检索结果-内蒙古大学图书馆

Local deformable 3d Reconstruction with Cartan's Connections

IEEE TRANSACTIONS ON PATTERN ANALYSIS ANd MACHINE INTELLIGENCE 2020年第12期42卷 3011-3026页

作者： Parashar, Shaifali Pizarro, daniel Bartoli, Adrien Univ Auvergne CNRS EnCoV F-63000 Clermont Ferrand France Univ Alcala GEINTRA Alcala De Henares 28801 Spain

3d reconstruction of deformable objects using inter-image visual motion from monocular images has been studied under Shape-from-Template (SfT) and Non-Rigid Structure-from-Motion (NRSfM). Most methods have been developed for simple deformation models, primarily isometry. They may treat a surface as a discrete set of points and draw constraints from the points only or they may use a non-parametric representation and use both points and differentials to express constraints. We propose a differential framework based on Cartans theory of connections and moving frames. It is applicable to SfT and NRSfM, and to deformation models other than isometry. It utilises infinitesimal-level assumptions on the surfaces geometry and mappings. It has the following properties. 1) It allows one to derive existing solutions in a simpler way. 2) It models SfT and NRSfM in a unified way. 3) It allows us to introduce a new skewless deformation model and solve SfT and NRSfM for it. 4) It facilitates a generic solution to SfT which does not require deformation modeling. Our framework is complete: it solves deformable 3d reconstruction for a whole class of algebraic deformation models including isometry. We compared our solutions with the state-of-the-art methods and show that ours outperform in terms of both accuracy and computation time.

关键词： deformable models Three-dimensional displays Surface reconstruction Image reconstruction Solid modeling Mathematical model computer vision 3d reconstruction NRSfM deformable reconstruction connections 3d computer vision

来源：评论

学校读者我要写书评

暂无评论

Total quality control for automotive raw foundry brake disks

引用

INTERNATIONAL JOURNAL OF AdVANCEd MANUFACTURING TECHNOLOGY 2005年第3-4期27卷 359-371页

作者： Lerones, PM Fernández, JL García-Bermejo, JG Zalama, E CARTIF Boecillo 47151 Valladolid Spain Univ Valladolid Dept Automat Control ETSII E-47011 Valladolid Spain

In this paper, a solution for the automatic raw foundry brake disk dimensional characterisation and visual inspection for the automotive industry is presented. Three different computer vision techniques are used: a calibrated 3d structured-light vision technique, for dimensional characterisation and inspection;a 3d uncalibrated structured-light vision technique for local fault detection;and a common 2d-vision technique for a further local fault recognition. A new and fully automated 3d-calibration procedure for piece dimensional characterisation is also described. The whole system is an accurately synchronised blending of mechanics, automation, computer vision and robotics. Results from industrial implementation are presented.

关键词： 3d computer vision automatic calibration automotive industry automatic surface inspection brake disk robotics

来源：评论

学校读者我要写书评

暂无评论

3d Morphable Face Models-Past, Present, and Future

引用

ACM TRANSACTIONS ON GRAPHICS 2020年第5期39卷 157-157页

作者： Egger, Bernhard Smith, William A. P. Tewari, Ayush Wuhrer, Stefanie Zollhoefer, Michael Beeler, Thabo Bernard, Florian Bolkart, Timo Kortylewski, Adam Romdhani, Sami Theobalt, Christian Blanz, Volker Vetter, Thomas Univ York Dept Comp Sci Deramore Lane York YO10 5GH N Yorkshire England MIT 77 Massachusetts AveRoom 46-4066 Cambridge MA 02139 USA Max Planck Inst Informat Saarland Informat Campus D-66123 Saarbrucken Germany Univ Grenoble Alpes Inria CNRS Grenoble INPLJKInst Engn F-38000 Grenoble France Saarland Informat Campus Saarbrucken Germany Stanford Univ 353 Serra Mall Stanford CA 94305 USA Disney Res Studios Zurich Switzerland Max Planck Inst Intelligent Syst Max Planck Ring 4 D-72076 Tubingen Germany Johns Hopkins Univ 3400 North Charles St246 Malone Hall Baltimore MD 21218 USA IDEMIA 2 Pl Samuel Champlain F-92400 Courbevoie France Univ Basel Basel Switzerland Univ Siegen Hoelderlinstr 3 D-57076 Siegen Germany

In this article, we provide a detailed survey of 3d Morphable Face Models over the 20 years since they were first proposed. The challenges in building and applying these models, namely, capture, modeling, image formation, and image analysis, are still active research topics, and we review the state-of-the-art in each of these areas. We also look ahead, identifying unsolved challenges, proposing directions for future research, and highlighting the broad range of current and future applications.

