Defogging method based on anisotropic filtering

Abstract

The invention relates to a defogging method based on anisotropic filtering, belonging to the field of image processing and computer vision, and providing the defogging method based on anisotropic filtering which can improve the visual effect of images taken in a foggy day, a rainy day and the like. The method comprises the following steps: 1) calculating an atmospheric light value through image guide; 2) utilizing an anisotropic filter to calculate a transfer function diagram through image guide; 3) dynamically and self-adaptively adjusting a lower threshold of defogging; and 4) according to an observed foggy image, the atmospheric light value and the atmospheric light value diagram in an atmospheric scattering physical model, recovering a scene image. The method can self-adaptively process various images taken in the foggy day, the rainy day and the like on the basis of the atmospheric scattering physical model, the defogged images have the ideal contrast and visual effect, and the overall enhancement effect is superior to the traditional image defogging method.

Description

technical field [0001] The invention relates to the fields of image processing and computer vision, in particular to a method for defogging based on anisotropic filtering. Background technique [0002] Many outdoor users of computer vision systems, such as urban traffic, video surveillance, image recognition, etc., require robust detection of image features. However, the contrast of images collected under conditions such as foggy and rainy days is very low, and the visibility is very poor, which has caused certain obstacles to the extraction of image information. Therefore, removing the influence of rain and fog on captured images will bring about effective improvements in convenience and safety to the society. [0003] In the field of computer vision, at present, the fog degraded image restoration algorithm based on the dark channel prior and the transfer function estimation using the Guided Filter filter, and the fog degraded image restoration algorithm based on the estim...

AI Technical Summary

Problems solved by technology

The fog degraded image restoration algorithm based on the dark channel prior and using the Guided Filter to estimate the transfer function can effectively estimate the transfer function, but there is a deviation in the restoration of the scene color, and the halo phenomenon appears in the sky; the estimation based on the atmospheric film The degraded image restoration algorithm in foggy weather restores the color better, but the sky area appears gray, and the visual effect is not good

[0004] Chinese patent 201010139441.1 discloses an automatic image defogging method based on dark channels. This method uses dark channel priors to obtain transmission images, and multi-scale Retinex to obtain brightness component images. The processing speed is slow, and the lower threshold of the transfer function cannot be dynamically adjusted. Adaptive adjustment, the presence of halos in the processed sky area

Images