Generating Multi-Depth 3D Holograms Using a Fully Convolutional Neural Network.

Efficiently generating 3D holograms is one of the most challenging research topics in the field of holography. This work introduces a method for generating multi-depth phase-only holograms using a fully convolutional neural network (FCN). The method primarily involves a forward-backward-diffraction framework to compute multi-depth diffraction fields, along with a layer-by-layer replacement method (L2RM) to handle occlusion relationships. The diffraction fields computed by the former are fed into the carefully designed FCN, which leverages its powerful non-linear fitting capability to generate multi-depth holograms of 3D scenes. The latter can smooth the boundaries of different layers in scene reconstruction by complementing information of occluded objects, thus enhancing the reconstruction quality of holograms. The proposed method can generate a multi-depth 3D hologram with a PSNR of 31.8 dB in just 90 ms for a resolution of 2160 × 3840 on the NVIDIA Tesla A100 40G tensor core GPU. Additionally, numerical and experimental results indicate that the generated holograms accurately reconstruct clear 3D scenes with correct occlusion relationships and provide excellent depth focusing.

The U-Net-based phase-only CGH using the two-dimensional phase grating.

In this paper, the phase-only holograms with clear first diffraction orders have been generated based on the U-Net and the two-dimensional phase grating. Firstly, we proved the modulation effect of two-dimensional phase grating on diffraction field, and came to a conclusion that it could move the diffraction pattern of the hologram to the odd-numbered diffraction orders' center of that. Then we changed the generation process of phase-only holograms and the training strategy for U-Net according to this conclusion, which converted the optimization target of the U-Net from the zeroth diffraction order in the center of diffraction field to the first diffraction order in the edge of that. And we also used a method called "phase recombination" to improve the structure of U-Net for less memory footprint and faster generating speed. Finally, the holograms with the 4K resolution have been generated in 0.05s, and the average peak signal to noise ratio (PSNR) of the reconstructed images is about 37.2 dB in DIV2K-valid-HR dataset.

Phase-only hologram generated by a convolutional neural network trained using low-frequency mixed noise.

A phase-only hologram generated through the convolution neutral network (CNN) which is trained by the low-frequency mixed noise (LFMN) is proposed. Compared with CNN based computer-generated holograms, the proposed training dataset named LFMN includes different kinds of noise images after low-frequency processing. This dataset was used to replace the real images used in the conventional hologram to train CNN in a simple and flexible approach. The results revealed that the proposed method could generate a hologram of 2160 × 3840 pixels at a speed of 0.094 s/frame on the DIV2K valid dataset, and the average peak signal-to-noise ratio of the reconstruction was approximately 29.2 dB. The results of optical experiments validated the theoretical prediction. The reconstructed images obtained using the proposed method exhibited higher quality than those obtained using the conventional methods. Furthermore, the proposed method considerably mitigated artifacts of the reconstructed images.

Measurement and correction of the macrolens array's position error in integral imaging.

The macrolens array is generally used in the large-format integral imaging display system, and the position error of the lens will affect the result of the integral imaging display. In this paper, the average value and variance of the distances from the distributed reconstructed points to their ideal points are used to measure the size of the position error, and the performance of the measurement methods is discussed. The one-to-one relationship of the pixels of the elemental image under the ideal case and the position error are analyzed, and the corresponding correction methods of these position errors are analyzed. Then the pixels are rearranged to compensate for the position error, thereby alleviating the influence of the microlens array's (MALA's) position error on the imaging quality, which is verified by theoretical simulation and experimental results.

Implementation of the real-virtual 3D scene-fused full-parallax holographic stereogram.

This work focuses on the generation of three-dimensional (3D)-scene information as well as the fusion of real and virtual 3D scene information for the full-parallax holographic stereogram based on the effective perspective images' segmentation and mosaicking (EPISM) method. The improved depth-image-based rendering (DIBR) method was used to generate the virtual viewpoint images of the real 3D scene, and the regularization and densification processing models of the degraded light field were established; as a result, the real sampling-light field was reconstructed. Combined with the computer-rendered virtual 3D scene information, a "real + virtual" light-field fusion method based on a pixel-affine-projection was proposed to realize the fusion of the real and virtual 3D scene. The fusion information was then processed by the EPISM encoding and was then holographically printed. The optical experiment results showed that the full-parallax holographic stereogram with the real-virtual scene-fused 3D scenes could be correctly printed and reconstructed, which validated the effectiveness of our proposed method.

