基于LBP和深度学习的手写签名识别算法

J4 ›› 2017, Vol. 34 ›› Issue (1): 23-31.

基于LBP和深度学习的手写签名识别算法

马小晴，桑庆兵

江南大学物联网工程学院，江苏无锡 214122

收稿日期:2016-01-15 修回日期:2016-03-01 出版日期:2017-01-28 发布日期:2017-01-28
通讯作者: 桑庆兵(1973-)，安徽明光人，博士，副教授，主要研究方向为图像视频质量评价。 E-mail:439860266@qq.com
作者简介:Supported by National Natural Science Foundation of China(国家自然科学基金,61170120）, Prospective Research Project of Jiangsu Province(江苏省产学研前瞻性联合研究项目,BY2013015-41）, Fundamental Research Funds for the Central Universities(江南大学自主科研计划重点项目,JUSRP51410B)

Handwritten signature verification algorithm based on LBP and deep learning

MA Xiaoqing, SANG Qingbing

School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, China

Received:2016-01-15 Revised:2016-03-01 Published:2017-01-28 Online:2017-01-28
About author:马小晴(1989-)，女，河北辛集人，研究生，研究方向为手写签名识别。Email：xiaoqingxjzx@126.com；

摘要/Abstract

摘要： 为优化手写签名识别算法性能，提出了一种局部二值模式( LBP)和深度学习相结合的手写签名识别算法。针对签名图像进行预处理、维纳滤波去除噪声；将预处理签名图像分为3×4子块，LBP应用于分块后的每个子图像，并将每个子块的纹理直方图特征连接起来，形成全局直方图特征；将得到的特征向量作为深度信念网络(DBN)的输入，逐层训练网络，并在顶层形成分类面，对签名图片进行识别。在GPDS、MCYT、及原创数据库进行实验，识别率误差分别为5.85%、9.3%、1.17%，有效提高了手写签名的识别精度，符合实际应用的要求。

关键词: 图像处理；手写签名识别；深度学习；深度信念网络；局部二值模式

Abstract: In order to improve the performance of handwritten signature verification algorithm, a handwritten signature verification algorithm based on local binary pattern (LBP) feature and deep learning is presented. Aiming at signature image, preprocessing and Wiener filtering are used to get rid of noise. The preprocessed signature image was divided into 3×4 blocks and LBP is used to each sub-block. The texture histogram characteristics of each sub-block are connected to form a global histogram characteristics. The feature vectors obtained are used as inputs of deep belief network (DBN) , DBN is trained layer by layer, and the classification plane is formed at the top to recognize the signature image. Experiments are conducted on GPDS，MCYT and the original database, and the recognition error rates are 5.85%, 9.3% and 1.17%, respectivly. The handwritten signature recognition accuracy is effectively improved, which meets the requirements of practical application.

Key words: image processing; handwritten signature verification; deep learning; deep belief network ; local binary pattern

马小晴，桑庆兵. 基于LBP和深度学习的手写签名识别算法[J]. J4, 2017, 34(1): 23-31.

MA Xiaoqing, SANG Qingbing. Handwritten signature verification algorithm based on LBP and deep learning[J]. J4, 2017, 34(1): 23-31.

参考文献

[1]Lin Daozhi.Study on Sparse Representation Based Off-line Signature Verification[D], SUN YAT-SEN University(广州：中山大学)，2014（in Chinese）. [2]Zhang Jian.Research on Off-line Chinese Signature Automatic Verification Method Based on Bipartite Graph Matching[D], Huazhong University Of Science And Technology(武汉：华中科技大学)，2014(in Chinese). [3]M.A. Ferrer,JF. Vargas,A. Morales,and A. Ordonez. .Robustness of Offline Signature Verification Based on Gray Level Features Information[J].Forensics and Security, 2012, 7(3):996-997 [4]Kaur R, Choudhary P.Offline Signature Verification in Punjabi Based On SURF Features and Critical Point Matching Using HMM[J].International Journal of Computer Applications, 2015, 111(16):4-11 [5] Samaneh Abdoli.Offline Signature Verification using Geodesic Derivative Pattern[C]. The 22nd Iranian Conference on Electrical Engineering (ICEE 2014), May 20-22, 2014, Shahid Beheshti University [6]B.H. Shekar and RK. Bharathi.DCT-SVM-Based Technique for Off-line Signature Verification[J].Emerging Research in Electronics, Computer Science and Technology, 2012, 248(1):843-853 [7]Vargas J F, Ferrer M A, Travieso C M, et al.Off-line signature verification based on grey level information using texture features[J].Pattern Recognition, 2011, 44(2):375-385 [8]Jing Wen, MoHan Chen, and JiaXin Ren.Off-Line Signature Verification Based on Local Structural Pattern Distribution Features[J].Pattern Recognition, 2014, 484(2):499-507 [9]Aini Najw Azmi.Freeman Chain Code (FCC) Representation in Signature Fraud Detection Based On Nearest Neighbour and Artificial Neural Network (ANN) Classifiers[J].International Journal of Image Processing, 2014, 8(6):434-434 [10]Das, Sanghamitra, and Abhinab Roy.Signature Verification Using Rough Set Theory Based Feature Selection[J].Computational Intelligence in Data Mining, 2016, 2(1):153-161 [11]Hatkar P V, Salokhe B T, Malgave A A.Offline Handwritten Signature Verification using Neural Network[J].Methodology, 2015, 2(1):1-5 [12]Guo Lili, Ding Shifei.Research Progress on Deep Learning[J].计算机科学, 2015, 42(5):28-33 [13]Hinton G E, Osindero S, Teh Y W.A fast learning algorithm for deep belief nets[J].Neural computation, 2006, 18(7):1527-1554 [14]Hu G, Yang Y, Yi D, et al.When Face Recognition Meets with Deep Learning: an Evaluation of Convolutional Neural Networks for Face Recognition[J].arXiv preprint arXiv, 2015, 1(1):1-8 [15]Lv Qi，Dou Yong，Niu Xin，et al.Remote Sensing Image Classification Based on DBN Model[J].计算机研究与发展, 2014, 51(9):1911-1918 [16]Song Zhijian, Yu Rui.Research on the Classification of Handwriting Number Based on Deep Learning[J].重庆工商大学学报: 自然科学版, 2015, 32(8):49-53 [17]Ojala T, Pietik?inen M, Harwood D.A comparative study of texture measures with classification based on feature distribution[J].Pattern Recognition, 1996, 29(1):50-51 [18]Ojala T, Pietik?inen M, M?enp?? T.Multiresolution gray-scale and rotation invariant texture classification with local binary patterns[J].Pattern Analysis and Machine Intelligence, 2002, 24(7):971-987 [19]Vargas F, Ferrer M, Travieso C, et al.Off-line handwritten signature GPDS-960 corpus[C]//icdar. IEEE, 2007: 764-768. [20]Ortega-Garcia J, Fierrez-Aguilar J, Simon D, et al.MCYT baseline corpus: a bimodal biometric database[J].IEE Proceedings-Vision, Image and Signal Processing, 2003, 150(6):395-401