基于HSV改进的Retinex算法的图像去雾研究  

Research on Image Dehazing Based on Improved HSV Retinex Algorithm

作　　者：马潇菲 MA Xiaofei(Teaching and Development Center,Fuzhou University of International Studies and Trade,Fuzhou 350202,China)

机构地区：[1]福州外语外贸学院教学发展中心,福州350202

出　　处：《景德镇学院学报》2024年第3期53-57,共5页Journal of JingDeZhen University

摘　　要：针对传统Retinex算法在处理雾化图像时面临的计算量大和耗时长等问题,文章提出了一种改进的Retinex算法。首先,将传统的颜色空间模型转换为HSV颜色空间模型进行计算。然后,在处理过程中保持色调H不变,最后,对亮度分量V进行Retinex去雾处理,从而显著降低图像去雾算法的数学计算量。通过与多尺度(MSR)的Retinex算法对比,测试基于HSV改进的算法在示例雾化图像上的去雾性能。实验结果表明,相较于MSR算法,基于HSV算法在去雾效果和处理时间上均有所提升,其中局部雾化样本与整体雾化样本的去雾效率分别提高了158%和162%,且亮度分布较为均匀。An improved Retinex algorithm is proposed to address the issues of computation-intensive and time-consumption in traditional Retinex algorithms for atomized images.Firstly,the traditional color space model is converted into the HSV color space model for calculation;Then,the color tone H is kept unchanged during the processing;and finally,the brightness component V is subjected to Retinex dehazing processing,significantly reducing the mathematical computation of the image dehazing algorithm.The dehazing performance of the improved algorithm based on HSV is test on example hazy images,compared to the MSR algorithm.The results show that the improved HSV algorithm exhibits better dehazing effect,shorter processing time,with dehazing efficiencies for local and overall hazy samples improving by 158% and 162%,respectively,as well as relatively uniform brightness distribution.

关 键 词：RETINEX 颜色空间转换 图像滤波 图像去雾 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]