河北师范大学学报—

期刊信息

刊名：河北师范大学学报（自然科学版）Journal of Hebei Normal University (Natural Science)
主办：河北师范大学
ISSN： 1000-5854
CN： 13-1061/N
中国科技核心期刊
中国期刊方阵入选期刊
中国高校优秀科技期刊
华北优秀期刊
河北省优秀科技期刊

协同采编系统

友情链接

基于深度卷积网络的红外图像人体姿态识别方法

(淮北师范大学物理与电子信息学院，淮北235000)
DOI： 10.13763/j.cnki.jhebnu.nse.202302026

Human Pose Recognition Method Based on Deep Convolutional Network in Infrared Images

YUE Yurong,ZHAO Dan,DONG Xuan,GUO Shanshan,CUI Shaohua,SHAN Wei

PDF
下载
HTML

摘要/Abstract

摘要：

针对传统红外图像行人姿态识别的问题，在经典LeNet-5模型的基础上，提出一种改进型LeNet-5的网络模型.网络设定输入红外图像尺寸为256×256×1，选取4层卷积计算增加网络深度，以Leaky ReLu为激活函数并加入dropout层，最后以1×1卷积代替全连接，减小模型参数尺寸，防止过拟合.实验将改进型LeNet-5与经典LeNet-5模型进行比对，结果表明改进型LeNet-5效果最好.与流行的ShuffleNet，NasNet-mobile，EfficientNet-b0和MobileNetV2算法进行对比，实验结果表明，所得测试集的准确率达到97.5%，mean average precision,average recall 和 F_1-score性能指标均优于其他算法.

Abstract：

Aiming at the problem of pedestrian pose recognition in traditional infrared images,we proposed an improved LeNet-5 network based on the classic LeNet-5 model.The input infrared image size was set as 256×256×1.Four layers of convolution was selected to deepen the network depth,Leaky ReLu was used as the activation function and adds the Dropout layer was added.Finally,it uses 1×1 convolution instead of full connection was used to reduce the model parameter size and prevent overfitting.Compared the improved LeNet-5 model with the classic LeNet-5 model,the experimental results show that the improved LeNet-5 model has the best performance.Compared it with popular ShuffleNet,NasNet-mobile,EfficientNet-b0 and MobileNetV2 algorithms,the results show that the proposed network had better mean average precision,average recall,and F_1-score.

关键词

关键词： 改进型LeNet5；红外图像；姿态识别；卷积神经网络；深度学习

Key words： improved LeNet5 ; infrared image ; pose recognition ; convolutional neural network ; deep learning

