Computer generated hologram with geometric occlusion using GPU-accelerated depth buffer rasterization for...

Area of Science:

• Optics and Photonics
• Computer Graphics
• Computational Imaging

Background:

• Generating realistic computer-generated holograms (CGHs) with geometric occlusion is challenging.
• Existing methods often struggle with complex object models and computational efficiency.

Purpose of the Study:

• To develop a rapid method for producing CGHs with accurate geometric occlusion.
• To leverage parallel processing capabilities for efficient hologram generation.

Main Methods:

• A ray-casting approach is employed, shooting rays from hologram samples into the object volume.
• Z-buffering is utilized to determine the nearest object intersection for each ray.
• The complex field contribution of the nearest object point is added to the corresponding hologram sample.
• Parallel computation on Graphics Processing Units (GPUs) is exploited for speed.

Main Results:

• The method efficiently handles complex object models by depending on fixed system parameters rather than model complexity.
• Analysis of the Wigner distribution function provides an upper limit for the ray bundle cone angle.
• Experimental results show acceptable reconstruction quality with an angular sampling distance of 0.01 degrees for a 2.66 degrees cone angle.

Conclusions:

• The proposed GPU-accelerated ray-casting method enables rapid and efficient generation of holograms with geometric occlusion.
• This approach offers a significant improvement over traditional methods for complex holographic scenes.
• The findings contribute to advancements in real-time holographic display and simulation.