Research On Face Recognition Hybrid Algorithm Based On Wavelet And SVM

With the continuous development of the internet and the popularity of various high-end electronic products,face recognition technology has been gradually applied to various fields and become a very important part of people's life.Traditional face recognition technology has brought many inconvenience to people's life because of its low accuracy and high time-consuming,so how to extract face features accurately and classify them accurately has become a research focus.In view of the above problems,this paper proposes a more effective face recognition method based on PCA algorithm,wavelet transform and support vector machine.The contents of this paper are as follows:(1)fast PCA dimensionality reduction.Traditional PCA is a dimensionality reduction technique,which can transform high-dimensional face image data into low-dimensional data through linear transformation.these low-dimensional data can represent the information of the original image to a large extent,so PCA can extract features from face images.However,because the dimension of face image is too high,the traditional PCA feature extraction technology has the problems of time-consuming and complicated calculation,and the text uses fast PCA dimension reduction technology.(2)Weighted wavelet transform.Wavelet transform is applied to face recognition.the face image is decomposed into a low frequency component and three high frequency components.the low frequency component represents the approximate information of the original image,and the three high frequency components represent the horizontal edge information,vertical edge information and oblique edge information of the original image.Traditional wavelet transform uses low-frequency components and abandons high-frequency components,resulting in the loss of some features in face images.therefore,this paper proposes a weighted wavelet transform,which weights the coefficients of four components and fuses the information of four components.(3)Face image preprocessing.The traditional face image is affected by external factors such as illumination,noise and so on,which will reduce the image quality and affect the face recognition effect.Therefore,the face image preprocessing operation can effectively avoid these problems so as to highlight the face image features.the face image preprocessing operation includes grayscale operation,denoising operation,correction operation and normalization operation.(4)Theta transformation.The traditional face image will be blurred and dim by illumination,which will reduce the image quality and affect the face recognition effect.although the face image can be purified by preprocessing,it also changes the brightness of the image to some extent.therefore,this paper proposes that ?transformation can change the brightness of the image to facilitate recognition,and the other face image has the risk of being stolen.therefore,before face recognition,? transformation can be used to encrypt the face image,which reduces the risk of face image being stolen.