基于结构纹理融合的水下图像增强算法  

Underwater image enhancement algorithm based on structure-texture fusion

作　　者：马悦[1] MA Yue(Information Construction Management Division,Shaanxi University of Chinese Medicine,Xianyang 712046,Shaanxi Province,China)

机构地区：[1]陕西中医药大学信息化建设管理处,陕西咸阳712046

出　　处：《信息技术》2022年第12期102-106,共5页Information Technology

摘　　要：针对水下环境中采集的图像存在颜色失真、雾化现象等问题,提出一种基于结构纹理融合的水下图像增强算法。首先,采用自动色彩增强方法对图像进行颜色校正;然后,结合水下图像的特点,通过二次引导滤波将图像分解为低频结构层和高频纹理层;其次,利用红色暗通道先验来非局部消除雾化对低频层进行增强操作,同时利用最小化梯度残留的方法对高频层进行去噪处理;最后,将增强后的结构层与去噪后的纹理层相融合,得到清晰的结果。实验表明,该方法能够有效增强图像质量,同对比算法在主观和客观评价上均有显著提升。In order to solve the problems of color distortion and atomization in the underwater complex environment,an underwater image enhancement algorithm based on structure-texture fusion is proposed.Firstly,the automatic color enhancement method is used to correct the color of the image,then,combined with the characteristics of underwater image,the image is decomposed into low-frequency structure layer and high-frequency texture layer by secondary guided filtering.Secondly,the red dark channel prior is used to nonlocally eliminate the atomization to enhance the low-frequency layer,and the method of minimizing gradient residue is used to denoise the high-frequency layer.Finally,the enhanced structure layer is fused with the denoised texture layer to get a clear result.Experiment results show that the proposed method can effectively enhance the image quality,and has significant improvement in subjective and objective evaluation compared with other algorithms.

关 键 词：水下图像增强 图像分解 图像融合 结构纹理层 暗通道先验 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]