Study On Human Face Detection Algorithm Based On Skin Color

Recently, human face detection, as a key technology in human face information processing, is becoming a big problem that gathers more attention in the field of pattern recognition and computer vision. And now it has been applied into other application such as entrance security, video coding, video surveillance and tracking and content based image retrieval etc. The existing face detection methods can be classified into four categories: Knowledge-based methods, Feature invariant approaches, Template matching methods and Appearance-based methods. In this paper, a novel face detection approach, which combines skin classification and cell-divided feature verifying method, is presented to detect the frontal faces of multi-angle and multi-scale. Optimized illumination compensation can optimize the quality of the image to some extent. In regard to facial segmentation we utilize clustering algorithm. Because the hue component plays an important role as well as saturation component, we take the two factors into consideration during our segmentation work. In order to decrease the insufficient or excessive segmentation effectively, we add mosaic algorithm, edge extraction using sobel operator and region combination. And we combine intensity and wavelet information to extract human eyes features. This method can crop eyes feature from the facial images with shadow, high light or gloom characters and save the feature details when we do the extraction job. Using high-boost filtering, mouth region can be found more easily under a wider lighting conditions. The proposed face detection system can detect human faces of multiangle and multiscale under different lighting conditions in color images with complex background, and expression change is tolerated. Experiments demonstrate that an approximately 85% success rate is reached.