J4 ›› 2012, Vol. 50 ›› Issue (03): 527-.

• 计算机科学 • 上一篇    下一篇

基于视频的全天候多特征融合火焰检测算法

李文辉, 刘培勋, 王莹, 肖林厂, 王聪   

  1. 吉林大学 计算机科学与技术学院, 长春 130012
  • 收稿日期:2011-09-27 出版日期:2012-05-26 发布日期:2012-05-28
  • 通讯作者: 刘培勋 E-mail:liupeixun_jlu@163.com

RoundtheClock Video Flame Detection AlgorithmBased on Multifeature Fusion

LI Wenhui, LIU Peixun, WANG Ying, XIAO Linchang, WANG Cong   

  1. College of Computer Science and Technology, Jilin University, Changchun 130012, China
  • Received:2011-09-27 Online:2012-05-26 Published:2012-05-28
  • Contact: LIU Peixun E-mail:liupeixun_jlu@163.com

摘要:

针对传统可视化火灾探测技术采用单独分析可见光摄像设备或红外摄像设备采集的图像进行火灾探测, 不能全天候监控火焰的问题, 提出一种全天候多特征融合的检测算法. 首先通过视频类型判断算法确定摄像头采集的每帧图像类型, 然后利用与图像类型相对应的算法提取疑似火焰区域, 最后对疑似区域进行基于DCT的火焰频率检测和帧间相关性检测, 判断火焰是否存在. 实验表明, 该检测算法解决了传统可视化探测技术不能全天候监控火焰的问题, 且能在保持火焰高检测率的同时降低误检率.

关键词: 视频类型判断; 全天候; 火焰检测; 特征融合; 红外视频

Abstract:

A multifeature fusion flame detection algorithm using videotype judgment algorithm was proposed. In this algorithm, the videotype of each frame was first judged. Then candidate flame region was extracted by the algorithm that corresponds image types, at last the flame frequency feature based on DCT and image correlation feature between frames in the candidate region were detected to determine the existence of flame. The traditional visualization fire detection technology is basically based on analyzing the images from infrared cameras or visible camera to detect flame. The method can not meet the requirements of roundtheclock monitoring flame in engineering application. Experiments show that the proposed detection algorithm can resolve this problem, detect flame fast and reduce the error rate.

Key words: video type judgment, roundtheclock, fire detection, feature fusion, infrared video

中图分类号: 

  • TP391.41