近年来,随着城市化进程的飞速发展,火灾事故频频发生,导致经济和能源大量损失。为了避免和降低火灾事故的发生,火灾探测系统应运而生。传统感温感烟探测器有着在大空间建筑中探测距离短、受环境因素影响大、误报率高的缺点,使其应用受到限制。可视火灾烟雾探测系统不受传统感温感烟探测器缺陷的限制,这种类型的检测系统,使用摄像头捕捉图像序列的监控区域,通过电脑来处理序列图像,使用机器视觉方法来决定是否有 In recent years, with the rapid development of urbanization, fire accidents occur frequently, resulting in huge economic and energy losses. In order to avoid and reduce the occurrence of fire accidents, fire detection system came into being. The traditional temperature smoke detectors have the shortcomings of short detection range, large influence on environmental factors and high false positive rate in large space buildings, which limits their application. The visual fire smoke detection system is not limited by the defects of the traditional thermal smoke detector. This type of detection system uses a camera to capture the surveillance area of ​​the image sequence, processes the sequence image through a computer, and uses a machine vision method to determine whether there is a