检索结果-内蒙古大学图书馆

3d Point Cloud Matching Based Selfie Generation for Chang'e-5

JOURNAL OF COMPUTER SCIENCE ANd TECHNOLOGY 2025年第1期40卷 85-98页

作者： Chen, Xiao-Rui Yang, Meng-Fei Zhang, Gao Zhang, Wu Peng, Jing Gu, Zheng deng, Xiang-Jin Yang, Liu-Zhi Yang, Fei Yang, Yun Sun, Shou-Qian Tong, Ruo-Feng Tang, Min Zhejiang Univ Coll Comp Sci & Technol Hangzhou 310007 Peoples R China China Acad Space Technol Beijing 100094 Peoples R China Beijing Inst Spacecraft Syst Engn Beijing 100094 Peoples R China

Generating selfie images on the surface of a celestial body poses several challenges, including the position of the robotic arm, camera field of view, and limited shooting time. To address these challenges, the PCMIS (3d Point Cloud Matching Based image Stitching) algorithm is designed, along with a corresponding shooting plan. This algorithm establishes a correspondence between depth and color information, enabling the generation of stitching views under any given view parameter. Furthermore, the algorithm is accelerated using GPU processing, resulting in a significant reduction in stitching time. The algorithm is successfully applied to generate selfie images for the Chang'e-5 mission.

关键词： image stitching 3d point cloud matching Chang'e-5 GPU computing selfie generation

来源：评论

学校读者我要写书评

暂无评论

Biometric Recognition in 3d Medical images: A Survey

引用

IEEE ACCESS 2023年 11卷 125601-125615页

作者： Fu, Biying damer, Naser Fraunhofer Inst Comp Graph Res IGD D-64283 Darmstadt Germany RheinMain Univ Appl Sci D-65195 Wiesbaden Germany Tech Univ Darmstadt TU Darmstadt Dept Comp Sci D-64283 Darmstadt Germany

This work explores the potential applications of biometric recognition in 3d medical imaging data. We investigate various 3d imaging techniques commonly used in the medical domain, including 3d ultrasound imaging, magnetic resonance imaging (MRI), computer tomography (CT) scans, and 3d near-infrared (NIR) imaging. For each technique, we provide an overview of its working principle and discuss the advantages of integrating biometrics into 3d medical imaging data. Major advantage of using biometrics in this context is motivated by the research that using biometrics could not only increase data security but, more importantly, decrease the mix-up errors in patient's medical data and thus improving patient safety and patient care. Our analysis uncovers certain weaknesses in current algorithms and limitations in existing research. Possible reasons include insufficient data availability, the under-utilization of deep-learning-based approaches to enhance accuracy and performance, and the absence of standardized benchmarking databases to support research. Our survey frames existing works and lead to practical recommendations and motivates efforts to improve the current state of research. Beyond exploring the utilization of biometrics in 3d medical imaging data, our study touches on further potential interactions between them, such as extracting health information from biometric captures. This work is thus the first work to survey and present an overview on works proposing the use of medical images for biometric recognition.

关键词： Biometrics (access control) Biomedical imaging Medical diagnostic imaging Three-dimensional displays Imaging Surveys image recognition Biometrics biometric recognition 3d medical imaging biometric in 3d medical data

来源：评论

学校读者我要写书评

暂无评论

Position Encoding for 3d Lane detection via Perspective Transformer

引用

IEEE ACCESS 2024年 12卷 106480-106487页

作者： Zhang, Meng Li Wang, Ming Wei deng, Yan Yang Lei, Xin Yu Shaanxi Univ Sci & Technol Shaanxi Joint Lab Artificial Intelligence Xian 710021 Peoples R China

3d lane detection from the input monocular image is a basic but indispensable task in the environment perception of automatic driving. Recent work uses modules such as depth estimation, coordinate system transformation, and time series tracking to achieve the correspondence of 2d to 3d information. However, inaccurate depth information generated by perturbations during conversion poses a challenge to lane detection methods that rely only on monocular images. To solve the above problems, we propose a PELd model, a proxy transformation using BEV aerial view, to explicitly give 3d lane detection results. Specifically, when sampling feature information, feature flipping is proposed to supplement the global context information before view conversion, and the 3d position encoding information generated by the forward-looking features enhances the depth information. After the 3d position encoding information is combined with the feature information, the cross-attention module is used as a value for adaptive supervision of BEV queries. On the one hand, we use deformable attention to sample forward looking features and generate explicit lane representation;on the other hand, we supplement supervised lane line generation by supplementing forward looking features and enhancing 3d spatial information. PELd implements a more advanced approach than ever before on OpenLane and Apollo datasets.

