Three-dimensional computer holography enabled from a single 2D image

Automotive Augmented Reality Head-Up Displays

As the next generation of in-vehicle intelligent platforms, the augmented reality heads-up display (AR-HUD) has a huge information interaction capacity, can provide drivers with auxiliary driving information, avoid the distractions caused by the lower head during the driving process, and greatly improve driving safety. However, AR-HUD systems still face great challenges in the realization of multi-plane full-color display, and they cannot truly achieve the integration of virtual information and real road conditions. To overcome these problems, many new devices and materials have been applied to AR-HUDs, and many novel systems have been developed. This study first reviews some key metrics of HUDs, investigates the structures of various picture generation units (PGUs), and finally focuses on the development status of AR-HUDs, analyzes the advantages and disadvantages of existing technologies, and points out the future research directions for AR-HUDs.

Liquid lens based holographic camera for real 3D scene hologram acquisition using end-to-end physical model-driven network

With the development of artificial intelligence, neural network provides unique opportunities for holography, such as high fidelity and dynamic calculation. How to obtain real 3D scene and generate high fidelity hologram in real time is an urgent problem. Here, we propose a liquid lens based holographic camera for real 3D scene hologram acquisition using an end-to-end physical model-driven network (EEPMD-Net). As the core component of the liquid camera, the first 10 mm large aperture electrowetting-based liquid lens is proposed by using specially fabricated solution. The design of the liquid camera ensures that the multi-layers of the real 3D scene can be obtained quickly and with great imaging performance. The EEPMD-Net takes the information of real 3D scene as the input, and uses two new structures of encoder and decoder networks to realize low-noise phase generation. By comparing the intensity information between the reconstructed image after depth fusion and the target scene, the composite loss function is constructed for phase optimization, and the high-fidelity training of hologram with true depth of the 3D scene is realized for the first time. The holographic camera achieves the high-fidelity and fast generation of the hologram of the real 3D scene, and the reconstructed experiment proves that the holographic image has the advantage of low noise. The proposed holographic camera is unique and can be used in 3D display, measurement, encryption and other fields.

End-to-end real-time holographic display based on real-time capture of real scenes

Holographic display is considered as a promising three-dimensional (3D) display technology and has been widely studied. However, to date, the real-time holographic display for real scenes is still far from being incorporated in our life. The speed and quality of information extraction and holographic computing need to be further improved. In this paper, we propose an end-to-end real-time holographic display based on real-time capture of real scenes, where the parallax images are collected from the scene and a convolutional neural network (CNN) builds the mapping from the parallax images to the hologram. Parallax images are acquired in real time by a binocular camera, and contain depth information and amplitude information needed for 3D hologram calculation. The CNN, which can transform parallax images into 3D holograms, is trained by datasets consisting of parallax images and high-quality 3D holograms. The static colorful reconstruction and speckle-free real-time holographic display based on real-time capture of real scenes have been verified by the optical experiments. With simple system composition and affordable hardware requirements, the proposed technique will break the dilemma of the existing real-scene holographic display, and open up a new direction for the application of real-scene holographic 3D display such as holographic live video and solving vergence-accommodation conflict (VAC) problems for head-mounted display devices.

Compact reconstruction of a Fourier hologram for a 3D object by scaling compensation

The Fourier holographic projection method is compact and computationally fast. However, since the magnification of the displayed image increases with the diffraction distance, this method cannot be used directly to display multi-plane three-dimensional (3D) scenes. We propose a holographic 3D projection method of Fourier holograms by scaling compensation to offset the magnification during optical reconstruction. To achieve a compact system, the proposed method is also used to reconstruct 3D virtual images with Fourier holograms. Different from traditional Fourier holographic displays, images are reconstructed behind a spatial light modulator (SLM) so that the observation position can be placed close to the SLM. The effectiveness of the method and the flexibility of combining it with other methods are confirmed by simulations and experiments. Therefore, our method could have potential applications in the augmented reality (AR) and virtual reality (VR) fields.

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.