检索结果-内蒙古大学图书馆

diffraction model-driven neural network with semi-supervised training strategy for real-world 3d holographic photography

引用

OPTICS EXPRESS 2024年第26期32卷 45406-45420页

作者： Fang, Qiwei Zheng, Huadong Xia, Xinxing Peng, Junchang Zhang, Tengfei Lin, Xingyu Yu, Yingjie Shanghai Univ Dept Precis Mech Engn Shanghai 200444 Peoples R China Minist Educ Key Lab Adv Display & Syst Applicat Shanghai 200072 Peoples R China

Compared to traditional 2d displays, 3d display technology provides richer information to the viewer. Learning-based computer-generated holography (CGH) has shown great potential in realizing real-time holographic 3d displays. However, most of the current learningbased CGH algorithms cannot quickly complete the training stage and produce high-quality holograms due to insufficient constraints in the training stage of the neural network. In this paper, we propose a diffractive model-driven neural network trained using a semi-supervised training (SST-holo) strategy and incorporate a state-of-the-art monocular depth estimation algorithm to achieve the fast generation of holograms of real-world 3d scenes. Compared to the supervised training strategy, our proposed semi-supervised training strategy does not require high-quality labeled datasets, but can significantly improve the imaging quality and generalization of the algorithm. Incorporating the Res-MSR block in SST-holo to adaptively learn image features of different scales enhances the learning capability of the network. In addition, we adopt a random splicing processing strategy to preprocess the dataset to ensure that the original features in the dataset are not corrupted. SST-holo can generate high-quality 3d phase-only holograms with 2 K resolution in 0.015 seconds. Both monochrome and color optical experiments show that the proposed algorithm has good 3d effect and generalization ability and can effectively improve the quality of reconstructed images.

关键词： Near eye displays Neural networks Real time imaging Spatial light modulators Stochastic gradient descent Three dimensional displays

来源：评论

学校读者我要写书评

暂无评论

image Reconstruction for Accelerated MR Scan With Faster Fourier Convolutional Neural Networks

引用

IEEE TRANSACTIONS ON image processing 2024年 33卷 2966-2978页

作者： Liu, Xiaohan Pang, Yanwei Sun, Xuebin Liu, Yiming Hou, Yonghong Wang, Zhenchang Li, Xuelong Tianjin Univ Sch Elect & Informat Engn Tianjin Key Lab Brain Inspired Intelligence Techno Tianjin 300072 Peoples R China Tiandatz Technol Co Ltd Tianjin 301723 Peoples R China Capital Med Univ Beijing Friendship Hosp Beijing 100050 Peoples R China China Telecom Corp Ltd Inst Artificial Intelligence TeleAI Beijing 100033 Peoples R China

High quality image reconstruction from undersampled ${k}$ -space data is key to accelerating MR scanning. current deep learning methods are limited by the small receptive fields in reconstruction networks, which restrict the exploitation of long-range information, and impede the mitigation of full-image artifacts, particularly in 3d reconstruction tasks. Additionally, the substantial computational demands of 3d reconstruction considerably hinder advancements in related fields. To tackle these challenges, we propose the following: 1) A novel convolution operator named Faster Fourier Convolution (FasterFC), aims at providing an adaptable broad receptive field for spatial domain reconstruction networks with fast computational speed. 2) A split-slice strategy that substantially reduces the computational load of 3d reconstruction, enabling high-resolution, multi-coil, 3d MR image reconstruction while fully utilizing inter-layer and intra-layer information. 3) A single-to-group algorithm that efficiently utilizes scan-specific and data-driven priors to enhance ${k}$ -space interpolation effects. 4) A multi-stage, multi-coil, 3d fast MRI method, called the faster Fourier convolution based single-to-group network (FAS-Net), comprising a single-to-group ${k}$ -space interpolation algorithm and a FasterFC-based image domain reconstruction module, significantly minimizes the computational demands of 3d reconstruction through split-slice strategy. Experimental evaluations conducted on the NYU fastMRI and Stanford MRI data datasets reveal that the FasterFC significantly enhances the quality of both 2d and 3d reconstruction results. Moreover, FAS-Net, characterized as a method that can achieve high-resolution (320, 320, 256), multi-coil, (8 coils), 3d fast MRI, exhibits superior reconstruction performance compared to other state-of-the-art 2d and 3d methods.

