一种Petri网优化的验证码识别方法

doi:10.19678/j.issn.1000-3428.0055178

计算机工程 ›› 2020, Vol. 46 ›› Issue (7): 277-285. doi: 10.19678/j.issn.1000-3428.0055178

一种Petri网优化的验证码识别方法

马金林, 陈德光, 马自萍, 魏麟

北方民族大学计算机科学与工程学院, 银川 750021

收稿日期:2019-06-11 修回日期:2019-07-22 发布日期:2019-07-25
作者简介:马金林(1976-),男,副教授、博士,主研方向为机器学习、智能信息处理;陈德光,硕士研究生;马自萍,副教授、博士;魏麟,硕士研究生。
基金资助:
国家自然科学基金（61462002，61762003，61862001）；北方民族大学研究项目（2018XYZJK02）；北方民族大学教育教学重大研究项目（2018ZHJY01）；宁夏自然科学基金项目（2020AAC03215）；"图像与智能信息处理"民委创新团队项目（PY1805）。

A CAPTCHA Identification Method Optimized by Petri Net

MA Jinlin, CHEN Deguang, MA Ziping, WEI Lin

School of Computer Science and Engineering, North Minzu University, Yinchuan 750021, China

Received:2019-06-11 Revised:2019-07-22 Published:2019-07-25

摘要/Abstract

摘要： 针对AlexNet网络对验证码（CAPTCHA）多目标分类问题效果不理想、模型参数量与浮点数计算量过大的问题，提出一种基于Petri网优化的CAPTCHA识别方法。利用Petri网理论对AlexNet和DenseNet-BC建模，并通过所建模型优化网络结构和参数。同时，根据模型参数量与浮点数计算量的关系，提出超活性概念，对Petri-ANPP-net、Petri-ANPS-net、Petri-DNBC-net模型进行灵敏度分析。实验结果表明，经过Petri网优化后，Petri-ANPP-net模型的最高准确度为60.40%，且超活性较小，模型灵敏度较差，Petri-ANPS-net模型的最高准确度为97.50%，但超活性较小，模型灵敏度较差，Petri-DNBC-net模型的最高准确度达到99.24%，且超活性较大，模型灵敏度较高。说明Petri网能在一定程度上优化网络模型结构和参数，且超活性对于评价模型的灵敏度具有一定的优越性。

关键词: Petri网, 神经网络, 图像识别, 超活性, 模型优化

Abstract: AlexNet does not perform well in the multi-target classification of verification codes due to the large number of parameters and heavy floating point computation.To address the problem,this paper proposes a CAPTCHA recognition method optimized by Petri net.The method uses the Petri net theory to model AlexNet and DenseNet-BC,and optimizes the network structure and parameters with the built models.At the same time,according to the relationship between the number of model parameters and the amount of floating point computation,the concept of hyperactivity is proposed.Then sensitivity analysis is carried out on Petri-ANPP-net,Petri-ANPS-net,and Petri-DNBC-net models.Experimental results show that after the Petri net-based optimization,the highest accuracy of the Petri-ANPP-net model is 60.40%,and its super activity as well as the model sensitivity is poor.The highest accuracy of the Petri-ANPS-net model is 97.50%,but its superactivity and the model sensitivity is poor.The highest accuracy of the Petri-DNBC-net model is 99.24%,and its superactivity as well as the model sensitivity is high.The results show that Petri net can optimize the network model structure and parameters to a certain extent,and the hyperactivity has certain advantages in evaluating the sensitivity of the model.

Key words: Petri net, neural network, image recognition, superactivity, model optimization

中图分类号:

TP316

马金林, 陈德光, 马自萍, 魏麟. 一种Petri网优化的验证码识别方法[J]. 计算机工程, 2020, 46(7): 277-285.

