摘要
为避免在处理掌纹识别时人工提取掌纹特征,提出使用卷积神经网络(CNN)来处理掌纹识别问题。首先根据掌纹的几何形状特点进行预处理,切割出掌纹的感兴趣区域(ROI);然后将感兴趣区域进行归一化并组成一个二维矩阵作为卷积神经网络的输入;再使用批量随机梯度下降算法对网络进行训练,得到最优的网络参数;最后对测试掌纹进行分类识别,分类器使用Softmax。应用于香港理工大学掌纹数据库(v2)的掌纹识别率达到99.15%,单张掌纹的识别时间小于0.01 s,验证了方法的有效性。
To avoid extracting palmprint features when solved the palmprint recognition problem,Convolution Neural Network( CNN) to deal with it was attempted to use. First of all,according to the geometric features of palmprint to preprocess palmprint image,so that extracted region of interest( ROI). And then normalized ROI to form a two-dimensional matrix as the input of CNN. Secondly,mini_batch stochastic gradient descent algorithm was used to train the network to get the optimal network parameters. Finally,the softmax classifier is used to classify the palmprint. The results of experiments show that proposed network achieves 99. 15% recognition accuracy on Poly U Palmprint Database( 2 nd Version),and single palmprint image recognition time is in less than 0. 01 s. The results demonstrate that proposed algorithm can be used to improve the recognition accuracy.
出处
《科学技术与工程》
北大核心
2017年第35期272-276,共5页
Science Technology and Engineering
基金
国家自然科学基金重点项目(61136002)资助
关键词
卷积神经网络
掌纹识别
深度学习
convolutional neural network
palmprint recognition
deep learning