关键词： Three-dimensional displays image reconstruction Convolution Magnetic resonance imaging Feature extraction Task analysis Interpolation 3d MRI reconstruction magnetic resonance imaging faster Fourier convolution

来源：评论

学校读者我要写书评

暂无评论

PointMCd: Boosting deep Point Cloud Encoders via Multi-View Cross-Modal distillation for 3d Shape Recognition

引用

IEEE TRANSACTIONS ON MULTIMEdIA 2025年 27卷 754-767页

作者： Zhang, Qijian Hou, Junhui Qian, Yue City Univ Hong Kong Dept Comp Sci Hong Kong Peoples R China

As two fundamental representation modalities of 3d objects, 3d point clouds and multi-view 2d images record shape information from different domains of geometric structures and visual appearances. In the current deep learning era, remarkable progress in processing such two data modalities has been achieved through respectively customizing compatible 3d and 2d network architectures. However, unlike multi-view image-based 2d visual modeling paradigms, which have shown leading performance in several common 3d shape recognition benchmarks, point cloud-based 3d geometric modeling paradigms are still highly limited by insufficient learning capacity due to the difficulty of extracting discriminative features from irregular geometric signals. In this article, we explore the possibility of boosting deep 3d point cloud encoders by transferring visual knowledge extracted from deep 2d image encoders under a standard teacher-student distillation workflow. Generally, we propose PointMCd, a unified multi-view cross-modal distillation architecture, including a pretrained deep image encoder as the teacher and a deep point encoder as the student. To perform heterogeneous feature alignment between 2d visual and 3d geometric domains, we further investigate visibility-aware feature projection (VAFP), by which point-wise embeddings are reasonably aggregated into view-specific geometric descriptors. By pair-wisely aligning multi-view visual and geometric descriptors, we can obtain more powerful deep point encoders without exhausting and complicated network modification. Experiments on 3d shape classification, part segmentation, and unsupervised learning strongly validate the effectiveness of our method.

关键词： Three-dimensional displays Shape Point cloud compression Solid modeling Visualization Feature extraction Task analysis 3d point cloud multi-view images cross-modal knowledge distillation 3d shape recognition

来源：评论

学校读者我要写书评

暂无评论

research on a Multiscale U-Net Lung Nodule Segmentation Model Based on Edge Perception and 3d Attention Mechanism Improvement

引用

IEEE ACCESS 2024年 12卷 165458-165471页

作者： Ming, Hui Li, Yuqin Hu, Tianjiao Lan, Yihua Nanyang Normal Univ Sch Artificial Intelligence & Software Engn Nanyang 473061 Peoples R China Henan Engn Res Ctr Intelligent Proc Big Data Digit Nanyang 473061 Peoples R China Nanyang Normal Univ Sch Life Sci & Agr Engn Nanyang 473061 Peoples R China

Lung nodule semantic segmentation using deep learning has achieved good results. However, problems such as information loss on lesion edges, boundary segmentation blurring, lung nodule misdection, and low segmentation accuracy remain in lung CT (Computed Tomography) detection using deep learning due to the high degree of heterogeneity and a wide variety of nodule sizes, shapes, and locations, as well as the characteristics of convolutional localized feature extraction and the limitations of the continuous downsampling receptive field. So, a new model called EMC-UNet (Edge-aware _ Multiscale feature extraction residual _ 3d CA-Net attention module _ 3d U-Net), which integrates edge-awareness, 3d attention (3d CA-Net, Three-dimensional coordinate attention mechanism network), and multiscale techniques for segmenting lung nodules, is introduced. The model first uses an edge-aware module to accurately locate lesion edges, extract key edge features in the image, and increase the perception of lesion edge features by the model. Then, a 3d attention mechanism is added to focus the network on important lesion image features, emphasizing that the lesion features can improve segmentation performance. In conclusion, the 3d multiscale feature extraction module enhances the network's perceptual range by processing information at various scales simultaneously, capturing features to offer a more comprehensive object context. This approach achieves notable results, with a dice coefficient of 87.95% and an IoU value of 78.5% on the publicly available LIdC-IdRI(The Lung image database Consortium and image database Resource Initiative) dataset, outperforming existing lung nodule segmentation models.

关键词： Lungs image segmentation Three-dimensional displays Feature extraction Attention mechanisms image edge detection Accuracy Solid modeling Lung cancer data models Biomedical image processing Lung nodule segmentation deep learning attention mechanism 3d multiscale feature extraction edge enhancement biomedical image processing

来源：评论

学校读者我要写书评

暂无评论

Learning Virtual View Selection for 3d Scene Semantic Segmentation

引用

IEEE TRANSACTIONS ON image processing 2024年 33卷 4159-4172页

作者： Mu, Tai-Jiang Shen, Ming-Yuan Lai, Yu-Kun Hu, Shi-Min Tsinghua Univ Minist Educ Key Lab Pervas Comp Beijing 100084 Peoples R China Tsinghua Univ Dept Comp Sci & Technol Beijing 100084 Peoples R China Cardiff Univ Sch Comp Sci & Informat Cardiff CF24 4AG Wales

