Foggy day image reconstruction method and system

Abstract

The invention discloses a foggy day image reconstruction method and system; the method comprises the following steps: obtaining target polarization images of a plurality of different polarization angles under foggy day weather; determining the original foggy day target image and atmosphere light polarization images respectively corresponding to the target polarization images one by one according to the plurality of target polarization images; determining an atmosphere light polarization degree image and an atmosphere light intensity difference value according to the plurality of atmosphere light polarization images; determining an infinite distance atmosphere light intensity value image according to a bright channel prior rule; finally using the original foggy day target image, the atmosphere light polarization degree image, the atmosphere light intensity difference value and the infinite distance atmosphere light intensity value image to determine a fog-free target image. The method and system can solve the problems that a conventional method only consider the atmosphere light polarization characteristics, cannot consider the target light polarization characteristics, thus reducing the image sharpness.

Description

technical field [0001] The invention relates to the technical field of target detection, in particular to a foggy image reconstruction method and system. Background technique [0002] Most outdoor vision systems need to extract image features clearly and accurately, but vision systems are very sensitive to weather conditions, especially in foggy conditions. Because the light energy radiated by scene points is scattered by atmospheric particles, the performance of the system is greatly reduced or even unable to work, and the characteristics of the collected images are severely attenuated, which greatly limits and affects the effectiveness of the system. Therefore, in order to make the system robust and reliable in bad weather, it is necessary to dehaze the degraded images in foggy weather. [0003] The reasons for image degradation in foggy weather mainly include two aspects: one is that the light intensity of the target itself is absorbed and scattered by suspended particle...

AI Technical Summary

Problems solved by technology

[0004] In the prior art, although a lot of research has been done on reconstructing fog-free images from foggy images, only the polarization characteristics of atmospheric light are considered when reconstructing fog-free target images, and the polarization characteristics of target light are not considered, thus lead to certain errors in the determination of atmospheric light

Images