MA Jinlin, CHEN Deguang, MA Ziping, WEI Lin. A CAPTCHA Identification Method Optimized by Petri Net[J]. Computer Engineering, 2020, 46(7): 277-285.

http://www.ecice06.com/CN/Y2020/V46/I7/277

参考文献

[1] VON A L,BLUM M,HOPPER N J,et al.CAPTCHA:using hard AI problems for security[C]//Proceedings of International Conference on Theory and Applications of Cryptographic Techniques.Berlin,Germany:Springer,2003:294-311.
[2] ZHANG Zheng,WANG Shunfan,DONG Lei.CAPTCHA recognition based on deep learning[J].Journal of Hubei University of Technology,2018,33(2):5-8,25.(in Chinese)张铮,王顺帆,董雷.基于深度学习的验证码识别[J].湖北工业大学学报,2018,33(2):5-8,25.
[3] CHEN Jun,LUO Xiangyang,LIU Yingying,et al.Selective learning confusion class for text-based CAPTCHA recognition[J].IEEE Access,2019,7:22246-22259.
[4] MA Wentao,QIN Jiaohu,XIANG Xuyu,et al.Adaptive median filtering algorithm based on divide and conquer and its application in CAPTCHA recognition[J].Computers,Materials and Continua,2019,58(3):665-677.
[5] ARAIN R H,SHAIKH R A,MAITLO A,et al.A deep learning model for recognition of complex text-based CAPTCHA[J].International Journal of Computer Science and Network Security,2018,18(2):103-107.
[6] QIN Bo,GU Naijie,ZHANG Xiaoci,et al.Image CAPTCHA recognition based on convolutional neural network[J].Computer Systems & Applications,2018,27(11):144-150.(in Chinese)秦波,顾乃杰,张孝慈,等.基于卷积神经网络的图像验证码识别[J].计算机系统应用,2018,27(11):144-150.
[7] ROBERTSON S,PENN G,WANG Y.Exploring spectro-temporal features in end-to-end convolutional neural networks[EB/OL].[2019-05-10].https://arxiv.org/abs/1901.00072?context=cs.LG.
[8] LI Wenxiang.Modeling and verification of a temperature control system based on temporal Petri nets[J].Journal of Shandong University of Technology(Natural Science Edition),2018,32(6):24-28.(in Chinese)李文翔.温控系统的时序Petri网建模与验证[J].山东理工大学学报(自然科学版),2018,32(6):24-28.
[9] VAN D A W,WEIJTERS T,MARUSTER L.Workflow mining:discovering process models from event logs[J].IEEE Transactions on Knowledge and Data Engineering,2004,16(9):1128-1142.
[10] SMIRNOV S,WEIDLICH M,MENDLING J.Business process model abstraction based on behavioral profiles[C]//Proceedings of International Conference on Service-Oriented Computing.Berlin,Germany:Springer,2010:1-16.
[11] SOBOCINSKI P,STEPHENS O.Penrose:putting compositionality to work for petri net reachability[C]//Proceedings of International Conference on Algebra and Coalgebra in Computer Science.Berlin,Germany:Springer,2013:346-352.
[12] ZHAO Yang,LI Tong,LIU Qing.A performance analysis method for software process modeling based on the extended Petri net[J].Computer Engineering and Applications,2004,40(26):70-72.(in Chinese)赵杨,李彤,柳青.一种基于扩展Petri网的软件过程模型性能分析方法[J].计算机工程与应用,2004,40(26):70-72.
[13] TAX N,SIDOROVA N.Discovering more precise process models from event logs by filtering out chaotic activities[J].Journal of Intelligent Information Systems,2019,52(1):107-139.
[14] XU Yuyin.Discussion on Petri net activity determination method[D].Jinan:Shandong University of Science and Technology,2006.(in Chinese)徐誉尹.Petri网活性判定方法的探讨[D].济南:山东科技大学,2006.
[15] ZHANG Jianping.Application of Petri net dynamic workflow in OA system of mining company[J].China's Manganese Industry,2019,37(2):107-110.(in Chinese)张建平.Petri网动态工作流在矿业公司OA系统中的应用[J].中国锰业,2019,37(2):107-110.
[16] REISIG W.Petri nets:an introduction[M].Berlin,Germany:Springer,2012.
[17] MARSAN M A,BALBO G,CONTE G,et al.Modelling with generalized stochastic Petri nets[M].[S.1.]:John Wiley & Sons,Inc.,1994.
[18] GUERAR M,MERLO A,MIGLIARDI M.Completely automated public physical test to tell computers and humans apart:a usability study on mobile devices[J].Future Generation Computer Systems,2018,82:617-630.
[19] GOTO M,SHIRATO T,UDA R.Text-based CAPTCHA using phonemic restoration effect and similar sounds[C]//Proceedings of the 38th IEEE International Computer Software and Applications Conference Workshops.Washington D.C.,USA:IEEE Press,2014:270-275.
[20] HUANG G,LIU Z,VAN D M L,et al.Densely connected convolutional networks[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2017:4700-4708.
[21] PLEISS G,CHEN D,HUANG G,et al.Memory-efficient implementation of densenets[EB/OL].[2019-05-10].https://www.researchgate.net/publication.
[22] HE Kaiming,SUN Jian.Convolutional neural networks at constrained time cost[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2015:5353-5360.

