基于双域协同驱动的水下图像增强

Underwater Image Enhancement Based on Dual-Domain Collaborative Drive

摘要:

水下捕获的图像易受到光的吸收与散射影响，导致水下图像存在颜色失真、对比度降低和细节模糊等问题。针对上述退化现象，提出一种双域协同驱动的水下图像增强方法。该方法首先在 CIELab 空间对 a、b 颜色通道进行中心化处理，并结合标准差约束拉伸色彩分布，以校正颜色失真问题。随后，借助像素亮度的累计直方图对全局亮度分布进行重映射，并引入基于局部统计的反馈机制，根据区域的曝光状态动态调节增强强度，实现对比度的自适应增强。实验结果表明，该方法在颜色校正、对比度提升和细节增强方面均表现出优异的效果。

Abstract:

Images captured underwater are susceptible to the absorption and scattering of light，resulting in problems such as color distortion，reduced contrast，and blurred details in underwater images. In response to the above-mentioned degradation phenomenon，we propose a dual-domain cooperative driven underwater image enhancement method. This method first centralizes the a and b color channels in the CIELab space. It combines the standard deviation constraint to stretch the color distribution to correct the color distortion problem. Subsequently， the global luminance distribution is remapped using the cumulative histogram of pixel luminance，and a feedback mechanism based on local statistics is introduced to dynamically adjust the enhancement intensity according to the exposure state of the region，achieving adaptive contrast enhancement. The experimental results show that this method exhibits excellent effects in color correction，contrast improvement，and detail enhancement.