Compress-image quality measures in image-processing applications

In distributed multimedia applications, an information compression stage is usually necessary. The objective here is to obtain a compression rate that makes the application viable. In this stage, reductions in image quality are often produced in order to reach high rates of compression. This image degradation must be measured and known in order to ensure that it can be limited within an acceptable range of values. The most commonly used metrics are MSE and PSNR. In recent years other metrics have been developed with the aim of obtaining values closer to human subjective evaluation. In this paper, metrics of image compression degradation for image processing applications are analyzed. We propose degradation metrics when the images are used in image processing applications. These are tested on two types of application which show a more than acceptable correlation between the measures and vision algorithms.