基于视频数据特性的动态手势识别

doi:10.13190/j.jbupt.2020-056

北京邮电大学学报 ›› 2020, Vol. 43 ›› Issue (5): 91-97.doi: 10.13190/j.jbupt.2020-056

基于视频数据特性的动态手势识别

谢晓燕¹, 赵欢¹, 蒋林²

1. 西安邮电大学计算机学院, 西安 710121;
2. 西安科技大学集成电路设计实验室, 西安 710054

收稿日期:2020-06-07 发布日期:2021-03-11
通讯作者: 赵欢(1995-),男,硕士生,E-mail:750746730@qq.com. E-mail:750746730@qq.com
作者简介:谢晓燕(1972-),女,教授,硕士生导师.
基金资助:
国家自然科学基金项目（61834005，61772417，61602377）；陕西省国际科技合作计划项目（2018KW-006）；榆林市科技计划项目（2019-133）

Dynamic Gesture Recognition Based on Characteristics of Encoded Video Data

XIE Xiao-yan¹, ZHAO Huan¹, JIANG Lin²

1. School of Computer, Xi'an University of Posts&Telecommunications, Xi'an 710121, China;
2. Integrated Circuit Design Laboratory, Xi'an University of Science and Technology, Xi'an 710054, China

Received:2020-06-07 Published:2021-03-11

摘要/Abstract

摘要： 针对现有动态手势识别方法环境适应性低、计算复杂的问题，提出了一种基于视频数据特性的动态手势识别方法.使用基于密度的聚类算法DBSCAN直接从视频编码数据中的运动矢量提取出运动趋势特征，再通过随机森林分类运动趋势，结合卷积神经网络（CNN）提取的手型特征识别动态手势.实验结果表明，该方法对剑桥大学和美国西北大学数据集中动态手势的平均识别率分别达到94.22%和94.48%，并且与CNN结合长短期记忆网络的识别方法相比，手势识别时间减少了85%.在背景图像复杂且光照条件不足时，该方法仍然能够维持较高的识别率，表现出较好的鲁棒性.

关键词: 动态手势识别, 运动矢量, DBSCAN, 随机森林, 卷积神经网络

Abstract: Aiming at the challenges to scene adaptability and computational complexity of dynamic gesture recognition, a method based on characteristics of encoded video data is proposed. Firstly, density-based spatial clustering of applications with noise is used to extract motion trend features from motion vectors. Then, the motion trends are classified by random forest. Finally, combined by the hand shape features extracted by convolutional neural network(CNN), the dynamic gesture is recognized. The experiment shows that the proposed method has an average recognition rate of 94.22% and 94.48% respective for university of Cambridge and Northwestern University hand gesture data sets. Compared with the scheme combine of CNN and long short-term memory, the gesture recognition time is reduced by 85%. It can still maintain a higher recognition rate for the complex background with insufficient illumination, represents a higher robustness.

Key words: dynamic gesture recognition, motion vector, density-based spatial clustering of applications with noise, random forest, convolutional neural network

中图分类号:

TP391.4

谢晓燕, 赵欢, 蒋林. 基于视频数据特性的动态手势识别[J]. 北京邮电大学学报, 2020, 43(5): 91-97.

XIE Xiao-yan, ZHAO Huan, JIANG Lin. Dynamic Gesture Recognition Based on Characteristics of Encoded Video Data[J]. Journal of Beijing University of Posts and Telecommunications, 2020, 43(5): 91-97.

