Improved image light fog synthesis processing method based on atmospheric diffusion model

The invention discloses an improved image light and fog synthesis processing method based on an atmospheric diffusion model, and the method comprises the steps: calculating and generating a global transmissivity matrix corresponding to a to-be-processed image through a light and fog synthesis matrix algorithm, and inputting the to-be-processed image and the corresponding global transmissivity matrix into an image light and fog synthesis model, directly synthesizing and processing to obtain a light fog image corresponding to the to-be-processed image; therefore, an additional digital index step is avoided in the operation of a light fog synthesis matrix algorithm, an intermediate variable processing process is avoided, a global transmissivity matrix is directly formed by using a block matrix synthesis mode, and then a corresponding light fog image is directly synthesized by means of an image light fog synthesis model. According to the method, repeated operation of Euclidean distance and indexes in an algorithm