Graphics processing unit-based implementation of a one-dimensional novel-look-up-table for real-time computation of Fresnel hologram patterns of three-dimensional objects

Abstract

Abstract. A one-dimensional novel-look-up-table (1-D N-LUT) has been implemented on the graphics processing unit of GTX 690 for the real-time computation of Fresnel hologram patterns of three-dimensional (3-D) objects. For that, three types of optimization techniques have been employed, which include the packing technique of input 3-D object data and the managing techniques of on-chip shared memory and registers. Experimental results show that the average hologram calculation time for one object point of the proposed system has been found to be 0.046 ms, which confirms that the proposed system can generate almost 3 frames of Fresnel holograms with 1920 × 1080 pixels per second for a 3-D object with 8000 object points. © The Authors.Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

Object tracking mask-based NLUT on GPUs for real-time generation of holographic videos of three-dimensional scenes

“…In this paper, the NLUT method is implemented as the base CUDA kernel function. Since the CGH can be generated with PFPs on each depth in the NLUT method [17,22], the base CUDA kernel function may be called on each depth. Actually, the computational performance of the GPU depends on the number of base CUDA kernel function calls, and the conventional NLUT methods have different numbers of base CUDA kernel function calls.…”

“…Some of them include temporal redundancy-based NLUT (TR-NLUT) [18], motion compensationbased NLUT (MC-NLUT) [19] and MPEG-based NLUT (MPEG-NLUT) [20] methods. In practice, these NLUTs would be eventually implemented on FPGAs or GPUs for their practical application to real-time CGH generation of 3-D scenes [21,22]. For efficient implementation of NLUTs on GPU boards, they must be compatible with the software and memory structures of the GPU boards.…”

“…For fast loading and accessing the sub-PFPs data of the 1-D NLUT and the 3-D data of the input object on the GPU boards, three types of optimization techniques are also employed. These include the packing technique of the input 3-D object data for efficient storing in the on-chip shared memory, and the managing techniques of the on-chip shared memory for fast computation of the CGH, and the on-chip registers for quick storing of the calculated hologram data [22]. Fig.…”

“…However, because the CGH buffer is too big to be saved on the shared memory, it is loaded and accessed directly from the global memory when the GPU board calculates the CGH. Here, the global memory has long access latencies and a finite access bandwidth, which cause traffic congestion in the global memory, and eventually degrade the calculation performance of the GPU even though the GPU board has many threads available for parallel execution [22].…”

High-Performance Computer-Generated Hologram by Optimized Implementation of Parallel GPGPUs

“…Kim et al developed a one-dimensional novel-look-up-table (1-D N-LUT) which was implemented on the graphics processing unit of GTX 690 for the real-time computation of Fresnel hologram patterns of three-dimensional (3-D) objects. This system can generate almost 3 frames of Fresnel holograms with 1920×1080 pixels per second for a 3-D object with 8K object points [9]. Smithwick et al of rapid hologram generation for realistic 3D image reconstruction based on the angular tiling concept using a new graphic rendering approach integrated with a previously developed layer-based method for hologram calculation.…”