PCA expression recognition based on false geodesic distance

Expression recognition is a branch of face recognition. This paper describes how a 3D model is rebuilt from 2D static face image through specific algorithm and how the principle component better representing face characteristics is obtained by collecting geometrical characteristic information manifesting emotional variations through false geodesic distance in 3D space and eliminating the relativity of the characteristics extracted through Principal Component Analysis (PCA). At last the emotional mode of the samples to be tested shall be identified through distance test method by computing the Mahalanobis Distance between the test sample to be tested and the training sample.