ISSN: 2455-2631 © March 2021 IJSDR | Volume 6, Issue 3
IJSDR2103083 International Journal of Scientific Development and Research (IJSDR) www.ijsdr.org 498
Type of Skin Disease Identification by Machine
Learning Using Python
Dr. R. Muthalagu
1
, M. Mohammed Musheer
2
, K. Nandhini
3
, N. Saranraj
4
Professor
1
, Student
2
, Student
3
, Student
4
Department of Biomedical Engineering,
Agni College of Technology, Chennai, Tamilnadu, India
Abstract: It is a challenging factor for doctors even with the existence of emerging technology, to diagnose the skin disease
symptoms. Because many people are exposed to serious skin diseases that require them to go to hospitals and go through a
number of different expensive medical examinations which takes up to days. The proposed work can solve the above
problem to an approachable extent, through the design of a program by Python Machine Learning. A method is based on
vectors and pixels classification of the images, and is proposed to identify the five various types of skin diseases: The
diseases are namely Melanoma, Psoriasis, Rosacea, Vitiligo and Xanthelasma. It is a process to detect the type of disease in
just a few seconds, making the diagnosis more fast and realistic. The aim of this project is to classify the different diseases
based on images given as input. The project is purely based on python software platform. The images are collected from
various publicly available databases like DermWeb, Dermnet etc. Here the proposed method is the use of Machine Learning
with tensor flow for training the dataset and the SVM Algorithm to classify the five types of skin diseases in Python software.
The types of diseases like Psoriasis, Melanoma, Rosacea, Vitiligo and Xanthelasma can be identified and shows the output
as the name of the disease in python Software as the output.
Keywords: Diseases, Fast Classification, Python, SVM, Machine Learning.
I.INTRODUCTION
In the day to day life the skin diseases is the disease which is the most common disease of all human health afflictions
and it affects millions of people in the world at any time. The ranking of the skin disease in the world is about fourth because it is a
common cause of human illness, but many affected people do not consult doctors. To include people who check with the doctors at
the time screening examinations were performed randomly on participating visitors. According to the study for about 2700
individual people, at least one skin abnormality was observed in 1,660 of the participants that is for about 64.5 percent [4]. The
Psoriasis (26.6 percent), Rosacea (25.5 percent), Xanthelasma (15.2 percent), Vitiligo (23.3%) and Melanoma (11.7 percent). Now
a day’s the skin diseases are increased with the age due to the atmospheric condtions and they are frequent in men (71.9 percent)
than in women (59.5 percent). Most of the affected participants were unaware of their abnormal findings of their skin. The skin
diseases are the one varying from small problems like psoriasis to high level skin diseases like cancerous diseases like melanoma
[12] and so on. It occurs mostly and affects the diabetic’s patients like drying of skin.
The Image processing is one of the emerging technologies which has been growing rapidly in field of medicine. The
following equipment like magnetic resonance imaging (MRI), computed tomography (CT) and digital subtraction angiography
(DSA) was same of the equipment that are wildly used on people’s life that is based on digital image technology. [13]
The aim of the project is to identify the five types of skin diseases which normally occurs to the common human beings
and leads to severity and causes death too. The need of the project is the database which consists of two categories: training and
testing. Because the training dataset images are used to train the computer and make the computer to learn about the types of
diseases. The testing dataset is used for verification and evaluation of the disease [14].
The flow is to collect sample data which will be trained using Python with help of dependencies modules tensor flow with
its classification using the SVM Algorithm [12] and processed through the image processing in Python software with the testing
data then the types of diseases like Psoriasis, Melanoma, Rosacea, Vitiligo and Xanthelasma can be identified and shows the output
as the name of the disease in python Software.
The main objective of this paper is to design a program using the python to detect human skin diseases [14] at an early stage. The
outcomes are stated below:
1. To be able to classify the images based on the type of disease by SVM Classifier.
2. To come up with a training model that can be used for testing.
3. To create a Graphical User Interface (GUI) for making it easy to understand.
II. PROPOSED SYSTEM
The objective of this project is to classify the different diseases based on the test image which is given as input. The project is based
on python software platform. The dataset images are collected from publicly available databases like Dermnet, DermWeb, etc. In
the First Step, the sample images of five skin diseases need to be pre-processed using the SVM [11] and tensor flow. Secondly, the
SVM is needs to train with the tensor flow and made corresponding geometric transformation pixels. Based on this, five types of