SURVEY OF PERFORMANCE AND ANALYSIS COLOR SPACES FOR FACE DETECTION

Abstract

Skin color segmentation is a technique of discrimination between the skin and non-skin pixels of an image, Color modeling is one of the methods used for face detection. However robust detection techniques are facing difficulties as skin segmentation is still an ongoing hard problem to be sorted out.  Many research works have been submitted to work on the simplest color model for face detection, as there is no consensus on a suitable color model for color segmentation, so we present a comparison between the different color spaces used in skin modeling and study reference materials and techniques related to this field and present the best color model for facial detection using color segmentation. The objective of this paper is to analyst and a comparison of the various color spaces used in skin modeling and discover the color model that is best for detecting faces using color segmentation. The analysis is based on a literature survey of studies on color segmentation from recent publications reputable journals dan publication repositories. The study is relied on conducting a comprehensive review of the published color spaces and then classifying these color models and their advantages and limitations based on the influence of external factors affecting skin color detection and face detection. the YCbCr color model appeared as the best color used in face detection, which includes luminance and coloration in which colors are represented by a region and one component luminance (y), and coloring (Cr, Cb). In the future, this result will be used to help researchers develop and improve skin tone detection through color segmentation for face detection.