Title :
Perceptually adaptive filtering for stereoscopic video compression
Author :
Maryam Homayouni;Payman Aflaki;Miska M. Hannuksela;Moncef Gabbouj
Author_Institution :
Department of Signal Processing, Tampere University of Technology, Finland
Abstract :
Advances in multiview video coding aim to fulfil the required bandwidth and storage capacity of the 3D content. However, despite these evolutions, further compression efficiency can be achieved by taking into account the perceptual characteristics of 3D video. This paper presents a regionally adaptive filtering scheme for 3D video, exploiting the sensitivity of the human visual system to the perceptual characteristics of texture video. The location and the strength of the smoothing filter are determined based on the degree of brightness of each individual pixel within the image. A series of systematic subjective tests were conducted, confirming that no quality degradation is introduced by application of such filtering and hence, keeping the perceived quality of the filtered content intact. At the same time, objective measurements show a Bjontegaard delta bitrate reduction of up to 33.55% and with an average of 21.35%.
Conference_Titel :
3D Imaging (IC3D), 2015 International Conference on
DOI :
10.1109/IC3D.2015.7391824