Face Recognition Algorithm Improvements

Face recognition is a biometrc technology possessing great developable potential, researching on the face recognition technology has great the oretical and practical values.The original face image captured by the recognition system usually is denoted the grey values of grid pixels.Isolated grey values of pixels cannot reflect the characteristics contained in human face directly.Two-dimensional Gabor wavelets transform can link the pixels in an adjacent region together,and reflect the changes of the grey values of pixels in a local area of an image from different frequency scales and orientations.This dissertation researches into the theory and technology of face recognition through two-dimensional Gabor wavelets transform,the main content of the dissertation are as follows:(1)The preprocessing of face image is researched.The aim of face image preprocessing is to regularize the face image which is captured by image collecting devices to normalized image,it includes three steps mainly: face image light compensation,eyes detection and location,geometry normalization and grey value normalization.This dissertation emphasizes on the research of eyes detection and location method which is based on improvement symmetry transformation algorithm.(2)Two-dimensional Gabor wavelets transform is researched.Two-dimensional Gabor wavelets transform is realized by computing the convolutions of a bank of two-dimensional Gabor filters and the grey values of pixels in an area aroud a given position in an image,and then researches the two-dimensional Gabor filter parameter choice and its significance.(3)Classical elastic bunch graph matching algorithm is improved. The classical elastic bunch graph matching algorithm matches the face image to a predefined face bunch graph in order to obtain the rough positions of the feature points firstly;then,every feature point is located through elastic fine adjustment;lastly, the two-dimensional Gabor wavelets transform coefficients are computed at the feature points and these coefficients are used for face classification and recognition.the computation of classical elastic bunch graph matching algorithm is prohibitive.In this dissertation, the characteristic point suitable has been taken and weighted the value of choice,seven representative grid structures are obtained through the clustering of the grid structures of the face labeled graphs of many training images,these grid structures are used to constitute a template bunch of the face bunch graph.During the matching stage,the elastic bunch graph matching algorithm is combined with improvement symmetry transformation algorithm:firstly,the eyes are located by improvement symmetry transformation algorithm,using eye coordinates as the datum marks,the input image is geometrically normalized;then,the most appropriate grid structure is selected from the template bunch to determine the geometrical feature of a face, and precise matching is per formed further based on the outcome.The matching computation is simplified largely after improvements. The experimental results indicates that the improvement elastic bunch graph matching algorithm has certain enhancement comparing to the original algorithm in the recognition rate,and in overcoming aspects that are light,expression,posture and so on also have the very big improvement.