选择文件类型/文献管理软件名称

选择包含的内容

一种Petri网优化的验证码识别方法

A CAPTCHA Identification Method Optimized by Petri Net

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	靳雁霞, 史志儒, 杨晶, 刘亚变, 乔星宇, 张翎. 布料与精细建模物体间的碰撞检测算法研究[J]. 计算机工程, 2023, 49(7): 269-277.
[2]	曹坪, 杨怀志, 薄一军, 尤嘉, 张淳杰, 李丹勇. 面向低质量裂缝图像的多知识蒸馏分类[J]. 计算机工程, 2023, 49(7): 204-213.
[3]	白明昌. 基于折叠路径聚合的属性网络节点嵌入方法[J]. 计算机工程, 2023, 49(7): 76-84.
[4]	赵世豪, 毛国君, 熊保平, 黄山, 林江宏. 基于图小波卷积神经网络的时空图挖掘模型[J]. 计算机工程, 2023, 49(7): 85-93.
[5]	廖涛, 孙皓洁, 张顺香. 基于跨度和特征融合的实体关系联合抽取模型[J]. 计算机工程, 2023, 49(6): 107-114.
[6]	于海洋, 景鹏, 张文涛, 谢赛飞, 滑志华, 宋草原. 基于残差与注意力机制的道路裂缝检测U-Net改进模型[J]. 计算机工程, 2023, 49(6): 265-273.
[7]	代祖华, 刘园园, 狄世龙. 语义增强的图神经网络方面级文本情感分析[J]. 计算机工程, 2023, 49(6): 71-80.
[8]	沈学利, 田桂源, 姜彦吉, 马琳琳. 基于双阶段Conv-Transformer的时频域语音增强算法[J]. 计算机工程, 2023, 49(6): 123-130.
[9]	丁子轩, 俞雷, 张娟, 李想, 王新宇. 基于深度残差自适应注意力网络的图像超分辨率重建[J]. 计算机工程, 2023, 49(5): 231-238.
[10]	区展华, 李翠然, 杨茜. 基于ANN的能量采集无线传感器网络中继选择策略[J]. 计算机工程, 2023, 49(5): 215-222,230.
[11]	陈治旭, 靳雁霞, 芦烨, 杨晶, 刘亚变, 史志儒. 基于子图卷积神经网络的多精度服装建模方法[J]. 计算机工程, 2023, 49(4): 174-181.
[12]	安志国, 彭政, 易满成, 刘健欣, 俞思帆. 神经网络滤波器竞争训练[J]. 计算机工程, 2023, 49(4): 120-124.
[13]	徐康, 李霏, 姬东鸿. 结合依存图卷积与文本片段搜索的方面情感三元组抽取[J]. 计算机工程, 2023, 49(4): 61-67.
[14]	陈锐, 孙羽菲, 郭强, 隋轶丞, 周振辉, 石昌青, 张玉志. OclDNN:一种可应用于TensorFlow的通用DNN库[J]. 计算机工程, 2023, 49(4): 138-148.
[15]	衡红军, 苗菁. 语义与句法信息加强的二元标记实体关系联合抽取[J]. 计算机工程, 2023, 49(4): 77-84.

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

一种Petri网优化的验证码识别方法

A CAPTCHA Identification Method Optimized by Petri Net

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价