PQMET: A digital image quality metric based on human visual system

In this paper we propose a digital image quality metric based on human visual system. This metric considers the coefficients of discrete cosine transform related to each 8×8 block of an image. To perform image evaluation PQMET detects a distortion applied to the image and then selects the corresponding matrix of minimum visibility thresholds capable to produce a visible signal. Therefore it uses the information about the distortion applied, with a consequent improvement in the image quality evaluation. The experimental results show that the proposed metric achieves competitive performance with other well known metrics and outperforms them in some cases.