Underwater image enhancement method based on dark channel prior and white balance

Abstract

The invention relates to an underwater image enhancement method in the field of digital image processing. It aims to better realize image enhancement by modifying the background light, improve image clarity and contrast, and achieve the purpose of image optimization. The technical solution adopted by the present invention is an underwater image enhancement method based on dark channel prior and white balance. The steps are as follows: A. Background light preprocessing stage B. Dark channel prior stage. The first step of the dark channel prior stage is to first use dark channel prior. The channel prior algorithm obtains the refined transmittance t(x) after matting and the image after dehazing: the second step of the dark channel prior stage is to derive the depth map d(x) from the transmittance t(x); dark The third step of the channel prior stage is to obtain the desired mask; the fourth step of the dark channel prior stage is to extract the area of ​​interest; C white balance stage. The invention is mainly used in underwater image enhancement situations.

Description

technical field [0001] The invention relates to an underwater image enhancement method in the field of digital image processing, in particular to an underwater image enhancement method based on dark channel prior and white balance. Background technique [0002] With the continuous growth of population and the rapid development of economy, land resources can no longer meet the growing needs of people, and marine technology has increasingly become the focus of economic and technological competition around the world. Underwater image enhancement is an important part of underwater detection, and it also provides strong support for the study of marine related subjects. However, in the process of underwater imaging, due to the absorption of light by water bodies and the scattering of light by suspended particles, underwater images often have insufficient resolution and strong attenuation, resulting in color distortion in underwater images and a sharp drop in contrast. , and the l...

AI Technical Summary

Problems solved by technology

However, in the process of underwater imaging, due to the absorption of light by water bodies and the scattering of light by suspended particles, underwater images often have insufficient resolution and strong attenuation, resulting in color distortion in underwater images and a sharp drop in contrast. , at the same time, the low visibility underwater will blur the image, and the underwater image quality is often poor

Some of today's underwater image enhancement algorithms are either aimed at the color degradation caused by different attenuation of light of different wavelengths when it is transmitted underwater, or for the blurring effect of underwater images caused by the scattering of water molecules and suspended particles. At the same time, some algorithms still have problems of low practicability, such as: image acquisition equipment is expensive, a large number of input pictures are required, and the processing method does not consider the particularity of the underwater environment, etc. Therefore, the underwater image enhancement algorithm still needs further in-depth research

[0003] In 2006, Bazeille et al. [1] proposed a new type of image preprocessing algorithm, which consists of multiple filtering steps. First, high-pass homomorphic filtering is performed on the degraded image, and then wavelet transformation is used to remove noise. Anisotropic filtering further removes noise, then stretches the brightness of the image to a range of 0-255, and finally uses the average intensity of each color channel to equalize the color of the image, but this algorithm ignores the particularity of the underwater environment and takes a long time , not suitable for underwater image processing

In 2007, Hou et al. [2] combined the point spread function and the modulation transfer function, and used it to remove the blurring effect of underwater images, but the image acquisition equipment used by this algorithm is relatively complicated, and it is difficult to apply it in practice

Images