Convolutional symmetric compressed look-up-table method for 360° dynamic color 3D holographic display.

Area of Science:

• Optics
• Computer Science
• Digital Holography

Background:

• Computer-generated holography (CGH) is crucial for 3D display technologies.
• Conventional methods for CGH computation can be computationally intensive, limiting real-time applications.
• Look-up-table (LUT) based methods offer acceleration but can face limitations in quality and speed.

Purpose of the Study:

• To propose and validate a novel, accelerated method for computer-generated hologram computation.
• To enhance the speed and quality of color hologram generation.
• To enable real-time holographic display capabilities.

Main Methods:

• Development of a convolutional symmetric compressed look-up-table (CSC-LUT) method.
• Utilizing dynamic convolution operations and coordinate symmetry for data compression.
• Resampling of test objects to ensure convolutional translation invariance.

Main Results:

• The CSC-LUT method significantly accelerates CGH computation compared to the conventional SC-LUT method.
• Achieved higher reconstruction quality and wider applicability.
• Further optimization enabled real-time color holographic display (>24fps) with multiple perspectives via GPU parallel acceleration.

Conclusions:

• The proposed CSC-LUT method offers a significant advancement in CGH computation speed and quality.
• This method paves the way for practical, high-performance holographic display systems.
• Real-time holographic display is achievable through efficient algorithms and hardware acceleration.