A Study On Solving The Problem Of Illumination Occlusion In Face Recognition Using Image Gradient Method

Face recognition is an extremely important branch in the field of pattern recognition,and its application in the field of vision technology is irreplaceable.With the development of information technology,the combination of computer vision technology and mathematics,the application of face recognition is becoming more and more mature.However,in real life,it is often impossible to control the occlusion of the shadow caused by the light to provide a particularly ideal face image,so that the general face recognition performance is not ideal and cannot meet the actual needs,so how to solve the problem caused by the light Being obscured by shadows and resulting in poor recognition results has become a problem to be solved in the field of face recognition and even pattern recognition in practical applications.Based on the comprehensive study and analysis of the current research situation at home and abroad,combined with the relevant knowledge of pattern recognition,we conducted an in-depth study on how to obtain a complete face image in the shadow.The specific research contents are as follows:First,the basic concepts of face recognition and related information are introduced,and the face recognition method used in this topic and the problem of shadow occlusion on face images is briefly introduced.Secondly,according to the characteristic of the face recognition algorithm of the characteristic face method that is sensitive to shadow occlusion caused by lighting,in order to study the effect of the algorithm on the shadow image enhancement,the principal component analysis is used to process the image information for dimensionality reduction classification,in order to improve the efficiency of the algorithm recognition process,In the image pre-processing stage,the image is compressed,and the weighted average compressed image method is used to compress the image.On the one hand,it retains the feature information of the original image before compression,and on the other hand,it has the effect of smoothing the image,which can reduce the interference of noise on the compressed image.Again,an image enhancement method based on gradient optimization is proposed.This method uses histogram equalization to increase the contrast of the image,so that the pixels at different gray levels are more evenly distributed on the gray scale,and uses the Sobel operator with Gaussian filtering to convolve on the image,which is reduced by Gaussian kernel convolution.The noise in the image makes the influence of noise in the process of obtaining the image gradient smaller,and finally obtaining the face-to-image gradient plays a role in enhancing the facial features of the image.Finally,the proposed method is experimentally verified,and the proposed method is compared with the experimental results of the existing methods for solving the shadow occlusion problem.