Binary hologram generation of real objects by combination of color and depth map

Abstract

Holograms can be captured optically or computed numerically. Optical methods enable to record information of real objects as they are, but high stability and dark room conditions for experimental set-up is required. It often restricts the capture to only small shape objects. On the other hand, Computer Generated Hologram (CGH) approach enables to calculate complex hologram in a flexible way. The scale of the scene can be set as well as the viewing distance or the resolution. The main challenges are to overcome the heavy computational burden and to provide a good sense of reality. CGHs can be calculated from point clouds or polygon models. With this approach, the contributions of every element should be computed for every pixels of the hologram. Holographic data can also be computed by applying propagation formula on 2D images, but the phase should be determined and as well as the depth position at every pixel. We exploited a Kinect camera that has raised a lot of interest recently since it enables to scan a scene for 3D modeling, or to capture both depth map and RGB frame of a scene. Depth map can be used to extract phase information. Combined with the amplitude image, complex field can be determined in a given plane. Then, CGH can be computed in a fast way. For the display part, Direct Binary Search (DBS) algorithm was applied to convert CGH of real scene into binary format so that bandwidth of the SLM can be fully exploited.