PROJECT TITLE
A MINOR PROJECT REPORT
[INTERNSHIP REPORT]
Submitted by
STUDENT1 NAME [REG NUM]
STUDENT2 NAME [REG NUM]
Under the guidance of
Guide Name
(Guide Affiliation)
in partial fulfillment for the award of the degree
of
BACHELOR OF TECHNOLOGY
in
COMPUTER SCIENCE & ENGINEERING
of
FACULTY OF ENGINEERING AND TECHNOLOGY
S.R.M. Nagar, Kattankulathur, Chengalpattu District
NOVEMBER 2021
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
(Under Section 3 of UGC Act, 1956)
BONAFIDE CERTIFICATE
Certified that 18CSP107L minor project report [18CSP108L internship
report] titled “PROJECT TITLE ” is the bonafide work of “STUDENT1 NAME
[REG NUM], STUDENT2 NAME [REG NUM]” who carried out the minor project
work[internship] under my supervision. Certified further, that to the best of my
knowledge the work reported herein does not form any other project report or
dissertation on the basis of which a degree or award was conferred on an earlier
occasion on this or any other candidate.
SIGNATURE
SIGNATURE
GUIDE NAME
HOD NAME
GUIDE
HEAD OF THE DEPARTMENT
Guide Affiliation
Professor
Dept. of Computing Technologies
Signature of the Panel Head
Panel Head Name
Panel Head Affiliation
Internship offer letter or completion letter for internship students
only
ABSTRACT
Abstract must be a single paragraph in times new roman 14pt with a maximum
of 300 words.
iii
TABLE OF CONTENTS
ABSTRACT iii
TABLE OF CONTENTS iv
LIST OF FIGURES v
LIST OF TABLES vi
ABBREVIATIONS vii
1 INTRODUCTION 1
1.1 subtitle 1 2
1.2 subtitle 2 3
1.3 Software Requirements Specification 4
2 LITERATURE SURVEY 5
2.1 subtitle 1 5
2.2 subtitle 2 10
3 SYSTEM ARCHITECTURE AND DESIGN 15
3.1 subtitle 1 15
3.1.1 subsection 1 16
3.1.2 subsection 2 17
3.2 Design of Modules 18
4 METHODOLOGY 21
4.1 subtitle 1 21
4.1.1 subsection 1 23
4.1.2 subsection 2 25
4.2 subtitle 2 28
5 CODING AND TESTING 30
6 RESULTS AND DISCUSSIONS 40
6.1 subtitle 1 41
6.2 subtitle 2 43
7 CONCLUSION AND FUTURE ENHANCEMENT 45
REFERENCES 46
APPENDIX
A CONFERENCE PUBLICATION 50
B JOURNAL PUBLICATION 51
C PLAGIARISM REPORT 52
LIST OF FIGURES
2.1 Thresholding segmentation in action on the skin lesion image input . . 4
2.2 Computer Vision Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1 ROC curve CNN and dermatologists . . . . . . . . . . . . . . . . . . . . 8
3.2 confusion matrix with CNN vs doctors . . . . . . . . . . . . . . . . . . . 10
3.3 Adam optimizer error rate . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.1 Sample lesions from each class . . . . . . . . . . . . . . . . . . . . . . . 13
4.2 File structure of dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.3 Diagnosis Techniques Plot . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.4 Number of data points of each class . . . . . . . . . . . . . . . . . . . . 16
4.5 Effect of SMOTE and number of variables on KNN (Euclidean Distance) 17
4.6 YOLO CNN Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.7 ANN sample architecture . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.8 Convolution Layer simulation . . . . . . . . . . . . . . . . . . . . . . . 25
4.9 Pooling layer simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.10 CNN complete architecture . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.11 CNN pre assigned weights for each class . . . . . . . . . . . . . . . . . . 27
4.12 Sample layers of our CNN model . . . . . . . . . . . . . . . . . . . . . . 28
5.1 Overall architecture of Product. . . . . . . . . . . . . . . . . . . . . . . 32
5.2 Difference between UI & UX . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3 Ionic Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.4 Ionic Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.5 Back end Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.6 Components of API . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.7 REST API Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.8 SOAP API Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
