[1]朱陶,杜治国,洪卫军.一种基于深度卷积神经网络的摄像机覆盖质量评价算法[J].江西师范大学学报(自然科学版),2015,(03):309-314.
 ZHU Tao,DU Zhiguo,HONG Weijun.The Camera Coverage Quality Evaluation Algorithm Based on Deep Convolution Neural Network[J].,2015,(03):309-314.
点击复制

一种基于深度卷积神经网络的摄像机覆盖质量评价算法()
分享到:

《江西师范大学学报》(自然科学版)[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2015年03期
页码:
309-314
栏目:
出版日期:
2015-05-31

文章信息/Info

Title:
The Camera Coverage Quality Evaluation Algorithm Based on Deep Convolution Neural Network
作者:
朱陶;杜治国;洪卫军
中国人民公安大学,北京 100038
Author(s):
ZHU TaoDU ZhiguoHONG Weijun
关键词:
视频监控摄像机 覆盖质量 深度卷积神经网络 支持向量回归机
Keywords:
video surveillance camera coverage quality deep convolution neural network support vector regression
分类号:
TP 391.41
文献标志码:
A
摘要:
随着视频监控系统的大规模普及,视频监控系统的效用评价成为一个重要的研究课题.当前视频监控系统评价只考虑了摄像机的覆盖率,缺少对摄像机覆盖质量的量化评价.该文提出了一种基于深度卷积神经网络的监控摄像机覆盖质量评价算法.将摄像机覆盖质量评价问题转化为对摄像机所采集视频帧的质量评价问题,探讨了基于视频帧的摄像机覆盖质量等级的分级策略,标注了一个摄像机视频帧质量等级数据集; 设计了一种新颖的多维标签赋值方法,利用深度卷积网络学习鲁棒的视频帧表示,进一步基于支持向量回归机(SVR)学习视频质量回归函数,从而实现对摄像机覆盖质量的鲁棒估计.实验结果表明:该算法能够准确地对监控摄像机的覆盖质量进行自动评测,有效监测了摄像机监控质量的实时变化.
Abstract:
Along with the popularity of video surveillance system,effect evaluation of video surveillance system becomes an important research item.Current evaluation of video surveillance system only takes camera coverage rate into consideration without quantitative evaluation of camera coverage quality.The article provides a surveillance camera coverage quality evaluation algorithm based on deep convolution neural network.The problem of camera coverage quality evaluation algorithm is transformed into the problem of quality evaluation on video frames collected by cameras.A classification strategy based on camera coverage quality levels of video frames is provided and a data set of quality levels of camera video frames is labeled.A multi-dimension label assignment method is designed for utilizing deep convolution neural network to learn a robust video frame indication,and furthermore,to learn a video quality regression function based on Support Vector Regression(SVR),thus a robust evaluation on video coverage quality is performed.The experiment result shows that the algorithm of the article can perform an automatic evaluation on the surveillance camera coverage quality precisely,and effectively monitors the real-time change of camera surveillance quality.

参考文献/References:

[1] 吕世良,王晓茜,刘金国.数字视频监控系统设计与实 现 [J].测控技术,2014,33(2):80-82.
[2] Yan Lecun,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition [J].Proceedings of the IEEE,1998,86(11):2278-2324.
[3] Krizhevsky A,Sutskever I,Hinton G E.Imagenet classification with deep convolutional neural networks [C]∥Advances in Neural Information Processing Systems,2012:1097-1105.
[4] Taigman Y,Yang M,Ranzato M A,et al.Deepface:Closing the gap to human-level performance in face verification [C]∥Computer Vision and Pattern Recognition(CVPR),2014:1701-1708.
[5] Girshick R,Donahue J,Darrell T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation [EB/OL].
[2014-10-19].http:∥arxiv.org/abs/1311.2524.
[6] Basak D,Pal S,Patranabis D C.Support vector regression [J].Neural Information Processing-Letters and Reviews,2007,11(10):203-224.
[7] Chang Chih-chung,Lin Chih-jen.LIBSVM:a library for support vector machines [J].ACM Transactions on Intelligent Systems and Technology(TIST),2011,2(3):27.
[8] Jia Yangqing,Shelhamer E,Donahue J,et al.Caffe:Convolutional architecture for fast feature embedding [EB/OL].
[2014-10-20].http:∥arxiv.org/abs/1408.5093.
[9] Guo Zhenhua,Zhang D.A completed modeling of local binary pattern operator for texture classification [J].Image Processing,IEEE Transactions on,2010,19(6):1657-1663.
[10] Zhu Qiang,Yeh M C,Cheng Wangting,et al.Fast human detection using a cascade of histograms of oriented gradients [C]∥Computer Vision and Pattern Recognition,2006 IEEE Computer Society Conference on IEEE,2006,2:1491-1498.

备注/Memo

备注/Memo:
国家“863”计划(2013AA0146042014);中国人民公安大学基本科研业务费科技类课题(2014JKF02205)
更新日期/Last Update: 1900-01-01