一种基于高阶统计量的图像质量客观评价方法

A Method of Objectively Assessing Image Quality Based on High Order Statistics

提出一种新的图像质量的客观评价方法 .该方法用平均高阶细节累积量作为衡量图像质量的指标 ,无需原始图像作参考 ,特别适合于多媒体图像通信及无法得到原始图像的场合对图像质量进行评价 ,且很容易去除高斯噪声对图像质量评价的影响 .仿真实验结果表明 ,该方法所得结果大多数情况下与峰值对数信噪比指标相符合 ,并且在峰值对数信噪比不能很好地反映图像质量的情况下 ,该方法所用的指标也能比较准确地反映图像的质量 。

A new method of objectively assessing image quality is presented. This method is based on high order statistics. It does not need the original image for reference, and can remove the effect of Gauss noise on the assessment of image quality. So it is suitable for use in the situation that the original image can not be obtained. Experimental results have shown the effectiveness of the proposed scheme. It is in accordance with the conventional assessment method PSNR in most of the cases, and it is also in good accordance with the image quality even in the cases that the PSNR is not.