Publication Details
Overview
 
 
Geert Braeckman, Adriaan Barri, Gabor Fodor, Ann Dooms, Joeri Barbarien, Peter Schelkens, András Bohó, Li Weng
 

Chapter in Book/ Report/ Conference proceeding

Abstract 

The visual quality expected by the end-user of today's multimedia applications is ever increasing due to the raising availability of high-resolution cameras, HD and 3D screens and televisions. During production, transmission and rendering of the image/video material it is liable for quality degradation due to, amongst others, lossy (re)compression, transmission errors and suboptimal rendering. It is therefore essential to guard the perceptual quality of multimedia from design to exploitation phase. The most accurate form of quality monitoring consists of subjective evaluations with test sequences, which is costly and time-consuming. Therefore, the research community now strives for the design of objective perceptual quality metrics that can quantify the degree of degradation following the Human Visual System. This is a truly challenging task and currently research mostly focuses on full-reference metrics, meaning that comparison with the original material is required. However, this is not feasible in most multimedia applications such as live streaming and other cloud or Internet of Things applications etc. In this paper, we present a flexible modular and extendable XML-based software tool for reduced-reference visual quality assessment and professional benchmarking, which is based on image and video watermarking and perceptual hashing.

Reference 
 
 
Link