Open Access
Add to BibSonomy
Abstract
A true-color light-field display system with a large depth-of-field (DOF) is demonstrated. Reducing crosstalk between viewpoints and increasing viewpoint density are the key points to realize light-field display system with large DOF. The aliasing and crosstalk of light beams in the light control unit (LCU) are reduced by adopting collimated backlight and reversely placing the aspheric cylindrical lens array (ACLA). The one-dimensional (1D) light-field encoding of halftone images increases the number of controllable beams within the LCU and improves viewpoint density. The use of 1D light-field encoding leads to a decrease in the color-depth of the light-field display system. The joint modulation for size and arrangement of halftone dots (JMSAHD) is used to increase color-depth. In the experiment, a three-dimensional (3D) model was constructed using halftone images generated by JMSAHD, and a light-field display system with a viewpoint density of 1.45 (i.e. 1.45 viewpoints per degree of view) and a DOF of 50 cm was achieved at a 100 ° viewing angle.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In recent years, 3D displays have seen increasing adoption in various fields, including medicine, education, art, and military [1–3]. Holography, autostereoscopic display, integrated imaging, light-field display technology, and other naked-eye 3D display technologies are booming, and various technologies seek to transmit natural and real 3D image information to the human eye [4–10].
Among these, light-field displays have emerged as one of the fastest-growing technologies due to their ability to provide natural perception. The light-field system constructs a large number of voxels to realize the spatial reproduction of 3D objects [11–23].
Light-field displays are challenged with limited DOF. When the depth exceeds a certain threshold, there are problems with voxels constructed from a limited number of controllable beams, leading to a discontinuous viewing parallax and ghosting in the display. Fourier transform is applied to transform light-field images into the frequency domain for a more rigorous analysis [10–16,5,17–21,14].
The frequency domain analysis reveals that the phenomenon is due to the constraint on the viewing-angle, viewpoint density and spatial resolution under the condition of limited display information [22–25].
To attain large depth of field (DOF) displays, it is crucial to enhance the angular resolution by increasing viewpoint density, while simultaneously considering the spatial resolution [26,27]. One solution is to use information redundancy in the time dimension to increase the amount of displayed information [28–32]. For example, with three groups directional backlights and a fast-switching liquid crystal display (LCD) panel, a time-multiplexed light-field display with a wide viewing angle is demonstrated [28]. A 3D display device combined with a directional-sequential light distribution to increase the number of viewpoints [29,30]. A multi-projection time-multiplexed 3D display has been proposed by using a steering screen to improve the angular resolution [32].
Another solution is to make a trade-off between angular resolution and spatial resolution through system design without increasing the information content [25,8,33–35]. Most recently, the retinal imaging principle (foveated vision strategy) is used to adjust the angular resolution of different regions of the display device [35].
In order to solve the DOF problem of light-field display, the angular resolution and spatial resolution need to be adjusted and balanced by taking advantage of the characteristics of the microscopic display structure. From a micro perspective, a light-field display device is comprised of periodic LCU that are capable of emitting directional light beams in multiple directions.The size of each LCU represents the spatial resolution, while the number of beams that each LCU can control represents the viewpoint density. Therefore, when the size of the LCU is smaller than the resolution scale of human eyes, the more beams controlled by the LCU, the greater the DOF of the display.
Here, a true-color light-field display system with a large DOF is proposed. By adopting 1D light-field coding algorithm, the light-field display system based on halftone ink dots can achieve high-definition display with a viewpoint density of 1.45 (i.e. 1.45 viewpoints per degree of view) and a DOF of 50 cm.
The collimated backlight and the oppositely placed ACLA are used to suppress the aliasing and crosstalk between the beams, so that the LCU can control more than 100 beams. As a result, existing high-PPI LCD screens cannot meet the required beam density. For a long time, screen-based 3D display systems have been limited by the shape and size of pixels, so the number of controllable light beams and the size of the LCU are restricted by the screen pixels. The halftone image constructed by high-precision printing uses 4000 dpi high-resolution halftone ink dots to realize light-field display, and the halftone pixel size is close to 6 microns.
