Analysis of Human Skin Texture by using Machine Learning Approaches
Keywords:Machine Learning, Skin texture analysis, Gray-Level Co-Occurrence Matrix, Decision Tree Classifier
The production and improvement of cosmetics, skin texture modeling, facial recognition in security applications, as well as computer-assisted dermatology diagnostic, all benefit greatly from an understanding of skin texture analysis. Several types of skin conditions that affect humans will increase the problem in daily life. Own research shows that the methods have significant limitations such as lack of accuracy and also the methods for texture analysis can’t be in a specific category. So, to overcome the problem, the method and technique were improved to solve the problem of human skin texture. The Gray Level Co-Occurrence Matrix features a method and Decision Tree methods as a classifier to analyze the human skin texture used in this project. The features will be used to train a machine learning algorithm that will learn and analyze the skin features extracted in the previous process. Using this data, the machine learning algorithm can easily categorize the skin texture based on its level of dryness. The experiment result shows that the overall framework achieved a satisfactory result in recognizing and categorizing skin dryness based on skin texture.