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

摘要

A variety of video-phone systems have begun to appear in thenconsumer market. The quality of the images generated by these systems isnseverely affected by the bandwidth restriction of the twisted cablenpair. Typically, some of these systems operate with a transmission bitnrate less than 19.2 kbps. In comparing different systems with respect tonimage quality, one finds it useful to assess the qualify in terms ofndescriptors that would represent the subjective assessment morennaturally. In this paper, we investigate the possible use of fuzzy setndescriptor for the video-phone image. In order to do this, we examinenthe types of image distortion processes that are obtained in suchnvideo-phone systems. We consider three physical effects, namely blurringndue to defocusing, blurring due to motion of subject, additivenelectrical noise in the transmission systems. In addition to thenphysical distortion processes, we consider distortion due to softwarenartifacts, mainly arising from the use of lossy compression algorithmsnas well as the associated image blocking effects