A fast smog removal method for a single image based on independent component analysis

Abstract

The invention discloses a fast haze removal method for a single image based on independent component analysis. After morphological erosion of smog degraded image, In view of the fact that the non-fogimage should satisfy the prior statistical law of dark channel of the sunny day image, A smog propagation characteristic surface is generated according to the dark channel map, and then the color space components of the smog image and the smog propagation characteristic surface are fused and de-mixed as independent components based on independent component analysis to obtain a clear image after smog is eliminated. The haze atmospheric particle scatter model is taken as a processing basis, the haze removal method generate a haze atmospheric particle propagation characteristic curved surface, which has a correct theoretical model basis; Independent component analysis (ICA) is used to separate smog component from smog-free color space component, which eliminates time-consuming operations suchas fine soft matting and image segmentation, and greatly improves the computational efficiency. Aiming at the smog degraded image, the smog removal effect is good, the image details are enhanced significantly, and the color fidelity is high.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a method for quickly removing haze from a single image based on independent component analysis. Background technique [0002] In hazy weather, due to the scattering effect of atmospheric particles, on the one hand, part of the reflected light on the surface is lost due to scattering, which reduces the light intensity reaching the observation point, and decays exponentially with the increase of the propagation distance; on the other hand, The scattering effect of atmospheric particles also comes from the atmospheric light attached to the target image, so that the atmospheric light shows the characteristics of the light source, and the intensity of the ambient light gradually increases with the increase of the propagation distance; the effects of the above two aspects lead to the haze image Contrast, color and other features attenuate significantly. [0003] Th...

AI Technical Summary

Problems solved by technology

[0004] In view of the above-mentioned deficiencies in the prior art, the single image haze fast removal method based on independent component analysis provided by the present invention solves the lack of theoretical model support and the need for model-based haze removal methods in non-model-based haze removal methods. Carry out large-scale sparse matrix operations, problems with high time complexity and space complexity

Images