Description and Coding of Human Face Images Based on Active Appearance Models
Haiyan Wu · Jisuanji fangzhen · 2009
Active Appearance Model(AAM) is used to describe and code human face images.After a certain number of iteration,matching the model with face images,the face images are synthesized.This statistical model reflects the facial shape and grey-level appearance.This method searches and explains the target image by the statistical information modeling.Because of using optimal algorithm,this algorithm makes synthetic model and target image more and more similar.At last,it obtains the synthetic model which represents both the shape and the texture of the target image.Experiment indicates that AAM algorithm is feasible for describing and coding human face images.