[1]郑怀兵,翟济云.基于视频分析的森林火灾烟雾检测方法[J].南京理工大学学报(自然科学版),2015,39(06):686.
 Zheng Huaibing,Zhai Jiyun.Forest fire smoke detection based on video analysis[J].Journal of Nanjing University of Science and Technology,2015,39(06):686.
点击复制

基于视频分析的森林火灾烟雾检测方法
分享到:

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

卷:
39卷
期数:
2015年06期
页码:
686
栏目:
出版日期:
2015-12-31

文章信息/Info

Title:
Forest fire smoke detection based on video analysis
作者:
郑怀兵1翟济云2
1.南京森林警察学院 森林消防系,江苏 南京 210023;
2.南京航空航天大学 自动化学院,江苏 南京 210016
Author(s):
Zheng Huaibing1Zhai Jiyun2
1.Forest Fire Control Department,Nanjing Forest Police College,Nanjing 210023,China;
2.College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China
关键词:
视频分析 森林火灾 烟雾检测 高斯混合模型 静态特征 动态特征 支持向量机 运动方向 高频能量 紧凑度
Keywords:
video analysis forest fire smoke detection Gaussian mixture model static features dynamic features support vector machine moving direction high frequency energy compactness
分类号:
TP391.41
摘要:
为了提高火灾烟雾识别的实时性和准确性,提出了一种基于视频分析的森林火灾烟雾检测方法。采用高斯混合模型进行背景建模,通过背景差法检测运动目标。充分考虑烟雾的特点,设计了多种静态特征和动态特征。设计了基于支持向量机的分类器,对检测出的运动目标区域进行分类识别,确定其是否为烟雾。分别针对正常情况和雾天情况进行了实验。实验结果显示,该文方法能有效地应用于森林火灾烟雾的检测且对天气的影响具有一定的鲁棒性。检验了使用不同特征组合的识别效果。结果显示选取运动方向、高频能量、紧凑度3个特征组成的特征向量进行识别具有最优的效果,在正常天气和雾天情况下,正确率分别达到了92.7% 和76.3%。
Abstract:
In order to improve the real-time and accuracy of fire smoke detection,a forest fire smoke detection method based on video analysis is proposed here.Gaussian mixture model is employed for background modeling,and moving target detection is conducted by the comparison with the background model.Multiple static and dynamic features are designed considering the characteristics of smoke.A classifier is designed based on a support vector machine to detect smoke from moving targets.Experiments are conducted for the normal case and foggy case.The results show that the proposed method can detect forest fire smoke effectively and is robust.The performances of different combinations of features are compared.The results show that the smoke recognition feature vector composed of moving direction,high frequency energy and compactness has the best performance,correct recognition rates are 92.7% in normal case and 76.3% in foggy case.

参考文献/References:

[1] 李彬,朱宁.山区森林火灾频发的原因及对策[J].安徽农学通报,2012,18(6):136-137.
Li Bin,Zhu Ning.Causes and countermeasures of forest fire in mountain areas[J].Anhui Agriculture Science Bulletin,2012,18(6):136-137.
[2]游勇.基于视频动态特征分析的烟雾检测算法研究[D].成都:西南交通大学信息科学与技术学院,2013.
[3]曹丹华,邹伟,吴裕斌.基于背景图像差分的运动人体检测[J].光电工程,2007,34(6):107-111.
Cao Danhua,Zou Wei,Wu Yubin.Motion human detection based on difference background image[J].Opto-Electronic Engineering,2007,34(6):107-111.
[4]Duda R O,Hart P E,Stork D G.Pattern classification[M].2nd ed.New York,USA:John Wiley & Sons,2000.
[5]Gubbi J,Marusic S,Palaniswami M.Smoke detection in video using wavelets and support vector machines[J].Fire Safety Journal,2009,44(8):1110-1115.
[6] Nguyen D T,Zong Z,Ogunbona P,et al.Object detection using non-redundant local binary patterns[A].2010 17th IEEE International Conference on Image Processing(ICIP)[C].Hong Kong,China:IEEE,2010:4609-4612.
[7]Toreyin B U,Dedeoglu Y,Gudukbay U,et al.Computer vision based system for real-time fire and flame detection[J].Pattern Recognition Letters,2006,27(1):49-58.
[8]Markus A S,Markus O.Similarity of color images[J].Proc of SPIE,1995,2420:381-392.
[9]Sweldens W.The lifting scheme:A custom-design construction of biorthogonal wavelets[J].Applied and Computational Harmonic Analysis,1996,3(2):186-200.
[10]王哲峰.基于提升小波变换的图像匹配算法研究[D].长春:吉林大学通信工程学院,2007.
[11]袁非牛,张永明,刘士兴,等.基于累积量和主运动方向的视频烟雾检测方法[J].中国图象图形学报,2008,13(4):808-813.
Yuan Feiniu,Zhang Yongming,Liu Shixing,et al.Video smoke detection based on accumulation and main motion orientation[J].Journal of Image and Graphics,2008,13(4):808-813.
[12]Vapnik V.Statistical learning theory[M].New York,USA:Wiley,1998.
[13]王洪波.单分类支持向量机的学习方法研究[D].杭州:浙江大学电气工程学院,2012.

备注/Memo

备注/Memo:
收稿日期:2015-10-08 修回日期:2015-11-06
基金项目:江苏省林业三新工程项目(LYSX[2014]07)
作者简介:郑怀兵(1970-),男,副教授,主要研究方向:森林防火,E-mail:42636627@qq.com。
引文格式:郑怀兵,翟济云.基于视频分析的森林火灾烟雾检测方法[J].南京理工大学学报,2015,39(6):686-691.
投稿网址:http://zrxuebao.njust.edu.cn
DOI:10.14177/j.cnki.32-1397n.2015.39.06.009
更新日期/Last Update: 2015-12-31