Depth map up-sampling based on edge layers

Abstract

Depth map images are characterized by large homogeneous areas and strong edges. It has been observed that efficient compression of the depth map is achieved by applying a down-sampling operation prior to encoding. However, since high resolution depth maps are typically required for view synthesis, an up-sampling method that is able to recover the loss of information is needed within this framework. In this paper, an up-sampling algorithm that recovers the high frequency content of depth maps using a novel edge layer concept is proposed. This algorithm includes a method for extracting edge layers from the corresponding texture images, which are then used as part of a non-linear interpolation filter for depth map up-sampling. In the present work, the up-sampling is applied as a post-processing operation to generate multiview output for display. Views synthesized with our up-sampled depth maps show the efficiency of our proposed technique relative to conventional interpolation filters.