Conference Program:

3DSA1/3D1: Holography (1)

Dec. 6  13:00-14:20  Room 301 A

Chair: Kyoji Matsushima (Kansai University)
Co-Chair: Tomohiro Yendo (Nagaoka Univ. of Tech.)

3DSA1/3D1-1
13:00
Invited A Novel Transparent 3D Display Based on Integration of Passive and Active Optical Elements
*Hoang Yan Lin Lin1, Chia-Yuan Chang1
1. National Taiwan University (Taiwan)
We want to create a floating 3D image with integral imaging based 3D light field display and with an optical transparent display. We develop the Computer Generated Elemental Image (CGEI) method to create an image with a different depths from a two-dimensional image and a depth map. We create a new weight function to increase the image qualities. Also, creating an optical transparent light-field display ...
3DSA1/3D1-2
13:20
Invited Holographic Image Reconstruction Using a Magneto-Optical Spatial Light Modulator with Its Pixel Pitch of 1 μm and its Layout of 10k × 5k
*Ken-ichi Aoshima1, Nobuhiko Funabashi1, Ryo Higashida1, Mayumi Kawana1, Yuta Yamaguchi1, Kenji Machida1
1. Japan Broadcasting Corporation (Japan)
We have demonstrated successful holographic images reconstruction using a magneto-optical spatial light modulator (MO-SLM) which magnetization is controlled by current induced domain wall motion. The MO-SLM has the pixel pitch of one micron for horizontal direction showing large viewing zone angle of 30 degrees.
3DSA1/3D1-3
13:40
Invited Digitally Printed Holograms for Efficient Generation of Structured Light
*Jackin Boaz Jessie1, Kenji Kinashi1, Naoto Tsutsumi1, Wataru Sakai1
1. Kyoto Institute of Technology (Japan)
Here we combine analog hologram recording material and digital hologram printing technique, for efficient generation of structured light beams. A set of 100 vortex beams were generated from a hologram of size 1mm x 1mm with high diffraction efficiency. This method can aid future optical communication and information processing applications.
3DSA1/3D1-4
14:00
Using Time Multiplexing with Spatial Multiplexing to Reduce Speckle for AR-HUD
*Hung-I Chang1, Hoang-Yan Lin1, Yen-Jung Chen1
1. National Taiwan University (Taiwan)
The AR-HUD system based on LCoS and using laser as a light source has a speckle issue severely affects the image quality.This paper proposes using time and spatial multiplexing, a configuration effective method to reduce the speckle problem in the experiment.

 

3D2/3DSA2: Human Factor on Realistic Display

Dec. 6  14:50-16:10  Room 301 A

Chair: Chien-Yu Chen (National Taiwan University of Science and Technology)
Co-Chair: Takashi Yamazoe (Seikei University)

3D2/3DSA2-1
14:50
Relationship between Field of View Size and Head Motion in Eyepiece Displays
*Sumio YANO1,2, Shuichi Ojima3
1. Shimane University (Japan), 2. RITECS Inc. (Japan), 3. Sojo University (Japan)
We examined the relationship between the field of view and head motion on an eyepiece display and found that head motion is narrower in a wide field of view than in a narrow field of view, while velocity depends on the frequency component of the image.
3D2/3DSA2-2
15:10
HEVC Subjective and Objective Image Quality Assessment Considering Cross Reality Based on 360 Degrees Camera
*Norifumi Kawabata1
1. Kanazawa Gakuin University (Japan)
In this paper, we generated 360 degrees camera images by using RICOH THETA Z1. Next, we processed of H.265/HEVC for generated images, and estimated fundamentally how the coded image quality is controlled by changing of image region or coding parameter, finally, we considered application for eXtended reality using Xperia View.
3D2/3DSA2-3
15:30
Effects of Motion Parallax Smoothness and Head Moving Range on Reduction of The Cardboard Effect
*Kosuke Takahashi1, Haruki Mizushina1, Shiro Suyama2, Kenji Yamamoto1
1. Tokushima University (Japan), 2. Utsunomiya University (Japan)
Smoother motion parallax and wider head movements contribute to reduce the cardboard effect in binocular stereoscopic display. The cardboard effect, a perceptual distortion in stereoscopic displays that makes objects appear flat, can be reduced by adding motion parallax.
3D2/3DSA2-4
15:50
Multi-view DFD (Depth-Fused 3D) Rendering with Linear Blending
*Yuto Morishita1, Munekazu Date2, Haruki Mizushina1, Shiro Suyama1, Kenji Yamamoto1
1. Tokushima University (Japan), 2. Tokushima Bunri University (Japan)
The usefulness of Multi-view DFD rendering with linear blending is demonstrated. A combination of Multi-view DFD rendering and linear blending was considered to reduce the number of cameras in stacked multi-view display. The depth of 3D images was evaluated at several camera-to-camera distance.

 

3DSA3/3D3: Light Field/Multiview/Stereo/Aerial (1)
Special Topic of Interest on AR/VR and Hyper Reality

Dec. 6  16:40-17:40  Room 301 A

Chair: Hoang Yan Lin (National Taiwan University)
Co-Chair: Masaru Tsuchida (NTT)

3DSA3/3D3-1
16:40
Integral Imaging Using a Lens Array Composed of Fine Interleaved Fresnel Lenses
*Hiroto Omori1, Hideki Kakeya1
1. University of Tsukuba (Japan)
This paper proposes integral imaging comprising fine interleaved elemental lenses, which consist of elemental prisms whose slope angle corresponds to two adjacent convex lenses. By using fine elemental lenses, parallax between adjacent elemental images becomes small and continuity of image increases even when the depth of image is large.
3DSA3/3D3-2
17:00
Improved Eye-Tracking Function by Combining Inertial Sensor Information for Integral 3D Display
*Hisayuki Sasaki1, Masanori Kano1, Naoto Okaichi1, Hayato Watanabe1, Tomoyuki Mishina1
1. NHK (Japan)
When eye-tracking is compromised in integral 3D display owing to the deterioration of the lighting conditions of the observer's face, motion blur, or movement of the face outside the camera view, a discontinuous viewing zone is formed. Combining face image detection with inertial sensor information can lead to a robust eye-tracking function.
3DSA3/3D3-3
17:20
Depth Enhancement of Time Multiplexed Light Field Display by Multilayering Display Surface Using Two Lens Arrays
*Tatsuya Shiratori1, Kengo Fujii1, Tomohiro Yendo1
1. Nagaoka University of Technology (Japan)
One of the problems with light field displays is that the resolution of the displayed object decreases with distance from the display surface. To solve this problem, we propose a multilayer display using two lens arrays and a high-speed projector. The resolution improvement is confirmed by simulation.