In order to improve the angular resolution of the system, 1D light-field encoding is proposed. 1D light-field coding represents a controllable light beam through a column of halftone ink dots, increasing the number of controllable light beams of the LCU. Based on the light-field display theory, all 3D objects to be displayed can be split into a large collection of voxels. By backtracking all the light beams emitted by each voxel, the brightness value of the multi-angle light beams under each LCU can be determined. However, the use of 1D light-field encoding leads to a decrease in the color-depth of the light-field display system. For example, to realize the three-color high color-depth display of the device, 255 × 3 halftone dots need to be used to form a beam. In the case of 1D light-field encoding, the size of the LCU is limited by the spatial resolution, which cannot satisfy the construction of beams with a large number of halftone ink dots.
To make the light-field display more closely resemble a natural display, high color-depth is a necessary aspect of true color display [36,37]. The high color-depth display is realized through the JMSAHD. Halftone ink dots cannot adjust the gray scale, so the gray scale adjustment of the device requires the joint action of multiple halftone dots. In the experiment, there was a difference between the pixels of the computer-coded halftone image and the ink dots of the actual printed halftone image. Utilizing the difference between the two to adjust the position of the printed ink dots can make the light beam have a color-depth difference, so that the positional relationship between the ink dots can be used to increase the amount of displayed information. Based on this, a JMSAHD for field display can be developed, which can increase the number and color-depth of directional light beams. That is, without changing the physical structure, JMSAHD improves the spatial bandwidth product of the display system, so as to realize the display of large DOF and true color display.
2. System design
In the display architecture, LCUs emit directional beams to construct spatial voxels. Figure 1(a) shows the schematic of the light-field display system and its internal optical path diagram. The hardware of the system is composed of five main components: a collimated beams generator, a halftone film, an RGB optical filter, an ACLA, and a vertical diffusion film (VDF). These components work together to form the hardware of the light-field display system.
Fig. 1. (a) Structure diagram and optical path diagram of light-field display system. (b) Structure diagram and optical circuit diagram of system optical control unit.
In Fig. 1(a), the pathway of collimated beam in the light-field display system under the macro state is described. The beam undergoes intensity modulation through a halftone film, color modulation through RGB optical filter, and angular modulation through the ACLA. Finally, it passes through a VDF to broaden the beam in the vertical direction, ensuring a uniform viewing angle in the vertical direction. After passing through the above structures, the collimated beam achieves multi-angle emission in the horizontal direction and forms a uniform parallax within the viewing angular range, providing hardware support for the optical design and light-field encoding of subsequent systems.
Figure 1(b) shows the micro optical path of a beam containing three color components and the structure of a LCU. As shown in the schematic illustration of the microscopic controlled light unit in Fig. 1(b), the beams passing through the aspheric column lenses converge at the lens main axis, and after the converged beams diverge, different directional beams are formed.
After the system design is completed, in order to prevent lens distortion from affecting the uniformity of parallax distribution or low spatial resolution and angular resolution affecting DOF, its optical design needs to consider both spatial resolution and angular resolution, and solve the problem of uneven parallax distribution caused by distortion through lens surface design.
3. Optical design
In order to maintain the spatial resolution and angular resolution at the same time and avoid the uneven distribution of parallax caused by lens distortion, the design of the aspheric cylindrical lens unit is crucial for the creation of the light-field display system. By setting the orientation and surface data of the cylindrical lens unit, the optical design expands the collimated beam into multiple beams with the same width, enabling voxel formation and improved angular resolution.
3.1 Selection of imaging optical path
Generally, the convex surface of the cylindrical mirror unit utilized in LCD-based stereoscopic displays faces the direction of the light beam. The light scattered from the LCD screen is refracted by the cylindrical lens unit and emitted from multiple angles to form multi-angle parallax and thus, viewpoints and stereo vision. Figure 2(a) depicts the beam refraction optical path in a typical LCD-based stereoscopic display. The crosstalk of different light beams in the optical path is difficult to eliminate, resulting in more adverse effects when increasing the number of controllable light beams, resulting in ghosting or blurring of the viewing effect.
Fig. 2. Comparison of optical path between stereoscopic display system and light-field display system. (a) Schematic diagram of optical path of stereoscopic display cylindrical mirror grating. (b) Schematic diagram of light-field display cylindrical mirror grating optical path.