关键词： 3d computer vision computer graphics statistical modeling analysis-by-synthesis generative models

来源：评论

学校读者我要写书评

暂无评论

3d geometric structures and biological activity: Application to microbial soil organic matter decomposition in pore space

引用

ECOLOGICAL MOdELLING 2008年第3-4期216卷 291-302页

作者： Monga, Olivier Bousso, Mamadou Garnier, Patricia Pot, Valerie Ctr IRD Dakar UR GEODES UCAD2 Lab MAT Dakar 18524 Senegal INRA Ctr Grignon UMR Environm & Grandes Cultures F-78850 Thiverval Grignon France

during the past 10 years, soil scientists have started to use 3d Computed Tomography in order to gain a clearer understanding of the geometry of soil structure and its relationships with soil properties. We propose a geometric model for the 3d representation of pore space and a practical method for its computation. Our basic idea consists in representing pore space using a minimal set of maximal balls (delaunay spheres) recovering the shape skeleton. in this representation, each ball could be considered as a maximal local cavity corresponding to the "intuitive" notion of a pore as described in the literature. The space segmentation induced by the network of balls (pores) was then used to spatialize biological dynamics. Organic matter and microbial decomposers were distributed within the balls (pores). A valuated graph representing the pore network, organic matter and distribution of micro-organisms was then defined. Microbial soil organic matter decomposition was simulated by updating this valuated graph. The method was implemented and tested using real CT images. The model produced realistic simulated results when compared with data in the literature in terms of the water retention curve and carbon mineralization. A decrease in water pressure decreased carbon mineralization, which is also in accordance with findings in the literature. From our results we showed that the influence of water pressure on decomposition is a function of organic matter distribution in the pore space. As far as we know, this is the approach to have linked pore space geometry and biological dynamics in a formal way. Our next goal will be to compare the model with experimental data of decomposition using different soil structures, and to define geometric typologies of pore space shape that can be attached to specific biological and dynamic properties. (C) 2008 Elsevier B.V. All rights reserved.

关键词： computational geometry soil science pore space modelling microbial decomposition simulation 3d computer vision

来源：评论

学校读者我要写书评

暂无评论

Improved Adaptive-Support Stereo Matching Using Local Features

引用

JOURNAL OF INFORMATION SCIENCE ANd ENGINEERING 2015年第5期31卷 1557-1575页

作者： Yang, Shih-Hsuan Wang, Bossun Natl Taipei Univ Technol Dept Comp Sci & Informat Engn Taipei 106 Taiwan

Stereo matching is a process of estimating the disparity for every pixel in a pair of stereo images. Among the various stereo-matching algorithms, the adaptive support-weight (ASW) approach is the most representative local method that determines the disparity by using statistical correlation within a local support window. In the ASW approach, the degree of correspondence between two pixels within the support window is jointly measured using color similarity and geometric proximity. Although ASW is simple and typically effective, it does not provide satisfactory matching results for regions with ambiguous patterns or no identifiable image features. This study improved ASW by incorporating numerous local image features. First, an alternative adaptive cross-based support window, instead of the fixed square window, was used to increase the matching accuracy if a sparse-textured region surrounds the anchor pixel. Second, Canny edges in dense-textured regions were used for correspondence matching. Third, a smoothness constraint on disparity continuity was imposed to shrink the errors on object boundaries. Fourth, a disparity-refinement process was performed on the inconsistent pixels from the left-right consistency check. Experimental results on the Middlebury test bed confirmed the superiority of the proposed method over the ASW. Compared with other state-of-the-art ASW-based algorithms, the proposed method is among the most effective and achieves stable and efficient performance with only a slight increase in processing time.

