节点文献

基于3D卷积神经网络的人体动作识别算法

Human Action Recognition Algorithm Based on 3D Convolution Neural Network

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 张瑞李其申储珺

【Author】 ZHANG Rui;LI Qishen;CHU Jun;School of Information Engineering,Nanchang Hangkong University;Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition;

【机构】 南昌航空大学信息工程学院江西省图像处理与模式识别重点实验室

【摘要】 由于人体动作的多样性、场景嘈杂、摄像机运动视角多变等特性,导致人体动作识别的难度增加。为此,基于3D卷积神经网络,提出一种新的人体动作识别算法。以连续的16帧视频为一组输入,采用视频图像的灰度、x方向梯度、y方向梯度、x方向光流、y方向光流做多通道处理,训练网络参数,经过5层3D卷积、5层3D池化增加提取特征中时间维度的动作信息,最终通过2层全连接与softmax分类器得到识别分类结果。在UCF101数据库上进行实验,结果表明,相比iDT、P-CNN、LRCN算法,该算法具有较高的识别准确率,且运行速度更快。

【Abstract】 Human action diversity,scene noise,the camera motion angle changes and other factors increase the difficulty of human action recognition. This paper proposes a human action recognition algorithm based on 3D convolution neural network. Firstly,successive 16 frames of the video are divided into a group as the input. Secondly,the input data is multichannel processed using the gray,gradient-x,gradient-y,optflow-x and optflow-y,w hich effectively trains the network parameters. Thirdly,the extracted features are obtained using 5-layer 3D convolution and 5-layer 3D pooling to increase time dimension information,Finally,the recognition results are obtained by two full connection layers and the softmax classifier. Experiment is made on the UCF101 database,and the results show that compared with iDT,P-CNN,LRCN algorithms,the proposed algorithm has a higher accuracy of human action recognition and a faster running speed.

【基金】 国家自然科学基金(61663031);江西省自然科学基金(20132BAB201046);南昌航空大学研究生创新专项资金(YC2016009)
  • 【文献出处】 计算机工程 ,Computer Engineering , 编辑部邮箱 ,2019年01期
  • 【分类号】TP391.41;TP183
  • 【网络出版时间】2017-12-27 11:01:54
  • 【被引频次】25
  • 【下载频次】1041
节点文献中: 

本文链接的文献网络图示:

本文的引文网络