Research on Facial Feature Detection Based on Correlation Analysis

Abstract

In order to enhance the extraction efficiency of facial feature, the paper explores a novel defect evaluation method that uses combined features and modified method classifiers to characterize and classify the defects of facial expression. It provides a good approach to implement facial expression recognition both in 2D and 3D images. Innovative methods which are aimed at reducing the computational complexity and improving the accuracy of expression recognition are proposed.The experiments result showed the proposed method achieved lower error rate than other method.