2d-3d joint learning is essential and effective for fundamental 3d vision tasks, such as 3d semantic segmentation, due to the complementary information these two visual modalities contain. Most current 3d scene semantic segmentation methods process 2d images "as they are", i.e., only real captured 2d images are used. However, such captured 2d images may be redundant, with abundant occlusion and/or limited field of view (FoV), leading to poor performance for the current methods involving 2d inputs. In this paper, we propose a general learning framework for joint 2d-3d scene understanding by selecting informative virtual 2d views of the underlying 3d scene. We then feed both the 3d geometry and the generated virtual 2d views into any joint 2d-3d-input or pure 3d-input based deep neural models for improving 3d scene understanding. Specifically, we generate virtual 2d views based on an information score map learned from the current 3d scene semantic segmentation results. To achieve this, we formalize the learning of the information score map as a deep reinforcement learning process, which rewards good predictions using a deep neural network. To obtain a compact set of virtual 2d views that jointly cover informative surfaces of the 3d scene as much as possible, we further propose an efficient greedy virtual view coverage strategy in the normal-sensitive 6d space, including 3-dimensional point coordinates and 3-dimensional normal. We have validated our proposed framework for various joint 2d-3d-input or pure 3d-input based deep neural models on two real-world 3d scene datasets, i.e., ScanNet v2 and S3dIS, and the results demonstrate that our method obtains a consistent gain over baseline models and achieves new top accuracy for joint 2d and 3d scene semantic segmentation.

关键词： Three-dimensional displays Semantic segmentation Solid modeling Semantics Geometry Task analysis deep reinforcement learning Virtual view selection 2d-3d joint learning deep reinforcement learning 3d semantic segmentation

来源：评论

学校读者我要写书评

暂无评论

VoxelSky-3d: A Weather Radar Visualization Prototype for Air Traffic Control

引用

IEEE ACCESS 2025年 13卷 16252-16262页

作者： Zhan, Zhenhao Lim, Tek-Yong Ramakrishnan, Kannan Multimedia Univ Fac Comp & Informat Cyberjaya 63100 Selangor Malaysia

This paper presents VoxelSky-3d, a new 3d weather radar visualization prototype for civil aviation air traffic control. While previous research has explored text-based, image-based, and some three-dimensional meteorological displays, none have fully integrated voxelization techniques to deliver real-time, comprehensive 3d weather visualizations specifically tailored for Air Traffic Control Officers (ATCOs). The lack of traditional 2d weather radar displays vertical atmospheric structures for ATCOs, which are critically important for safety and efficiency during flight. VoxelSky-3d is an auxiliary system that uses voxelization to transform weather radar data into 3d voxel-based representations. The methods used in this study include the conversion of coordinates, voxelizing of the geographic coordinate system, and integration with Geographic information Systems for rendering via the Web Graphics Library. The experimental results verified that VoxelSky-3d is an effective application for visualizing weather radar data in 3d, which directly provides ATCOs with intuitive views of the vertical atmospheric conditions. The performance test results revealed that the system could build real-time complex 3d weather scenes and provide high frame rates under relatively high volumes of incoming data. The VoxelSky-3d prototype is a considerable step forward in the field of meteorological data visualization for air traffic control, which improves the situational awareness and decision-making capabilities of ATCOs.

关键词： Meteorology Radar Three-dimensional displays Meteorological radar data visualization Two-dimensional displays doppler radar Safety Prototypes Air traffic control geographic information systems 3d displays weather radar web

来源：评论

学校读者我要写书评

暂无评论

High-quality reproduction method for three-dimensional light-field displays using parallax-view information synthesis and aberration precorrection

引用

OPTICS ANd LASERS IN ENGINEERING 2024年 173卷

作者： Xie, Xinhui Yu, Xunbo Fu, Bangshao Shen, Sheng Zhang, Hui Zhong, Chongli Pei, Xiangyu Sang, Xinzhu Beijing Univ Posts & Telecommun BUPT State Key Lab Informat Photon & Opt Commun Beijing 100876 Peoples R China

Limited by the resolution of display panel and residual aberrations of optical components, it is difficult for the three-dimensional light-field displays (3d LFds) to provide viewers with awesome light field reproduction. Starting from the optical process of reproducing a 3d scene by LFd, a light field display degradation model is constructed to measure the display effect observed by viewer at a specific perspective. Considering the prevalent usage of deep learning techniques in the realm of light field image processing, a novel method based on parallaxview information synthesis and aberration precorrection is proposed to elevate the display performance. This method leverages the complementary parallax information between different perspectives parallax images to enhance the image quality of single-view and introduces a convolutional neural network to estimate the optical components aberrations. A series of optimized parallax images in multiple perspectives are generated and loaded on the display panel through coding, realizing the reproduction of original 3d light field. Simulation experiments and experiments on an LFd prototype verify the effectiveness of the proposed method which can achieve higher quality 3d reproduction.