View-flipping effect reduction and reconstruction visualization enhancement for EPISM based holographic stereogram with optimized hogel size.

To reduce the view-flipping effect and enhance the viewing resolution, the modulation characteristics of the hogel based holographic stereogram is constructed and validated. The performance of the view-flipping effect is analyzed, and the results indicate that decreasing the size of hogel is beneficial to the reduction of the view flipping, however, which will result in significant diffraction effect which can degrade the reconstruction quality. Furthermore, a diffraction-limited imaging model of the hogel based holographic stereogram is established, where both the limited aperture of the hogel and the defocused aberration of the object point are introduced, and the effective resolvable size of the reconstructed image point is simulated. The theory shows that there is an optimal hogel size existed for the certain depth of scene. Both the numerical and optical experiments are implemented, and the results are well agreed with the theoretical prediction, which demonstrates that the view-flipping reduction and reconstruction visualization enhancement for EPISM based holographic stereogram can be achieved when the proper size of hogel is utilized.

Centered-camera-based effective perspective images' segmentation and mosaicking method for full-parallax holographic stereogram printing.

The centered-camera-based effective perspective images' segmentation and mosaicking (CCEPISM) method is proposed to improve the previous EPISM-based printing of full-parallax holographic stereograms. The idea of pixel mapping from target camera images to centered camera images is analyzed, and backward pixel mapping as well as forward pixel mapping are modeled and formulated to transform the perspective images captured by different camera strategies. The principle of the proposed CCEPISM is introduced in detail along with its specific implementation. The experimental results validate the proposed method and demonstrate that CCEPISM is an effective method for printing full-parallax holographic stereograms, and the quality of optical reconstruction is the same as that of EPISM-based holographic stereograms. The whole 3D scene is captured by a centered camera image, and the pixel waste caused by simple camera capture and the number of sampled perspective images is reduced significantly.

Analysis on the reconstruction error of EPISM based full-parallax holographic stereogram and its improvement with multiple reference plane.

To reduce the reconstruction error of holographic stereogram fabricated by effective perspective images' segmentation and mosaicking method (EPISM), a multiple-reference-plane (MRP) approach is proposed and validated. The reconstruction error for traditional EPISM is analyzed, and the results indicate that the distortion as well as the blur will be involved for object points located far away from the reference plane. A new method by introducing multiple reference planes is proposed, which divides the 3D scene into several parts along its depth direction, and sets a reference plane for each of the object part. By resynthesizing all the effectively synthetic perspective images referred to their own reference planes of the object parts, the finally effectively synthetic perspective image exposed to one holographic elemental by only once exposure is generated. The optically experimental results demonstrate the validity of the proposed method, and the reconstruction error of full-parallax holographic stereogram printed by MRP based EPISM can be reduced evidently while the displayed depth range of 3D scene can be extended, compared to the traditional EPISM approach.

Integral imaging based light field display with holographic diffusor: principles, potentials and restrictions.

Under the framework of light field, the diffusing angle and spatial location of holographic diffusor is investigated for the integral imaging based light field display. These two parameters are considered in terms of continuous light field reconstruction of the object point and blind visual spots elimination for the viewer. The concept of joint view reconstruction as well the essence of this view interpolation is analyzed theoretically, and a new phenomenon caused by the additional holographic plane called double window violation is also deduced. To the best of our knowledge, this is the first report on the quantitative analysis of the two key parameters of the holographic diffusor, which is deduced theoretically and verified experimentally.

Characteristic and improvement on the reconstructed quality of effective perspective images' segmentation and mosaicking-based holographic stereogram.

