A Hybrid Skin Color Model for Face Detection
Vandana S. Bhat · 2014
For a human vision system performing the process of face detection is an easy task compared to an intelligent machine. This paper contains a system for human Face detection by combining various skin color models. It's an amalgamation of three diverse skin color models specifically the RGB, YCbCr and HSV. A hybrid Skin color model has been proposed for varying illumination conditions. The main objective for using is to overcome the problem of illumination conditions availability in an arbitrary image. Skin detection can be defined as the process of finding skin-colored pixels and regions in a given image. Skin detectors classically transform a given pixel into an appropriate color space and the classification is used to label whether it is a skin or non-skin pixel. The extraction of skin region is carried out using a set of bounding rules based on skin color distribution. Later the segmented face regions are classified using combination of morphological operations. Experimental results on the benchmark face databases such as FERET and the acquired images showed that the proposed model is able to achieve an accurate detection for near-frontal face orientation and skin color.