Dust removal and smoke rapid recovery method for high-definition image under mine

Abstract

The invention discloses a dust removal and smoke rapid recovery method for a high-definition image under a mine. The method comprises the steps of obtaining a to-be-processed image; calculating a dark primary color image of the to-be-processed image to generate a coarse transmissivity graph; performing image edge processing on the to-be-processed image by using an HED edge detection algorithm to generate an edge enhancement guide graph; superposing the edge enhanced guide graph and the coarse transmissivity graph to generate an edge enhanced transmissivity graph; and using a dust fog degradation model to carry out image restoration on the edge enhanced transmittance image to generate a dust-removed smoke image. According to the technical scheme, the problem that the operation time of a dark channel prior principle on a high-resolution image is greatly increased is solved through a scaling interpolation method, the operation efficiency of a dark channel prior theory algorithm on a high-definition image is effectively improved, and part of edge features of a dust smoke image are enhanced; and the actual working effect of dust and smoke removal treatment of the high-definition image is improved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a fast recovery method for removing dust and smoke from high-definition images in a mine. Background technique [0002] Underground images are affected by dust, smoke, low illumination and uneven illumination distribution, resulting in blurred images, submerged image features, and reduced image quality. Among them, different dusts are very different in size, shape and composition, but no matter what kind of dust and smoke will scatter and absorb the incident light, which will weaken the color and contrast of the image, cause image degradation, and cause many useful features to be covered. , hindering the promotion and application of machine vision relying on image feature recognition in dusty environments. [0003] There are many research algorithms for image dehazing, but they are mainly divided into two categories: dehazing algorithms based on image enhancement and de...

AI Technical Summary

Problems solved by technology

In practical applications, when processing high-resolution images to remove dust and smog, when using an image enhancement algorithm based on dark channel prior to remove image dust and smog, the image processing operation time will increase significantly. With the improvement of camera resolution, resulting in The original calculation process for removing dust and smoke is not applicable, practical, or usable

At present, the dark channel prior method is a statistical result, mainly the statistical result of a large number of outdoor fog-free photos. If there is dust, smoke and dim light in the target scene, it will not be possible to obtain a satisfactory result because the preconditions are not established. Effect

[0004] In practical applications, when processing high-resolution images (high-definition images above 960P) to remove dust and smog, the image processing operation time will be large when using an image enhancement algorithm based on dark channel prior (dark channel prior) to remove image dust and smog. With the increase of the camera resolution, the original dust and smoke calculation process is not applicable, practical, and unusable.

[0005] Traditional edge detection methods use first-order or second-order differential operators for detection, such as Sobel, Prewitt, Canny, LOG, etc. These differential operators only consider local sharp changes, especially changes in color, brightness, and gradient. To detect the edge, there are low-level features that are difficult to reflect, and more complex scenes, such as the internal texture of the object, the background interference in the image, etc., the scene effect is not good.

Images