Improved look-up table method of computer-generated holograms.

Area of Science:

• Optics and Photonics
• Computer Graphics
• Display Technology

Background:

• Computer-generated holograms (CGHs) are crucial for 3D displays but face computational and memory challenges.
• Existing methods often struggle with heavy computation loads and vast memory requirements, limiting their practical application.
• The coherent ray-trace (CRT) method offers high quality but is computationally intensive.

Purpose of the Study:

• To develop an efficient and high-quality method for CGH reconstruction.
• To address the bottlenecks of computation load and memory usage in CGH generation.
• To improve the performance of CGH methods on graphics processing units (GPUs).

Main Methods:

• An improved look-up table (LUT) method was proposed, suitable for arbitrarily sampled object points.
• The method was implemented on a graphics processing unit (GPU) for accelerated computation.
• A 'distance factor' was defined, pre-computed off-line, and stored in the LUT.

Main Results:

• The improved LUT method achieved reconstruction quality comparable to the computationally expensive coherent ray-trace (CRT) method.
• On-line computation time was dramatically reduced compared to existing LUT methods on GPUs.
• Memory usage was significantly lower than previously proposed novel LUT approaches.

Conclusions:

• The proposed improved LUT method effectively overcomes the computational and memory limitations of CGH generation.
• This method offers a practical solution for high-quality, efficient 3D holographic display reconstruction.
• Optical experiments validated the effectiveness and performance improvements of the proposed technique.