Research on Fire Recognition Method in Large Indoor Space Based on Edge Gradient Features

With the construction of domestic airports, railway stations, large warehouses and other large buildings, the fire protection problem of large indoor spaces has become more and more prominent. Indoor large-space fire monitoring is increasingly carried out by video fire detection. Compared with traditional detection methods, this method has the advantages of fewer installations, larger detection space, and faster response speed. This paper aims at the problem of difficult fuel fire identification under special interference sources such as car lights and searchlights that may exist in the monitoring site. Based on the analysis of existing fire identification algorithms, the difference between the edge gradient features of the fire image and the interference source is used for fire image recognition. It eliminates the interference of the above-mentioned interference sources on fire identification, and improves the accuracy of fire identification. T.H. Chen’s image color.