (s), Co-ord
(mention the Papers Years’ T.E. Mini you have you have inator
Site-Address (mention the Project Project approached approached ( with
& other Title of the Ideas (mention for Getting a for Getting Date )
details) Paper(s) & (mention Details ) Project idea a Project
its Author(s) the Titles) ( Prof.) idea

1. https://thes Nurasyeera
ai.org/Public Rohaziat, Mohd
ations/ViewP Razali Md
Tomari, Wan
aper?Volume Nurshazwani Wan
=11&Issue=1 Zakaria, Nurmiza
0&Code=IJA Othman “White
CSA&SerialN Blood Cells
Detection using
-- -- Prof. Mrudul
o=58 YOLOv3 with Dixit
CNN Feature
Extraction
Models'',Internatio
nal Journal of
Advanced
Computer Science
and Applications,
Vol. 11, No. 10,
2020.

2. https://www Lipi
.researchgate B.Mahanta,Kangk
.net/publicat ana Bora,Sourav
ion/3374905 Jyoti
47_Automate Kalita,Priyangshu
Prof. Mrudul
d_Counting_ Yogi “Automated
Counting of -- -- Dixit
of_Platelets_a Platelets and
nd_White_Bl White Blood Cells
ood_Cells_fro from Blood Smear
m_Blood_Sm Images” ,
Springer Nature
ear_Images Switzerland AG
,2019

3.
 https://www Navid Shaikh,
.irjet.net/arc Karansing
hives/V6/i4/ Rajaput, Rahul
Pawar, Dr. A. S. -- -- Prof. Mrudul
IRJET-V6I45 Vibhute Dixit
02.pdf “Counting of
RBCs and WBCs
using Image
Processing
Technique'',Intern
ational Research
Journal of
Engineering and
Technology
(IRJET),April
2019.
.

( NOTE : - Attach Copy of the SAME Work-Plan thathas been enclosed with the Fair Synopsis. Tick-mark in the
Box provided after Completion of each Module.)
Month&wk JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY
1 2 3 4 5 6 7 8 9 9 10

task mod
1.Concept
Development

2.Initializatio
n
a) Literature
review
b)Block
Schematics

3.Project
definition
scope and
Goal
establishme
nt

4.Presentatio
n with
internal and
external
guide
5.Preprocess
ing and
training

6.Validation/
Testing

7.Deploymen
t
(Optional)

8.Report
Sr. Month Date Discussion Task given by Task Task Task Sign.
no. Details Int.Guide given by Completion Completio of the
Ext. Sign. of the n Sign. of Co-ordi
(Industry Int.Guide the nator
) Guide ( with Date ) Int.Guide ( with
( with Date )
Date )
1. JULY 1/7/20 Topic Explore Delete 30/8/2021 Delete
research different this this
topics column column
across the repeated
domain
2.
AUG-SEP 1/9/2020 Literature Explored 30/9/2020
Survey number of
papers
3.
Architect Possible 31/10/202
October
1/10/202 ure architecture 0
0 finalizatio s available
n,scope
research
4.
November 1/11/202 Exploring Presentatio 31/12/202
0 the n 0
finalized
model,coll
ecting
resources Hyperpara 28/2/2021
5. to meter
January 1/1/2021 implemen discuss
t
6.
March 1/3/2021 Coding- Discussion 31/3/2021
preproces of training
sing, results and
training validation
output
7.
1/4/2021 Validation Research 30/4/2021
Report paper
Report Discussion
submissio
n
 Day : - Friday Date : - 8/5/2021

● Purpose -
To provide better and accurate results for the medical diagnosis of patients.

● Scope -
Dataset for the training is expected to have more number of images in order to provide results accurately
without any overfitting.

● Software Overview -
The software used is Google Colaboratory wherein it was coded using Python language using different
libraries suitable for the project and GPU was used to implement it faster.

● Infrastructure requirements - This part lies in the optional implementation(if time permits) , wherein we
would deploy it into an application which could be used on a suitable device.

only those students who have taken part in this activity : -

Sign.(s) of
Srushti Shinde Jui Oak Kajal Shrawagi
 Sign. of Int. Guide : - Dr.Prachi Mukherji Sign. of Ext. Guide : -
Sign. of Co-ordinator : - Prof. Milind Patankar
( Note : - The matter can be continued on a NEW page if NOT fitting in this single page . )

Day : - Friday Date : - 8/5/2021

METHODOLOGY
A. DATASET

This work involves the use of a dataset consisting of blood smear images taken from BCCD .There are three image
categories which are WBC’s , RBC’s and Platelets.Dataset is split up into training, testing and validation . The number
of training images ,testing images and validation images are 255 ,36 and 73 respectively..There are 4888 labels across
3 classes. The labelling process is already done in the dataset for each image.After successful completion of dataset
preparation,training process was continued.
B. YOLOV5

‘You Only Look Once’ in short YOLO is an object detection classification algorithm.For faster prediction of image and
location only one forward propagation is required across the network which splits the image into regions and predicts
bounding boxes and probabilities for each region.Grid cell which is responsible for detecting a particular object should
have it center falling into the respective grid cell .The dataset images are already resized to 416 x 416 dimension

C. EVALUATION OF VALIDATION DATA

The values of mAP(mean Average Precision ), Precision and Recall were calculated and the model’s performance was
determined.

