Advanced Computational Holography for Perceptual Enhancement and Application in HUDs

This thesis explores computational holography with a focus on perceptual contrast enhancement, aiming to enable more practical and feasible holographic displays for both 2D and 3D applications. It begins with an overview of holographic displays and their advantages. Based on the core components of a holographic display system, including the laser source, spatial light modulator (SLM), and computer-generated hologram (CGH) algorithms, the challenges and limitations of conventional hologram reconstruction for real-world applications are identified. These limitations motivate the exploration of drawbacks and the development of corresponding solutions. This thesis focuses on three aspects of the system: software, hardware, and applications. First, we propose a state-of-the-art CGH algorithm designed to enhance perceptual contrast and extend the dynamic range of reconstructed holographic images while maintaining overall luminance. The proposed scheme incorporates a non-linear merit function into the classical Gerchberg–Saxton (GS) framework, where energy redistribution is coupled with exponent-based intensity weighting and a dynamic range shift that selectively suppresses noise in perceptually sensitive areas. As a result, the method achieves superior contrast ratios and produces sharper and more vivid reconstructed holographic images with faster convergence. The optical reconstruction results indicate a 6.25% improvement in peak signal-to-noise ratio (PSNR) and a 67.85% improvement in dynamic range index (DRI). Second, recognizing that the inevitable imperfections of SLM hardware constrain the optimal system performance, we introduce a subpixel-level optimization method that accounts for practical issues such as fringing effects. The core advantage lies in simulating subpixel behaviour through a frequency-domain shift principle without altering the computational grid size, which markedly improves efficiency compared with conventional direct expansion, reducing the computational time by 39% for a 3×3 configuration and 59% for a 5×5 subpixel lattice. Furthermore, the zero-order contribution is analysed and effectively suppressed. In parallel, the kernel shape is investigated to capture the physical characteristics of liquid crystals in LCoS devices, where anisotropy arises from the rubbed alignment layer. The analysis reveals that the effective spread of fringing field effect follows a Laplacian-like profile, with energy preferentially extending along the rubbing axis and exhibiting a positive shift due to the pretilt angle. Experimental validation demonstrates that the proposed framework enables a more accurate simulation of realistic scenarios and achieves enhanced visual fidelity with improved diffraction efficiency in zero-order–free holographic displays. Finally, the thesis presents a 3D head-up display (HUD) as a representative application of holographic display technology aimed at enhancing driving safety and user experience. An LCOS-based holographic engine, integrated with a relay lens system and windshield combiner, is implemented with design parameters optimized to achieve a wide field of view, a practical eyebox, and multi-depth representation. This configuration offers true phase-based depth cues that enable accurate spatial registration of virtual signs within the ambient environment, reduce visual strain, and improve driver immersion. Experimental validation, supported by enhanced CGH algorithms, demonstrates multi-layer depth reproduction with high contrast and diffraction efficiency. Furthermore, the pronounced impact of speckle noise on HUD clarity is mitigated through a time-multiplexing strategy, while zero-order diffraction is effectively suppressed. Collectively, these advances establish a viable pathway toward next-generation holographic HUDs for automotive displays.

Name dc:type.qualificationname

Doctor of Philosophy (PhD)

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2025

Author dc:creator

• Liu, Jiaqi

Advisor dc:contributor.advisor

• Pivnenko, Mike

• Computer-Generated Holography
• Head-Up Displays

Licence

• https://www.repository.cam.ac.uk/bitstreams/33642e59-890a-4390-9cb0-a92c18b69d7f/download
• http://purl.org/NET/rdflicense/allrightsreserved

Language dc:language

eng

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.125089

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/395651