3D Content Acquisition and Coding

Abstract

This chapter starts by addressing the impact of the inaccurate camera system alignment on the spatial reconstruction accuracy and stereo perception. An experimental study is described, using a stereoscopic camera setup and its deterministic relations derived by trigonometry, spatial model, and basic stereoscopic formulas. The significance of errors that can occur for possible cameras system setup is analyzed in order to find the appropriate settings and physical constraints of the camera system, which minimize the error. Then, the chapter presents an overview of the compression tools used in current stereoscopic and multiview video encoders. It includes the stereoscopic frame-compatible formats using spatial multiplex in the side-by-side and top-and-bottom fashion; the video plus depth representation, the layered coding approach, and multiview encoding. Furthermore, an extension of multiview video compression for the arbitrary camera arrangements is presented. The current status of MPEG exploration experiments of next-generation video codec technologies is also outlined in the last section. First, the UltraHD compression performance beyond HEVC is presented and second, the recent developments in HDR/WCG format conversion and coding are presented. Finally, the testing procedures and 3D projection formats for 360° video are addressed.