Training method and device of image defogging model suitable for underground coal mine

Abstract

The invention provides a training method and device suitable for an underground coal mine image defogging model, and relates to the technical field of image processing. The method comprises the following steps: acquiring a first dust fog sample image in a dust fog environment and a fog-free sample image in a non-dust fog environment; carrying out fogging processing on the fogless sample image to generate a fogged image; acquiring a depth information graph of the first dust fog sample image; obtaining a second dust fog sample image based on the depth information graph and the fog adding image; and training an image defogging model based on the first dust fog sample image and the second dust fog sample image to generate a target image defogging model. In the embodiment of the invention, aiming at the characteristics of fuzzy monitoring image, low contrast and unclear details in dust fog environments such as underground coal mines, the problems of high fog dust concentration and non-uniform illumination are solved by adopting a deep learning method, the definition of the image can be improved, and visual observation and recognition feature extraction are facilitated.

Description

technical field [0001] The present application relates to the field of image processing, and in particular, to a training method and a device for an image dehazing model suitable for underground coal mines. Background technique [0002] With the continuous popularization of intelligent video surveillance systems and digital image acquisition equipment, the use of image technology for safe production plays an important role in the development of the coal mine industry. In the related art, the lighting in coal mines is uneven. Compared with outdoor light, the illumination is very low. The limitation of point light sources makes the image brightness distribution very uneven. When the method of dust reduction is used to reduce dust, the dust is reduced after spraying, but there is a lot of fog in the air, which will seriously affect the quality of the captured image, which may cause the captured image to be blurred, unclear, and overall dark. , which is not conducive to the ext...

AI Technical Summary

Problems solved by technology

In the related technology, the underground illumination of the coal mine is not uniform. Compared with the outdoor light, the illumination is very low. The limitation of the point light source makes the image brightness distribution very uneven. In addition, a large amount of dust will be generated during the coal mining work in the coal mine. When the method of dust reduction is used to reduce dust, the dust is reduced after spraying, but there is a lot of fog in the air, which will seriously affect the quality of the captured image, which may lead to defects such as blurred, unclear details, and overall darkening of the collected image , which is not conducive to the extraction of visual observation and recognition features, which brings a lot of troubles to the application of image technology

Images