关键词： Three-dimensional displays Feature extraction Lane detection Encoding decoding Task analysis Convolution deep learning Machine learning 3d lane detection position embedding view conversion autonomous vehicle deep learning machine learning

来源：评论

学校读者我要写书评

暂无评论

Improving small intestinal stromal tumor detection using 3d context and auxiliary priori cues

引用

BIOMEdICAL SIGNAL processing ANd CONTROL 2025年 102卷

作者： Ju, Jianguo Qiu, dandan Liu, Ying Zhang, Tongtong Tu, Huijuan Xu, Zuming Cheng, Xiaochun Xian Univ Sci & Technol Coll Comp Sci & Technol Xian Shaanxi Peoples R China Northwest Univ Sch Informat Sci & Technol Xian Shaanxi Peoples R China Kunshan Hosp Chinese Med Dept Radiol Kunshan Jiangsu Peoples R China Xidian Univ Sch Comp Sci & Technol Xian Shaanxi Peoples R China Swansea Univ Dept Comp Sci Swansea Wales

Small Intestinal Stromal Tumor (SIST) is a potentially malignant neoplasm characterized by a lack of specific clinical manifestations. Accurately detecting SIST from computed tomography (CT) scans is a crucial but challenging endeavor within precision oncology imaging. Manual annotation currently serves as the prevailing clinical practice for tumors' longitudinal assessment, proving to be both highly time-consuming and prone to inconsistencies. Effectively training an automatic detection algorithm is desirable but relies heavily on the availability of abundant pixel-level labeled data. How to exploit limited medical image features to obtain accurate feature representation is a mainstream research topic. Embedding 3d context into 2d networks is acknowledged to be advantageous for this differential task. Nevertheless, current detection algorithms integrate 3d contextual information at the backbone backend, resulting in the loss of such information in the early stages. This limitation hampers the algorithm's capability to effectively learn more intricate 3d features. In this work, we focus on SIST detection in CT images and propose a novel SIST detection algorithm that leverages 3d contextual information and incorporates weakly supervised segmentation as an auxiliary method to automatically identify and classify lesions. Specifically, our algorithm takes slice grouping as input to introduce contextual information and designs a 3d multi-scale feature fusion network (3dMF Net) to fully integrate 3d contextual features and multi-scale features, effectively enriching the feature expression of SIST. Furthermore, to alleviate a substantial number of false positive problems in SIST prediction results, we design a weakly supervised semantic segmentation auxiliary branch, using box-level annotations as initial pseudo-masks, and optimized by a graph theoretical network. Finally, our segmentation branch provides auxiliary prior cues to the detection branch, achieving more robust

关键词： Small intestinal stromal tumor Tumor detection 3d context Weakly supervised segmentation auxiliary

来源：评论

学校读者我要写书评

暂无评论

3d rock fragmentation analysis using lidar, based on point cloud deep learning segmentation and synthetic data

引用

POWdER TECHNOLOGY 2025年 456卷

作者： Hafshejani, Mojgan Faramarzi Esmaeili, Kamran Univ Toronto Dept Civil & Mineral Engn 170 Coll St Toronto ON M5S 3E3 Canada

Accurate online measurement of particle size distribution is crucial in mining, tunnelling, and mineral processing industries to enable intelligent process control and optimization, ultimately enhancing efficiency and productivity. The current method for rock fragmentation relies on 2d image analysis, which is highly dependent on optimal lighting conditions, limiting its applicability and robustness in the challenging lighting environments commonly found in mining. This study diverges from the prevalent 2d image and photogrammetry approaches in rock fragmentation analysis, and pioneers a novel approach by harnessing laser scanner data for point cloud segmentation, offering a promising solution to overcome the limitations of image analysis techniques. By leveraging laser scanner data, a robust framework for rock fragmentation analysis is developed that is tailored to the specific challenges related to lighting situations. To avoid the laborious task of collecting and labelling point cloud datasets, this research introduces an innovative approach of using synthetic labeled datasets of scanned rockpiles. A platform is developed to automatically create and scan labeled point clouds of rock piles, facilitating the utilization of transfer learning. The synthetic 3d dataset was used to train a deep learning model for precise segmentation of rock instances in three-dimensional coordinates, providing an accurate representation of the rock object in 3d. The accuracy of the developed predictive model was tested and validated on experimental laser scanning data of three different rock piles. The proposed method depends on coordinate data instead of RGB information, rendering it particularly applicable in challenging conditions such as underground mining, night shifts, or situations where maintaining optimal lighting conditions is difficult or costly. The findings present a significant leap forward in rock fragmentation analysis, opening avenues for enhanced practices in diverse

