Face Recognition And Research Based On SURF

At present,with the development of pattern recognition technology,people in the direction of face recognition technology research more and more in-depth,for the face image information description of the details,specific,can better accelerate and enhance the face The efficiency of recognition is more and more concerned by people,so this paper presents a more detailed description of the face image information and can be classified and identified algorithm.First of all,the input of the face image related to the pretreatment,specifically for the first face image histogram equalization,light compensation,and then affect the face of the image noise to do the appropriate filtering,using the Value filtering method,and then the face image of the positive picture of a normalized processing,so that it can meet the standard face form.Then,in the process of feature extraction,we find that the feature description information is simple and its rotation is not good.In this paper,we add the rotating invariant LTP feature to consider the local texture feature of the image in the feature point description process,The LTP operator with the invariant rotation can describe the characteristics around the feature point,and then a SURF-based feature description algorithm which can describe the image feature information more in detail is proposed.Then,the improved SURF algorithm is applied to the feature point detection of face images.Compared with Harris corner detection,it is found that the SURF algorithm has better stability under the influence of changing factors.And the improved SURF algorithm and SIFT algorithm in the face image matching efficiency on the experimental verification,found that the algorithm in its matching frame rate is higher than the SIFT algorithm matching results,the efficiency of the algorithm to achieve the effect of strengthening.Secondly,in the face recognition,it is found that the face recognition rate is low and the recognition time is long.Therefore,in order to improve the efficiency of face recognition and recognition,and in order to shorten the time,the Euclidean distance and support vector machine Advantages and disadvantages,so that it can better serve the face recognition technology,proposed based on the texture feature of Euclidean distance and support vector machine combined two-level classifier of this method.Specifically,the Euclidean distance is used as the classifier of the first level of the face image recognition,and this method of the support vector machine is used as the classifier of the second stage so that they can cooperate with each other and finally achieve the desired result.Finally,through the relevant experiments to verify the efficiency of face recognition,reflecting the effectiveness of the method.