[1]程换新,刘文翰,郭占广,等.基于胶囊网络在复杂场景下的行人识别[J].计算机技术与发展,2021,31(02):75-79.[doi:10. 3969 / j. issn. 1673-629X. 2021. 02. 014]
　CHENG Huan-xin,LIU Wen-han,GUO Zhan-guang,et al.Pedestrian Recognition in Complex Scenes Based on Capsule Network[J].,2021,31(02):75-79.[doi:10. 3969 / j. issn. 1673-629X. 2021. 02. 014]
点击复制

基于胶囊网络在复杂场景下的行人识别()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
31
期数:
2021年02期
页码:
75-79
栏目:
图形与图像
出版日期:
2021-02-10

文章信息/Info

Title:
Pedestrian Recognition in Complex Scenes Based on Capsule Network
文章编号:
1673-629X(2021)02-0075-05
作者:
程换新; 刘文翰; 郭占广; 张志浩
青岛科技大学 自动化与电子工程学院,山东 青岛 266061
Author(s):
CHENG Huan-xin; LIU Wen-han; GUO Zhan-guang; ZHANG Zhi-hao
School of Automation and Electronic Engineering,Qingdao University of Science and Technology,Qingdao 266061,China
关键词:
大数据; 深度学习; 胶囊网络; 行人识别; TensorFlow
Keywords:
big data; deep learning; capsule network; pedestrian recognition; TensorFlow
分类号:
TP391. 9
DOI:
10. 3969 / j. issn. 1673-629X. 2021. 02. 014
摘要:
大数据环境下,对行人检测的需求度不断提高,然而视频中的信息越来越丰富,视频中所获取的场景也愈加复杂。在如此背景下,目前大多使用卷积神经网络进行识别,但识别率不高。 在原有的胶囊网络模型的基础上,增加了两层卷积层并将胶囊维度进行了扩展,同时使用了动态路由迭代算法,提出了一种基于改进胶囊网络的行人识别模型(PRMICN),该网络能够更有效地减少复杂背景中多余信息的干扰。 实验在 TensorFlow 框架下使用三个国际知名且有一定难度的公开通用数据集 CUHK01、CUHK03 和 Market-1501 上进行验证,并将结果与 PRM-AlexNet 和 PRM-VGG-16 两个著名的行人识别网络相对比。 实验结果表明在三个数据集上,所提出的网络模型在 CMC 曲线和 MAP 指标下都要优于其他两个网络,证明了所提模型在复杂场景下识别效果的优越性。
Abstract:
In the context of big data,the demand for pedestrian detection is constantly increasing. However,the information in the video is get-ting more and more abundant,and the scenes acquired in the video are also becoming more and more complicated. Under such background,convolutional neural network is mostly used for recognition at present,but the recognition rate is not high. Based on the original capsule network model, two con-volutional layers are added and the capsule dimension is extended. At the same time, according to dynamic routing iteration algorithm,a pede-strian recognition model based on the improved capsule network (PRM-ICN) is proposed,which can more effectively reduce the interference of redundant information in the complex background. The experiments are verified under the TensorFlow framework using three publicly available datasets CUHK01,CUHK03 and Market-1501, and the results are compared with two famous pedestrian recognition networks,PRM-AlexNet and PRM-VGG-16. Experiment shows that on the three data sets,the proposed network model is greater than another two networks under the CMC curve and the MAP index,which proves its superiority in complex scene recognition.

相似文献/References:

[1]严霄凤,张德馨.大数据研究[J].计算机技术与发展,2013,(04):168.
　YAN Xiao-feng,ZHANG De-xin.Big Data Research[J].,2013,(02):168.
[2]王雷,陈彦先,袁哲,等. 面向预拌混凝土行业的云计算[J].计算机技术与发展,2014,24(08):14.
　WANG Lei,CHEN Yan-xian,YUAN Zhe JI Xu. Research on Cloud Computing for Ready-mixed Concrete Industry[J].,2014,24(02):14.
[3]金宗泽,冯亚丽,文必龙,等. 大数据分析流程框架的研究[J].计算机技术与发展,2014,24(08):117.
　JIN Zong-ze,FENG Ya-l,WEN Bi-long,et al. Research on Framework of Big Data Analytic Process[J].,2014,24(02):117.
[4]张也弛,周文钦,石润华. 一种面向云的大数据完整性检测协议[J].计算机技术与发展,2014,24(09):68.
　ZHANG Ye-chi,ZHOU Wen-qin,SHI Run-hua. A Big Data Integrity Checking Protocol for Cloud[J].,2014,24(02):68.
[5]谢怡,王航,刘新瀚,等. 大数据环境下数据读取关键技术研究[J].计算机技术与发展,2015,25(02):113.
　XIE Yi,WANG Hang,LIU Xin-han,et al. Research on Data Reading Techniques Based on Big Data Environment[J].,2015,25(02):113.
[6]付燕平,罗明宇,刘其军. 大数据三维模型快速显示技术研究[J].计算机技术与发展,2015,25(05):87.
　FU Yan-ping,LUO Ming-yu,LIU Qi-jun. Research on Fast Display Technology for Big Data Three-dimensional Model[J].,2015,25(02):87.
[7]赵震,任永昌. 大数据时代基于云计算的电子政务平台研究[J].计算机技术与发展,2015,25(10):145.
　ZHAO Zhen,REN Yong-chang. Research on E-government Platform Based on Cloud Computing in Big Data Era[J].,2015,25(02):145.
[8]胡存刚,程莹. 基于粒子群算法的大数据智能搜索引擎的研究[J].计算机技术与发展,2015,25(12):14.
　HU Cun-gang,CHENG Ying. Research on Big Data Intelligent Search Engine Based on PSO[J].,2015,25(02):14.
[9]陈强锐,谢世朋.基于深度学习的肺部肿瘤检测方法[J].计算机技术与发展,2018,28(04):201.[doi:10.3969/ j. issn.1673-629X.2018.04.043]
　CHEN Qiang-rui,XIE Shi-peng.Lung Cancer Detection Method Based on Deep Learning[J].,2018,28(02):201.[doi:10.3969/ j. issn.1673-629X.2018.04.043]
[10]孔钦,叶长青,孙赟.大数据下数据预处理方法研究[J].计算机技术与发展,2018,28(05):1.[doi:10．3969/j．issn．1673－629X．2018．05．001]
　KONG Qin,YE Changqing,SUN Yun.Research on Data Preprocessing Methods for Big Data[J].,2018,28(02):1.[doi:10．3969/j．issn．1673－629X．2018．05．001]

更新日期/Last Update: 2020-02-10