Research On Technology Of Face Recognition Based On Grey Relational Analysis

The technology of face recognition is a research topic with broad prospects and practical value. The technology of face recognition generally includes pre-processing of face image, feature extraction and face recognition. This dissertation presents the studies as follows:face image pretreatment, the method of feature extraction based on K-L transform and singular value feature vector, and face recognition based on grey relational analysis.Adopt correlation coefficient to obtain the weights of some area and integrate the features of every area such as eye area, nose area, mouth area, and the whole face area. Grey relational analysis is utilized in recognizing face image.The experimental results acted on Olivetti Research Laboratory (ORL) face database and Yale face database show that the ten most singular value feature vectors of each area in face image contain over 99.89% of the total information of the whole area, and the correlation of each area and the whole face area reached more than 99.7%. It is reliable that using the ten most singular value feature vectors of each area in face image to recognize the face image. The experiments of face recognition based on grey relational analysis in this dissertation show that the method has better recognition ability in the instance of dealing with less number of feature vectors.