基于多尺度融合的水下图像对比度增强算法  

A Contrast Enhancement Algorithm for Underwater Images Based on Multi-Scale Fusion

作　　者：江巨浪 张高兴 纪娟娟 刘娟 JIANG Julang;ZHANG Gaoxing;JI Juanjuan;LIU Juan(School of Electronic Engineering and Intelligent Manufacturing,Anqing Normal University,Anqing 246133,China)

机构地区：[1]安庆师范大学电子工程与智能制造学院,安徽安庆246133

出　　处：《安庆师范大学学报(自然科学版)》2024年第4期62-66,共5页Journal of Anqing Normal University(Natural Science Edition)

基　　金：安徽省自然科学基金(2108085MF196,2108085QD182);安徽省高校重点科研项目(2023AH050474,2022AH051038)。

摘　　要：针对水下图像存在的颜色偏移、能见度低与细节模糊等问题,本文提出了一种多尺度融合的对比度增强算法。对于颜色偏移,采用基于最小颜色损失的自适应校正算法来进行色彩校正。对于全局对比度退化与局部细节模糊,首先分别采用归一化的幂律变换方法与自适应局部直方图均衡化方法以进行复原处理,然后采用拉普拉斯金字塔对以上两种复原图像进行多尺度融合。仿真实验结果表明,该算法能够正确复原水下图像颜色,与其他算法相比,进一步改善了图像全局对比度,并显著提高了图像局部细节可见性。A global and local contrast collaborative enhancement algorithm for underwater images is proposed to address the issue of color deviation,low visibility,and detail blur.For color deviation,an adaptive correction algorithm based on minimum color loss is adopted.To tackle global contrast degradation and local detail blur,normalized power law transformation methods and adaptive local histogram equalization methods are used for restoration,respectively.Finally,the Laplace pyramid is used to perform multi-scale fusion of the above two restored images.Simulation results indicate that this method can correctly restore the color of underwater images and further improve the global contrast of the image compared to other algorithms,significantly enhancing the visibility of local details.

关 键 词：图像对比度增强 水下图像 颜色校正 多尺度融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]