Research And Application Of Key Features Of Human Face Location

Facial feature point location is one of the fundamental and crucial problems in the field of facial recognition, computer vision and graphics. It aims to locate the facial feature points and the shape information of eyes, mouth and so on based on the facial detect. The state-of-art AFR system can perform identification successfully under well-controlled environment. However, evaluation results and practical experience have shown that AFR technologies are currently far from mature. A great number of challenges are to be solved before one can implement a robust practical AFR application, especially the accurate facial feature location problem, which is the precondition of facial recognition.Active Shape Model (ASM) is one of the popular methods in facial feature point location. ASM is based on PDM (Point Distribution Model). In PDM, the original shape vectors of objects that similar in shape are composed of Landmarks coordinates. We first align the training set, and then perform PCA (Principal Component Analysis) to model the shape variation. We first get the best location of Landmarks using Local Texture Model during ASM search, and then adjust the shape parameters according to the restriction of global shape iteratively. In this thesis, classical ASM is profoundly researched. Based on the classical ASM algorithm, novel and creative improvement is proposed. The main discussion of this thesis is as follows:1) System overview of biometric identification technology and facial feature point location and introduce the current prevailing algorithms, comparing the advantages and disadvantages of various algorithms.2) Introduce the classical ASM algorithm in detail. The ASM is mainly composed of modeling global shape variation and local texture, and the target search based on Active Shape Model.3) Introduce the basis theory of wavelet, and the measurement of Gabor similarity.4) Introduce four improvements based on classical ASM algorithm: Local Texture Model with edge restriction, Multi-Resolution search, Landmarks local Gabor feature, Initialization of mean shape.5) We combine ASM and Gabor wavelet. During search, we use both local texture feature and local Gabor feature of Landmarks to improve the precision of feature point location.6) We introduce the application of Improved ASM algorithm in Face 3D reconstruction and Face recognition.