Research Of Image Feature Extraction Algorithm In The Application Of Palmprint Recognition

As a kind of biometrics reorganization,palmprint recognition is a method of identifying the texture by extracting the texture features of the palm of your hand.Compared with the commonly used fingerprints and other methods of identification,palmprint image acquisition is more likely to be accepted,and the palm print area is relatively contains more identification information.Therefore,palmprint identification with other identification will have a broader application prospects.In addition,palmprint collection equipment requirements are low,do not need high-resolution imaging equipment.As a result,palmprint recognition is becoming another hot research direction.There are many kinds of palmprint recognition methods,such as recognition method based on palmprint texture,mainline recognition method,direction-based recognition method,correlation-based recognition method and subspace-based recognition method.Similar to other image recognition,palmprint recognition process includes: palmprint image acquisition,image preprocessing,feature extraction,classification and decision-making and other major parts.Among them,feature extraction is the key of palmprint recognition and occupies a very important position in the whole recognition process.The feature extraction of the palmprint image directly determines the merits of the palmprint recognition result,and how to extract the feature of the palmprint image is the key to the palmprint recognition.With the Fourier transform,wavelet transform and the use of filters for palmprint image transformation and other methods,palmprint feature extraction method has been rapid development.Some of these extraction algorithms can even extract features from the original palmprint images without preprocessing,reduce the computational complexity and improve the recognition speed.These feature extraction algorithms not only improve the response speed of palmprint recognition,but also increase the robustness of recognition and the accuracy of matching.Commonly used image features,including the image point,line,surface and other features,including statistical features of images,image color features and texture features.Classical palmprint recognition algorithm including feature extraction algorithms based on line and area,and frequency domain feature extraction algorithms based on Fourier transform and Gabor transform.The method based on statistic features mainly uses the common statistical features of gray image,such as variance,standard deviation,etc.To define the original features of the image.Firstly,the method of image feature extraction is based on the feature of point,And wavelet transform.The subspace-based feature extraction,also known as the image representation-based recognition method,obtains the local image by the Fourier transform,the Gabor transform and the wavelet transform.The main feature of the spatialFeature,using this local feature of the image to construct the feature vector to carry on the recognition of the image,and realize the transformation from N dimension to M dimension.Common methods include principal component analysis and Fisher linear discriminant.SIFT(Scale Invariant Feature Transform)algorithm has good performance in image feature extraction field,but it has little research on palmprint image recognition.So in this paper we proposes to use SIFT algorithm for palmprint feature extraction.The key of SIFT algorithm is to find the feature points and generate the feature descriptors of the feature points.In this paper,the SIFT algorithm has relatively mature feature generation method.In this paper,feature vector generation is carried out on the palmprint image,and then the Euclidean distance is used to perform the feature matching calculation,and a better image matching result is achieved.In this paper,we introduce the key techniques of image feature extraction and the matching algorithm in detail,and a typical image feature extraction algorithm is explored in detail.Various algorithms are found in the feature extraction of palmprint image and the palmprint recognition method based on SIFT feature extraction is proposed based on the understanding of these algorithms.This paper introduces the technology of fingerprint image feature extraction to provide the technical reference for the developers in this field.The key technology used in this field will lay the foundation for the research and development of related applications.