Performance of Acne Type Identification Using GLCM and SVM

Autor: Syifa Fitratul M, Anif Hanifa Setianingrum, Siti Ummi Masruroh
Rok vydání: 2020
Předmět:
Zdroj: 2020 8th International Conference on Cyber and IT Service Management (CITSM).
DOI: 10.1109/citsm50537.2020.9268797
Popis: Acne is a common skin disease in teenage to adult populations. Acne is caused by intrinsic factors, such as genetics, race and hormonal, as well as extrinsic factors such as stress, diet, cosmetics and drugs. Treatment by the doctor is generally determined by the type and severity of the patients. The identification process begins with image acquisition and acne segmentation process using Multi-Level Thresholding. The feature extraction using GLCM is applied to the image of the segmentation result to get the value of contrast, correlation, energy and homogeneity. Classification is based on previously extracted features. The best result of accuracy level for testing data is 89% of the total 18 acne images. Meanwhile, for two images non-acne, the accuracy of the testing data reached 50%.
Databáze: OpenAIRE