Based on the effective perspective images' segmentation and mosaicking (EPISM) printing method, the influence of the holographic element (hogel) size on the reconstructed quality is analyzed to improve the reconstructed quality of holographic stereogram. The flipping effect and diffraction effect in the spatial domain are also discussed, and the optical transfer function of the holographic stereogram is used in the spectrum domain to evaluate the reconstructed quality. Theoretical analysis and optical experimental results show that the hogel size plays an important role on the flipping effect as well as the clarity of the reconstrued 3D perspective images of the EPISM-based holographic stereogram, and the flipping effect can be improved significantly with optimized hogel size.

Post-calibration compensation method for integral imaging system with macrolens array.

Three-dimensional display with large-format is an inevitable and foreseeable trend for the future display technology, integral imaging is one of the most powerful and promising candidates to achieve this goal with full-parallax, true-color, acceptable viewing resolution and viewing angle. To obtain a 3D display with high quality, calibration is needed to correct optical misalignment and optical aberrations, which is often challenging and time consuming. We propose a post-calibration compensation method for the integral imaging system with macrolens array, the inter-lens position misalignment is corrected by forcing it to image in a regular ideal reference grid. Our method distinguishes itself from previous ones by finding the correct pixel-to-ray correspondence with a relatively simple setup and acceptable precision. A prototype is fabricated to evaluate the feasibility of the proposed method. Furthermore, the proposed method is evaluated in terms of the geometrical accuracy and quality of the reconstructed images.

Improvement of printing efficiency in holographic stereogram printing with the combination of a field lens and holographic diffuser.

In this paper, we use a field lens and a holographic diffuser together to improve the printing efficiency of a holographic stereogram printing system based on the effective perspective images' segmentation and mosaicking method. The light rays' regulation function of the field lens and the modulation function of the holographic diffuser are analyzed. Holographic diffusers with different expanding angles are optimized by numerical simulations and verified by optical experiments. We can achieve a better holographic stereogram reconstruction effect as well as a better printing efficiency when adopting a field lens and a 10° holographic diffuser together. With the proposed method, the energy efficiency can be improved, and the printing time can be reduced greatly.

Nonlinear mapping method for the generation of an elemental image array in a photorealistic pseudoscopic free 3D display.

Limited available elemental image array resources may be the most severe bottleneck for the promotion and application of integral-imaging-based 3D display. We propose a nonlinear mapping method for the generation of an elemental image array to get a photorealistic pseudoscopic free 3D display based on the parallel light field reconstruction nature of the integral imaging system. All the light rays emitted from the display panel are classified into a corresponding parallel light field according to its direction, and all the parallel light fields are captured as orthogonal projections of the scene before synthesizing all the orthogonal projections using a nonlinear mapping method to form the final elemental image array. Preliminary optical experiments as well as ray optical analysis are conducted to prove the feasibility and validity of the proposed method. The proposed method can exploit most of the current 3D platform. It is an effective and efficient way to generate an elemental image array.

Characteristic and optimization of the effective perspective images' segmentation and mosaicking (EPISM) based holographic stereogram: an optical transfer function approach.

Based on our proposed method for holographic stereogram printing using effective perspective images' segmentation and mosaicking (EPISM), we analyze the reconstructed wavefront errors, and establish the exit pupil function model of proposed printing system. To evaluate the imaging quality, the optical transfer function (OTF) of the holographic stereogram is modelled from the aspect of frequency response. The characteristic of the OTF with respect to the exit pupil size and the aberration are investigated in detail. We also consider the flipping effect in spatial domain. The optimization of hogel sizes, i.e., the sampling interval of original perspective images and the printing interval of synthetic effective perspective images, are given for the optimized reconstruction. Numerical simulations and optical experiments are implemented, and the results demonstrate the validity of our analysis, and the optimized parameters of hogel sizes can improve the imaging quality of full parallax holographic stereogram effectively.

One-step real-virtual combined reflection hologram: a 4f relay approach.

The production of conventional optical reflection holograms can be classified into the one-step method or the two-step method. In the one-step method, only the diverging light of an object can be recorded, and the reconstructed scene is a virtual one behind the recording medium. In the two-step method, the diverging light or the converging light can be recorded alternatively. However, the process is complicated considering double exposures. The object is first imaged by a 4f system, and then the interference patterns are recorded by single exposure. The reconstructed image can be either a virtual image behind the recording medium or a real image in front of the recording medium. The ideal imaging property of a 4f system has been demonstrated theoretically, and the proposed method has been verified experimentally.