v
LIST OF TABLES
2.1 Thresholding segmentation in action on the skin lesion image input . . 4
2.2 Computer Vision Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.3 ROC curve CNN and dermatologists . . . . . . . . . . . . . . . . . . . . 8
3.4 confusion matrix with CNN vs doctors . . . . . . . . . . . . . . . . . . . 10
vi
ABBREVIATIONS
AES Advanced Encryption Standard
ANN Artificial Neural Network
CNN Colvonutional Neural
Network CSS Cascading Style Sheet
CV Computer Vision
DB Database
DNA Deoxyribo Neucleic Acid
GCP Google Cloud Platform
HAM Human Against Machine
HTML Hyper Text Markup
Language HTTP Hyper Text Transfer
Protocol JS Javascript
KNN K Nearest Neighbours
MNIST Modified National Institute of Standards and Technology
PWA Progressive Web App
RNA Ribo Neucleic Acid
ROC Receiver Operating Characteristic
SASS Syntactically Awesome Style Sheets
SMOTE Synthetic Minority Oversampling
Technique SQL Structured Query Language
SVM Support Vector Machine
UI User Interface
UV UltraViolet
UX User Experience
YOLO You Only Look Once
Vii
CHAPTER 1
INTRODUCTION
1.1 Subtitle 1
Font style for entire report must be times new roman. Chapter number and title must be
capitalized with font size of 16pt bold. Subtitle 1 is 16pt with each word capitalized with bold.
Subsection under a subtitle is 12pt with bold, each letter capitalized. All the content of the
document is 12pt size with 1.5 Spacing. Left margin 1 inch and right margin 0.5 inch. Use justify
option for both left and right alignment. A paragraph may contain maximum of 12 lines and an
empty space to be left between each paragraph.
All figures must be numbered with corresponding chapter number like Fig 1.1, Fig 1.2 etc for
chapter 1 and Fig 2.1, Fig 2.2 etc for chapter 2 and so on.. with figures in center alignment and the
caption to be mentioned below the figure.
Fig 3.1: ROC curve CNN and dermatologists
All tables must be numbered with corresponding chapter number like Table 1.1, Table 1.2 etc
for chapter 1 and Table 2.1, Table 2.2 etc for chapter 2 and so on.. with tables in center alignment
and the caption to be mentioned above the Table.
Every chapter must start in new page. Page numbers are started from chapter 1 and to be
included till the end of the report. Page numbers can be given at the bottom of the page with center
alignment.
1
All references must be cited inside the text with sequential numbers[1] and to be listed in the
same order in references. Two articles can be cited as [4,5] and multiple references can be cited as
[7-10]. Minimum 15 references to be included.
Proofs must be included for all publications and plagiarism report to be generated using turnitin
with the help of your guide with similarity index less than or equal to 10 percent. Total number of
pages in the report is minimum 30 excluding coding and screenshots.
REFERENCES
[1] Hernández, E.; Sanchez-Anguix, V.; Julian, V.; Palanca, J.; Duque, N. Rainfall prediction: A
deep learning approach. In International Conference on Hybrid Artificial Intelligence Systems;
Springer: Cham, Switzerland, 2016; pp. 151–162.
[2] Goswami, B.N. The challenge of weather prediction. Resonance 1996, 1, 8–17.
[3] Nayak, D.R.; Mahapatra, A.; Mishra, P. A survey on rainfall prediction using artificial neural
network. Int. J. Comput. Appl. 2013, 72, 16.
[4] Kashiwao, T.; Nakayama, K.; Ando, S.; Ikeda, K.; Lee, M.; Bahadori, A. A neural network-
based local rainfall prediction system using meteorological data on the internet: A case study
using data from the Japan meteorological agency. Appl. Soft Comput. 2017, 56, 317–330.
[5] Mislan, H.; Hardwinarto, S.; Sumaryono, M.A. Rainfall monthly prediction based on artificial
neural network: A case study in Tenggarong Station, East Kalimantan, Indonesia. Procedia
Comput. Sci. 2015, 59, 142–151.
