【摘要】 关于掌纹特征提取要求提高识别率,局部二值模式(LBP)掌纹识别,提取的特征维数高,特征之间存在一定冗余,导致掌纹识别率较低。为了提高掌纹识别率,提出一种主成份分析(PCA)的LBP的掌纹特征提取方法(PCA-LBP)。首先对掌纹图像进行灰度预测,采用LBP算法计算灰度直方图,得到256个灰度对应的像素数据,将其作为掌纹图像的原始特征,用主成分分析方法消除各特征之间的高度冗余性,并有效地降低了特征集的维数,得到了最有利于识别的最佳特征。根据最小欧式距离判别法对掌纹图像的进行识别,对PolyU标准库中的掌纹进行仿真实验,结果表明,相比传统的LBP算法和离散小波变换提取算法,可以提较少的特征维数取得了更高的的识别率,说明改进算法既不会丢失掌纹图像的原有信息,提高了识别率。更多还原

【Abstract】 n recent years,local binary pattern(LBP) has been successfully applied to face recognition.However,the dimensionality of the feature vector extracted by LBP is usually very high,then prediction accuracy is very low.In order to improve the accuracy of palmprint recognition,a new method was proposed based on principal component analysis(PCA) and local binary pattern..Firstly,the pixel values of the palmprint diagram were predicted and used as the original features palmprint image,and then he principal component analysis method was used to eliminate the high redundancy in the characteristics extracted,which effectively reduced the dimensions of the characteristics and obtained the features for identification.Finally,minimum Euclid distance was used to implement the palmprint image classification.Simulation experiment results based on the standard library PolyU show that this method not only has advantages in terms of time,but complete identification with less palmprint images and the recognition rate is significantly enhanced.更多还原