参考文献 19

[1] 周啸辉,余磊,何茜,等.基于改进ResNet-18的红外图像人体行为识别方法研究[J].激光与红外,2021,51(9):1178-1184.doi:10.3969/j.issn.1001-5078.2021.09.011 ZHOU Xiaohui,YU Lei,HE Xi,et al.Research on Human Behavior Recognition Method in Infrared Image Based on Improved ResNet-18[J].Laser & Infrared,2021,51(9):1178-1184.
[2] NANDA H,DAVIS L.Probabilistic Template Based Pedestrian Detection in Infrared Videos[C]//Intelligent Vehicle Symposium.IEEE：IEEE,2002:15-20.doi:10.1109/IVS.2002.1187921
[3] 邵延华,郭永彩,高潮.基于稠密轨迹特征的红外人体行为识别[J].光电子·激光,2015,26(4):758-763.doi：cnki.sun.gdzj.0.2015-04-025 SHAO Yanhua,GUO Yongcai,GAO Chao.Infrared Human Action Recognition Using Dense Trajectories-based Feature[J].Guangdianzi Jiguang/Journal of Optoelectronics Laser,2015,26(4):758-763.
[4] 朱大炜.基于深度学习的红外图像飞机目标检测方法[D].西安：西安电子科技大学,2018.doi:cnki.cdmd.2.1019.003410 ZHU Dawei.Infrared Image Plane Target Detection Method Based on Deep Learning[D].Xi′an：Xi′an University of Electronic Science and Technology,2018.
[5] 张汝榛,张建林,祁小平,等.复杂场景下的红外目标检测[J].光电工程,2020,47(10):128-137.doi:10.12086/oee.2020.200314 ZHANG Ruzhen,ZHANG Jianlin,QI Xiaoping,et al.Infrared Target Detection and Recognition in Complex Scene[J].Opto-Electron Eng,2020,47(10):128-137.
[6] 廖莎莎.基于筛选深度特征的红外图像目标识别方法[J].红外与激光工程,2022,51(5):20210372.doi:10.3788/IRLA20210372 LIAO Shasha.Infrared Image Target Recognition Method Based on Selected Deep Features[J].Infrared and Laser Engineering,2022,51(5):20210372.
[7] LECUN Y，BOTTOU L，BENGIO Y，et al.Gradient-based Learning Applied to Document Recognition[J].Proceedings of the IEEE，1998,86（11）：2278-2324.doi:10.1109/5.726791
[8] KAREN S,ANDREW Z.Very Deep Convolutional Networks for Large-scale Image Recognition[J].CoRR,2014,abs:1409.1556.doi:10.48550/arXiv.1409.1556
[9] KRIZHEVSKY A,SUTSKEVER I,HINTON G.ImageNet Classification with Deep Convolutional Neural Networks[J].Advances in Neural Information Processing Systems,2012,25(2):84-90.doi:10.1145/3065386
[10] 王立扬,张瑜,沈群,等.基于改进型LeNet-5的苹果自动分级方法[J].中国农机化学报,2020,41(7):105-110.doi:10.13733/j.jcam.issn.2095-5553.2020.07.016 WANG Liyang,ZHANG Yu,SHEN Qun,et al.Automatic Detecting and Grading Method of Apples Based on Improved LeNet-5[J].Chinese Journal of Agricultural Machinery Chemistry,2020,41(7):105-110.
[11] 刘东来,崔亚飞,罗辉,等.基于改进型LeNet-5的工业机器人工件自动识别研究[J].制造技术与机床,2021,710(8):103-107.doi:10.19287/j.cnki.1005-2402.2021.08.005 LIU Donglai,CUI Yafei,LUO Hui,et al.Research on Automatic Recognition of Industrial Robot Workpiece Based on Improved LeNet-5[J].Manufacturing Technology and Mnachine Tools,2021,710(8):103-107.
[12] 张力超,马蓉,张垚鑫.改进的LeNet-5模型在苹果图像识别中的应用[J].计算机工程与设计,2018,39(11):3570-3575.doi:10.16208/j.issn1000-7024.2018.11.048 ZHANG Lichao,MA Rong,ZHANG Yaoxin.Application of Improved LeNet-5 Model in Apple Image Recognition[J].Computer Engineering and Design,2018,39(11):3570-3575.
[13] LIN M,CHEN Q,YAN S.Network in Network Computer Vision and Pottern Recognition[J].IEEE,2013，1312:1-10.doi:10.48550/arXiv.1312.4400
[14] LEE E J,KO B C,NAM J Y.Recognizing Pedestrian′s Unsafe Behaviors in Far-infrared Imagery at Night[J].Infrared Physics & Technology,2016,76:261-270.doi:10.1016/j.infrared.2016.03.006
[15] 郝帅,高山,马旭,等.基于跨尺度特征聚合与分层注意力映射的红外行人检测[J].光子学报,2022,51(6):419-435.doi:10.3788/gzxb20225106.0610006 HAO Shuai，GAO Shan，MA Xu，et al.Infrared Pedestrian Detection Based on Cross-scale Feature Aggregation and Hierarchical Attention Mapping[J].Acta Photonica Sinica，2022，51（6）：419-435.
[16] ZHANG X,ZHOU X,LIN M,et al.ShuffleNet:An Extremely Efficient Convolutional Neural Network for Mobile Devices[J].CoRR,2017,abs:1707.01083.doi:10.48550/arXiv.1707.01083
[17] ZOPH B,VASUDEVAN V,SHLENS J,et al.Learning Transferable Architectures for Scalable Image Recognition[J].CVPR，2017,1707:1-14.doi:10.1109/CVPR.2018.00907
[18] TAN M,LE Q V.EfficientNet:Rethinking Model Scaling for Convolutional Neural Networks[J].IEEE,2019,1905:1-11.doi:10.48550/arXiv.1905.11946
[19] SANDLER M,HOWARD A,ZHU M,et al.MobileNetV2:Inverted Residuals and Linear Bottlenecks[C]//CVF Conference on Computer Vision and Pattern Recognition(CVPR).IEEE:IEEE,2018:4510-4520.doi:10.1109/CVPR.2018.00474