针对现有的掌纹识别算法对掌纹图像的旋转、尺度和亮度变化缺乏足够的鲁棒性,而且识别速度较慢的问题,本文通过LDP算子进行特征提取,将掌纹分成若干子区域后,然后通过连接这些子区域的LDP直方图生成掌纹特征向量,使得发生变化的同一类掌纹图像的相似性变大。为了能够提高识别精度且加快识别速度,通过概率神经网络(PNN)来进行分类。实验表明该算法对掌纹图像的旋转、尺度和亮度的变化有良好的鲁棒性,且提高识别率,识别速度较快。 To alleviate the limitations that the existing palmprint recognition methods are time-consuming, and their robustness to the variations of orientation, position and illumination is insufficient, this paper uses LDP operator to get feature extraction. The paimprint image is divided into sub-regions. Then connecting these sub-regions LDP histogram to generate palmprint feature vector, this can increase the similarity of the same type of palmprint image. In order to improve the recognition accuracy and accelerate the recognition speed, the classification is performed by Probabilistic Neural Networks (PNN). It is also shown that the proposed approach is robust to the variations of orientation, position and illumination and improves the recognition rate, and accelerates the recognition speed.
假设训练样本数量为n,记为 X = ( x 1 , x 2 , ⋯ , x n ) ,每一个输入样本 x i 的维数都是m维,将输入样本不需要任何改变的直接输入到输入层中,输入层中的输入节点数和神经元个数等于输入样本 x i 的维数m。模式层计算输入特征向量与训练集中各个样本的匹配程度,也就是相似度,一般该非线性算子取高斯函数:
M i j ( X ) = exp ( − ∑ i = 1 n ( ( X i − w i j ) 2 / ( X i + w i j ) ) σ ) (2)
图3. 特征直方图的联合过程
图4. 概率神经网络结构图
其中, x i 是指训练集中的第i个训练样本, w i j 是输入层的第i个神经元和模式层的第j个神经元的权重,σ是高斯核函数的平滑参数,取决于用户选择。求和层负责将各个样本类的模式层单元连接起来,即对向量M进行加权求和,这一层的神经元个数是样本的类别数目。
S i ( X ) = ∑ i = 1 N 1 w i j M i j ( X ) , i = 1 , 2 , ⋯ , n (3)
其中 ∑ i = 1 N 1 w i j = 1 , i = 1 , 2 , ⋯ , n 且 w i j ∈ [ 0 , 1 ]
实验3:本实验的目的是测试所提出方法的计算成本。本次仿真实验环境为Intel®CeleronCPU 1.99G Hz,1.99GB内存,Microsoft Windows XP操作系统。选用第一次采集样本进行试验,计算实验1中所提及方法的平均运行时间,实验结果见表3。实验证明,本文算法的运行时间较好,这是由于LDP进行特
The classification performance in different approaches (1
方法
正确识别率(%)
本文方法
95.41
Kong的方法
92.88
基于PCA的方法
94.25
基于LDA的方法
92.05
基于双树复小波变化的方法
92.35
表1. 不同算法的正确识别率(一)
The classification performance in different approaches (2
方法
正确识别率(%)
本文方法
91.21
Kong的方法
88.67
基于PCA的方法
83.25
基于LDA的方法
84.05
基于双树复小波变化的方法
87.25
表2. 不同算法的正确识别率(二)
The average running time(s) in different algorithm
周萍萍,王 晅. 基于LDP和PNN的掌纹识别算法 Research of Palmprint Identification Algorithm Based on LDP and PNN[J]. 计算机科学与应用, 2018, 08(04): 464-471. https://doi.org/10.12677/CSA.2018.84051
参考文献ReferencesJain, A.K., Flynn, P. and Ross, A. (2007) Handbook of Biometrics. Springer, New York, 1-22.Yue, F., Zuo, W.-M. and Zhang, D.P. (2010) Survey of Palmprint Ecognition Algorithm. Acta Automatica Sinica, 36, 353- 365. <br>https://doi.org/10.3724/SP.J.1004.2010.00353Kong, A., Zhang, D. and Kamel, M. (2009) A Survey of Palmprint Recogni-tion. Pattern Recognition, 42, 1408-1418.