Computational integral imaging reconstruction of perspective and orthographic view images by common patches analysis.

A novel method to computationally reconstruct perspective and orthographic view images with full resolution of a recording device from a single integral photograph is proposed. Firstly, a group of image slices that contain full yet redundant information to reconstruct the view image are generated, and the object surface is divided into pieces by the points that correspond to the centers of image slices. Secondly, the image slices that contribute to the pieces are extracted and redundant information embedded in them are figured out by common patches analysis. Finally, the view image is reconstructed by excluding the redundant information and resampling with maximum sampling rate. Each piece of the object surface is represented with 9 patches at most from 4 adjacent elemental images, and view images with high quality are reconstructed. Both simulations and experiments verify the validity of the method.

Method of single-step full parallax synthetic holographic stereogram printing based on effective perspective images' segmentation and mosaicking.

With the principle of ray-tracing and the reversibility of light propagation, a new method of single-step full parallax synthetic holographic stereogram printing based on effective perspective images' segmentation and mosaicking (EPISM) is proposed. The perspective images of the scene are first sampled by a virtual camera and the exposing images, which are called synthetic effective perspective images, are achieved using the algorithm of effective perspective images' segmentation and mosaicking according to the propagation law of light and the viewing frustum effect of human eyes. The hogels are exposed using the synthetic effective perspective images in sequence to form the whole holographic stereogram. The influence of modeling parameters on the reconstructed images are also analyzed, and experimental results have demonstrated that the full parallax holographic stereogram printing with the proposed method could provide good reconstructed images by single-step printing. Moreover, detailed experiments with different holographic element sizes, different scene reconstructed distances, and different imaging planes are also analyzed and implemented.

High-power 266 nm ultraviolet generation in yttrium aluminum borate.

A yttrium aluminum borate [YAl(3)(BO(3))(4)] (YAB) crystal with UV cutoff wavelength of 165 nm is used as the nonlinear optical crystal for fourth harmonic generation. The fundamental frequency laser at 1064 nm from an Nd:YVO(4) master oscillator power amplifier laser was frequency doubled to 532 nm. Using the type I phase-matching YAB crystal, a 5.05 W average power 266 nm UV laser was obtained at the pulse repetition frequency of 65 kHz, corresponding to the conversion efficiency of 12.3% from 532 to 266 nm. The experimental results show great potential for the application of using YAB as a nonlinear optical crystal to get high-power fourth harmonic generation.

Combined guiding effect in the end-pumped laser resonator.

A theoretical model as well as the experimental verification of the combined guiding mechanism for the transverse mode formation in the end-pumped laser resonator are investigated. The nonlinear Schrödinger-type wave equation in the gain medium is derived, in which the combined guiding mechanism: the thermal induced refractive index guiding effect as well as the gain guiding effect, is taken into account. The gain saturation and spatial hole burning are considered. The split step Fourier method is used to solve the nonlinear wave equation. A high power end-pumped Nd:YVO4 laser resonator is built up. After establishing the pump absorption model of our laser resonator, the temperature distribution in the gain medium is obtained by the numerical solving of the heat diffusion equation. The combined guiding effect is first observed in the end-pumped Nd:YVO4 laser resonator, and the experimental transverse mode profiles well agree with the theoretical prediction from the derived nonlinear Schrödinger-type wave equation. The geometric design criterion of the TEM00 mode laser is compared with our wave theory. The experimental- and theoretical- results show that our wave theory with the combined guiding mechanism dominates the transverse mode formation in high power end-pumped laser resonator.

Symmetric TEM00 output from Q-switched quasi-concentric laser resonator with line-shaped end-pumping profile.

We report a 12.2 W pulsed TEM(00) output from Q-switched quasi-concentric laser resonator with line-shaped end-pumping profile, with the repetition rate of 30 kHz and optical-optical efficiency of 27.1%. The laser output mode is symmetrized in two directions in terms of beam quality, waist radius, and waist position.