期刊文献+

一种基于视频图像相关性的火灾火焰识别方法 被引量:24

Renovated method for identifying fire plume based on image correlation
下载PDF
导出
摘要 由于火焰燃烧时气体羽流的卷吸特性和空气流动的影响,燃烧的火焰表现出不停的振荡特性,基于这种振荡性提出利用图像的相关性来探测火灾的发生。火灾视觉特征的提取是视觉火灾探测的重要问题,首先确定运动区域,再利用火焰振荡特性提出了相关性算法,并分析了R、G、B3种分量对相关性的影响。实验结果表明利用相关性来分析时,火焰的相关性系数图是剧烈振荡的,而非火焰的相关性系数图是相对比较平坦的,为可视化火灾探测系统提供了一种有效的判据;R、G、B对相关性的影响是一致的,在实际应用中只需利用其中的1个分量即可,可有效提高火灾探测系统的响应时间。 This article is inclined to present its authors' renovated method for identiicing fire plume based on the plume image correlation. As in the fire-fighting practice in recent years, it has become a latest developmental trend of identifying the fire burning strength by judging its burning color, texture and time features, for a fire flame is characteristic of oscillation of an air flow, which itself is different from other rigid matters. Thus, we have brought forward the so-called oscillation-identifying method whose main feature is to identify and measure the correlation coefficients between the time of burning and the appearance of the flame. In our experiments, location was presented for surveying ordinary CCD between the first frame and the corresponding subsequent frames. First of all, it is to determine the flame- moving area based on picking-up the commix background. Secondly, it is to make use of the image correlation for fire detection based on its moving region. In the meantime, we have to analyze the fireplume image correlation of its intensity titled as R, G and B for the identification based on the fire oscillation. The present method is inclined to be used for reducing the fire disaster likely to cause casualty and material damage due to the failure to get reported so as to boost up the robust of the detection system. Experimental results indicate that the algorithm is effective and useful. The method can also be used for non-fire detection purposes. All in all, the image of the correlation coefficients of the flames is characteristic of oscillation, whereas those of the non-flame matters are rather smooth, whereas the mutual influence of the R, G or B intensities proves to be fairly stable. When the method is put for practical application, it is recommended to choose a sort of intensity among the three intensities of R, G or B to reduce the operation time and enhance the system's capability.
出处 《安全与环境学报》 CAS CSCD 2007年第6期96-99,共4页 Journal of Safety and Environment
基金 国家科技支撑计划项目(2006BAK06B07)
关键词 安全工程 视觉特征 相关性 safety engineering vision character correlation
  • 相关文献

参考文献12

  • 1YAMAGSHI H, YAMAGUCHI J. Fire flame detection algorithm using a eolor camera[C]// Proceedings of the 1999 International Symposium on Micromechatronics and Human Science,Nagoya. Japan, 1999 : 255 - 260. 被引量:1
  • 2YAMAGSHI H. YAMAGUCHI J. A contour fluctuation data pronessing method for fire flame detection using a color camera [ C]//IEEE. Proceedings of 26th Annual Conference on IECON of the hldustrial Electronics Society, Nagoya, Japan, 2000:824-829. 被引量:1
  • 3NODA S, UEDA K. Fire detection in tunnels using an image processing method[ C ]// Proceedings of Vehicle Navigation and Information Systems Conference, Yakohama, Japan, 1994:57 - 64. 被引量:1
  • 4PHILIPS Ⅲ W, SHAH M, DA VITORIA LOBO N. Flame recognition in video[ C ]// IEEE 5th IEEE Workshop on Application of Computer Vision and Patter Recognition, Palm Springs, America, 2000:224 - 229. 被引量:1
  • 5董华,程晓舫,范维澄.早期火灾图像监测技术的应用与比较[J].光学技术,1997,23(5):51-53. 被引量:19
  • 6卢结成,吴龙标,宋卫国.一种火灾图像探测系统的研究[J].仪器仪表学报,2001,22(4):437-440. 被引量:24
  • 7袁宏永,刘炳海,陈晓军,宗封仪.图像型火灾智能自动探测与空间定位技术[J].消防科技,1998,22(2):2-4. 被引量:21
  • 8袁非牛,廖光煊,张永明,刘勇,于春雨,王进军,刘炳海.计算机视觉火灾探测中的特征提取[J].中国科学技术大学学报,2006,36(1):39-43. 被引量:52
  • 9OLLERO A, ARRUE B C, MARTINEZ J R, et al. Techniques for reducing false alarms in infrared forest-fire automatic detection systems [J]. Control Engineering Practice, 1999, 7:121 - 131. 被引量:1
  • 10STAUFFER C, GRIMSON W E L, Adaptive background mixture models for real-time tracking [J]. Proc IEEE Computer Society Conf on Computer Vision and Pattern Recognition, 1999, 2: 246-252. 被引量:1

二级参考文献15

  • 1林家明.黑白CCD摄像机的近红外特性和应用[J].光学技术,1996,22(6):33-34. 被引量:6
  • 2Yamagishi H,Yamaguchi J.Fire flame detection algorithm using a color camera[C]∥Proceedings of 1999 International Symposium on Micromechatronics and Human Science.Nagoya,Japan,1999:255-260. 被引量:1
  • 3Yamagishi H,Yamaguchi J.A contour fluctuation data processing method for fire flame detection using a color camera[C]∥IEEE 26th Annual Conference on IECON of the Industrial Electronics Society,Nagoya,Japan,2000,(2):824-829. 被引量:1
  • 4Noda S,Ueda K.Fire detection in tunnels using an image processing method[C]∥Proceedings of Vehicle Navigation and Information Systems Conference.Yokohama,Japan,1994:57-62. 被引量:1
  • 5Phillips Ⅲ W,Shah M,Da Vitoria Lobo N.Flame recognition in video[C]∥Fifth IEEE Workshop on Applications of Computer Vision.California:Palm Springs,2000:224-229. 被引量:1
  • 6Healey G,Slater D,Lin T,et al.A system for real-time fire detection[C]∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition.New York,1993:605-606. 被引量:1
  • 7Video smoke detection[EB/OL].http://www.chubb.com.au/vsd.asp. 被引量:1
  • 8CHEN Thou-ho,KAO Cheng-liang,CHANG Sju-ma.An intelligent real-time fire-detection method based on video[C]∥Processings of IEEE 37th Annual 2003 International Carnahan Conference on Security Technology.Taipei,2003:104-111. 被引量:1
  • 9Stauffer C,Grimson W E L.Learning patterns of activity using real-time tracking[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.2000,22(8):747-757. 被引量:1
  • 10Pavlidis I,Morellas V,Tsiamyrtzis P,Harp S.Urban surveillance systems:from the laboratory to the commercial world[C]∥Proceedings of the IEEE,2001,89(10):1 478-1 497. 被引量:1

共引文献91

同被引文献190

引证文献24

二级引证文献156

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部