An image defogging method based on dark channel information

Abstract

The invention discloses an image defogging method based on dark channel information, which belongs to the technical fields of image processing and computer vision. The method proposed by the present invention first uses the minimum value filter to calculate the minimum value of each pixel for each color channel of the input image, and then calculates the statistical value of the dark channel of the image under different scale parameters, and calculates the dark channel statistics corresponding to each scale parameter Gaussian smoothing filtering is performed on the value, the filtered dark channel statistical value is assigned different weight values ​​according to different scale parameters, and the transmission coefficient of the scene is calculated by optimizing the weighted channel statistical value to realize the dehazing process of the image. The application of the present invention can avoid complex soft matting optimization steps and computational complexity of dehazing processing, and the image quality of dehazing processing is high, and can meet the requirements of real-time processing applications.

Description

technical field [0001] The invention belongs to the technical field of image processing and computer vision, and in particular relates to a single image defogging processing method based on dark channel information. Background technique [0002] Foggy image restoration is a key technology in the technical fields of image processing and computer vision. Because the classic computer vision algorithms often assume that the input image is the reflected light on the surface of the object, for some foggy images, the mathematical models of these methods will inevitably fail to a certain extent, thus affecting the final image defogging effect. Therefore, it is of great significance to study the imaging degradation model and dehazing method of foggy images. [0003] For a single foggy image, the existing image defogging methods are usually based on the prior information rules of the foggy image. Typical methods are Tan method (local contrast maximization method), Fattal method (sce...

AI Technical Summary

Problems solved by technology

Although the dehazing quality of the image recovered by the He method is relatively ideal, only one scale parameter is considered, and the calculation is extremely complex and the soft matting optimization method is used to estimate the transmission coefficient of the scene, and the calculation amount will appear as the image size increases. The geometric level increases, and it is difficult to achieve real-time processing of fog image removal on the basis of existing hardware, and it is difficult to apply it to practical applications

Images