[6] Muka, Z.; Maraj, E.; Kuka, S. Rainfall prediction using fuzzy logic. Int. J. Innov. Sci. Eng.
Technol. 2017, 4, 1–5.
APPENDIX A
CONFERENCE
PRESENTATION
Our paper on Hybrid application based skin lesion analyzer using deep
neural networks was presented at ICIOT 2020 conference held at SRM. 200+
shortlisted teams presented their papers on various fields in the conference. Our
paper got accepted as paper id : 25 with a plagiarism of just 2 %.
Figure A.1: ICIOT 2020 Acceptance
On presenting the paper in this international conference held at SRM KTR
campus, we received positive remarks and suggestion from the judging panel.
We were then awarded the best paper award at the same conference.
Figure A.2: ICIOT 2020 Best Paper award
83
APPENDIX B
PUBLICATION DETAILS
We submitted our research paper for publication at IJPR publication house puducherry. We
had selected the journal International Journal of Psychosocial Rehabilitation (ISSN:
1475- 7192). We got the acceptance notification from the IJPR stating our paper has been
published in the April Issue of the same journal. Proof of publication is attached in figure B.1
The research
Figure B.1: Publication Notification
paper cover page has been attached below.
International Journal of Psychosocial Rehabilitation, Vol. 24, Issue 08,
2020 ISSN: 1475-7192
Hybrid Application Based Skin Lesion Analyzer
Using Deep Neural Networks
1
S. Poornima, 2Shivang Kaul, 3Yash Aggarwal, 4M. Pushpalatha
Abstract--Skin cancer with more than 5 million cases reported every year. Early detection can increase
the probability of survival. In recent study it was shown neural networks outperform medical board certified
doctors in classifying lesions as cancerous. We intend to build a whole system encompassing Image capturing
processing it by neural net , sending the response back to the device and formulating a report for the user. We
intent to use CNNs to classify the image of skin lesion into 7 categories of cancerous lesions: Melanoma, Benign
Keratosis, Actinic Keratoses, Dermatofibroma, Vascular skin lesion and Basal Cell Carcinoma. Our goal is to
make the system easily usable by untrained users and make detecting skin cancer easy with higher efficiency.
Key words--Neural Networks, Image Processing, Convolu-tional Neural Networks, Skin Cancer
Detection, Skin Lesion Imaging, App Development, Localization Algorithms, Cloud Computing, GCP, Compute
Engine, App Engine.
I. INTRODUCTION
Skin Cancer is a major kind of cancer with around 5 million reported cases worldwide every year. The
major cause of skin cancer is exposure to UV rays. Diagnosing skin cancer generally included the skin lesion
being examined by a doctor. Recent studies have shown neural networks to be more efficient in classifying lesion
as cancerous as compared to trained doctors. Misdiagnosing or late detection of cancer can lead to a higher
mortality rate and less chance of cure. The goal of this project is making detection and classification of lesions on
the skin easier. Not all the marks on skin are a matter of concern but early detection and treatment of cancer can
save lives. So this gives the user a way to check if there’s a chance of the mark on your skin being cancerous. The
aim of this project is to detect and analyse such a correlation using neural networks. It is expected that the
outcome of this project will lead to automated classification of skin lesions.
II. LITERATURE SURVEY
The following papers were read and analysed for the refer-ence of this paper. A brief image has been
presented here.
1) Andre Esteva et al. 2017,” Dermatologist-level classification of skin cancer with deep neural networks.”
Contribution: Claimed to classify skin lesions at par with board trained dermatologists. Methodology used:
1
Assistant Professor, CSE Department, SRMIST, Chennai, India
2
Assistant Professor, CSE Department, SRMIST, Chennai, India, [email protected]
3
Assistant Professor, CSE Department, SRMIST, Chennai, India,
[email protected] 4Assistant Professor, CSE Department, SRMIST, Chennai,
India
DOI: 10.37200/IJPR/V24I8/PR280274
Received: 21 Jan 2020 | Revised: 08 Feb 2020 | Accepted: 14 Mar 2020 2545
APPENDIX C
PLAGIARISM REPORT