 

3D4/3DSA4: Light Field/Multiview/Stereo/Aerial (2)
Special Topic of Interest on AR/VR and Hyper Reality

Dec. 7  9:00-10:20  Room 301 A

Chair: Toshiaki Fujii (Nagoya Univ.)
Co-Chair: Yongjun Lim (ETRI)

3D4/3DSA4-1
9:00
Invited VR Noh “GHOST IN THE SHELL” ~Immersive Style Stage Direction without HMD !
*Shutaro Oku1
1. The University of Tokyo (Japan)
Immersive style stage direction without HMD for VR Noh “GHOST IN THE SHELL” will be achieved by applying a holographic screen on the stage and implementing multi-layer projection, low-resolution effects and gaze guidance. This foundational technology will open up the next stage as a new metaverse interface.
3D4/3DSA4-2
9:20
Development of 3D Glasses Using Aspherical Phase Plates for Endoscopic Surgery
*Kotomi Takatsu1, Yasushi Miyajima2, Kazutoshi Kato2, Yuki Kondo3, Hiroharu Kobayashi3, Yasuhiro Takaki1
1. Tokyo University of Agriculture and Technology (Japan), 2. Itoh Optical Industrial Co., Ltd. (Japan), 3. Seirei Hamamatsu General Hospital (Japan)
Three-dimensional (3D) glasses were developed for the 3D endoscopic surgeries using third-order aspherical phase plates to reduce visual fatigue. To assess their effectiveness, a subjective evaluation was conducted using a simulator sickness questionnaire in the hospital. The results of the evaluation conducted over four months are reported.
3D4/3DSA4-3
9:40
Pixel-Density Enhanced Integral Three-Dimensional Display by Time-Division Multiplexing of Two-Dimensional Image
*Hayato Watanabe1, Naoto Okaichi1, Masanori Kano1, Hisayuki Sasaki1, Jun Arai1
1. NHK (Japan)
We propose a method for displaying three-dimensional (3D) images with high maximum pixel density by synthesizing a two-dimensional image with an integral 3D image using time-division multiplexing. We developed a prototype display system and verified its display performance through display experiments.
3D4/3DSA4-4
10:00
Development of Vertical Diffusive Reflection Screen with Background Transparency for AR 3D Display
*Sei Sato1, Kengo Fujii1, Tomohiro Yendo1
1. Nagaoka University of Technology (Japan)
We propose an AR 3D display with multiple horizontal viewpoints. The background-transmitting part and the diffuse-reflecting part are alternately arranged on a screen to realize this display. The screen was engraved lines on a transparent board. The screen characteristics were also compared by changing the engraved line conditions.

 

3DSA5/3D5: Light Field/Multiview/Stereo/Aerial (3)
Special Topic of Interest on AR/VR and Hyper Reality

Dec. 7  10:50-12:10  Room 301 A

Chair: Sumio Yano (Shimane Univ.)
Co-Chair: Keita Takahashi (Nagoya University)

3DSA5/3D5-1
10:50
Embossing Volume 3D Effect by Using DCRA Plate
*Shi-Hwa Huang1, Chih-Hao Chuang2, Chien-Yu Chen1
1. National Taiwan University of Science and Technology, Taipei, Taiwan (Taiwan), 2. Feng Chia University (Taiwan)
Our research involves utilizing the DCRA to slice objects into multiple images, enabling us to display their volume. By leveraging the persistence of the visual effect in the human eye, we aim to achieve an embossed volumetric display.
3DSA5/3D5-2
11:10
Point Light Source Array Based Integral 3D Display Using Eye-Tracking and Time-Division Technologies
*Naoto Okaichi1, Hayato Watanabe1, Masanori Kano1, Hisayuki Sasaki1, Jun Arai1
1. NHK (Japan)
This article proposes an integral three-dimensional (3D) display system based on a point light source array using eye-tracking and time-division technologies. The prototype system has wide viewing angles of 100.2° horizontally and 68.6° vertically with eye tracking and a time-division display that doubles the 3D resolution both horizontally and vertically.
3DSA5/3D5-3
11:30
Distortion Reduction of Transparent Display Perceived Beyond the Surface of the Display Using Small Mirror Array Screen
*Ryosuke Kida1, Kengo Fujii1, Tomohiro Yendo1
1. Nagaoka University of Technology (Japan)
We propose an AR display oriented to applications such as store windows. We reduced the image distortion caused by moving the viewpoint, which was a problem with the method of projecting images onto a half-mirror array.
3DSA5/3D5-4
11:50
Reduction of Image Distortion for Vertically Shifted Viewpoints on Superimposed 3D Display Viewable from 360 Degrees
*Yuto Inaba1, Kengo Fujii1, Tomohiro Yendo1
1. Nagaoka University of Technology (Japan)
In the previous superimposed 3D display viewable from 360 degrees, it has a problem that stereoscopic images are distorted by viewpoint shifts. In this paper, we propose a method to improve the distortion of stereoscopic images caused by vertical viewpoint shift by changing the mirror tilt of the previous system.

