The Research On Face Detecting And Tracking In Video

Face-tracking is an important research field of computer vision. As a key technology, face tracking has received lots of attentions in automatic human face recognition, retrieval and other fieldsIn this thesis several research works are given as follows:Firstly, recent developments of face-tracking are introduced in this thesis. In fact, face-tracking is the face-detecting technology and target-tracking technology combined with each other. In the part of face-detecting, a new white-balance algorithm is proposed. On this basis, a skin model in the suitable color space is constructed. A input color image is segmented in skin-like and non skin-like regions based the model. Candidate face regions are got after using dilate and erosion. Face verification are performing based on a set of rules about shape and size information of the rectangle around human face, and it is used to locate the human faces in candidate face regions.Main algorithms of face-tracking are introduced in this thesis. CAMShift algorithm is discussed in detail. Since the algorithm can robustly track target of different shape and size with the immunity against illuminant fluctuation and noise inference, and it is very simple. However, CAMShift suffers when flesh-like interference and occlusion occur.Some methods, such as removing the flesh-like background and accessory information, are proposed to enhance the robust of CAMShift, and an face detecting based skin is used to initiate the searching window automatically. The improved CAMShhift algorithm can track an object robustly and automatically. The experimental results demonstrate the robustness and efficiency of the proposed algorithm in real-time face tracking.