In response to this issue, we flipped the cylindrical lens array and employed collimated backlighting. As shown in Fig. 2(b), the collimated beam converges on the main optical axis of the lens after passing through the aspherical cylindrical lens and is emitted from multiple angles. The visible angle of the LCU is greatly improved compared to the display device in Fig. 2(a), and the crosstalk between the beams is also optimized. As a result, under the same size of the LCU, the light-field display system can accommodate more controllable beams, improving the system's angular resolution and improving the display device's DOF.
3.2 Parameter design of cylindrical lens
In order to improve the angular resolution, the LCU of the light-field display system need to be able to control the densely arranged multi-directional light beams, which reduces the tolerance to the distortion caused by the aberration of the cylindrical lens, so that the spherical lens cannot meet the requirements. Hence, the surface shape of the cylindrical lens needs to be designed as aspherical. An optimized compound aspherical lens can effectively correct distortions caused by aberrations, ensuring uniform light distribution of different angles during voxel construction.
Figure 3(a) is a schematic diagram of the light control path after the aspheric surface of the cylindrical lens is designed. The collimated light beam is incident on the reversely placed cylindrical lens, and the cylindrical lens can control the refracted light beam within a visible angle of 100°. Figure 3(a) is enlarged one thousand times from Fig. 3(a). Figure 3(a) shows the beam distribution in the macroscopic state, and the beams are evenly and approximately equidistantly distributed on the light receiving surface.
The surface shape formula for the aspherical lens design is given by:
After optical simulation and actual manufacturing experiments, the non-spherical lens surface profile formed by the data in Table 1 can significantly reduce the distortion interference caused by aberrations, and its point spread function meets the requirements of light-field display systems.
Table 1. Parameters of the aspheric surface
4. 1D light-field encoding algorithm
With the optical design of aspherical cylindrical lens, the diffuse spot formed on the object surface after the beam is reverse traced is less than 6 microns. Based on parameters such as diffuse spot radius and printing accuracy, 1D light-field encoding is designed, that is, the number of controllable beams is increased in the LCU. Since ACLA only provides parallax in the horizontal direction, in a square LCU, the limit of the number of beams that can be controlled is the number of halftone ink dots in the horizontal direction of the LCU. 1D light-field encoding uses a column of ink dots on the film of the LCU to modulate the intensity of a beam of controllable beams, and the number of controllable beams of a single LCU is the number of ink dots that the LCU can accommodate in the horizontal direction. The intensity of different light beams is determined by the ratio of the light-transmitting and shading area of the film it passes through. Therefore, the number of gray levels (the intensity of the controllable light beam) of the voxel is determined by the number of ink dots that the LCU can accommodate in the horizontal direction.
Figure 4 shows 1D light-field encoding algorithm diagram after the light-field is discretized. The discretization of the light-field takes the radius of the diffuse spot of the cylindrical lens as the sampling period. According to the needs of 1D light-field encoding, the beam obtained after sampling corresponds to a column of halftone ink dots in the LCU. The beam expression is:
The light-field display realizes intensity modulation, color modulation, angle modulation and longitudinal diffusion through halftone film, RGB filter, cylindrical lens grating and VDF respectively, so that the light beam emitted in all directions carries correct optical information. In this way, the precise and stable construction of voxels in space is realized. These voxels converge to form a 3D image whose composition expression in space is as follows:
Among them, $IMF({x,y} )$ is the light intensity modulation formula, which adjusts the light intensity of the sample collimated beam by controlling the number of the collimated beam transmission area of the black ink dot on the halftone film. $CMF(y )$ is the color modulation formula. The RGB filter presents a single color distribution along the horizontal direction and RGB alternating distribution along the vertical direction. Three RGB color components will appear when a collimated beam passes through the RGB filter. $\sum {AM{F_i}} ({x,y,\theta } )$ is the angle modulation formula, which adjusts the collimated beam to different angles through the cylindrical mirror grating, and reduces the aberration through the aspheric design, so that the beam divergence is uniform, facilitating the positioning and construction of voxels. $VDF(\varphi )$ is a vertical diffusion formula. Through the VDF, the emitted light beam is vertically diffused, so that the viewer can observe the 3D graphics from a non-horizontal direction. The horizontal area of the LCU can be fully utilized when constructing the voxels. As shown in Fig. 4, the collimated beam after discrete sampling has RGB components and presents strip distribution. After discrete sampling using the light-field construction algorithm, a halftone dot array will be used to represent a collimated beam with a controllable emission direction. An array of halftone dots, framed in blue, can modulate the intensity of a steerable beam of light. The light beam is divided into three color channels, and the color modulation is completed through a filter to form a display pixel after the three colors are superimposed. The display pixels of a controllable light beam are angularly modulated by ACLA, and together with other controllable light beams of different angles, constitute a voxel in space. Viewers watch voxels from different viewpoints, and can see different sides of voxels, thereby forming stereoscopic vision.
