PSNR vs . quality assessment metrics for image and video codec performance evaluation

One of the most important performance metrics for evaluation and comparison of image or video codecs is the Rate/Distortion (R/D) where quality is measured in terms of PSNR (Peak Signal-toNoise Ratio). It is well known that PSNR does not always rank quality of an image or video sequence in the same way that a human being. There are many other factors considered by the human visual system and the brain. So, a lot efforts were performed to define an objective video quality metric that is able to measure the quality distortion close to the one perceived by the destination user. We analyze the behaviour of some of the most relevant objective quality metrics when they are applied to video sequences, compressed by different video codecs at different bit-rates taking as reference the classical PSNR metric. So we try to find if there is a more accurate metric in terms of human quality perception that could substitute PSNR in the R/D plots used in the performance evaluation of different coding proposals.