Fusion Enhancement Of Underwater Images Based On Multi-scale Retinex Theory

With the rapid development of economy and the continuous growth of population,the existing resources on the land are becoming increasingly scarce,which can not meet the needs of human beings.Therefore,Research scholars and researchers at home and abroad gradually shift their attention to the exploration and development of Marine resources.Therefore,the degree of understanding of underwater information largely determines the degree of development of the ocean.Due to the physiological conditions of human beings,we can not go underwater for a long time,which requires us to obtain underwater information through other ways.Optical imaging is one of the most effective and direct ways at present.Due to the optical characteristics of water medium,the image acquired under water is degraded seriously,and the problems of color bias,low contrast and unclear details are common.The main purpose of this paper is to enhance the underwater image,enhance the contrast,highlight the edge details of the image,and restore the real situation of the underwater target.The specific content of this paper is as follows:1.Analyzes the characteristics of the transmission of light in the water,learned that the cause of underwater image degradation is water dielectric absorption and scattering of light,underwater imaging model is the absorption and scattering of light by the water medium,we studied the several classic underwater image enhancement algorithms,through simulation experiment resembles the various algorithm enhanced the results and analyzes the advantages and disadvantages of various algorithms.2.In view of the degradation of underwater image quality and the advantages and disadvantages of several classical algorithms,this paper proposes an underwater image fusion enhancement algorithm based on the multi-scale Retinex theory.First,perform CLAHE processing on the original image taken to improve the overall brightness and contrast of the image.Then,the preprocessed image was decomposed by wavelet transform to obtain the high and low frequency components of the image.The low frequency components were processed by the improved MSRCR algorithm of guided filtering to improve the brightness of the image and remove part of the noise.Then the bilateral filtering algorithm is used to de-noise the high frequency information,and finally the enhanced image is reconstructed by the inverse wavelet transform of the processed high and low frequency components,which effectively improves the problems existing in the underwater image.The experimental platform of underwater imaging is built,and the algorithm proposed in this paper is compared with several other classical underwater image algorithms.The enhancement effect of underwater images in this paper is better,which can effectively enhance the contrast and detail information of underwater images.