D.DETAIL DESIGN
The project involves the use of the yolov5 model for detection of different blood cells.Yolo V5 is implemented with
283 hidden layers, including input and output layer.Different values of batches and epochs were passed and tested and
the output was observed.

only those students who have taken part in this activity : -

Sign.(s) of
Srushti Shinde Jui Oak Kajal Shrawagi
Sign. of Int. Guide : - Dr.Prachi Mukherji Sign. of Ext. Guide : -
Sign. of Co-ordinator : - Prof. Milind Patankar
( Note : - The matter can be continued on a NEW page if NOT fitting in this single page . )

Day : - Friday Date : - 8/5/2021

Name of the Module : - You Look Only Once (YOLO-V5)

 For Individual SOFTWARE Modules
Day : - Friday Date : - 8/5/2021

Name of the Module : - Yolo v5

Equipment used : - Google colaboratory , GPU

Observations : - [ Note : - In case of Software Projects , Computer printouts of the Result(s)

may be attached . ]
Sr. Input (s) given Expected Output (s) Actual Output (s) Comment on the
no obtained matching of
Expected O/p &
Actual O/p
 1. Images segmentation of blood cells segmentation of the accuracy
with bounding boxes blood cells with obtained was 93%.
bounding boxes

only those students who have taken part in this activity : -

Sign.(s) of
Srushti Shinde Jui Oak Kajal Shrawagi
Sign. of Int. Guide : - Dr.Prachi Mukherji Sign. of Ext. Guide : -
Sign. of Co-ordinator : - Prof. Milind Patankar

As Our project is software based, as such no money is required

to execute the entire project.
 the students : -
Sign.(s) of
Sign. of Int. Guide : - Dr.Prachi Mukherji
Sign. of Ext. Guide : - ___________
Sign. of Co-ordinator : - Prof. Milind Patankar

Sr. Details of the Information Name of the Sign. of

no. source Module in your Int. Guide
Author(s) Title of the Publisher Project for which
( in case Book / ( in case of this information
of a Book Paper / a Book ) is searched .
or Paper ) Website
address
Nurasyeera “White ,International Journal CNN
Rohaziat, Blood of Advanced
1. Mohd Razali Cells Computer Science and
Md Tomari, Detection Applications
Wan using
Nurshazwani YOLOv3
Wan Zakaria, with CNN
Nurmiza Feature
Othman Extraction
Models''
Zhengfen “Improved 2021 Optical Society DEEP
LEARNING
Jiang,Xin detection of America,
2.
Liu,Zhuangzh performanc
i Yan,Wenting e in blood
Gu,and Jiehui cell count
Jiang by an
attention-g
uided deep
learning
method”,
Navid Shaikh, “Counting ,International COUNTING OF
BLOOD CELLS
Karansing of RBCs Research Journal of
3.
Rajaput, and WBCs Engineering and
Rahul Pawar using
 Image Technology
Processing (IRJET),April 2019.
Technique''
,
Mohammad “Machine ,Healthcare MACHINE
4. LEARNING
Mahmudul learning Technology Letters,
Alam, approach of 2019,
Mohammad automatic
Tariqul Islam identificati
on and
counting of
blood
cells”
Akshaya P. “COUNTI International Journal COUNTING OF
5. RBC,WBC,
Sahastrabudd NG OF of Research in
PLATELETS
he RBC AND Engineering and
WBC Technology
USING
IMAGE
PROCESSI
NG: A
REVIEW”
Lipi “Automate Automated Counting COUNTING OF
RBC,WBC,
B.Mahanta,K d Counting of Platelets and White
6. PLATELETS
angkana of Platelets Blood Cells from
Bora,Sourav and White Blood Smear Images
Jyoti Blood
Kalita,Priyan Cells from
gshu Yogi Blood
Smear
Images”
Khaled “Improved Hindawi CLASSIFICATIO
N RBC,WBC,
Almezhghwi Classificati Computational
7. PLATELETS
on of
 and Sertan White Intelligence and
Serte Blood Neuroscience
Cells with
the
Generative
Adversarial
Network
and Deep
Convolutio
nal Neural
Network”,
Simge Çelebi “ Red and AIMS CLASSIFICATIO
N USING
and Mert white blood Bioengineering,
ARTIFICIAL
8. Burkay Çöteli cell INTELLIFENCE
classificati
on using
Artificial
Neural
Networks”
Mehdi “ White Springer-Verlag CNN
9.
Habibzadeh, Blood Cell Berlin
Adam Krzy Differential Heidelberg,2013.
̇zak CountsUsin
g
Convolutio
nal Neural
Networks
for Low
Resolution
Images”
Thanh Tran “ Blood Preprints,September SEMANTIC
10. SEGMENTATIO
Thi Phuong, cell count 2019.
N
Lam Binh using Deep
Minh, Learning
 Suk-Hwan Semantic
Lee and Segmentati
Ki-Ryong on”
Kwon

Sign. of the Coordinator :-Mr. Milind Patankar

( Note : - The matter can be continued on a NEW page if NOT fitting in this single page . )