Skin Type Classification using Machine Learning

Surenthran, Priyaganessri (2020) Skin Type Classification using Machine Learning. [Final Year Project] (Submitted)

[thumbnail of 24578_Priyaganessri Surenthran.pdf] PDF
24578_Priyaganessri Surenthran.pdf
Restricted to Registered users only

Download (3MB)


Skin is the largest organ of the body which protects us from microbes and helps
to regulate body temperature. Skin types are determined by genetics. However, it can
vary according to internal and external factors. Most people are not aware of their skin
types and use multiple medication with no results. The identification of skin type has
been controversial in the skincare industry because it is the primary factor to cure skin
issue. Manual skin type identification has few drawbacks like human error and time
consuming. The advancement of technology has made it easier to perform this
diagnosis. However, most of the research on image classification only focuses on skin
issues like acne, eczema and so on. There are very few which focuses on skin type
classification. Even when that it requires large number of datasets. This research was
done to analyse suitable method using image recognition and machine learning that
can classify skin type with small number of datasets. It also evaluates proposed method
to ensure it tackles the issue resulting in a testable prototype. The results obtained will
be visualised using Rstudio to help user understand better. The methodology of this
research would be data acquisition, pre-processing the data, image segmentation,
classification and testing the model. As of for future work, the accuracy should be
improved alongside number of datasets. The classification variety could be expanded
as well.

Item Type: Final Year Project
Subjects: Q Science > Q Science (General)
Departments / MOR / COE: Sciences and Information Technology > Computer and Information Sciences
Depositing User: Mr Ahmad Suhairi Mohamed Lazim
Date Deposited: 23 Sep 2021 23:44
Last Modified: 23 Sep 2021 23:44

Actions (login required)

View Item
View Item