Abstract
dc:description.abstractReal-time obstacle scanning and projection are crucial to decrease road accident rates worldwide. Current display advances in modern vehicles increase the risk of driver distraction and endanger safety. Conventional head-up displays require the driver to shift the field of view from the far field of the road towards a small region on the windscreen. Holographic 3D Head-Up Displays (HUDs) can provide a basis for the transportation sector to build on accessible and inclusive design strategies. However, current holographic HUDs lack a full-parallax effect, the number of pixels needed to recreate duplicate Augmented Reality (AR) images of the original object, and the display capabilities to reach real-time projections. Panoramic holographic projections that provide different depths of objects in creating a virtual reality experience in the driver’s field of view prevent driver distraction. In this thesis, holographic setups were developed to display 2D and 3D Ultra-High Definition (UHD) projections using Light Detection and Ranging (LiDAR) data in the driver’s field of view (eye box). Different light sources such as a HeNe laser and an Nd:YVO<sub>4</sub> laser were compared in terms of accuracy, precision, and ease of use in the HUD applications. A UHD Spatial Light Modulator (SLM) with a panel resolution of 3840×2160 for UHD and a ferroelectric reflective SLM were utilised for the replay field projections. Both setups were compared, and the laser sources were used to illuminate Computer-Generated Holograms (CGHs) and generate the projection by reconstructing the object image in the far (replay) field. Graphics Processing Unit (GPU)-accelerated real-time holograms 16.6 times faster than the equivalent CPU processing time. A virtual Fresnel lens was used to enlarge the driver’s eye box to 25 mm × 36 mm. Real-time scanned road obstacles from different perspectives provide the driver a full view of risk factors such as generated depth in 3D mode and the ability to project any scanned object from different angles across 360°. The 3D holographic projection technology allows the driver to maintain focus on the road instead of the windshield. It enables assistance by projecting road obstacles hidden from the driver’s field of view. A multicolour 3D reconstruction method accelerated with GPU is introduced into the 360° LiDAR-based HUD architecture with single-mode fibre pigtailed laser diodes. The important aspect of this work is the leverage of driver integration into the method. The LiDAR data collection and real-time processing were based on the iPhone-integrated LiDAR sensors to be utilised during driving. Architectures including light sources, detectors, and rotating transmitters were previously accommodated to capture a panoramic scene with LiDAR. The data processing method was optimised with a point cloud algorithm accelerating computation time. Furthermore, experimental results of dynamic AR reconstructions were discussed, verifying the application for real-time colour 3D HUDs.
Degree
thesis:*- Name dc:type.qualificationname
- Doctor of Philosophy (PhD)
- Level dc:type.qualificationlevel
- Doctoral
- Grantor dc:publisher.institution
- University of Cambridge
- Year dc:date.issued
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Skirnewskaja, Jana
- Advisor dc:contributor.advisor
-
- Wilkinson, Timothy
Subjects
dc:subject × 4Rights
dc:rights- Licence
- Language dc:language
- eng
Identifiers
dc:identifier.*- DOI dc:identifier.doi
- https://doi.org/10.17863/CAM.106922
- OAI identifier oai:identifier
- oai:www.repository.cam.ac.uk:1810/365838