Face Detection Based on Skin Color Segmentation | Computer Vision and Image Processing. DE

Abstract: This paper explores the identification of human face and eye instill frontal color pictures. First; the appropriate pre-processing stage is done. It requires image refresh, RGB to grayscale conversion, image binarization, noise reduction, and removal of small artifacts. The proposed algorithm is then extended to the position of the face by detecting the edges of the face using the detection of pixel color variations in the binary image. Finally, a normalized cross-relation is applied to determine the precise location of the eyes within the localized region of the face. the proposed algorithm is robust enough to achieve approximately 94% accuracy and can locate a face in single face images.