Low-resolution contactless online palmprint matching method

Abstract

The invention discloses a low-resolution non-contact online palmprint matching method, which mainly solves the problems of different sensitivities of anisotropic Gabor filters to different angles and low matching speed when multiple angle values ​​are taken in the existing methods. The implementation steps are: 1. Read in the palmprint image to be matched; 2. Extract the region of interest of the palmprint image to be matched; 3. Use the isotropic Gabor filter to convolve the region of interest and perform convolution on the convolution result. Binary encoding; 4. Comparing the binary encoding images corresponding to the regions of interest to obtain the matching similarity score of the palmprint image to be matched. The invention can improve the accuracy of low-resolution palmprint matching under the condition of maintaining high robustness and low computational complexity, and can be used in a low-resolution non-contact palmprint recognition system.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and in particular relates to an online palmprint matching method, which can be used in a low-resolution contactless palmprint recognition system. Background technique [0002] With the progress of society and the urgent needs of the current security situation, palmprint recognition, especially low-resolution contactless online palmprint recognition technology, has become another biometric research hotspot after fingerprints, iris, etc. Palmprint recognition has effective The advantages of large area, rich detailed information, convenient collection, good user experience, and high accuracy that other biometric identification methods do not have, so they have received more and more attention. [0003] The code-based palmprint matching method is a relatively successful method in the low-resolution contactless online palmprint matching method, which has strong robustness, high matchi...

AI Technical Summary

Problems solved by technology

[0007] 1. Since the PalmCode method uses a special direction to extract palmprint features, it loses palmprint features in other directions, thus limiting the improvement of its accuracy

[0008] 2. Although the FusionCode method obtains more comprehensive palmprint feature information by increasing the number of Gabor filter templates, the convolution operation is a relatively time-consuming operation, so it is sacrificed in order to extract more comprehensive palmprint feature information. Increased matching speed and increased computational complexity

[0009] 3. These two matching methods use real part and real part matching, imaginary part and imaginary part matching in the matching stage, and finally directly calculate the matching similarity score strategy, so the real part and imaginary part of the same region of interest are ignored The connection between partial coding results causes the loss of palmprint feature information and affects the improvement of matching accuracy.

Images