Android Application based Melanoma Detection System

Computer Engineering

Mobile devices including smartphones are being used by billions of people all around the world thus creating an oppurtunity to design a wide range of mobile imaging applications. Among many imaging applications, health care applications have drawn a lot of attention recently.Traditional methods for early detection of melanoma rely upon a dermatologist to visually assess a skin lesion using the ABCDE (Asymmetry, Border-irregularity, Color variation, Diameter, Evolution) criteria before confirmation can be done through biopsy or pathologist. In the proposed system, we have researched on a mobile imaging system for the early diagnosis of melanoma. Our work focuses on smart phone-captured visible light images. Capturing of images under loosely-controlled conditions puts forth challenges for melanoma detection as the smart phone is subject to computation and memory constraints. Thus to address these challenges, we propose to localize the skin lesions by using a fast segmentation algorithm- K means segmentation. Further a feed forward back propagation neural network is designed to improve the quality of classification.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.