Improvement of Face Recognition Technology Based on Intelligent Image

Traditionally, the face recognition algorithm based on principal component analysis (PCA) is not robust to the change of skin color and pose. In order to solve this problem, a dynamic face recognition algorithm based on block sample feature matching is proposed According to the principal component feature decomposition algorithm, each block feature quantity is projected onto the base coordinates of the test face sample set. The average mutual information entropy of each module vector quantization coding is used to represent the weight of the module, and the corner detection and texture matching of the face are carried out. The weight of the module is given to the feature face projection of the module, and face recognition is optimized The simulation results show that the algorithm has good feature matching and high recognition accuracy.