REAL-TIME SKIN SEGMENTATION ON LOW RESOLUTION AND GRAY IMAGES

Abstract

In this study, the skin segmentation problem, which is frequently used in the literature, will be detected on a real-time webcam with semantic segmentation deep learning algorithms. Skin segmentation generally requires high processing power to distinguish it from images with high resolution. The deep learning algorithm and the proposed image processing method used in this study provide real-time, very low CPU usage, and lag-free skin detection via webcam.