Research On Face Recognition Based On Local Binary Pattern Operator

Along with the rapid development of computer vision and pattern recognition technology, many recognition methods based on the biometrics have been developed. They play important roles in our real life and other fields. Automatic Face Recognition (AFR) holds an important position in various biometrics techniques for its superiority, which has important theoretical research value and broad application prospects. Face recognition is a typical problem in image pattern analysis, understanding and classification compute area. In the past four decades, great achievement has been made in AFR. In close to ideal condition, face recognition system can accomplish certain work. Feature extraction is the crux of face recognition problem, which is one of the most important and challenging aspect of the study. Recently, the LBP has been successfully applied to face recognition as texture descriptor and excellent result has achieved. However, there are still many limitations in the basic LBP operator and the LBP-based face recognition algorithm, when the dimension of samples exceeds thousands even reaches ten thousand. In this paper, the LBP features space is studied and analyzed from the texture features of images, and discuss the application possibility of the access control system which based on face recognition in the army's important places. The main works are as follows:1. We proposed an improved algorithm after studying and analyzing the basic principle of LBP and the deficiency of original LBP operator. First LBP image has been computed and divided into blocks. After this, the LBP histograms have been built up into a high dimension eigenvectors, then the classical PC A has been used to reduce the dimensions, finally classifier has been adopted for classification and recognition. This algorithm divides different images in different manner, so it makes dimension reduction more flexible and improves the accuracy of feature extraction.2. In this paper, we discuss three kinds of LBP operators and adopt many partition methods for LBP images, then experiment with images from three different face databases. Supported by a large number of experimental data, we compare the impact of various methods and steps on the final recognition performance.3. This paper discuss the application of access control system based on face recognition in the important places in the army, and provide suggestions on the structures, functions and realization methods of the access control system in basic unit army, which plays an important role on normalization and modernization construction of the troops.This paper research on LBP operator and its application in face recognition, LBP dimension reduction, LBP block and PCA dimension reduction deeply. The method is introduced in the army, which is valuable for further development of automatic face recognition.