Rapid defogging method and system for single image

Abstract

The invention relates to the technical field of image processing, and provides a rapid defogging method and system for a single image. The method comprises the following step: carrying out minimum value filtering on an inputted fogged image I(x), calculating to obtain an initial dark primary color image M(x), calculating an atmospheric transmission function t(x) according to the calculated initial dark primary color image M(x), combining a projection method to calculate an atmospheric light value A, and recovering a fog-free image J(x) according to an atmospheric scattering module of the fogged image and the calculated parameters of the atmospheric light value A and atmospheric transmissivity. Therefore, the fogged image is rapidly defogged, the details of the defogging effect are clear, the color is real and natural, the method is extensively applicable, the balance between the recovery quality and the recovery speed of the image is guaranteed, and the method can be applied to an occasion operating in real time.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method and system for quickly defogging a single image. Background technique [0002] In the haze scene, the scattering effect of atmospheric particles will seriously degrade the information captured by the optical sensor, and the contrast and color fidelity in the image will be attenuated to varying degrees, which directly affects the perception of human vision and the performance of the machine vision system. Therefore, it is of great significance to study image defogging methods. [0003] At present, the single image dehazing algorithm based on assumptions or prior knowledge has become a research hotspot. For example, based on the prior knowledge that the haze-free image has a higher contrast than the fogged image, Tan realizes image dehazing by maximizing the contrast of the restored image. Dehazing, but the algorithm is prone to oversaturation of the dehazing ima...

AI Technical Summary

Problems solved by technology

[0003] At present, the single image dehazing algorithm based on assumptions or prior knowledge has become a research hotspot. For example, based on the prior knowledge that the haze-free image has a higher contrast than the fogged image, Tan realizes image dehazing by maximizing the contrast of the restored image. Dehaze, but the algorithm is prone to oversaturation of the dehaze image and color distortion; Fattal assumes that the object surface chromaticity and medium propagation have local statistical irrelevance, and uses independent component analysis to estimate the scene albedo

This method can produce a good defogging effect when the fog is thin, but for dense fog images, the estimation will fail due to insufficient color information; He et al. proposed a method based on dark channel A priori dehazing method, the algorithm uses the minimum value filter to estimate the medium propagation function and optimizes the estimation to achieve a better dehazing effect; Tarel et al. use the median filter method to estimate the atmospheric dissipation function, and then apply Tone mapping obtains the dehazed image, and the image restored by the algorithm is clear and realistic, but it is easy to cause the Halo effect at the sudden change of the depth of field; Kratz and Nishino assumed that the foggy image is composed of two independent layers of scene albedo and scene depth, and Using the FMRF idea to model, the more accurate depth information is finally obtained, but the color of some pixels in the image processed by this method is too saturated

[0004] Among the above algorithms, the fast defogging method for a single image proposed by He et al. has been studied by more people because of its simple principle and excellent effect. However, this method uses Softmatting to refine the transmittance, resulting in slow calculation speed.

Although a variety of methods to speed up the refinement of transmittance appeared later, such as guided filtering, bilateral filtering, anisotropic filtering, edge-preserving filtering, median filtering, etc., which improved the calculation speed to a certain extent, they still cannot be applied to real-time in the system

In addition, because the sky area and white objects do not satisfy the dark primary color assumption, there are block effects or serious color distortion in the restored image, which affects the overall visual effect of the image.

Images