基于天空区域分离和自适应中值滤波的偏振图像复原方法  

Polarization Image Restoration Method Based on Sky Region Separation and Adaptive Median Filtering

作　　者：彭文竹[1] PENG Wenzhu(Chengyi College,Jimei University,Xiamen Fujian 361021,China)

机构地区：[1]集美大学诚毅学院,福建厦门361021

出　　处：《曲靖师范学院学报》2023年第3期33-38,共6页Journal of Qujing Normal University

基　　金：福建省自然科学基金项目“基于高性能二硫化钼场效应晶体管的免标记生物传感器研究”(2021J011222);福建省教育厅中青年教师教育科研科技类基金项目“基于物理模型的多幅图像去雾方法及其应用研究”(JAT191166)。

摘　　要：针对传统偏振图像去雾方法参数计算不便、出现噪声和色彩失真等问题,提出了一种基于天空分离技术的偏振图像复原方法.该方法首先对两幅0°和90°的偏振图像进行融合处理;然后通过天空分离技术分离偏振雾霾图像景物和天空部分后计算大气光强,利用自适应中值滤波方法优化大气散射光并计算图像传输率;最后利用偏振图像去雾模型去除图像雾霾.对使用单反相机和KENKO CPL67MM偏振滤镜采集到的图像进行测试,采用本文方法复原的图像能够在信息熵和平均梯度比图像评价指标中取得较大改善,并较好地保留了图像中丰富的边缘细节信息,克服了天空区域中出现的噪声问题,解决了人工计算参数中的不便,有效改善雾霾图像质量.This paper proposes an improved method to restore polarization image based on based on sky separation technology in order to solve the problems of traditional polarization image defogging methods,such as inconvenient parameter calculation,noise and color distortion.Firstly,we use the method of image registration fusion to deal with polarization image fusion.Then,we separate the polarized haze image scene and sky region by sky separation technology and calculate the atmospheric light intensity.We optimize the atmospheric scattered light and calculate the image transmission rate by the method of adaptive median filter.Finally,we remove the image haze by the polarized image defogging model.We use the images obtained by SLR camera and KENKO CPL67MM polarizing filters for experimental recovery test.The image restored by the proposed method retains the rich edge details and makes great improvement in the evaluation indexes of information entropy and average gradient ratio.The method effectively improves the quality of haze image,overcomes the noise problem in the sky area,and solves the inconvenience in manual calculation of parameters.

关 键 词：偏振图像复原 大气散射模型 大气光 传输率 评价指标 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]