Information reduction for integral imaging and video coding for multi views

Abstract

We have been developing the most suitable video coding for an integral 3D image to implement a 3D television service. To reduce the amount of information needed to reconstruct an integral image, we evaluate the correlation between the quality of the integral imaging and the number of multi views. In this study, we show the results for the correlation and the compression performances when fewer multi views including the depth views are coded by 3D-HEVC. We compared compression performance when coding all views and coding fewer views including depth views. The results showed that coding fewer multi views including depth views enhanced the efficiency of the compression performance more than coding all views.