Klasifikasi Jenis Flek pada Kulit Wajah dengan Menggunakan Metode K–Means Clustering dan Convolutional Neural Network

Abstract

Skin is the largest organ of the human body that plays a role in protecting parts of the body, especially the face when exposed to sunlight. Facial skin, which is more prone to skin problems, often develops spots due to sun exposure that increases melanin production. These spots can be divided into three types, namely melasma, freckles, and acne inflammation spots, which are often difficult to distinguish by laypeople because of their similar characteristics. Therefore, this study aims to build a digital image processing system on facial skin to help laypeople identify types of facial spots and provide basic prevention and treatment information according to the type of spots experienced using the K–Means Clustering and Convolutional Neural Network (CNN) methods with three types of spots on facial skin, namely melasma, freckles, and acne inflammation spots. The data used amounted to 945 image data, including of 540 data of training, 360 data of validation, and 45 data of testing. K–Means Clustering is applied for image segmentation, while CNN for image classification, and this system can classify three types of spots on facial skin by achieving an accuracy of 93.33%.