<br>https://doi.org/10.1016/j.patcog.2009.01.018Huang, D.S., Jia, W. and Zhang, D. (2008) Palmprint Verification Based on Principal Lines. Pattern Recognition, 41, 1316-1328. <br>https://doi.org/10.1016/j.patcog.2007.08.016Leung, M.K.S., Fong, A.C.M. and Hui, S.C. (2007) Palmprint Verification for Controlling Access to Shared Computing Resources. IEEE Pervasive Compu-ting, 6, 40-47. <br>https://doi.org/10.1109/MPRV.2007.78Rafael Diaz, M., Travieso, C.M., Alonso, J.B. and Ferrer, M.A. (2004) Biometric System Based in the Feature of Hand Palm. Proceedings of 38th Annual International Carnahan Conference on Se-curity Technology, Albuquerque, 136-139.Wu, X.Q., Wang, K. and Zhang, D. (2002) Line Feature Extraction and Matching in Palmprint. Proceedings of the Second International Conference on Image and Graphics, Hefei, 31 July 2002, 583-590.
<br>https://doi.org/10.1117/12.477200Wu, X.Q., Zhang, D. and Wang, K. (2003) Fisherpalms Based Palmprint Recognition. Pattern Recognition Letters, 24, 2829-2838. <br>https://doi.org/10.1016/S0167-8655(03)00141-7Lu, G., Zhang, D. and Wang, K. (2003) Palmprint Recognition Using Eigenpalms Features. Pattern Recognition Letters, 24, 1463-1467.Connie, T., Jin, A.T.B., Ong, M.G.K. and Ling, D.N.C. (2005) An Automated Palmprint Recognition System. Image and Vision Computing, 23, 501-515. <br>https://doi.org/10.1016/j.imavis.2005.01.002Du, F., Yu, P., Li, H., et al. (2011) Palmprint Recognition Using Gabor Fea-ture-Based Bidirectional 2DLDA. Communications in Computer & Information Science, 159.Li, W., Zhang, D. and Xu, Z. (2002) Palmprint Identification by Fourier Transform. International Journal of Pattern Recognition and Artificial Intelligence, 16, 417-432. <br>https://doi.org/10.1142/S0218001402001757Arivazhagan, S., Ganesan, L. and Priyal, S.P. (2006) Texture Classification Using Gabor Wavelets Based Rotation Invariant Features. Pattern Recognition Letters, 27, 1976-1982. <br>https://doi.org/10.1016/j.patrec.2006.05.008Li, W., You, J. and Zhang, D. (2005) Texture-Based Palmprint Retrieval Using a Layered Search Scheme for Personal Identification. IEEE Transactions on Multimedia, 7, 891-898. <br>https://doi.org/10.1109/TMM.2005.854380Wu, X.Q., Wang, K. and Zhang, D. (2006) Palmprint Texture Analysis Using Derivative of Gaussian Filters. Proceedings of 2006 International Conference on Computational Intelligence and Security, Guangzhou, 3-6 November 2006, 751-754. <br>https://doi.org/10.1109/ICCIAS.2006.294235刘伟锋, 李树娟, 王延江. 人脸表情的LBP特征分析[J]. 计算机工程与应用, 2011, 47(2): 149-152.Jabid, T., Kabir, M.H. and Oksam, C. (2010) Local Directional Pattern (LDP) for Face Recognition. 2010 Digest of Technical Papers International Conference on Consumer Electronics (ICCE), Las Vegas, NV, 9-13 January 2010, 329-330. <br>https://doi.org/10.1109/ICCE.2010.5418801Jabid, T. and Kabir, M.H. (2010) Local Di-rectional Pattern (LDP)—A Robust Image Descriptor for Object Recognition. 7th IEEE International Conference on Advanced Video and Signal Based Surveillance, Boston, MA, 29 August-1 September 2010, 482-487.Specht, D.F. (1990) Probabilistic Neural Network. Neural Network, 3, 109-118.
<br>https://doi.org/10.1016/0893-6080(90)90049-QZhang, D., Kong, W.K., You, J. and Wong, M. (2003) Online Palmprint Identification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 25, 1041-1049. <br>https://doi.org/10.1109/TPAMI.2003.1227981Connie, T., Teoh, A., Goh, M. and Ngo, D. (2004) Palmprint Recognition with PCA and ICA. Proceedings of Image and Vision Computing, New Zealand, 227-232.Lu, G., Zhang, D. and Wang, K. (2003) Palmprint Recognition Using Eigenpalms Features. Pattern Recognition Letters, 24, 1463-1467. <br>https://doi.org/10.1016/S0167-8655(02)00386-0