The Face Detection Technology Research In Color Image Based On Skin Color Segmentation And Unascertained Clustering

The technology of face recognition is one of the biometric technologies. It is widely accepted because of its application in identification with non contact.Face detection is an important part of face recognition, its definition is:the input image or images, if there will be face, then detect the faces’location, size and other important information. At present, the face detection technology has been applied in many fields, such as image retrieval, digital video processing, access control monitoring system.The paper uses the skin color segmentation technique for the detection of skin color regions, then uses the algorithm of unascertained clustering on these skin color regions.That is the realization of face detection technology.Most of the current detection algorithms are to do that change a acquired color image into gray image for processing.The paper processes the original color image, improves the efficiency of detection. The color spaces that traditional color segmentation methods use mostly, the skin color segmentation effect are not good.In view of this, this article uses the new color space YCgCr and the Gauss model in the skincolor segmentation, at the same time using edge detection algorithm to extract the edge information of original image, then skin color segmentation effect graph and the edge information of superposition, get color map, to its through morphological denoising, using unascertained clustering algorithm classify the skin color regions, in order to detect human faces.