 

3D6/3DSA6: Holography (2)

Dec. 7  13:30-14:50  Room 301 A

Chair: Boas Jessie Jackin (Kyoto Institute of Technologies)
Co-Chair: Ken-ichi Aoshima (NHK)

3D6/3DSA6-1
13:30
Invited Wavefront Aberration Introduced by DMDs
*Jung-Young Son2, Tetiana Venkel3, Jung Kim2, Hyoung Lee1, Beom-Ryeol Lee1
1. ETRI (Korea), 2. Konyang University (Korea), 3. Chernivtsi University (Ukraine)
Wavefront aberrations introduced by DMDs of different pixel structures to the reconstructed images of holograms displayed on the DMDs are investigated with a Shack-Hartmann wavefront sensor. The aberration increases as the pixel size increases, and introduces distortions and blurring to the reconstructed image. This aberration can also result the reconstructed image of a point to be bigger for the DMD with larger pixel size than ...
3D6/3DSA6-2
13:50
Invited Depth-Varying AR HUD Display Method
*Chien-Yu Chen1, Tzu-An Chou1, Chih-Hao Chuang2
1. National Taiwan University of Science and Technology (Taiwan), 2. Feng Chia University (Taiwan)
The utilization of CGH display technology in multi-depth head-up display (multi-depth HUD) for augmented reality (AR) is currently a prominent area of research. However, existing implementations often involve the use of multiple projection devices or spatial light modulators with complex optical mechanisms, such as Freeform Mirror, to achieve multi-depth projection in automotive HUDs. These approaches pose challenges in terms of system size and cost. Therefore, ...
3D6/3DSA6-3
14:10
Improvement of Image Quality of Wide-Angle Holographic Display Using a Non-Periodic Photon Sieve
*Shunki Sasaki1, Yasuhiro Takaki1
1. Tokyo University of Agriculture and Technology (Japan)
This study improves the image quality of the reconstructed images generated through holographic techniques using a non-periodic photon sieve, which provides a wide viewing zone angle. We developed the optimization technique based on the Gerchberg-Saxton algorithm to calculate the phase distribution displayed on the phase-only spatial light modulator.
3D6/3DSA6-4
14:30
HOE Screen Generating Multiple Viewpoints for Transparent 3D Display
*Nobuyuki Murakoshi1, Kengo Fujii1, Tomohiro Yendo1
1. Nagaoka University of Technology (Japan)
We propose a transparent 3D display with a wide field of view using fewer projectors by combining a unique HOE screen that focuses the projector image in multiple directions with face tracking. In this paper, we built a hologram printer to confirm the optical function of the proposed HOE screen.

 

3DSA7/3D7: Emerging Technologies for 3D/Hyper-realistic Display

Dec. 8  9:00-10:20  Room 301 A

Chair: Hyoung Lee (ETRI)
Co-Chair: Takafumi Koike (Hosei Univ.)

3DSA7/3D7-1
9:00
Simulation and Verification of 1×13 Wavelength Selective Switch Using LCoS for 2D Light Splitting
*You-Si Lin1, Yu-Chia Cheng1, Hoang-Yan Lin1
1. National Taiwan University (Taiwan)
We designed a 1×12 wavelength selective switch with a 2D fiber array, used Zemax simulation to optimize the optical element parameters, and used the iterative Fourier transform algorithm to design the optimal phase modulated LCoS-SLM. Finally, experiment is designed to verify the efficiency and crosstalk of the system.
3DSA7/3D7-2
9:20
Light Field Acquisition Using Shack-Hartmann Wavefront Sensor and Its Application to Holographic Stereogram
*Koki Matsuoka1, Chihiro Tsutake1, Keita Takahashi1, Toshiaki Fujii1
1. Nagoya University (Japan)
We propose to use a Shack-Hartmann wavefront (SHW) sensor for acquiring a light field (LF) with high angular resolution, which is suitable for holographic stereogram. We present an end-to-end pipeline from LF capture to holographic display on real hardware.
3DSA7/3D7-3
9:40
Interactive Relit Realistic Paintings on a Digital Frame
*Ying-Hsuan Lee1, Chih Yang1, Tzung-Han Lin1
1. National Taiwan University of Science and Technology (Taiwan)
An innovative 3D interactive digital framework was proposed. We rendered oil-paintings with PBR textures on anti-reflective screens to reproduce realistic lighting effects. After detecting the viewer's position, the content and re-rendered dynamically to fit the scenario which emphasizes texture details, brushstrokes and colors of artworks due to lighting directions.
3DSA7/3D7-4
10:00
Encoding of Stereo Images Using Bilateral Filters and Different Compression Ratio for Limited Communication Bandwidth
*Naoto Nakayama1, Haruki Mizushina1, Kenji Yamamoto1
1. Tokushima University (Japan)
Stereo images should be highly compressed for limited communication bandwidth. In this experiment, we evaluated stereo images with different compression ratios on the left and right sides using a bilateral filter, and stereo images with the same compression ratio on the left and right sides.

 

3D8/3DSA8: Holography (3)

Dec. 8  10:50-12:10  Room 301 A

Chair: Yasuhiro Takaki (Tokyo University of Agriculture and Technology)
Co-Chair: Hisayuki Sasaki (NHK)