关键词： Rock fragmentation analysis deep learning Point cloud instance segmentation Synthetic data generation 3d size distribution analysis Computer vision Laser scanner applications

来源：评论

学校读者我要写书评

暂无评论

A Study on Vector-Based processing and Texture Application Techniques for 3d Object Creation and Visualization

引用

APPLIEd SCIENCES-BASEL 2025年第7期15卷 4011-4011页

作者： Kang, donghwi Kim, Jeongyeon Lee, Jongchan Lee, Haeju Kim, Jihyeok Byun, Jungwon WISEiTECH Res Inst Gwacheon 13824 South Korea

This study proposes a technique for generating 3d objects from Shapefile-based 2d spatial data and converting them to comply with the CityGML 3.0 standard. In particular, the proposed Wise Interpolated Texture (hereafter referred to as WIT) technique optimizes texture mapping and enhances visual quality. High-resolution Z-values were extracted using dEM data, and computational efficiency was improved by applying the constrained delaunay triangulation algorithm. This study implemented more realistic visual representations using high-resolution orthorectified imagery (hereafter referred to as orthoimages) TIF files and improved data retrieval speed compared to existing raster methods through vector-based processing techniques. In this research, data weight reduction, parallel processing, and polygon simplification algorithms were applied to optimize the 3d model generation speed. Additionally, the WIT technique minimized discontinuity between textures and improved UV mapping alignment to achieve more natural and uniform textures. Experimental results confirmed that the proposed technique improved texture mapping speed, enhanced rendering quality, and increased large-scale data processing efficiency compared to conventional methods. Nevertheless, limitations still exist in real-time data integration and optimization of large-scale 3d models. Future research should consider dynamic modeling reflecting real-time image data, BIM data integration, and large-scale texture streaming techniques.

关键词： vector-based processing 3d object creation CityGML 3.0 texture mapping digital twin

来源：评论

学校读者我要写书评

暂无评论

NeRF-dA: Neural Radiance Fields deblurring With Active Learning

引用

IEEE SIGNAL processing LETTERS 2025年 32卷 261-265页

作者： Hong, Sejun Kim, Eunwoo Chung Ang Univ Dept AI Seoul 06974 South Korea Chung Ang Univ Sch Comp Sci & Engn Seoul 06974 South Korea

Neural radiance fields (NeRF) represent multi-view images as 3d scenes, achieving a photo-realistic novel view synthesis quality. However, capturing multi-view images in real-world scenarios is not well aligned and often results in blur or noise. deblur-NeRF, which uses kernel deformation to improve sharpness, is effective but the quantity of training blur samples and imbalance significantly affect the overall results. In this study, we propose neural radiance fields deblurring with active learning (NeRF-dA), focusing on high-quality blurred images for 3d scene modeling. NeRF-dA uses pool-based active learning with uncertainty estimation to improve model efficiency with a high-quality training set. Subsequently, we deblur the data using the trained model and proceed with NeRF training by selecting the best-sharpened images for querying. Experiments on both camera motion blur and defocus blur demonstrate that NeRF-dA significantly enhances the quality of the existing deblur-NeRF.

关键词： Cameras Neural radiance field Training Entropy Kernel image reconstruction Light fields Three-dimensional displays image edge detection image color analysis Active learning deblurring neural radiance fields (NeRFs)

来源：评论

学校读者我要写书评

暂无评论

Bidirectional Hybrid LSTM Based Recurrent Neural Network for Multi-View Stereo

引用

IEEE TRANSACTIONS ON VISUALIZATION ANd COMPUTER GRAPHICS 2024年第7期30卷 3062-3073页

作者： Wei, Zizhuang Zhu, Qingtian Min, Chen Chen, Yisong Wang, Guoping Peking Univ Dept EECS Beijing 100871 Peoples R China

Recently, deep learning based multi-view stereo (MVS) networks have demonstrated their excellent performance on various benchmarks. In this paper, we present an effective and efficient recurrent neural network (RNN) for accurate and complete dense point cloud reconstruction. Instead of regularizing the cost volume via conventional 3d CNN or unidirectional RNN like previous attempts, we adopt a bidirectional hybrid Long Short-Term Memory (LSTM) based structure for cost volume regularization. The proposed bidirectional recurrent regularization is able to perceive full-space context information comparable to 3d CNNs while saving runtime memory. For post-processing, we introduce a visibility based approach for depth map refinement to obtain more accurate dense point clouds. Extensive experiments on dTU, Tanks and Temples and ETH3d datasets demonstrate that our method outperforms previous state-of-the-art MVS methods and exhibits high memory efficiency at runtime.

