Research On Underwater Image Enhancement Method Based On Pyramid Fusion

Underwater images are the principal means for humans to obtain marine information.Still,the particular physical and chemical characteristics of the underwater environment have severely affected the quality of underwater images.On the one hand,water absorbs different levels of light in different spectral regions,which often causes color distortion in underwater images.On the other hand,the scattering of light by fine particles suspended in water results in underwater images with low contrast and poor visual effects.Therefore,an underwater image enhancement method based on fusion technology is proposed for the phenomenon of marine image color reduction,low reduction,and poor visual effects.The main research content of this article includes the following aspects:(1)For the problem of color distortion of underwater image,the image is corrected by an automatic white balance algorithm based on dynamic thresholds.At the same time,in order to solve the problem of poor processing effect for image with large back scattering,color correction algorithm based on mean and mean square error is applied to further improve image color.The color of the processed image is greatly improved.(2)To better balance the global and local contrast of the image,a contrast enhancement method based on guided filter image layering is proposed to process the color-corrected image.That is,the color-corrected image is first decomposed into a base layer representing the low-frequency component of the image and a detail layer representing the high-frequency component of the image and then processed separately.The basic layer is converted to Lab space,and the CLAHE algorithm is used to process the L channel in the Lab space,which can not only improve the image contrast,but also reduce the impact on the color components of the image.The texture details are enhanced.At the same time,a weighted fusion algorithm based on local variance and global variance is used for the enhanced base layer and detail layer to obtain a contrast-enhanced image.After processing by the above method,the contrast of the image significantly improved.(3)Based on the idea of image pyramid fusion,the four weighted images such as exposure and salience are calculated for the color-corrected image and the contrast-enhanced image,and then they are normalized,and then the normalized weight maps are decomposed by the Gaussian pyramid.The color-corrected image and contrast-enhanced image are decomposed by the Laplacian pyramid.At the same time,in order to improve the fusion efficiency,only the top image representing the low-frequency contour information of the image and the bottom image representing the high-frequency detail information of the image are applied for fusion,the corresponding level images of the contrast-enhanced image are adopted for the images at other levels.Finally,the image is reconstructed to obtain the enhanced result image.The enhanced image looks more realistic and natural,and the visual effect is better.The experimental results show that the algorithm of this paper has achieved good results in four aspects of underwater recovery images in different scenes,such as subjective visual angle,objective quality indicators,color restoration effect,and feature point detection,which can effectively improve the original image color and contrast,improve image visuals.