3D8/3DSA8-1
10:50
Invited Expanding Space Bandwidth Product in Quantitative Phase Imaging by Comparing Transport of Intensity Equation with Fourier Ptychographic Microscopy
*Yongjun Lim1, Kwan-Jung Oh1
1. Electronics and Telecommunications Research Institute (Korea)
Phase information given by transport of intensity equation and Fourier ptychographic microscopy are compared with each other adopting incoherent light sources, and the expansion of space bandwidth product in quantitative phase imaging based on the spectral domain analysis is to be provided.
3D8/3DSA8-2
11:10
Aperture Control for Increasing Hologram Signal Amplitude in Incoherent Digital Holography
*Tetsuhiko Muroi1, Teruyoshi Nobukawa1, Yutaro Katano1, Kei Hagiwara1
1. NHK (Japan)
In incoherent digital holography (IDH), the hologram signal amplitude decreases as the object size relative to the image size increases. To address this issue, we implemented aperture control within the IDH optical configuration. This allowed us to increase the hologram signal amplitude and reduce random noise in the reconstructed image.
3D8/3DSA8-3
11:30
Holographic Phase Retrieval Based on PhaseMax
*Tatsuki Wada1, Chihiro Tsutake1, Keita Takahashi1, Toshiaki Fujii1
1. Nagoya University (Japan)
We propose a method for generating phase-only holograms. Our method is inspired by an optimization-based phase retrieval method, referred to as PhaseMax. In contrast to conventional methods, our method always produces globally optimal phase-only holograms. We demonstrate through our optical experiments that visually-plausible images can be produced from our hologram.
3D8/3DSA8-4
11:50
Invited Calculation of Computer-Generated Hologram Over 100K Resolution Range for Realistic 3D Meta-Space Reproduction
*Masahiro Yamaguchi1
1. Tokyo Institute of Technology (Japan)
This paper demonstrates the calculation of a computer-generated hologram that can reproduce deep and realistic 3D scenes in space, which is still or animated. The method, based on ray-wavefront conversion, enables hologram computation from the data rendered by a computer graphics technique or captured by a real camera.

 

MVS4/3DSA9/3D9: 3D Technologies for Metaverse Systems
Special Topic of Interest on AR/VR and Hyper Reality

Dec. 8  13:30-14:50  Room 301 A

Chair: Yuzo Hisatake (Shizuoka University)
Co-Chair: Taesoo Kwon (Hanyang Univ.)

MVS4/3DSA9/3D9-1
13:30
Invited Control of a Single-Rigid-Body Character in Various Environments
*Taesoo Kwon1, Yoonsang Lee1
1. Hanyang University (Korea)
Expressing the full-body character as a single rigid body, we can obtain a control policy that can adapt to various unobserved environmental changes and controller transitions. We demonstrate the ability of the policy with environments such as running on uneven terrain or pushing a box, and transitions between learned policies.
MVS4/3DSA9/3D9-2
13:50
Real-time MIV Player for Metaverse Service
*Hong-Chang Shin1, Gwangsoon Lee1, Won-Sik Cheong1, Hyun-gon Choo1
1. Electronics and Telecommunications Research Institute (Korea)
The MPEG-I Visual group is currently working on standardization efforts related to 3DoF+ video generation technology with the aim of providing immersive video services under the name MPEG Immersive Video (MIV). Immersive video technology, which combines multiple camera footage to synthesize scenes in virtual space and reproduce lifelike videos, is closely related to reality-based metaverse services. In this paper, we introduce the core technology of ...
MVS4/3DSA9/3D9-3
14:10
Head Motion-Based Accident Detection Model in Construction Site Using Simulation
*Kisub Lee1, Minsu Kyeon1, Gaeun Noh1, Jong-Il Park1
1. Hanyang University (Korea)
This study presents an accident detection model using a head-mounted display (HMD) device to prevent casualties caused by falls and material collisions at construction sites. The model utilizes simulations to generate data on head movements during accident scenarios. By comparing this data with animation data that represents normal conditions, the model performs binary classification to identify whether a situation is hazardous or safe. This study ...
MVS4/3DSA9/3D9-4
14:30
Ultra-Wide Field-of-View Display by Use of Aerial Stereoscopic Images Reconstructed Just Behind the Eyes
*Hiroki Takatuska1, Munekazu Date2, Masaki Yasugi1,3, Shiro Suyama1, Hirotsugu Yamamoto1
1. Utsunomiya University (Japan), 2. Tokushima Bunri University (Japan), 3. Fukui Prefectural University (Japan)
When an observer's eye is placed before the aerial image, an ultra-wide field-of-view is realized because the image is reconstructed just behind the eye. However, negative binocular disparity is reproduced. We use stereoscopic display technique to reproduce positive binocular disparity on the aerial images that are reconstructed behind the observer.

 

3DSAp1: Emerging technologies for 3D/Hyper-realistic Display (1)

