A defogging method based on image quality evaluation optimization

Abstract

The invention discloses a defogging method based on image quality evaluation optimization. The defogging method comprises the following steps: 1) obtaining a series of preliminary candidate defoggingimages by roughly valuing transmissivity; 2) calculating quality evaluation function values of a series of preliminary candidate defogged images, and selecting the candidate defogged image corresponding to the maximum value as a preliminary estimation result; 3) obtaining a series of accurate candidate defogged images according to the small-range value of the transmissivity; and 4) calculating a quality evaluation function value of each accurate candidate defogged image, and selecting the candidate defogged image corresponding to the maximum value as an optimal defogged result image. Accordingto the invention, a brand new image quality evaluation function is established; and on the basis, an optimal result is selected from the candidate defogged images as a defogged result image, so thatthe method can be applied to real-time defogged processing of the foggy unmanned aerial vehicle image, and particularly, a relatively good defogged effect can be generated for a foggy unmanned aerialvehicle forward shot image.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and relates to an optimal defogging method based on image quality evaluation. Background technique [0002] Image defogging methods have always been a research hotspot in recent years. Image defogging methods based on atmospheric scattering models have been vigorously developed, but unmanned aerial vehicle image defogging is still an unsolved problem. In recent years, UAV aerial photography, as a new observation method, is becoming more and more mature, which also puts forward higher requirements for the dehazing method and dehazing efficiency of UAV images. [0003] UAV images have their own characteristics compared with ordinary images: when the UAV image sensor shoots vertically from high altitude to the ground, the depth of all scene points on the acquired image is almost the same and equal to the flight height, because the flight height is mostly in the For hundreds of mete...

AI Technical Summary

Problems solved by technology

[0002] The image defogging method has always been a research hotspot in recent years. The image defogging method based on the atmospheric scattering model has been vigorously developed, but the unmanned aerial vehicle image defogging is still an unsolved problem.

Images