Constructing voxels with a large depth of field requires denser controllable beams to ensure continuous parallax jump and a suitable optical structure to reduce the influence of crosstalk between beams. The halftone points in the horizontal direction of the light control unit are fully utilized by the 1D light-field encoding algorithm, so the number of controllable light beams under one light control unit is the number of halftone points in the horizontal direction. In this way, the angular resolution of the light control unit is greatly improved. At the same time, spatial resolution is also extremely necessary for constructing large DOF voxels. Excessive size of the light control unit (insufficient spatial resolution) will lead to insufficient precision of voxel construction for large depth of field. The 1D light field encoding algorithm makes full use of the halftone encoding array without enlarging the size of the light control unit to improve the angular resolution and realize the display requirement of a large depth of field.
5. Joint modulation for size and arrangement of halftone dots (JMSAHD)
As shown in Fig. 5, the encoded pixels of spatial voxels are obtained after light-field coding. The encoded pixels are separated by RGB three channels, and the grayscale of the three channels are respectively screened to form the coding result of a 145.67 × 1 halftone film. Therefore, in order to make the 3D image meet the coding expectation, a LCU should be composed of 145.67 × 145.67 points. Therefore, the number of multi-angle beams controlled by a LCU is 145.67. Compared with the LCU of the common light-field display now, the number of horizontal beams has been greatly improved, thus improving the horizontal viewpoint density, and its DOF has been greatly improved. Detailed results are shown in Section 7.
Fig. 5. Schematic diagram of the production process of halftone phenanthroline map and the color error problems faced.
In order to maintain spatial resolution not lower than the human eye resolution, the size of the LCU cannot be expanded indefinitely. Additionally, the optimization and manufacturing of optical lenses are constrained by technology, leading to a lower limit on the radius of the diffuse spot. Moreover, the LCU must remain square. The combined effect of these factors is that, at a printing resolution of 4000 dpi, the number of controllable directional lights in the LCU does not meet 255 × 3, and the number of dots composing a single directional light also fails to meet 255 × 3. Consequently, in high color-depth displays, it is difficult for static light-field display devices to accurately load color information of directional lights.
Figure 5 clearly illustrates that the encoded halftone dot is square-shaped, which is advantageous for computer processing, whereas the printed halftone dot is circular due to the laser's working mode. This results in a difference in intensity modulation between the halftone encoded map and the actual printed halftone map. To address the issue of low color-depth in the display effect, the following section proposes a solution using JMSAHD coding, which takes advantage of the small color-depth difference between the encoded and printed halftone dots.
The proposed JMSAHD utilizes an analysis of the encoded stereo image to decompose the three-channel stereo image. By leveraging laser energy to regulate the size of the ink point and adjusting the print ink point's position, the algorithm enables more precise control over color-depth, resulting in a ten-fold increase in color-depth from 145.67/3 and achieving high color-depth display.
5.1 Physical basis of JMSAHD
Figure 5 shows the production process of halftone film and the resulting color difference problem. This section uses this color difference problem for color correction and color-depth improvement.
Before laser exposure, the size of the halftone dots printed on the phenanthrene film can be controlled by adjusting the laser energy. In this section, the size of the dots is represented by the diffusivity, which is equal to 1 when the dot spacing is equal to the dot diameter. Figure 6 illustrates that the diffusivity can be reduced or expanded by adjusting the laser energy, resulting in voids or overlaps of ink dots.
