在深入研究遥感图像与普通图像差别的基础上,对结构相似度算法进行改进,提出了一种新的无参考遥感图像质量评价方法.首先,采用非下采样轮廓波变换进行多尺度分解,将子图像进行K-means边缘检测及细化融合,最终得到边缘区域.采用一阶偏导有限差分计算遥感图像的梯度幅值,设置两个阈值,将满足区间的像素提取出来得到纹理区域.然后,使用边缘计算对比度分量,纹理计算结构相似度分量,对SSIM进行改进得到ET_SSIM方法.最后,采用“再降质”的方式构造出参考图像,将边缘纹理区域与平滑区域分开评价,ET_SSIM对边缘纹理区域评价,SSIM对平滑区域评价,根据人眼对边缘纹理及平滑区域的不同重视程度加权求平均,得到最终结果VSSIM.实验结果证明,本文方法VSSIM与主观评价值的线性相关度相对于MSE、PSNR、SSIM、GSSIM、BLIINDS-Ⅱ方法分别提高了22.2％,6.2％,0.8％,0.2％,1.3％.%Based on the study of the difference between remote sensing image and ordinary image,the structural similarity algorithm is improved,and a new method of non-reference remote sensing image quality assessment is proposed.Firstly,non-subsampled contourlet transform is used in this thesis to perform multi-scale decomposition.Secondly,the edge area is detected by K-means and thinned by Mathematical morphology.And texture region is extracted.Thirdly,we use the edge area to calculate the contrast component and used texture area to calculate the structural similarity component.Finally,we obtained the ET_SSIM method.In order to get a more general method of non-reference remote sensing image assessment,First of all,the reference image is constructed by " redescending",then the edge texture region and the smooth region are evaluated separately.the edge texture regions are assessed by ET_SSIM,smooth areas are assessed by SSIM,Last,According to different emphasis on the edge texture and smooth area of the human eye,weighted average to get the final result VSSIM.The experimental results show that the linear correlation between the method VSSIM and the subjective assessment value is 22.2％,6.2％,0.8％,0.2％,1.3％higher than that of MSE,PSNR,SSIM,GSSIM and BLIINDS-Ⅱ.
展开▼
