基于颜色校正和加权融合的水下图像增强

Underwater image enhancement based on color correction and weighted fusion

为解决水下图像由光照不均匀等原因引起的颜色偏差、对比度低、细节不清晰等问题,提出了一种基于颜色校正和加权融合的水下图像增强方法。首先,对补偿了红色通道的原始图像采用白平衡算法进行颜色校正,解决图像失真问题。并将颜色校正后的图像从RGB空间转换到Lab空间,然后采用对比度限制自适应直方图均衡比(CLAHE)算法处理的L亮度通道来增强图像的对比度和亮度。最后,将颜色校正后的图像和增强对比度后的图像进行加权融合增强图像细节的清晰度。实验结果证明,采用所提算法处理不同水下场景图像效果显著,4个图像质量评价指标信息熵、峰值信噪比(PSNR)、水下图像质量评价度量(UIQM)和水下彩色图像质量评级(UCIQE)值与其他算法相比均有明显提升。并且与原始图像相比,所提算法处理后的图像的信息熵至少提升了5.6%,UIQM值提高了1.48倍以上,UCIQE值提高了7.5%。

To solve the problems of color deviation,low contrast,and unclear details of underwater images caused by uneven illumination,this paper proposes an underwater image enhancement method based on color correction and weighted fusion.Firstly,the original image with red channel compensation is corrected by the white balance algorithm to solve the image distortion problem.Then the color-corrected image was converted from RGB space to Lab space,and the L brightness channel was processed by the CLAHE algorithm to enhance the contrast and brightness of the image.Finally,the image after color correction and the image after contrast enhancement is weighted and fused to enhance the clarity of image details.The experimental results show that the algorithm proposed in this paper is effective in processing images of different underwater scenes,and the values of information entropy,PSNR,UIQM,and UCIQE of the four image quality evaluation indexes are improved compared with other algorithms.Compared with the original image,the information entropy of the image processed by this algorithm is increased by at least 5.6%,the UIQM value is increased by more than 1.48times,and the UCIQE value is increased by 7.5%.