After determining the diffusivity of the ink dot, as illustrated in Fig. 7, it can be calculated through statistical analysis that when the number of halftone points corresponding to a single color component of the collimated beam is N, the total number of color scales can be generated by adjusting the ink dot position and the specific color scale value:
Figure 8 illustrates that the chromatic scale coordinate axis can be constructed based on the chromatic scale number calculated by the formula, and the specific chromatic scale distribution can be determined through the chromatic scale coordinate axis.
5.2 Process analysis and result analysis of JMSAHD
The JMSAHD utilizes the area difference between the square encoded halftone point and the circular printed halftone point, as well as the diffusivity difference of the circular halftone ink point caused by the change of laser energy. By adjusting the laser energy and the position of the halftone points that constitute the same beam, the intensity and color of each beam can be adjusted more accurately, thereby constructing more color scales using a small number of halftone points.
Figure 8 demonstrates that different diffusivities correspond to different color scale arrays, where one diffusivity corresponds to a set of color scales that can be generated. By analyzing the color information of the 3D image, the judgment of the JMSAHD encoding fitting can be made, and the most suitable color scale combination for the 3D image can be selected, thus adjusting the laser energy using the most appropriate diffusivity.
It can be seen from Fig. 9 that the production process of the halftone phenanthrene map of the light-field display system is: after acquiring the 3D image, it is processed by the JMSAHD, and then the halftone map is generated, and then the system assembly is completed for display. The JMSAHD is shown in Fig. 9(a). It analyzes the trichromatic composition of the RGB of the 3D image, selects the appropriate diffusivity according to the chromatic distribution curve, and completes the JMSAHD light-field coding.
The production process of the halftone film in the light-field display system, as depicted in Fig. 9, involves acquiring a 3D image, processing it using the JMSAHD to generate the halftone film, and completing the system assembly for display. The JMSAHD is illustrated in Fig. 9(a).
By analyzing the color distribution of the three colors in the 3D image, the weights of the color gradation of the three colors are determined and summarized. The diffusivity of the halftone exposure points suitable for the color gradation distribution of the image is calculated, so that the light intensity and color information of the light beams used to construct the spatial volume pixels in the system are more accurate. Figure 9(b) shows a comparison of the display effects of the light-field display system before and after processing with the hyperchromatic algorithm, revealing that the image generated using the hyperchromatic algorithm has a more uniform and coherent color gradation distribution.
6. Experimental results
In this paper, a true-color light-field display system with a large DOF is constructed, which is composed of collimated beams, a halftone film with a printing precision of 4000 dpi, an ACLA, a VDF and an RGB optical filter. RGB color filters are used to control the color of the light beams. Figure 10(a) illustrates the microstructure characteristics of the RGB filter used in this experiment. Compared with the reflective CMYK color, RGB color filters have a larger color gamut improvement space.
After testing, the color gamut of the light-field display system is shown in Fig. 10(b), which is approximately 103% of the sRGB color gamut width. The beam does not widen in the vertical direction before passing through the VDF, so the diffusion angle of the VDF in the vertical direction is the vertical visible angle of the display system. The VDF is an anisotropic diffusion film made by Luminit with a diffusion angle of 60° × 1°.
The pictures in Fig. 11 is are real shots of the monochrome light-field display system from different angles, and the pictures in Fig. 12 are real shots of the color light-field display system from different angles. The red box represents the depth range of the 3D model displayed by the system. Monochrome tigers and colored tigers are displayed as 3D virtual images. The display system has a depth of field of 50 cm and a viewing angle of 100°. According to the image analysis, the depth of field of the display device is about 50 cm, and the ratio between the DOF and the format is 1.25:1.
Fig. 11. Multi-angle display effect of monochromatic light-field display system. (see Visualization 1)
At present, the light field display system cannot meet the requirements of dynamic display. Because the current film cannot be dynamically displayed. Yet, we believe that this display technology can achieve dynamic 3D display in the future. We can design a device similar to a movie film projector to achieve dynamic display by rolling the film. And we may replace the film with a high-precision ink screen to achieve dynamic 3D display.
Funding
National Key Research and Development Program of China (2021YFB3600504); National Natural Science Foundation of China (62075016, 62175015).
Disclosures
The authors declare no conflicts of interest. This work is original and has not been published elsewhere.
