一种基于速度强度熵VMME与纹理特征的人群异常检测算法
PDF下载 (23145)李 斐,陈 恳 *,李 萌,郭春梅.一种基于速度强度熵VMME与纹理特征的人群异常检测算法[J].宁波大学学报(理工版),2017,30(4):63-67.DOI:
LI Fei,CHEN Ken *,LI Meng,GUO Chun-mei.An anomaly detection algorithm based on VMME and texture features[J].Journal of Ningbo University(Natural Science & Engineering Edition),2017,30(4):63-67.DOI:
| Title: | An anomaly detection algorithm based on VMME and texture features |
| 作者: | 李 斐, 陈 恳 *, 李 萌, 郭春梅 |
| Author(s): | LI Fei, CHEN Ken *, LI Meng, GUO Chun-mei |
| 关键词: | 人群异常检测; 纹理特征; 运动特征 |
| Keywords: | VMME; LBPCM; crowd anomaly detection; texture feature; motion feature |
| 分类号: | TP391 |
| 文献标识码: | A |
| 摘要: | 人群异常事件检测是智能视频监控领域的重要研究内容, 文章提出了一种融合速度强度熵VMME与纹理特征的人群异常行为检测算法. 该算法采用LBPCM算法提取图像纹理特征, 在视频帧计算光流基础上, 获得特征点速度强度图, 并以其熵VMME作为场景运动特征, 将场景纹理特征和运动特征送入支持向量机训练分类. 实验表明, 新算法可实现对人群异常行为的检测, 且有较高准确率. |
| Abstract: | In the field of intelligent video surveillance, the detection of abnormal events has remained an important subject in the research field. This paper proposes an algorithm for detecting abnormal behavior of crowds based on entropy of velocity magnitude map (VMME) and texture feature. Through the computation of optical flow in a video frame, the velocity magnitude map of feature points can be obtained, and the entropy of velocity magnitude map can be calculated as the feature of scene motion. LBPCM is first used to extract the texture features of the crowd, then the features of two kinds are fused into the support vector machine for training classification. Experiments show that the algorithm can effectively detect abnormal behavior and has high detection accuracy rate. |
| 参考文献 /References: | [1] HELD C, KRUMM J, MARKEL P, et al. Intelligent video surveillance[J]. Computer, 2012, 45(3):83-84. [2] 蔡利. 智能视频监控中运动目标的分类识别研究[D]. 南京: 南京邮电大学, 2015. [3] CHENG Y H, WANG J. A motion image detection method based on the inter-frame difference method[J]. Applied Mechanics & Materials, 2014, 490/491:1283-1286. [4] 屈晶晶, 辛云宏. 连续帧间差分与背景差分相融合的运动目标检测方法[J]. 光子学报, 2014, 43(7):213-220. [5] PATHAN S S, ALHAMADI A, MICHAELIS B. Incor- porating social entropy for crowd behavior detection using SVM[J]. Lecture Notes in Computer Science, 2010, 6453:153-162. [6] 董帅铭. 基于视频的行人检测及异常行为检测[D]. 哈尔滨: 哈尔滨工业大学, 2012. [7] ZHANG Y H, QIN L, YAO H X, et al. Abnormal crowd behavior detection based on social attribute-aware force model[EB/OL]. [2015-12-14]. https://www.researchgate.net/publication/261387575. [8] XIONG G, WU X, CHEN Y L, et al. Abnormal crowd behavior detection based on the energy model[C]//IEEE International Conference on Information and Automation, 2011:495-500. [9] LI Y, ZOU T, CHEN P. Estimation of crowd density based on adaptive LBP[J]. Advanced Materials Research, 2014, 998/999:864-868. [10] YE Z, WANG J, WANG Z, et al. Multiple features fusion for crowd density estimation[C]//International Conference on Internet Multimedia Computing and Service, 2012:42-45. [11] BAUER N, PATHIRANA P, HODGSON P. Robust optical flow with combined Lucas-Kanade/Horn-Schunck and automatic neighborhood selection[C]//International Conference on Information and Automation, 2006:378-383. [12] LI L, LI D. Fuzzy entropy image segmentation based on particle swarm optimization[J]. Progress in Natural Science, 2008, 18(9):1167-1171. [13] WU S D, MOORE B, SHAH M. Chaotic invariants of Lagrangian particle trajectories for anomaly detection in crowded scenes[C]//Proceedings of the 2010 IEEE Conference on Computer Vision and Pattern Recognition, 2010:2054-2060. [14] WU S, WONG H S, YU Z W. A Bayesian model for crowd escape behavior detection[J]. IEEE Transactions on Circuits & Systems for Video Technology, 2014, 24(1):85-98. [15] YANG C, YUAN J S, LIU J. Sparse reconstruction cost for abnormal event detection[C]//Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition, 2011:3449-3456. |
| 备注/Memo: | 收稿日期: 2016－11－16. 宁波大学学报（理工版）网址: http://journallg.nbu.edu.cn/ 基金项目: 国家自然科学基金（60972063）; 宁波市自然科学基金（2014A610065）; 宁波大学学科项目基金（XKXL1308）. 第一作者: 李斐（1992－）, 女, 河南驻马店人, 在读硕士研究生, 主要研究方向: 视频异常行为检测与分析. E-mail: m18892629127_1@163.com *通信作者: 陈恳（1962－）, 男, 重庆人, 博士/副教授, 主要研究方向: 图像与视频分析处理及智能控制. E-mail: chenken@nbu.edu.cn 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
