Image analysis model for skin disease detection using mobile application

Abstract

Skin disease is one of the common form of disease form in the world. Diagnosis of the skin diseases may requires a high level of dermatologist expertise. The advancement in computer and IT capabilities can offer a good opportunity and provide support to help in skin diseases diagnosis, which still suffer from shortage many countries and computer aided skin diseases diagnosis system can provide more objective and reliable solution to this issue. There are many researches in detecting skin disease like detection of skin cancer, and tumor skin. However the accurate recognition of disease is extremely challenging due to: low contrast between lesions and skin, visual similarity between disease and non-disease areas, and this study aims to detect skin disease by captured image and apply enhancement techniques in image for make it ready to diagnosis the skin diseases by apply segmentation the image by two of clustering algorithms for give more accurately result which there are K-Means clustering algorithm with the fixed number of clusters to do processing, and second one is Fuzzy C-Means algorithm which allows one piece of data to belong to two or more clusters so it is more flexible. By using two of segmentation algorithm, the study can achieve more accurately result in the project, whereas it reach to 94% within the dataset and 85% with the external data, making it a competitor for the rest of the projects especially it is processing four types of skin disease to detect which are: acne, psoriasis, melanoma, and heat rashes. The useful information that will help to detect the disease using mobile application with save time, its need to 5-8 minutes to get the final analysis result. The project is built successfully and the interface application is connecting properly database, and all the project functionalities is working properly, despite that there are some problems occurred through the image analysis with the data collected that are distorted by a watermark that obstacle the classification process. Moreover tips and instructions if possible to do them as a quick ambulatory are discussed all system functionalities is working properly, despite that there are some problems were we faced.