关键词： adaptive support-weight (ASW) depth map disparity estimation stereo matching 3d computer vision

来源：评论

学校读者我要写书评

暂无评论

HiSC4d: Human-Centered Interaction and 4d Scene Capture in Large-Scale Space Using Wearable IMUs and LidAR

引用

IEEE TRANSACTIONS ON PATTERN ANALYSIS ANd MACHINE INTELLIGENCE 2024年第12期46卷 11236-11253页

作者： dai, Yudi Wang, Zhiyong Lin, Xiping Wen, Chenglu Xu, Lan Shen, Siqi Ma, Yuexin Wang, Cheng Xiamen Univ Fujian Key Lab Sensing & Comp Smart Cities Xiamen 361005 Peoples R China Xiamen Univ Sch Informat Xiamen 361005 Peoples R China ShanghaiTech Univ Sch Informat Sci & Technol Shanghai 201210 Peoples R China

We introduce HiSC4d, a novel Human-centered interaction and 4d Scene Capture method, aimed at accurately and efficiently creating a dynamic digital world, containing large-scale indoor-outdoor scenes, diverse human motions, rich human-human interactions, and human-environment interactions. By utilizing body-mounted IMUs and a head-mounted LidAR, HiSC4d can capture egocentric human motions in unconstrained space without the need for external devices and pre-built maps. This affords great flexibility and accessibility for human-centered interaction and 4d scene capturing in various environments. Taking into account that IMUs can capture human spatially unrestricted poses but are prone to drifting for long-period using, and while LidAR is stable for global localization but rough for local positions and orientations, HiSC4d employs a joint optimization method, harmonizing all sensors and utilizing environment cues, yielding promising results for long-term capture in large scenes. To promote research of egocentric human interaction in large scenes and facilitate downstream tasks, we also present a dataset, containing 8 sequences in 4 large scenes (200 to 5,000 m(2)), providing 36 k frames of accurate 4d human motions with SMPL annotations and dynamic scenes, 31k frames of cropped human point clouds, and scene mesh of the environment. A variety of scenarios, such as the basketball gym and commercial street, alongside challenging human motions, such as daily greeting, one-on-one basketball playing, and tour guiding, demonstrate the effectiveness and the generalization ability of HiSC4d. The dataset and code will be publicly available for research purposes.

关键词： 3d computer vision dataset human-scene interaction IMU motion capture point cloud scene mapping

来源：评论

学校读者我要写书评

暂无评论

Using pore space 3d geometrical modelling to simulate biological activity: Impact of soil structure

引用

computerS & GEOSCIENCES 2009年第9期35卷 1789-1801页

作者： Monga, Olivier Bousso, Mamadou Garnier, Patricia Pot, Valerie IRD UR GEODES Ctr IRD Dakar Lab MAT IRD UCAD2 Dakar 18524 Senegal INRA Ctr Grignon UMR Environm & Grandes Cultures F-78850 Thiverval Grignon France

This study is the follow-up to a previous one devoted to soil pore space modelling. In the previous study, we proposed algorithms to represent soil pore space by means of optimal piecewise approximation using simple 3d geometrical primitives: balls, cylinders, cones, etc. In the present study, we use the ball-based piecewise approximation to simulate biological activity. The basic idea for modelling pore space consists in representing pore space using a minimal set of maximal balls (delaunay spheres) recovering the shape skeleton. In this representation, each ball is considered as a maximal local cavity corresponding to the "intuitive" notion of a pore as described in the literature. The space segmentation induced by the network of balls (pores) is then used to spatialise biological dynamics. Organic matter and microbial decomposers are distributed within the balls (pores). A valuated graph representing the pore network, organic matter and microorganism distribution is then defined. Microbial soil organic matter decomposition is simulated by updating this valuated graph. The method has been implemented and tested on real data. As far as we know, this approach is the first one to formally link pore space geometry and biological dynamics. The long-term goal is to define geometrical typologies of pore space shape that can be attached to specific biological dynamic properties. This paper is a first attempt to achieve this goal. (c) 2009 Elsevier Ltd. All rights reserved.

关键词： 3d computer vision Biological dynamics simulation Computed tomography Computational geometry Microbial decomposition Pore space modelling

来源：评论

学校读者我要写书评

暂无评论

Localization of furniture parts by integrating range and intensity data robust against depths with low signal-to-noise ratio

引用

ROBOTICS ANd AUTONOMOUS SYSTEMS 2014年第1期62卷 25-37页

作者： Meissner, Pascal Schmidt-Rohr, Sven R. Loesch, Martin Jaekel, Rainer dillmann, Ruediger FZI Res Ctr Informat Technol Karlsruhe Germany Karlsruhe Inst Technol Inst Anthropomat D-76021 Karlsruhe Germany

