Modelling of image distortion processes associated in video-phone systems and descriptors for image quality

Abstract

A variety of video-phone systems have begun to appear in the consumer market. The quality of the images generated by these systems is severely affected by the bandwidth restriction of the twisted cable pair. Typically, some of these systems operate with a transmission bit rate less than 19.2 kbps. In comparing different systems with respect to image qualiv, one j n & it useful to assess the quality in terms of descriptors that would represent the subjective assessment more naturally. In this paper, we investigate the possible use of fuzzy set descriptor for the video-phone image. In order to do this, we examine the types of image distortion processes that are obtained in such video-phone systems. We consider three physical effects, name+ blurring due to defocusing, blurring due to motion of subject, additive electrical noise in the transmission systems. In addition to the physical distortion processes, we consider distortion due to sofrware artifacts, mainly arising @om the use of lossy compression algorithms as well as the associated image blocking effects.