International Journal of Mathematics and Computer Sciences (IJMCS) ISSN: 2305-7661 Vol.21 September 2013 International Scientific Researchers (ISR)

Face detection is one of the important issues which nowadays many communities deal with it and beside, faces recognition is one of the most important areas of research in most of the industrial and research centers. Three main pillars has been considered in face detection which are: classified as the skin and non-skin color, design templates for face, and feature extraction from face images enable us to identify faces correctly.[1,2] One of the effective methods that have been proposed in recent years is face detection by using skin-color feature. In this proposed method, at first skin and non-skin areas are specified by using YCbCr model, then by matching the face template on skin areas that are likely to identify a face, the faces zone can be determined in images. The main advantage of the proposed method is using simple rules to detect the skin regions. In this method, the positive error rate is almost low, because the skin color modeling and identification of the skin regions, especially in images with low quality, noisy images and images from a far distance to the character is more efficient and it has ease of access and more fast speed. For comparison, a set of 105 images consists of 82 images include only one face and 23 images containing multiple faces, as a total of 134 figures has been used.[3,4,5]