Fire video smoke identification method based on time-space domain dual channels

The invention provides a fire video smoke identification method based on time-space domain dual channels. The method comprises the following steps: collecting and making a smoke data set containing cloud and fog interference; establishing a static feature extraction network and a dynamic feature extraction network, fusing and connecting the static feature extraction network and the dynamic feature extraction network, and constructing a video smoke recognition network model; training a video smoke identification network model by using the smoke data set to obtain an optimized video smoke identification network model; the optimized network model is utilized to process a smoke video collected in real time, the static feature extraction network extracts static features of an image in a spatial domain, the dynamic feature extraction network extracts dynamic features of the video in a time domain, the static features and the dynamic features are fused to generate smoke features, and the smoke features are extracted