Research On Face Recognition Algorithm Based On LBP

With the development of science and technology,face recognition technology has become a hot topic nowadays.Many scientific research institutions and scholars are committed to the practical research of this technology.At present,face recognition technology has made great breakthroughs and has been realized in many application fields,such as station detection,criminal pursuit,intelligent monitoring and so on.In this technology field,face detection,image classification,image feature extraction,image retrieval and so on all embody the practical value of this technology.And feature extraction is particularly important,which directly affects the accuracy of recognition.For studying how to acquire image features better,existing algorithms are analyzed and innovated in this paper,and propose a feature extraction algorithm based on LBP algorithm and Gabor wavelet transform technology.FPGA technology is used to simulate and analyze the algorithm in this paper,and test the performance of the algorithm through experiments.The traditional LBP algorithm is realized by acquiring local feature operators in the image.Its advantage is simple calculation and strong texture description.Although this method can acquire image texture when extracting image features and is robust to the uneven change of illumination,the acquired image texture does not have spatial characteristics.In order to solve this problem,this paper introduces the block-based LBP technology,which cuts the original image data into several regions,then LBP processes each region,counts the histograms in each region,and cascades the histograms in each domain.In this way,not only image features are obtained,but also feature of space features are obtained.In order to solve the problem of losing feature when LBP technology acquires image features,Gabor wavelet transform is introduced in this paper.Gabor wavelet transform can acquire multi-scale and multi-direction data featuresaccording to user's needs,which makes the acquired features more comprehensive and accurate,and improves the robustness of the algorithm to attitude.It can effectively improve the accuracy of the algorithm,and reduce the feature dimension of the data processed by Gabor algorithm.The algorithm with the function of filtering image features can reduce the computational complexity without reducing the recognition accuracy.FPGA technology uses hardware description language to realize software hardening technology.By making full use of the parallelism characteristics of this technology,it can improve the execution speed and reliability of the algorithm,reduce the computational complexity of the algorithm,and carry out simulation on the algorithm.The simulation results show that the proposed algorithm can effectively improve the recognition accuracy of the algorithm,improve the robustness of the algorithm to the irregular changes of illumination and attitude angle,reduce the running time of the algorithm and prove the reliability of the algorithm.