Dec. 7  15:40-17:00  Exhibition Hall B

3DSAp1-1 Performance Evaluation of Transmission System for Computer-Generated Holograms in Wired/Wireless Connections by Using Lossless Compression
*Misato Onishi1, Kazuhiro Yamaguchi1, Yuji Sakamoto2
1. Suwa University of Science (Japan), 2. Hokkaido University (Japan)
In this study, a transmission system for Computer-Generated Hologram (CGH) was developed. In the transmission system, data of CGHs were streamed through network, and 3D objects were reconstructed by outputting the received data on SLM in electro holography. Quality of optically reconstructed objects and performance of the transmission system were evaluated.
3DSAp1-2 Denisyuk-Type Wavefront Printer Developed for Printing Large-Scale Display CGVH
*Yuki Takemura1, Hirohito Nishi1, Kyoji Matsushima1
1. Kansai University (Japan)
The latest status of development is presented in a novel type of wavefront printer called the Denisyuk-type wavefront printer. The introduced 4K high-definition SLM and improved optics facilitates fabrication of large-scale full-parallax CGVHs with a wide viewing angle.
3DSAp1-3 Fast Hologram Generation Method for Holographic Video Display
*Chun Chen1, Minseok Chae Chae1, Yoonchan Jeong1, Byoungho Lee1
1. Seoul National University (Korea)
We propose a fast hologram generation method for holographic video display. In our approach, we utilize the previous frame's hologram and employ a mask to separate the background and foreground optimization to accelerate the hologram generation process.
3DSAp1-4 Evaluation on Usability of Aerial Display by Changing the Timing of Providing Multimodal Feedback
*Yasunori Terao1, Haruki Mizushina1, Kenji Yamamoto1
1. Tokushima University (Japan)
We revealed optimal relationship between position of the multimodal feedback (visual, auditory, and tactile using vibration to the sole of foot) and position of the aerial image for operating the aerial display. We also revealed the correlation between usability of aerial display and perceived depth position of aerial image.
3DSAp1-5 Increase in Size of Full-Color Stacked CGVH by Tiling Contact-Copy
*Shunsuke Fujiki1, Hirohito Nishi1, Kyoji Matsushima1
1. Kansai University (Japan)
A novel technique is presented for increasing in the size of stacked CGVHs (computer-generated volume hologram) that reconstruct high-quality full-color full-parallax 3D images. A tiling technique is introduced to extend the size without increasing the output power of the laser used for contact-copy.
3DSAp1-6 Real-Time 3D Video Communication System Using Stereo Video Capture and Autostereoscopic Display
*Kyosuke Yanagida1, Takafumi Koike1,3, Goro Hamagishi2,3, Hideya Takahashi2,3
1. Hosei University (Japan), 2. Osaka Metropolitan University (Japan), 3. RealImage Inc. (Japan)
We have developed a high-quality real-time 3D video communication system with the simplest configuration of stereo cameras and an autostereoscopic 3D display. The effectiveness of the system as a video communication system was discussed by conducting 2D videoconferences with the system for multiple users.
3DSAp1-7 Hollow Face Illusion Evoked Even by Binocular Vision of Aerial 3D Image
*Takumi Watanabe1, Hiroki Takatsuka1, Shiro Suyama1, Hirotsugu Yamamoto1
1. Utsunomiya University (Japan)
We have confirmed that hollow face illusion can be easily evoked even at binocular vision of an aerial 3D image by using the AIRR (aerial imaging by retroreflection). The aerial image of the concave surface can be tracked as depending on the angle of observation.
3DSAp1-8 Robust 3D Object Mesh Generation Based on NeRF with Image Brightness Equalization for Various Lighting Environments
*Hyeongil Nam1, Gaeun Noh1, Minsu Kyeon1, Jong-Il Park1
1. Hanyang University (Korea)
We propose a method for 3D mesh generation using NeRF and brightness equalization. Addressing lighting variations, we apply equalization to rendered images and utilize shared MLP features for improved volumetric rendering. Quantitative and qualitative comparisons validate effectiveness.
3DSAp1-9 Light Field Display Based on Monocular Stereo Vision
*Song Fang1, Haofeng Zhang1, Tianhong Wang1, Jing Zhang1, Jiangbo Yao2, Xin Zhang2
1. Shenzhen China Star Optoelectronics Semiconductor Display Technology Co. Ltd (China), 2. TCL China Star Optoelectronics Display Technology Co. Ltd (China)
The existence of monocular stereo vision can effectively alleviate vergence-accommodation-conflict, at the same time, it can also enrich the stereoscopic information and improve the quality of 3D performance. On the basis of the binocular parallax depth clue, the accommodation depth clue can be used to see a more natural 3D display effect. The light field display proposed in this paper is based on monocular stereo ...
3DSAp1-10L Evaluation of Simulated Image Quality for Wireless Transmission System of CGHs
*Kazuhiro Yamaguchi1, Misato Onishi1, Yuji Sakamoto2
1. Suwa University of Science (Japan), 2. Hokkaido University (Japan)
In this paper, a system models of wireless transmission system for computer-generated holograms (CGHs) is constructed, and quality of light waves and reconstructed images considering transmission errors of CGHs are analyzed by using computer simulations. SNR of light waves, PSNR and SSIM of computational reconstructed images were measured and evaluated.
3DSAp1-11L Diffraction Efficiency Analysis for a Large-Aperture Holographic Optical Element
*Sungjin Lim1, Kyung-Il Joo1, Jung Hun Choo1, Seon Kyu Yoon1, Jin Su Lee2, Kwang-Hoon Lee1
1. Korea Photonics Technology Institute (Korea), 2. Electronics and Telecommunications Research Institute (Korea)
Holographic optical elements are next-generation optical components that can implement complex optical systems in a thin film. However, as the size and shape of HOEs become more diverse, there is no measurement methods that can precisely analyze the optical properties of HOEs. In this paper, we propose a precision measurement method to analyze the optical properties of HOEs.
3DSAp1-12L A Method for Solving the Field of View Problem in Dihedral Corner Reflector Arrays
*Tsai Cheng Hsun1, Chuang Chih Hao, Huang Shi Hwa1, Chen Chien Yu1
1. National Taiwan University of Science and Technology (Taiwan)
DCRA relies on the second reflection of light on two mutually perpendicular surfaces to form an image at the location of the image. Therefore, we added a right-angle structure to the DCRA square column structure to achieve the purpose of expanding the viewing angle.
3DSAp1-13L A Study on the Effect of Total Variation Regularizer in Voxel Grid-based Radiance Field
*Sangwoon Kwak1,2, Jun-Young Jeong1, Joonsoo Kim1, Gwangsoon Lee1, Won-Sik Cheong1, Hyon-Gon Choo1, Munchurl Kim2
1. Electronics and Telecommunications Research Institute (Korea), 2. Korea Advanced Institute of Science and Technology (Korea)
In this paper, we consider a voxel grid-based radiance field framework for representing 3D static scene. One of the representative frameworks, Plenoxels, utilizes a total variation regularizer to effectively remove artifacts by encouraging similarity of features between neighboring voxels. This term is combined with the existing reconstruction loss through a weighted sum, but finding a well-balanced weight parameter can be challenging. To address this issue, ...
3DSAp1-14L Correlation Between XR Content Watching Time and Concentration/Stress According to Biometric Data Acquisition of XR Multimodal Devices Users
Jong-Bae Lee1, HyungKi Son1, Beom-Ryeol Lee1, Wook-Ho Son1, *Hyoung Lee1
1. Electronics Telecommunications Research Institute (ETRI) (Korea)
In this study, we develop a novel multimodal device capable of collecting biometric data from viewers of XR content, including EEG signals, pupil size changes, eye blink signals, head movement data, and hand movement data. Using the various biometric data collected, correlations between concentration and stress factors and feature information are verified using statistical analysis methods. Finally, the correlation of concentration/stress according to the ...
3DSAp1-15L Development of Virtual-Reality-Based Electromagnetism-Teaching Materials Incorporating Gamification
*Shunya Nakatani1, Tomoyoshi Shimobaba1, Tomoyoshi Ito1, Atsushi Shiraki1
1. Chiba University (Japan)
This study presents the enhancement of learning outcomes by developing high school physics electromagnetic materials using virtual reality (VR) incorporated with gamification elements. Specifically, we conducted electromagnetic-induction experiments and developed materials that integrate gaming components such as sensory representations and scoreboards.
3DSAp1-16L Partial Access using Group-based MPEG Immersive Video Coding Technology
*Gwangsoon Lee1, Sangwoon Kwak1, Hong-chang Shin1, Bong Ho Lee1, Won-sik Cheong1
1. ETRI (Korea)
In this paper, we introduce a proof of concept for partially decoding and rendering the immersive video, which is compliant to MIV (MPEG Immersive Video) standard. Specifically, we propose an immersive video coding system that can split source views into groups according to sub-viewing zones.

