Face Recognition Research Based On Local Binary Patterns

The challenging Face Recognition Technology covers wide fields of human brain cognitive science as well as subjects of image processing, pattern recognition, signal processing and artificial intelligence. In resent years, with the developing and popularizing of computer science and application, Face Recognition Technology, because of its extensive application prospect, has now become a hotspot in the fields of pattern recognition and image processing.Local Binary Patterns (LBP) is defined as a gray-scale invariant texture measure. This thesis makes a deep analysis and research on LBP. In order to overcome the limits of LBP, a method of face recognition is proposed, which is based on Multi-Threshold LBP (MTLBP). Firstly, the gray-scale difference is calculated between each pixel and its local neighborhoods of an image. Then different thresholds are chosen to code the foregoing gray-scale difference. Secondly, face image is divided into multi-regions. Histogram vectors extracted from multi-regions are adopted to describe the human face. Finally, face recognition will be completed by blurring the matching results. Experimental results show that the proposed method is robust to expression,background and distance variations.The application of block weighted technology in face recognition is discussed and a face recognition method based on Entropy-Weighted LBP (EWLBP) is proposed. For one thing, the human face image is separated into several sub-blocks after coding the whole image using LBP operators. For another thing, the method calculates the entropy value of each sub-block and endows them to the similarity in order to improve the recognition performance. The experiments demonstrate the proposed method's effectiveness and feasibility.