基于小波变换的多尺度SSIM算法

Multi-scale SSIM algorithm based on wavelet transforms

提出一种基于小波变换的多尺度结构相似度的图像质量评价(WWMS-SSIM)算法。首先对原始图像和失真图像分别进行小波变换,在对应的低频子带图像中比较亮度和相关度差异,在对应高频子带图像中比较对比度、相关度和结构差异,然后赋予不同子带图像不同的视觉加权值,最后得到归一化的图像质量评价值。通过对不同失真类型图像以及Live数据库图像的评价,证明了提出的算法(WWMS-SSIM)相对均方误差、峰值信噪比、结构相似度以及改进的结构相似度评价算法更符合人类的主观视觉感受。

This paper proposes a new algorithm composed of weighted multi-scale structural similarity（WWMS-SSIM）based on wavelet decomposition and HVS property.Wavelet decomposition is performed for the original and distorted images,respectively.Each image can be partitioned into one lowfrequency subband（LL）and a series of octave high-pass subbands（HL,LH and HH）.Different subbands are processed with different weighting factors.Based on the results of the above,modified WWMS-SSIM can be construced.In terms of experiment results,the new algorithm shows good feasibility comparing with MSE,PSNR,SSIM and improved SSIM image quality assessment methods.