一种失真模糊图像的无参考质量评价方法

针对现有的无参考质量评价方法对失真模糊图像不能有效评价这一现象,提出了一种针对失真模糊图像的无参考质量评价方法。该方法将结构相似度(structural Similarity,SSIM)全参考质量评价方法应用到无参考质量评价中,不仅扩宽了SSIM方法的应用范围,也解决了SSIM方法不能合理评价模糊图像质量这一缺陷。首先对失真模糊图像进行低通滤波得到参考图像,计算失真图像与参考图像的结构相似度;然后提取图像的纹理特征,计算失真图像与参考图像的纹理相似度;最后将这两个相似度指标作为输入,LIVE图像数据库提供的主观评价值(different mean objective score,DMOS)作为输出,建立一个29 1]单隐层BP(back propagation)神经网络预测模型。实验结果表明,方法的预测结果稳定且与人的主观评价分数偏差小,Pearson相关系数(correlation coefficient,CC)和Spearman等级相关系数(rank order correlation coefficient,ROCC)均达到了0.97以上。

A Kind of No Reference Image Quality Assessment Method for Blurred Images

Considering that many no reference image quality assessment methods cannot give better assessment results for blurred images,this paper proposes a no-reference image quality assessment method which has better assessment results. And at the same time the structural similarity (SSIM) index is introduced into no-reference image quality assessment for broadening the application scope of the method and solving the flaw that the SSIM cannot be reasonable evaluation the blurred image quality. This method firstly constructs a reference image by a low-pass filter and calculates the structural similarity between the blurred image and the reference image, and then extracts texture features of images to compute the texture similarity between the blurred image and the reference image. Finally, with these texture similarity and structural similarity as the input factors, the subjective assessment value DMOS provided by LIVE database as the output factor, a 2 9 1] back-propagation (BP) neural network prediction model is built. Experiments indicate that the prediction results show stability and little difference from the experimental data. Their Correlation Coefficients are all above 0.97.