Data availability
Underlying the results presented in this paper are not publicly available at this time but maybe obtained from the authors upon reasonable request.
References
1. B Javidi, A Carnicer, J Arai, et al., “Roadmap on 3D integral imaging: sensing, processing, and display[J],” Opt. Express 28(22), 32266–32293 (2020). [CrossRef]
2. Manuel Martínez-Corral and Bahram Javidi, “Fundamentals of 3D imaging and displays: a tutorial on integral imaging, light-field, and plenoptic systems,” Adv. Opt. Photonics 10(3), 512–566 (2018). [CrossRef]
3. Y Son J, H Lee, R Lee B, et al., “Holographic and light-field imaging as future 3-D displays[J],” Proc. IEEE 105(5), 789–804 (2017). [CrossRef]
4. W Zhao, B Liu, H Jiang, et al., “Full-color hologram using spatial multiplexing of dielectric metasurface[J],” Opt. Lett. 41(1), 147–150 (2016). [CrossRef]
5. S Yang, X Sang, X Yu, et al., “162-inch 3D light field display based on aspheric lens array and holographic functional screen[J],” Opt. Express 26(25), 33013–33021 (2018). [CrossRef]
6. H Hong S, S Jang J, and B Javidi, “Three-dimensional volumetric object reconstruction using computational integral imaging[J],” Opt. Express 12(3), 483–491 (2004). [CrossRef]
7. F Jin, S Jang J, and B Javidi, “Effects of device resolution on three-dimensional integral imaging[J],” Opt. Lett. 29(12), 1345–1347 (2004). [CrossRef]
8. Xunbo Yu, Xinzhu Sang, Xin Gao, Zhidong Chen, Duo Chen, Wei Duan, Binbin Yan, Chongxiu Yu, and Daxiong Xu, “Large viewing angle three-dimensional display with smooth motion parallax and accurate depth cues,” Opt. Express 23(20), 25950–25958 (2015). [CrossRef]
9. D Wang, C Liu, C Shen, et al., “Holographic capture and projection system of real object based on tunable zoom lens[J],” PhotoniX 1(1), 6–15 (2020). [CrossRef]
10. L Li Y, N Li N, D Wang, et al., “Tunable liquid crystal grating based holographic 3D display system with wide viewing angle and large size[J],” Light: Sci. Appl. 11(1), 188 (2022). [CrossRef]
11. H Adelson E and R Bergen J, Michael Landy and J Anthony Movshoneds., “The plenoptic function and the elements of early vision,” Computational Models of Visual Processing (1991).
12. A Isaksen, L McMillan, and S J. Gortler, Dynamically reparameterized light-fields//Proceedings of the 27th annual conference on computer graphics and interactive techniques. 2000: 297–306.
13. X Liu and H Li, “The Progress of Light-Field 3D Displays[J],” Inf. Disp. 30(6), 6–14 (2014). [CrossRef]
14. Xinzhu Sang, Xin Gao, Xunbo Yu, Shujun Xing, Yuanhang Li, and Yongle Wu, “Interactive floating full-parallax digital three-dimensional light-field display based on wavefront recomposing,” Opt. Express 26(7), 8883–8889 (2018). [CrossRef]
15. H. Huang and H. Hua, “Systematic characterization and optimization of 3D light-field displays,” Opt. Express 25(16), 18508–18525 (2017). [CrossRef]
16. C. Yao, D. Cheng, T. Yang, and Y . Wang, “Design of an optical see-through light-field near-eye display using a discrete lenslet array,” Opt. Express 26(14), 18292–18301 (2018). [CrossRef]
17. L. Y ang, X. Sang, X. Yu, B. Liu, L. Liu, S. Y ang, B. Y an, J. Du, and C. Gao, “Demonstration of a large-size horizontal light-field display based on the LED panel and the micro-pinhole unit array,” Opt. Commun. 414, 140–145 (2018). [CrossRef]
18. G. Chen, C. Ma, D. Zhao, Z. Fan, and H. Liao, “360 degree crosstalk-free viewable 3D display based on multiplexed light-field: theory and experiments,” J. Disp. Technol. 12(11), 1309–1318 (2016). [CrossRef]
