Aplikasi Android Pendeteksi Kerusakan Kulit Wajah Menggunakan Metode 2D Gabor Wavelet dan Gray Level Co-Occurrence

Abstract

Damage that occurs in the skin can be identified by looking at the texture and color of the skin surface. Symptoms that occur when you experience skin damage are changes that occur on the surface of the skin, such as the appearance of small red bumps on the face, red spots, rough skin texture and others. This is certainly very disturbing, besides affecting the appearance, especially if it occurs on the skin on the face, it can also give pain such as acne. In order to find out the type of skin damage experienced on the face, the author created an application that can detect the type of facial skin damage experienced. This application uses the 2D Gabor Wavelet filter method and the Gray-Level Co-Occurrence Matrix (GLCM). 2D Gabor Wavelet method is used for image extraction based on edge characteristics on skin damage images. Meanwhile, the Gray-Level Co-Occurrence Matrix (GLCM) method is used for the extraction of facial images based on the texture characteristics of skin damage / disease. Classification of the type of skin damage / disease using the K-NN algorithm.The test results in this application obtained the level of accuracy of each type of skin damage and normal skin is pustules 64.5%, papule 57.8%, nodule 76.8% and normal 70.6%. The accuracy results obtained from testing using the Receiver Operating Characteristic (ROC) algorithm by utilizing the area under curve (AUC). Keywords: Gray-Level Co-Occurrence Matrix (GLCM), 2D Gabor Wavelet, Facial Skin Damage , K-NN.