为了客观评价各种算法的重建效果，使用峰值信噪比(Peak Signal-to-Noise Ratio, PSNR)和结构相似度(Structural Similarity, SSIM)作为评价指标。
基于42个网页-相关网页
由频域函数、亮度函数和对比度函数计算得到结构相似度。
The structural similarity is calculated using the frequency domain, the brightness, and the contrast functions.
为了寻求评价准确的方法，设计一种改进的结构相似度评价方法。
In order to seek an approach which can assess distorted images more accurately, an improved SSIM image Quality Assessment method is proposed.
该文提出一种基于频域的结构相似度(FSSIM)的图像质量评价方法。
This paper presents a frequency domain-based structural similarity (FSSIM) method for image quality assessment.
应用推荐