In this article we present an approach for localizing planar parts of furniture in depth data from range cameras. It estimates both their six-degree-of-freedom poses and their dimensions. The system has been designed for enabling robots to autonomously manipulate furniture. Range cameras are a promising sensor category for this application. As many of them provide data with considerable noise and distortions, detecting objects, for example, using canonical methods for range data segmentation or feature extraction, is complicated. In contrast, our approach is able to overcome these issues. This is done by combining concepts of 2d and 3d computer vision as well as integrating intensity and range information in multiple steps of our processing chain. Therefore it can be employed on range sensors with both low and high signal-to-noise ratios and in particular on time-of-flight cameras. This concept can be adapted to various object shapes. It has been implemented for object parts with shapes similar to ellipses as a proof-of-concept. For this, a state-of-the-art ellipse detection method has been enhanced regarding our application. (C) 2012 Elsevier B.V. All rights reserved.

关键词： 3d computer vision Localization Range and intensity data Furniture Kinect Time-of-flight camera

来源：评论

学校读者我要写书评

暂无评论

3d Hand Pose Estimation From Monocular RGB With Feature Interaction Module

引用

IEEE TRANSACTIONS ON CIRCUITS ANd SYSTEMS FOR VIdEO TECHNOLOGY 2022年第8期32卷 5293-5306页

作者： Guo, Shaoxiang Rigall, Eric Ju, Yakun dong, Junyu Ocean Univ China Sch Comp Sci & Technol Qingdao 266005 Peoples R China Ocean Univ China Dept Informat Sci & Technol Qingdao 266005 Peoples R China

3d hand pose estimation from a monocular RGB image is a highly challenging task due to self-occlusion, diverse appearances, and inherent depth ambiguities within monocular images. Most of the previous methods first employ deep neural networks to fit 2d joint location maps, then combines them with implicit or explicit pose-aware features to directly regress 3d hand joints positions using their designed network structure. However, the skeleton positions and corresponding skeleton-aware content information located in the latent space are invariably ignored. These skeleton-aware contents effectively bridge the gap between hand joint and hand skeleton information by associating the relationship between different hand joints features and the hand skeleton positions distribution in 2d space. To address this issue, we propose a simple yet efficient deep neural network to directly recover reliable 3d hand pose from monocular RGB images with faster estimation process. Our purpose is the reduction of the model computational complexity while maintaining high precision performance. Therefore, we design a novel Feature Chat Block (FCB) to complete feature boosting, which enables the intuitively enhanced interaction between joint and skeleton features. First, this FCB module updates joint features effectively based on semantic graph convolutional neural network and multi-head self-attention mechanism. The GCN-based structure focuses on the physical hand joints included in a binary adjacency matrix and the self-attention part pays attention to hand joints located in a complementary matrix. Then, the FCB module employs query and key mechanisms respectively representing joint and skeleton features to further implement feature interaction. After a set of FCB modules, our model updates the fused features in a coarse-to-fine manner and finally outputs a predicted 3d hand pose. We conducted a comprehensive set of ablation experiments on the InterHand2.6M dataset to validate the effective

关键词： Three-dimensional displays Skeleton Pose estimation Solid modeling Feature extraction Task analysis Semantics Hand pose estimation graph CNN transformer feature interaction 3d computer vision

来源：评论

学校读者我要写书评

暂无评论

3d vision-based bolt loosening assessment using photogrammetry, deep neural networks, and 3d point-cloud processing

引用

JOURNAL OF BUILdING ENGINEERING 2023年 70卷

作者： Pan, Xiao Yang, T. Y. Univ British Columbia Dept Civil Engn Vancouver BC Canada

Structural bolts are essential structural elements. detection of structural bolt loosening is of great importance to provide earlier warnings of structural damages and prevent catastrophic system level collapse. Most existing studies about bolt loosening assessment were built in 2d computer vision, where the assessment may be restricted based on the camera views. In this paper, a novel 3d vision-based methodology is proposed for autonomous bolt loosening assessment. First, a 3d point cloud of bolted connection is created using readily available 2d images. Second, a new convolutional neural network (CNN)-based method is developed to localize structural bolts in the 3d point cloud. Further, a 3d point cloud processing algorithm is developed to recognize and quantify bolt loosening. Parameter studies were conducted to investigate the effectiveness of the proposed pipeline. Finally, a real-world implementation has been conducted to quantify bolt loosening on a steel column base connection with bolts. The results indicate that the proposed bolt loosening assessment methodology can effectively localize and quantify bolt loosening at high accuracy and low cost.

关键词： Structural health monitoring Structural bolt loosening 3d computer vision deep learning 3d point cloud processing

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：