 

3DSAp2: Emerging technologies for 3D/Hyper-realistic Display (2)

Dec. 7  17:30-18:50  Exhibition Hall B

3DSAp2-1 Double-Plane Image Reconstruction/Authentication Based on Double-Plane Phase Only Functions Located at Different Distances
*Hsuan-Ting Chang1, Mao-Siang Jhao1
1. National Yunlin University of Science and Technology (Taiwan)
In this paper, we propose a method of using double-plane phase only functions (DPPOFs) and applying Fresnel diffraction transform to achieve double-plane image reconstruction/authentication. Based on the principle of optical Fresnel diffraction, we use the double phase retrieval algorithm to encode two target images at two different reconstruction planes into DPPOFs. The simulation results show that the two target images can be successfully reconstructed ...
3DSAp2-2 Simplified Wide Viewing Angle Electronic Holography Using a Checkered Hologram Generated by a Lens Array
*Soma Fujimori1, Tomoyoshi Ito1, Tomoyoshi Shimobaba1
1. Chiba University (Japan)
Electronic holography has the problem of a small viewing angle. We propose wide viewing angle electronic holography using a checkered hologram generated by a lens array. The proposed method has the advantages of simplicity of the optical system and robustness to misalignment.
3DSAp2-3 Manifold Learning in the Hologram Domain: a Pipeline for Compressing and Reconstructing Complex Holograms
*Juhyun Lee1, Yoonchan Jeong1, Byoungho Lee1
1. Seoul National University (Korea)
We introduce a pipeline for manifold learning in the hologram domain. A traditional autoencoder architecture is employed to compress and reconstruct complex holograms. The pipeline utilizes MNIST data and demonstrates consistent shape reconstruction. The latent vector representation of the hologram exhibits well-separated manifolds, successfully capturing similarities between numbers.
3DSAp2-4 Increasing Vection Strength by Adding Optical Flow with Random Dots in the Peripheral Visual Field
*Kosuke Nakanishi1, Haruki Mizushina1, Kenji Yamamoto1
1. Tokushima University (Japan)
In this paper, we demonstrated that vection strength and duration can be increased, and latency can be decreased by presenting a stimulus video on the HMD, which is added by alpha blending optical flow with random dot images on the peripheral visual field of a live-action video.
3DSAp2-5 Lateral Viewing Zone of Depth-Fused 3D (DFD) Display by Use of Aerial Image in Front
*Takahiro Omoto1, Shiro Suyama1, Hirotsugu Yamamoto1
1. Utsunomiya University (Japan)
We derive the theoretical formulae for the viewing zone of a Depth-fused 3D (DFD) display that consists of a flat-panel display in the back side and an aerial image in front. We have developed a DFD display with an aerial image and confirmed our analyses on the lateral viewing zone.
3DSAp2-6 Occlusion-Aware SDFs with Inpainting for Hand-Object Reconstruction from a Single RGB Image
*Gaeun Noh1, Hyeongil Nam1, Jong-Il Park1
1. Hanyang University (Korea)
Hand-Object Reconstruction research aims to understand and accurately reconstruct hand-object interaction, addressing challenges in diverse shape representation and occlusion. We introduce GAN-based inpainting on occluded regions in monocular RGB image, extracting features that model missing areas for precise surface prediction in the SDF Decoder, facilitating clear boundary distinction for detailed shape representation.
3DSAp2-7 Voxel-Based 3D Object Reconstruction Using NeRF Robust Against Occlusion
*Minsu Kyeon1, Hyeongil Nam1, Jong-Il Park1
1. Hanyang University (Korea)
This paper introduces a technique to improve the robustness of 3D voxel reconstruction even when there is occlusion in the image. Previous studies on existing CNN and Transformer-based voxel reconstruction tend to be sensitive when there is occlusion in the image. To solve these problems, we present a framework that integrates a NeRF-based method and a 3D-CNN decoder to enable 3D voxel reconstruction. The proposed ...
3DSAp2-8 Influence of Material Appearance on Perception in Dyslexia
*Takashi Yamazoe1, Hanae Ikeshita2
1. Seikei University (Japan), 2. Sagami Women's University (Japan)
Dyslexia indicate person of who have difficulty in reading and writing without any other developmental disabilities. Dyslexia need cue of character recognition. In this paper, authors were evaluated material appearance of microfacet and macrofacet to use cue of character recognition. Result suggested dyslexia perceive unique impression on material appearance.
3DSAp2-9L Proposal of an 8×8 Holographic Directional Volumetric Display
*Tomoya Isayama1, Takashi Nishitsuji2, Naoto Hoshikawa3, Hirotaka Nakayama4, Tomoyoshi Shimobaba1, Tomoyoshi Ito1, Atsushi Shiraki1
1. Chiba University (Japan), 2. Toho University (Japan), 3. University of Tsukuba (Japan), 4. National Astronomical Observatory of Japan (Japan)
This study aims to realize a directional image display system using electronic holography to observe the projected image from two directions (front and side). We confirmed that it was possible to observe different holographic reconstruction images from both directions of a volumetric display.
3DSAp2-10L A Basic Study of the Parallel Polynomial Approximation-based Hologram Computation Method with a GPU
*Harutaka Shiomi1, Tomoyoshi Ito1, Tomoyoshi Shimobaba1
1. Chiba University (Japan)
Holography is a promising technology for three-dimensional images, but hologram computation is very time-consuming. To accelerate it, we propose a parallel polynomial approximation-based method with a graphical processing unit (GPU) and compare the computation time and quality of reconstruction. It achieved over 15 times faster computations without significant degradations.
3DSAp2-11L Virtual-Moving Microlens Arrays for 3D Imaging Application
