Study of a holographic TV system based on multi-view images and depth maps

Abstract

Electronic holography technology is expected to be used for realizing an ideal 3DTV system in the future, providing perfect 3D images. Since the amount of fringe data is huge, however, it is difficult to broadcast or transmit it directly. To resolve this problem, we investigated a method of generating holograms from depth images. Since computer generated holography (CGH) generates huge fringe patterns from a small amount of data for the coordinates and colors of 3D objects, it solves half of this problem, mainly for computer generated objects (artificial objects). For the other half of the problem (how to obtain 3D models for a natural scene), we propose a method of generating holograms from multi-view images and associated depth maps. Multi-view images are taken by multiple cameras. The depth maps are estimated from the multi-view images by introducing an adaptive matching error selection algorithm in the stereo-matching process. The multi-view images and depth maps are compressed by a 2D image coding method that converts them into Global View and Depth (GVD) format. The fringe patterns are generated from the decoded data and displayed on 8K4K liquid crystal on silicon (LCOS) display panels. The reconstructed holographic image quality is compared using uncompressed and compressed images.