Fig. 1. Key technology of high fidelity three-dimensional light field display

Fig. 2. Large-scale multi-projection light field 3D display system^[13]. (a) System structure propsoed in Ref. [13]; (b) girl pictures taken from different angles

Fig. 3. Structures of time division multiplexing display system^[15]. (a) Schematic diagram of proposed light-field display; (b) structure of multi-directional backlight unit

Fig. 4. 3D light field display of tri-colored glazed potteries in the Tang Dynasty and the ancient Buddha statues^[15]

Fig. 5. Tabletop integral imaging 3D display system based on annular point light source^[17]. (a) Structure diagram of proposed tabletop 3D display system; (b) experimental prototype of tabletop integral imaging 3D display; (c) display effects captured at different locations

Fig. 6. Schematic of proposed tabletop light field 3D display system based on vertically spliced light field Cave^[19]. (a) Structure of display system; (b) modulation principle of vertical compound lens array; (c) modulation principle of horizontal compound lens array

Fig. 7. Urban terrain 3D images at five different viewing angles (-50°, -25°, 0°, 25°, and 50°) along horizontal direction^[19]

Fig. 8. Optical see-through near-eye light field display based on discrete lens array^[21]. (a) Principle of optical see-through light-field near-eye display system; (b) optical component used in demonstration prototype; (c) images seen through OLED-based prototype

Fig. 9. System principle and comparison of optical designs of LC-LLU^[22]. (a) Schematic diagram of proposed light field display system; (b) light control principle of LC-LLU; (c) optimized compound lens structure and corresponding parameters; (d) comparison of spot diagrams of optimized compound lens and general single lens; (e) comparison of image reconstruction results of identical 3D scenes from two optical structures

Fig. 10. Comparison of 3D display effects reproduced by two view distribution methods^[22]. (a) Main parameters and 3D scene layout; (b) 3D effect based on proposed system prototype; (c) 3D effect based on parallax barrier

Fig. 11. Construction of voxel by proposed tabletop light-field display system^[28]. (a) Schematic of optical path of light field display unit; (b) schematic of microstructure of direction turning films; (c) comparison of 3D images at same position with and without directional diffusion film

Fig. 12. 360° desktop light field display^[32]. (a) Configuration of proposed 360° directional micro prism array for tabletop light field displays; (b) microstructure of prism array; (c) structure of light control unit

Fig. 13. Different perspectives of 3D moving car along annular direction^[32]

Fig. 14. Optimized aspheric lens^[33]. (a) Light path of optimized aspheric lens; (b) structure of optimized aspheric lens

Fig. 15. Dual-mode optical see-through integral imaging^[34]. (a) Schematic diagram of proposed system principle; (b) experimental setup of the first recording process; (c) experimental setup of the second recording process

Fig. 16. Design and optimization results of aspherical symmetric compound lens^[35]. (a) Structure of aspherical symmetric compound lens; (b) simulation results of imaging for 5 sets of parallel light beams using aspherical symmetric compound lens

Fig. 17. Optical experimental results of 3D light field fusion display system at larger viewing angle^[35]. (a) Viewpoint images observed at about -40°; (b) viewpoint images observed at about 40°

Fig. 18. Comparison of results of ideal lens and actual lens^[36]. (a) Optical path at 0° incidence angle without aberration; (b) optical path at 0° incidence angle in actual situation; (c) signal intensity distributions of ideal lens and actual lens at 0° incidence angle; (d) variation in DALB with increasing incidence angle

Fig. 19. Datasets used in simulation and comparison of corresponding encoding performance^[37]

Fig. 20. Model and results of pixel coding method^[41]. (a) Top view of mathematical model of coding algorithm; (b) elemental image array of light-field display; (c) elemental image

Fig. 21. Comparison of simulation results of pixel deviation thresholding algorithms^[41]. (a) Original image; (b) simulation result of backward ray-tracing algorithm; (c) simulation result of fitting algorithm; (d) simulation result of proposed algorithm

Fig. 22. Iterative optimization process of co-designed CNN^[43]

Fig. 23. Schematic diagram of production process of halftone phenanthroline map and color error problems^[45]

Fig. 24. Structure of RE-DNN^[46]

Fig. 25. Process of using co-designed CNN to fit multi-view light-field^[47]