19. Y . Zhu, X. Sang, X. Yu, P. Wang, S. Xing, D. Chen, B. Y an, K. Wang, and C. Yu, “Wide field of view tabletop light-field display based on piece-wise tracking and off-axis pickup,” Opt. Commun. 402, 41–46 (2017). [CrossRef]
20. L. Yang, X. Sang, X. Yu, B. Y ang, B. Y an, K. Wang, and C. Yu, “A crosstalk-suppressed dense multi-view light-field display based on real-time light-field pickup and reconstruction,” Opt. Express 26(26), 34412–34427 (2018). [CrossRef]
21. D. Chen, X. Sang, X. Yu, X. Zeng, S. Xie, and N. Guo, “Performance improvement of compressive light-field display with the viewing-position-dependent weight distribution,” Opt. Express 24(26), 29781–29793 (2016). [CrossRef]
22. N Balram and I. Tošić, “Light-field imaging and display systems[J],” Inf. Disp. 32(4), 6–13 (2016). [CrossRef]
23. C Kim, H Zimmer, Y Pritch, et al., “Scene reconstruction from high spatio-angular resolution light-fields[J],” ACM Trans. Graph. 32(4), 1–12 (2013). [CrossRef]
24. G Wu, B Masia, A Jarabo, et al., “Light-field image processing: An overview[J],” IEEE J. Sel. Top. Signal Process. 11(7), 926–954 (2017). [CrossRef]
25. D Nam, H Lee J, H Cho Y, et al., “Flat panel light-field 3-D display: concept, design, rendering, and calibration[J],” Proc. IEEE 105(5), 876–891 (2017). [CrossRef]
26. P Wang, S Xie, X Sang, et al., “A large depth of field frontal multi-projection three-dimensional display with uniform light field distribution[J],” Opt. Commun. 354, 321–329 (2015). [CrossRef]
27. M Zwicker, W Matusik, F Durand, et al., “Antialiasing for automultiscopic 3D displays[M]//ACM SIGGRAPH 2006 Sketches. 2006: 107-es.
28. B Liu, X Sang, X Yu, et al., “Time-multiplexed light-field display with 120-degree wide viewing angle[J],” Opt. Express 27(24), 35728–35739 (2019). [CrossRef]
29. H Ting C, C Chang Y, H Chen C, et al., “Multi-user 3D film on a time-multiplexed side-emission backlight system[J],” Appl. Opt. 55(28), 7922–7928 (2016). [CrossRef]
30. S Hwang Y, K Bruder F, T Fäcke, et al., “Time-sequential autostereoscopic 3-D display with a novel directional backlight system based on volume-holographic optical elements[J],” Opt. Express 22(8), 9820–9838 (2014). [CrossRef]
31. L Yang, X Sang, X Yu, et al., “Viewing-angle and viewing-resolution enhanced integral imaging based on time-multiplexed lens stitching[J],” Opt. Express 27(11), 15679–15692 (2019). [CrossRef]
32. X Xia, X Zhang, L Zhang, et al., “Time-multiplexed multi-view three-dimensional display with projector array and steering screen[J],” Opt. Express 26(12), 15528–15538 (2018). [CrossRef]
33. B Phillips D, J Sun M, M Taylor J, et al., “Adaptive foveated single-pixel imaging with dynamic super sampling[J],” Sci. Adv. 3(4), e1601782 (2017). [CrossRef]
34. A Cem, K Hedili M, E Ulusoy, et al., “Foveated near-eye display using computational holography[J],” Sci. Rep. 10(1), 14905 (2020). [CrossRef]
35. J Hua, E Hua, F Zhou, et al., “Foveated glasses-free 3D display with ultrawide field of view via a large-scale 2D-metagrating complex[J],” Light: Sci. Appl. 10(1), 213 (2021). [CrossRef]
36. T Fujine, T Kanda, Y Yoshida, et al., “Bit depth needed for high image quality TV-evaluation using color distribution index[J],” J. Disp. Technol. 4(3), 340–347 (2008). [CrossRef]
37. W Lu C and C Huang L, “A 10-bit LCD column driver with piecewise linear digital-to-analog converters[J],” IEEE J. Solid-State Circuits 43(2), 371–378 (2008). [CrossRef]
