Monte–Carlo Techniques Applied to CGH Generation Processes and Their Impact on the Image Quality Obtained

Abstract

Computer graphics aim to create visual representations for screens, where depth is simulated. In contrast, Computed Generated Holograms (CGH) focus on encoding and recreating light patterns to generate a true 3D holographic image that appears as a physical object in space. Therefore, although both use digital models, the computation of CGHs necessitates additional phase-related calculations, which in turn escalate computational demands. These calculations often result in excessively long development times or, at worst, render the process unfeasible. In order to reduce computational time, Partial Monte–Carlo Sampling (PMCS) techniques for CGH generation are presented, integrating them into the whole process of generating a CGH for a synthetic 3D scene, from design to rendering. PMCS is based on the random choice of a subset of rays used to compute the CGH and relates the computation time spent to the quality of the reconstructed scene. Quantitative analysis shows that PMCS does not significantly compromise image quality. Both simulated and in-laboratory image reconstruction from holograms demonstrates consistent trends, showcasing improved quality with higher numbers of rays and increased resolution. Furthermore, we establish a direct relationship between image quality and computational time, which effectively addresses specific requirements.