*Min-Kyu Park1, Kyung-Il Joo1, Tae-Hyun Lee1, Kwang-Hoon Lee1, Dong-Kil Lee1, Hak-Rin Kim2
1. Korea Photonics Technology Institute (Korea), 2. Kyungpook National University (Korea)
We propose an electrically controllable fast-switching virtual-moving microlens array (MLA). By controlling the polarization of incident lights without physical movement of the MLA, the periodic sampling position of the MLA can be switched fast with a polarization-switching layer. Using the virtual-moving MLA, spatial-resolution-enhanced light-field imaging system was demonstrated.
3DSAp2-12L Expanding the Viewable Area of Directional Volumetric Displays
*Kazuki Kamizuru1, Takeshi Nishitsuji2, Naoto Hoshikawa3, Hirotaka Nakayama4, Tomoyoshi Ito1, Tomoyoshi Shimobaba1, Astushi Shiraki1
1. Chiba University (Japan), 2. Toho University (Japan), 3. University of Tsukuba (Japan), 4. National Astronomical Observatory of Japan (Japan)
The existing directional volumetric display, which uses strings and a projector, allows for adjusting the viewing angle in the horizontal directions. However, it doesn't support manipulating the viewing angle in the vertical direction. Our goal is to enable the adjustment of the observation angle in the vertical direction.
3DSAp2-13L Enhancing Image Quality in Directional Volumetric Displays with Red-Net
*Seigou Takeshita1, Takahashi Nishitsuji2, Naoto Hoshikawa3, Hirotaka Nakayama4, Tomoyoshi Shimobaba1, Tomoyoshi Ito1, Atsushi Shiraki1
1. Chiba University (Japan), 2. Toho University (Japan), 3. University of Tsukuba (Japan), 4. National Astronomical Observatory (Japan)
The existing directional volumetric display, which employs threads and a projector, exhibits a significant problem: noise within the displayed image when juxtaposed with the original image. Herein, we harnessed deep learning techniques to expedite the process of enhancing image quality in contrast to conventional methodologies.
3DSAp2-14L Height Adjustment of Directional Volumetric Displays Using Deep Learning
*Yuta Fukumoto1, Takashi Nishitsuji2, Naoto Hoshikawa3, Hirotaka Nakayama4, Tomoyoshi Shimobaba1, Tomoyoshi Ito1, Atsushi Shiraki1
1. Chiba University (Japan), 2. Toho University (Japan), 3. University of Tsukuba (Japan), 4. National Astronomical Observatory of Japan (Japan)
The directional volumetric display, which uses threads, and a projector has an issue in that the irradiation position's height changes depending on the depth at which the thread is placed, even if the height of ray is the same. In this study, the afore-mentioned problem was solved using deep learning.
3DSAp2-15L Point Cloud Compression with Optimization on Point Density for Reconstruction of Visually Significant Structure
*Hideaki Kimata1
1. Kogakuin University (Japan)
We propose a compression method of a point cloud, which represents the shape of objects in the real world, while reducing compression noise that severely impairs object features, especially hole-popping. We introduce measures of point density and shape significance for optimizing coding rate and noise.
3DSAp2-16L Pose Estimation and Calibration System for Monocular Camera
Chung Wen Hung1, *Xuan Ni Chen1
1. National Yunlin University of Science and Technology (Taiwan)
A feature-based method for determining camera pose of monocular camera is proposed in this paper. This method helps experts establish image archives of artifacts for maintaining their state. A six-axis robotic arm is adopted to verify the efficiency of algorithm and the experimental results show accuracy of 0.1° and 2mm.
3DSAp2-17L Emergency-drill Online Visual System to Build Ability against Unexpected Situations
*Kyosuke Takahashi1, Akua Kawakami2, Masaki Umemoto2, Kousuke Takahashi2, Yuuya Sasai2, Yasunori Terao2, Kousuke Nakanishi2, Naoto Nakayama2, Yuuto Morishita2, Akito Inoue2, Shuto Ogura2, Kenta Saito2, Yuta Machigashira2, Kotomi Nakanishi2, Haruki Mizushina2, Kenji Yamamoto2, Hitoshi Inomo1
1. Kagawa University (Japan), 2. Tokushima University (Japan)
This paper introduces the emergency-drill system for building ability against unexpected situations. The system consists of main and local systems. Trainees at local location watch the video sent from a main location, and judge what to do on the unexpected situation that video shows. Several short scenarios are prepared before training, and trainer connects some of them according to the behavior of the trainees in ...
3DSAp2-18L Emergency-drill Scenarios to Build Ability against Unexpected Situations on Tsunami
*Kenji Yamamoto1, Masaki Umemoto1, Akua Kawakami1, Kousuke Takahashi1, Yuuya Sasai1, Yasunori Terao1, Kousuke Nakanishi1, Naoto Nakayama1, Yuuto Morishita1, Akito Inoue1, Shuto Ogura1, Kenta Saito1, Yuta Machigashira1, Kotomi Nakanishi1, Kyosuke Takahashi2, Hitoshi Inomo2, Haruki Mizushina1
1. Tokushima University (Japan), 2. Kagawa University (Japan)
This paper introduces the scenarios to build ability against unexpected situations between giant earthquake and tsunami.
3DSAp2-19L Emergency Drill Using Online Visual System
*Hitoshi Inomo1, Akua Kawakami2, Masaki Umemoto2, Kousuke Takahashi2, Yuuya Sasai2, Yasunori Terao2, Kousuke Nakanishi2, Naoto Nakayama2, Yuuto Morishita2, Akito Inoue2, Shuto Ogura2, Kenta Saito2, Yuta Machigashira2, Kotomi Nakanishi2, Kyosuke Takahashi1, Haruki Mizushina2, Kenji Yamamoto2
1. Kagawa University (Japan), 2. Tokushima University (Japan)
This paper reports on a disaster drill using an online visual system to prepare for earthquake and tsunami situations. The trainees walked with the students in the video to a nearby hill to avoid danger. Some unexpected situations happened along the way in the video. After experiment, trainees and trainer discussed whether the responses were appropriate.

