评价SAR图像去噪效果的无参考图像质量指标  

A new referenceless image quality index to evaluate denoising performance of SAR images

作　　者：张敏辉 杨剑[2] ZHANG Minhui;YANG Jian(School of Computer Science,Chengdu Normal University,Chengdu 611130,P.R.China;School of computer,Chengdu College,University of Electronic Science and Technology of China,Chengdu 711713,P.R.China)

机构地区：[1]成都师范学院计算机科学学院,成都611130 [2]电子科技大学成都学院计算机学院,成都611130

出　　处：《重庆邮电大学学报（自然科学版）》2018年第4期530-536,共7页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：四川省教育厅重点科研项目(17ZA0053)~~

摘　　要：为了客观、自动化地对合成孔径雷达(synthetic aperture radar,SAR)图像去噪效果进行评价,基于比例图像在空间域及小波域的统计特性,提出了一种无参考图像质量指标。先使用二维离散小波变换对比例图像和SAR原图进行小波分解,根据最高频域的小波系数计算得到相干斑噪声的标准差,然后利用针对SAR图像改进后的α边缘评价指标计算比例图像和SAR原图在空间域的边缘相似度系数,最后通过加权算法得出一种无参考图像质量指标。通过SAR仿真图和SAR原图的去噪评价实验,证明了新提出的质量指标能够对SAR图像去噪效果进行客观且自动化的评价。In order to objectively and automatically evaluate the image denoising performance of the filtered SAR(synthetic aperture radar)image,a new referenceless image quality index without reference image is proposed,based on the statistical properties of the ratio image.The two dimensional discrete wavelet transformation is firstly applied to decompose the ratio image and the original SAR image,while the standard deviation of speckle noise can be calculated according to the highest frequency of wavelet coefficients.Then the marginal similarity coefficient between the SAR ratio image and the original SAR image was calculated by using the improvedαedge evaluation index.Finally a new quality index is obtained by weighted formula.Experimental results show that the proposed quality index achieves the objective and automatic evaluation for the filtered SAR image.

关 键 词：SAR图像 质量指标 比例图像 小波变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]