主管:中华人民共和国应急管理部
主办:应急管理部天津消防研究所
ISSN 1009-0029  CN 12-1311/TU

Fire Science and Technology ›› 2020, Vol. 39 ›› Issue (2): 247-251.

Previous Articles     Next Articles

Recognition and extraction of flames based on visible light images

GE Jun1, WANG Li2, LI Wei2,3, ZHANG Bo3,4   

  1. (1.Huzhou Special Equipment Testing Research Institute, Zhejiang Huzhou 313000, China;2. School of Mechatronic Engineering, China University of Mining and Technology, Jiangsu Xuzhou 221006, China;3. Urban Emergency Technology and Equipment Research Center, China University of Mining and Technology, Jiangsu Xuzhou 221006, China;4. School of Computer Science and Technology, China University of Mining and Technology, Jiangsu Xuzhou 221006, China)
  • Received:2019-09-23 Online:2020-02-15 Published:2020-02-15

Abstract: The basic characteristics of flames were analyzed, and several representative flame characteristics were selected as criteria to identify and extract flame areas in the fire field image. Based on the luminance characteristics of flame, the suspended flame area was divided by region growing method. According to the corresponding relation of RGB and HIS color model, and by separating the R, G, B component of flame pixels, the color distribution rules of flame were verified. The interfere of regular light source was excluded by roundness criterion.

Key words: flame recognition, flame color characteristics, region growing method, segmentation of suspected flame regions, flame shape characteristics