摘要
研究掌纹准确识别问题,由于光照强度、位置移动、采集设备等影响,采集掌纹图像的分辨率较低。单一掌纹特征提取方法难以全面描述掌纹信息,导致掌纹识别率低。为了提高了掌纹识别率,提出一种基于Gabor滤波和LBP算法相融合的掌纹识别方法。首先对采集掌纹进行预处理,然后分别采用Gabor滤波和LBP算法进行特征提取,最后采用神经网络建立掌纹识别器。仿真结果表明,相对于单一特征提取算法,融合特征算法不仅提高了掌纹识别率,同时加快掌纹识别速度,能够很好满足实时掌纹识别系统的要求。
Research palmprint recognition problem. In order to improve the recognition rate, the paper proposed a palmprint recognition method based on the Gabor filter and LBP algorithm fusion. The palmprints were preprocessed firstly, and then the Gabor filter and LBP algorithm were usied for feature extraction. Finally, the neural network palmprint reeognizor was established. The simulation results show that, compared with the single feature extraction algorithm, the fusion feature algorithm can improve the recognition rate, accelerate palmprint recognition speed at the same time, and meet the real time requirements of palmprint identification system.
出处
《计算机仿真》
CSCD
北大核心
2012年第6期265-268,共4页
Computer Simulation
关键词
掌纹识别
特征提取
特征匹配
Palmprint recognition
Feature extraction
Feature matching