 

3Dp1: Emerging Technologies for 3D/Hyper-Realistic Display
Special Topic of Interest on AR/VR and Hyper Reality

Dec. 7  15:40-17:00  Exhibition Hall B

3Dp1-1 Comparison Between Various Homography Methods for Phase Profile Optimization of Holographic Display Using Lens Array
*Minseok Chae1, Chun Chen1, Siwoo Lee1, Yoonchan Jeong1, Byoungho Lee1
1. Seoul National University (Korea)
In this paper, we examine the optimization of spatial light modulator's phase profile in holographic displays, focusing on the effect of homography in distortion correction. Different homography methods are compared to find the most efficient one for handling complex distortions, especially in holographic displays employing lens arrays.
3Dp1-2 Development of a 3D Reconstruction Service System by Using Opensource for Smartphone Users
Matías Grijalba Acosta1, Danielle E. M. Krauch Bareiro1, Joel S. González Arrosse1, *Tzung-Han Lin2
1. Universidad Politécnica Taiwan-Paraguay (Paraguay), 2. National Taiwan University of Science and Technology (Taiwan)
This paper aims to create a system with a cloud server that facilitates the generation of 3D objects using smartphone cameras, empowering inexperienced users to create personalized 3D models. The backbone technology is a computer vision technique called “photogrammetry”, which infers geometric characteristics from the collection of photographs or videos.
3Dp1-3 Development of Holographic Optical Elements with an Expanded Eyebox
Ai Weng1, Hui-Chi Chen1, Hao-Ting Liao2, Chuan-Feng Kuo2, Tzu-An Chou2, *Chih-Hao Chuang3, Chien-Yu Chen2
1. Fu Jen Catholic University (Taiwan), 2. National Taiwan University of Science and Technology (Taiwan), 3. Feng Chia University (Taiwan)
This study aims to implement an optical system for near-eye displays based on holographic optical elements (HOEs) structure. The design and fabrication of the HOEs will involve beam splitters and pupil expanders. After the fabrication of the components, system integration and performance validation will be conducted.
3Dp1-4L Real-time Aerial 3D Display Using Holographic Projector and Resin 3D Screen
Junai Uchida1, Natsumi Furukawa1, Yuzuki Nakatani1, *Naoki Takada1
1. Kochi University (Japan)
We report on an aerial holographic 3D display. 4K-resolution holographic projector projects a focused 3D image on a resin 3D screen. The transmitted diffuse light generates an aerial 3D real image by retroreflectors. Finally, we realized a real-time aerial display of the 3D video comprising up to 21,972 points.
3Dp1-5L Very Wide FOV in Holographic AR Display Using a Large HOE Fabricated by Area Segmentation and Multiple Exposures
*Yusuke Sando1, Daisuke Barada2, Toyohiko Yatagai2
1. Osaka Research Institute of Industrial Science and Technology (Japan), 2. Utsunomiya University (Japan)
A large holographic optical element (HOE) has been fabricated by an originally developed recording system enabling the area segmentation and multiple exposures. An augmented reality (AR) 3D display with a very wide FOV of 34° and 25° has been successfully demonstrated based on the time division method.
3Dp1-6L Holography via Monocular Depth Estimation
*Ryota Kibune1, Chihiro Tsutake1, Keita Takahashi1, Toshiaki Fujii1
1. Nagoya University (Japan)
Holograms are usually computed from RGBD images or light fields, which requires special sensors. We demonstrate that holograms can be generated from an ordinary camera that captures monocular images alone via monocular depth estimation.
3Dp1-7L Controlling Speckle and Resolution of Reconstructed Images by Vibrating Multimode Optical Fiber with a Cylindrical Piezoelectric Transducer in Electro-Holography
*Koki Matsuno1, Yasuo Ito1, Yuji Sakamoto1
1. Hokkaido University (Japan)
To reduce the speckle of reconstructed images in electro-holography, reducing laser coherence is essential. We developed a device to vibrate a multimode optical fiber using a cylindrical piezoelectric transducer to reduce coherence. In addition to speckle, the resolution can be electronically controlled with this device.
3Dp1-8L Lateral Electric Field Driving of High-resolution Liquid Crystal Devices for Electronic Holography with Wide Viewing-Zone Angle
*Hiroto Tochigi1, Masakazu Nakatani1, Ken-ichi Aoshima2, Mayumi Kawana2, Yuta Yamaguchi2, Kenji Machida2, Nobuhiko Funabashi2, Hideo Fujikake1
1. Tohoku University (Japan), 2. Japan Broadcasting Corporation (Japan)
For an electronic holographic display with wide viewing-zone angle, we propose a lateral electric field drive system based on continuous potential-difference driving by lateral transparent electrodes. Light modulation was evaluated to confirm the independent driving of each pixel with high resolution of 1µm pitch. We also showed that by combination of driving and grounding electrodes, 2-dimensional driving is possible.