参考文献

[1] Deller M, Ebert A, Bender M, et al. Flexible gesture recognition for immersive virtual environments[C]//IV'06. London:IEEE, 2006:563-568.
[2] Elmezain M, Al-Hamadi A, Michaelis B. Hand trajectory-based gesture spotting and recognition using HMM[C]//ICIP. Cairo:IEEE, 2009:3577-3580.
[3] Barros P, Maciel-Junior N T, Fernandes B J T, et al. A dynamic gesture recognition and prediction system using the convexity approach[J]. CVIU, 2017, 155:139-149.
[4] Tang Hao, Liu Hong, Xiao Wei, et al. Fast and robust dynamic hand gesture recognition via key frames extraction and feature fusion[J]. Neurocomputing, 2019, 331:424-433.
[5] Patil A R, Subbaraman S. A spatiotemporal approach for vision-based hand gesture recognition using Hough transform and neural network[J]. Signal, Image and Video Processing, 2019, 13(2):413-421.
[6] Gatto B B, dos Santos E M, Da Silva W S. Orthogonal Hankel subspaces for applications in gesture recognition[C]//SIBGRAPI. Niteroi:IEEE, 2017:429-435.
[7] Molchanov P, Gupta S, Kim K, et al. Hand gesture recognition with 3D convolutional neural networks[C]//CVPR. Boston:IEEE, 2015:1-7.
[8] Lu Dongwei, Qiu Chu, Xiao Yi. Temporal convolutional neural network for gesture recognition[C]//ICIS. Singapore:IEEE, 2018:367-371.
[9] Muralidhar P, Rao C B R. Analysis of block matching motion estimation algorithms[C]//ICCCNT. Tiruchengode:IEEE, 2013:1-4.
[10] Singha J, Roy A, Laskar R H. Dynamic hand gesture recognition using vision based approach for human-computer interaction[J]. Neural Computing and Applications, 2018, 29(4):1129-1141.
[11] 杨真真, 匡楠, 范露, 等. 基于卷积神经网络的图像分类算法综述[J]. 信号处理, 2018, 34(12):1474-1489. Yang Zhenzhen, Kuang Nan, Fan Lu, et al. Review of image classification algorithms based on convolutional neural networks[J]. Journal of Signal Processing, 2018, 34(12):1474-1489.
[12] Kim T K, Cipolla R. Canonical correlation analysis of video volume tensors for action categorization and detection[J]. TPAMI, 2009, 31(8):1415-1428.
[13] Shen Xiaohui, Hua Gang, Williams L, et al. Dynamic hand gesture recognition:an exemplar-based approach from motion divergence fields[J]. Image and Vision Computing, 2012, 30(3):227-235.

[1]	余佳润王亚峰. 一种低信噪比下基于深度学习的DoA估计方法[J]. 北京邮电大学学报, 2022, 45(6): 119-125.
[2]	叶康张淑军郭淇李辉崔雪红. 基于CM-Transformer的连续手语识别[J]. 北京邮电大学学报, 2022, 45(5): 49-53,78.
[3]	武兴坤, 罗涛, 刘爱军, 杨明, 张文静. 面向主动脉缩窄辅助诊断的神经网络模型设计[J]. 北京邮电大学学报, 2022, 45(4): 1-6.
[4]	武兴坤罗涛刘爱军杨明张文静李剑峰李文秀. 面向主动脉缩窄辅助诊断的神经网络模型设计[J]. 北京邮电大学学报, 2022, 45(4): 1-6.
[5]	许同乐, 孟良, 孔晓佳, 苏元浩, 孙砚飞. 基于EEMD的ICNN故障诊断方法[J]. 北京邮电大学学报, 2022, 45(2): 110-116.
[6]	李晓辉, 王先文, 樊韬, 刘佳文, 万宏杰. 基于卷积神经网络的高性能5G下行同步算法[J]. 北京邮电大学学报, 2022, 45(2): 117-123.
[7]	王靖, 姜竹青, 门爱东, 郭晓强, 王智康. 阶梯式图像去噪方法[J]. 北京邮电大学学报, 2022, 45(1): 52-57.
[8]	刘永利, 朱亚孟, 晁浩. 多策略MRFO算法的卷积神经网络超参数优化[J]. 北京邮电大学学报, 2021, 44(6): 83-88,95.
[9]	普运伟, 郭江, 刘涛涛, 吴海潇. 基于多学习单元卷积神经网络的雷达辐射源信号识别[J]. 北京邮电大学学报, 2021, 44(6): 74-82.
[10]	王艺霏, 莫爽, 吴文睿, 范少华, 肖丁. 基于内外卷积网络的网络入侵检测[J]. 北京邮电大学学报, 2021, 44(5): 94-100.
[11]	易茜, 黎伟, 易树平, 谢俊东. 基于鼠标行为时频联合分析的用户可信认证[J]. 北京邮电大学学报, 2021, 44(4): 121-128.
[12]	刘国桢, 陈鸿龙. 卷积记忆图协同过滤[J]. 北京邮电大学学报, 2021, 44(3): 21-26.
[13]	彭鹏, 李卓, 梁纪峰, 马天祥, 刘开华. 支持泛洪攻击检测的命名数据网PIT[J]. 北京邮电大学学报, 2021, 44(2): 61-67.
[14]	曹毅, 黄子龙, 盛永健, 刘晨, 费鸿博. D-2-DenseNet噪音鲁棒的城市音频分类模型[J]. 北京邮电大学学报, 2021, 44(1): 86-91.
[15]	蒲悦逸, 王文涵, 朱强, 陈朋朋. 基于CNN-ResNet-LSTM模型的城市短时交通流量预测算法[J]. 北京邮电大学学报, 2020, 43(5): 9-14.

基于视频数据特性的动态手势识别

Dynamic Gesture Recognition Based on Characteristics of Encoded Video Data

RichHTML

PDF

可视化

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