Image defogging method and system, computer equipment, storage medium and terminal

Abstract

The invention belongs to the technical field of image defogging, and discloses an image defogging method and system, computer equipment, a storage medium and a terminal.The image defogging method comprises the steps that a cumulative distribution function is constructed by counting the adjacent gray probability of a haze image, and the average threshold value of pixels is obtained according to the concentrated distribution degree of the gray levels of the pixels; according to the maximum connectivity of the sky area, obtaining a sky area and a non-sky area which are roughly segmented; the difference between pixel gray levels of the sky and the non-sky area is enhanced through guiding filtering, and accurate segmentation of the sky area and the non-sky area is achieved through the Otsu algorithm; the transmissivity of the sky area is estimated through logarithmic adaptive transformation, the transmissivity of the non-sky area is estimated through an improved dark channel prior algorithm, and image defogging is achieved by synthesizing the transmissivity of corresponding pixels of the sky area and the transmissivity of corresponding pixels of the non-sky area. The atmospheric light value Ac of the pixel channel can be accurately obtained by calculating the gray value of the pixel point, and the defect that a dark channel prior algorithm fails in bright areas such as the sky is overcome.

Description

technical field [0001] The invention belongs to the technical field of image defogging, and in particular relates to an image defogging method, system, computer equipment, storage medium and terminal. Background technique [0002] At present, the gray value of the sky area in the haze image is generally higher than that of other areas, resulting in a low estimate of atmospheric light in the existing dehazing algorithm, and serious color cast and loss of details in the sky area after dehazing. , which affects the image dehazing effect. [0003] In the haze environment, various particles suspended in the atmosphere will absorb and scatter light, resulting in the deflection of the light propagation path, resulting in the degradation of the image quality obtained by the imaging device, loss of details, and decreased contrast and visibility, which seriously affects the follow-up. Analysis and processing of images in the fields of scene analysis, photography, computer vision, and...

AI Technical Summary

Problems solved by technology

[0007] (1) Since the gray value of the sky area in the haze image is generally higher than that of other areas, the estimated value of atmospheric light in the existing dehazing algorithm is low, and the sky area has serious color cast and loss of details after dehazing And other problems, affecting the effect of image defogging

[0008] (2) In the existing single-threshold segmentation method using grayscale histogram, the grayscale histogram of the image must have a bimodal type, and there are obvious differences between the object to be extracted in the image and the background. In actual situations, the grayscale value of the image pixel, The difference in contrast is large, and it is difficult to determine the global threshold to separate the object from the background

[0009] (3) In the existing method of directly using the Otsu algorithm to segment images, when the area of ​​the target in the image differs greatly from that of the background, the histogram has no obvious double peaks or the size of the two peaks differs greatly, the segmentation effect is not good , the more gray levels involved in the spatial distribution of pixels when processing images, the lower the rationality of segmentation; at the same time, it is also quite sensitive to noise

[0010] (4) In the existing defogging methods that apply the dark channel prior to a single image, it is easy to fail when the gray level of the target pixel in the processing scene is close to the atmospheric light, resulting in the overall dark picture and noise amplification in the defogged image and other phenomena, the restored image is prone to color distortion, etc.

[0011] (5) The existing method of obtaining the transmittance through the analysis of the inherent boundary constraints and iterative optimization using the context regularization method, due to the limitations of the boundary constraints, when the fog image scene has diversity or the brightness range is uncertain, Causes halos to appear in the sky area and bright areas of the image after defogging

[0012] (6) The existing method of estimating image transmittance based on fuzzy lines and using context regularization method, the physical parameters are not accurate when processing a single image, and halos still appear in bright areas of the image

[0013] (7) The existing method of dehazing images with different exposures by segmenting the sky area of ​​the image combined with multi-scale fusion has inaccurate segmentation of the local sky area, and the dehazed image will have excessive brightness and amplified noise.

[0014] (8) In the existing defogging method through dark channel fusion and fog density weighting, the color saturation of some images is too high after defogging, the overall color is dark, and partial distortion occurs, and the quality of the sky area after defogging is not high

Images