关键词： Costs Feature extraction Three-dimensional displays Runtime Point cloud compression image reconstruction Recurrent neural networks 3d reconstruction deep learning multi-view stereo recurrent neural network point clouds

来源：评论

学校读者我要写书评

暂无评论

Cross-Supervised LidAR-Camera Fusion for 3d Object detection

引用

IEEE ACCESS 2025年 13卷 10447-10458页

作者： Zuo, Chao Jie Gu, Cao Yu Guo, Yi Kun Miao, Xiao dong Nanjing Tech Univ Sch Mech & Power Engn Nanjing 211889 Peoples R China

Fusing LidAR and camera information is essential for achieving accurate and reliable 3d object detection in autonomous driving systems. due to inherent differences between different modalities, seeking an efficient and accurate fusion method is of great importance. Recently, significant progress has been made in 3d object detection methods based on lift-splat-shot (LSS-based) approaches. However, inaccurate depth estimation and substantial semantic information loss remain significant factors limiting the accuracy of 3d detection. In this paper, we propose a cross-fusion framework under a dual spatial representation, by integrating information in different spatial representations, namely bird's-eye view (BEV) and camera view, and establishing soft links to fully utilize the information carried by different modalities. It consists of two important components, gated LidAR supervised BEV (GLS-BEV) and multi-attention cross fusion (MACF) modules. The former achieves accurate depth estimation by supervising the transformation of LidAR data with clear depth into the image space, constructing point cloud features in vehicle's perspective. The latter utilizes three sub-attention modules with different roles to achieve cross-modal interaction within the same space, effectively reducing semantic loss. On the nuScenes benchmark, our proposed method achieves outstanding 3d object detection results with 71.8 mAP and 74.2 NdS.

关键词： Feature extraction Three-dimensional displays Laser radar Point cloud compression Cameras Object detection Semantics Accuracy Logic gates depth measurement 3d object detection LidAR-camera system multi-sensor fusion BEV

来源：评论

学校读者我要写书评

暂无评论

Rethinking Masked Representation Learning for 3d Point Cloud Understanding

引用

IEEE TRANSACTIONS ON image processing 2025年 34卷 247-262页

作者： Wang, Chuxin Zha, Yixin He, Jianfeng Yang, Wenfei Zhang, Tianzhu Univ Sci & Technol China Sch Informat Sci & Technol Hefei 230027 Peoples R China Univ Sci & Technol China Deep Space Explorat Lab Hefei 230027 Peoples R China

Self-supervised point cloud representation learning aims to acquire robust and general feature representations from unlabeled data. Recently, masked point modeling-based methods have shown significant performance improvements for point cloud understanding, yet these methods rely on overlapping grouping strategies (k-nearest neighbor algorithm) resulting in early leakage of structural information of mask groups, and overlook the semantic modeling of object components resulting in parts with the same semantics having obvious feature differences due to position differences. In this work, we rethink grouping strategies and pretext tasks that are more suitable for self-supervised point cloud representation learning and propose a novel hierarchical masked representation learning method, including an optimal transport-based hierarchical grouping strategy, a prototype-based part modeling module, and a hierarchical attention encoder. The proposed method enjoys several merits. First, the proposed grouping strategy partitions the point cloud into non-overlapping groups, eliminating the early leakage of structural information in the masked groups. Second, the proposed prototype-based part modeling module dynamically models different object components, ensuring feature consistency on parts with the same semantics. Extensive experiments on four downstream tasks demonstrate that our method surpasses state-of-the-art 3d representation learning methods. Furthermore, Comprehensive ablation studies and visualizations demonstrate the effectiveness of the proposed modules.

关键词： Point cloud compression Semantics Feature extraction Three-dimensional displays Representation learning Solid modeling Prototypes Shape Nearest neighbor methods image reconstruction Self-supervised point cloud representation learning optimal transport and part modeling

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：