关键词： 3d light-field display Parallax-view synthesis display effect Precorrection

来源：评论

学校读者我要写书评

暂无评论

A Tutorial on Single-Shot 3d Surface Imaging Techniques [Lecture Notes]

引用

IEEE SIGNAL processing MAGAZINE 2024年第1期41卷 71-92页

作者： Wang, ZhenZhou Univ Kentucky Lexington KY 40506 USA Chinese Acad Sci Beijing Peoples R China Univ Chinese Acad Sci Beijing Peoples R China Shandong Univ Jinan Peoples R China Huaibei Normal Univ Sch Comp Sci & Technol Huaibei 235000 Peoples R China

In recent years, noncontact 3d surface imaging techniques have made great progress due to advances in optical devices and image processing. The single-shot technique is essential for 3d surface imaging techniques to accurately calculate the 3d shape information of the irreversible transient surfaces in the high-speed dynamic scene. In this tutorial, we provide a review of recent advances in existing single-shot 3d surface imaging techniques. The recent advances in single-shot 3d surface imaging techniques lie in the abundant emergence of new methods, the increased measurement resolution, and the increased measurement accuracy. We particularly focus on the single-shot 3d surface imaging techniques that are based on structured light (SL) coding.

关键词： Surface reconstruction Three-dimensional displays image resolution image processing Tutorials Surface treatment Transient analysis Periodic structures Optical devices Encoding

来源：评论

学校读者我要写书评

暂无评论

Photonics Breakthroughs 2024: Large-Viewing-Angle Light Field 3d display Based on High-Precision Beam directionality

IEEE PHOTONICS JOURNAL

引用

IEEE PHOTONICS JOURNAL 2025年第3期17卷

作者： Zhao, Cheng-Bo Liu, Yi-Jian Huang, Yuan-Yi Zhang, Xiao-Tian Xing, Yan Wang, Qiong-Hua Beihang Univ Sch Instrumentat & Optoelect Engn Beijing 100191 Peoples R China

Light field 3d display technology provides viewers with a highly realistic stereoscopic visual experience and achieves authentic and comfortable glasses-free 3d display effects. In recent years, it has become a prominent research focus in the field of display technology. However, light field 3d display is still constrained by the challenges of limited viewing angle and resolution, which significantly impede its advancement. This paper presents a large-viewing-angle light field 3d display based on high-precision beam directionality. To enhance the resolution of 3d images within large viewing range, a highly directional light beam with a divergence angle of +/- 2.7 degrees is achieved by optimizing the display light sources with a compound microlens array. Additionally, a lenticular lens array is employed to reconstruct densely arranged light fields and achieve a large viewing angle. In the experiments, a 100 degrees viewing angle is achieved and it maintains correct geometric occlusion relationships and smooth motion parallax. In addition, the display has a compact architecture and is suitable for the full-color and switchable dynamic light field 3d display with refresh rate of 30 Hz. We expect the proposed display will contribute to the further development of light field 3d displays.

关键词： Three-dimensional displays Lenses Light fields Three-dimensional printing Two-dimensional displays Microoptics Compounds Crosstalk image reconstruction Stereo image processing Light field 3d display resolution beam directionality large-viewing-angle

来源：评论

学校读者我要写书评

暂无评论

Correction of the wavelength error in transmission of a full-color holographic 3d image

引用

OPTICS EXPRESS 2024年第10期32卷 17012-17027页

作者： Lee, Chang-Joo Nishi, Hirohito Matsushima, Kyoji Kansai Univ Dept Elect Elect & Informat Engn 3-3-35 Yamate Cho Suita Osaka 5648680 Japan

When a digital holographic image represented by a sampled wavefield is transmitted and the wavelength used in the three-dimensional (3d) display devices does not agree exactly with the wavelength of the original image data, the reconstructed 3d image will differ slightly from the original. This slight change is particularly problematic for full -color 3d images reconstructed using three wavelengths. A method is proposed here to correct the holographic image data and reduce the problems caused by wavelength mismatch. The effectiveness of the method is confirmed via theoretical analysis and numerical experiments that evaluate the reconstructed images using several image indices.

关键词： Chromatic aberration digital holography Head mounted displays Spherical aberration Three dimensional displays Three dimensional imaging

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：