From The late 90th, "Skin Detection" becomes one of the major problems inimage processing. If "Skin Detection" will be done in high accuracy, it can beused in many cases as face recognition, Human Tracking and etc. Until now somany methods were presented for solving this problem. In most of these methods,color space was used to extract feature vector for classifying pixels, but themost of them have not good accuracy in detecting types of skin. The proposedapproach in this paper is based on "Color based image retrieval" (CBIR)technique. In this method, first by means of CBIR method and image tiling andconsidering the relation between pixel and its neighbors, a feature vectorwould be defined and then with using a training step, detecting the skin in thetest stage. The result shows that the presenting approach, in addition to itshigh accuracy in detecting type of skin, has no sensitivity to illuminationintensity and moving face orientation.
展开▼
