5,556 Hits in 8.1 sec

Efficient Dense Point Cloud Object Reconstruction Using Deformation Vector Fields [chapter]

Kejie Li, Trung Pham, Huangying Zhan, Ian Reid
2018 Lecture Notes in Computer Science  
Extensive experiments demonstrate the efficiency and efficacy of our method on single-view 3D object reconstruction.  ...  Each surface comprises a depth-map and corresponding deformation-field that ensures every pixel-depth pair in the depth-map lies on the object surface.  ...  In this paper, we present a novel and highly efficient framework (shown in Fig. 1 ) to generate dense point clouds for representing the 3D shape of objects.  ... 
doi:10.1007/978-3-030-01258-8_31 fatcat:6nkbo5r2fbg2hlcc6beligrekm

VIRDO: Visio-tactile Implicit Representations of Deformable Objects [article]

Youngsun Wi, Pete Florence, Andy Zeng, Nima Fazeli
2022 arXiv   pre-print
VIRDO operates directly on visual (point cloud) and tactile (reaction forces) modalities and learns rich latent embeddings of contact locations and forces to predict object deformations subject to external  ...  Deformable object manipulation requires computationally efficient representations that are compatible with robotic sensing modalities.  ...  Deformation Module and Deformation Field: We define the deformation field as a 3D vector field that pushes a deformed object back to its original (nominal) shape.  ... 
arXiv:2202.00868v1 fatcat:f3beuu5dwve2faukgd2cpomo5m

FlowNet3D++: Geometric Losses For Deep Scene Flow Estimation [article]

Zirui Wang, Shuda Li, Henry Howard-Jenkins, Victor Adrian Prisacariu, Min Chen
2021 arXiv   pre-print
Inspired by classical methods, FlowNet3D++ incorporates geometric constraints in the form of point-to-plane distance and angular alignment between individual vectors in the flow field, into FlowNet3D.  ...  This is made possible through the contribution of a novel pipeline to integrate point-based scene flow predictions into a global dense volume.  ...  Section 3 describes our modifications to FlowNet3D, as well as our method to integrate the sparse scene flow vector field in dense dynamic reconstruction.  ... 
arXiv:1912.01438v3 fatcat:swwyodalwfdqlfu7d2fg7t5xui

Dynamic Reconstruction of Deformable Soft-Tissue With Stereo Scope in Minimal Invasive Surgery

Jingwei Song, Jun Wang, Liang Zhao, Shoudong Huang, Gamini Dissanayake
2018 IEEE Robotics and Automation Letters  
This paper proposes an innovative Simultaneous Localization and Mapping (SLAM) algorithm for deformable dense reconstruction of surfaces using a sequence of images from a stereoscope.  ...  The warping field is estimated by deforming the last updated model to the current live model.  ...  we directly acquire dense point cloud as data management.  ... 
doi:10.1109/lra.2017.2735487 dblp:journals/ral/SongWZHD18 fatcat:pq67bvv7h5d6jmwuw6p2bu6uvy

DeformNet: Free-Form Deformation Network for 3D Shape Reconstruction from a Single Image [article]

Andrey Kurenkov, Jingwei Ji, Animesh Garg, Viraj Mehta, JunYoung Gwak, Christopher Choy, Silvio Savarese
2017 arXiv   pre-print
Prior methods have tackled this problem through generative models which predict 3D reconstructions as voxels or point clouds.  ...  We introduce a new differentiable layer for 3D data deformation and use it in DeformNet to learn a model for 3D reconstruction-through-deformation.  ...  Figure 4 shows the comparison between reconstructed sparse and dense point clouds, using the model trained on sparse point clouds.  ... 
arXiv:1708.04672v1 fatcat:tdbjw6rn5bch3kuidapivqgsd4

Transferring Grasping Skills to Novel Instances by Latent Space Non-Rigid Registration [article]

Diego Rodriguez, Corbin Cogswell, Seongyong Koo, Sven Behnke
2018 arXiv   pre-print
The known object instances are modeled using a canonical shape and a transformation which deforms it to match the instance shape.  ...  object category.  ...  CPD provides a dense deformation field, allowing us to find deformation vectors for novel points, even those added after the field is created, which in turn allows us to apply the method for transferring  ... 
arXiv:1809.05353v1 fatcat:ykab5eqcrjd55mbdggx4c3srqq

Toward a real-time tracking of dense point-sampled geometry

F. Destelle, C. Roudet, M. Neveu, A. Dipanda
2012 2012 19th IEEE International Conference on Image Processing  
Experimental results obtained from both simulated and real reconstructed data sets demonstrate that the proposed method can handle efficiently the tracking process even for very large point clouds.  ...  In this paper, we address the problem of tracking temporal deformations between two arbitrary densely sampled pointbased surfaces.  ...  We propose a voxel-based method suited to the efficient tracking of deformable shapes over small-time step captures. This framework is designed to handle the tracking of arbitrary dense point clouds.  ... 
doi:10.1109/icip.2012.6466875 dblp:conf/icip/DestelleRND12 fatcat:rfrlaqerabdapn7ad2nhyv4kfa

Live dense reconstruction with a single moving camera

Richard A. Newcombe, Andrew J. Davison
2010 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition  
We take point-based real-time structure from motion (SFM) as our starting point, generating accurate 3D camera pose estimates and a sparse point cloud.  ...  We present a method which enables rapid and dense reconstruction of scenes browsed by a single live camera.  ...  We use real-time Structure from Motion to furnish a live camera pose estimate and a sparse point cloud which is the starting point for dense scene reconstruction.  ... 
doi:10.1109/cvpr.2010.5539794 dblp:conf/cvpr/NewcombeD10 fatcat:7b7yvrjlv5aipcurmtvj6xjxfa

RFNet-4D: Joint Object Reconstruction and Flow Estimation from 4D Point Clouds [article]

Tuan-Anh Vu, Duc Thanh Nguyen, Binh-Son Hua, Quang-Hieu Pham, Sai-Kit Yeung
2022 arXiv   pre-print
Object reconstruction from 3D point clouds has achieved impressive progress in the computer vision and computer graphics research field.  ...  To prove this ability, we design a temporal vector field learning module using unsupervised learning approach for flow estimation, leveraged by supervised learning of spatial structures for object reconstruction  ...  point cloud in space and time to implicitly represent trajectories of all the points in dense correspondences between occupancy fields.  ... 
arXiv:2203.16482v2 fatcat:3k4hspxebrfxfd3lbhmz76od44

Learning Free-Form Deformations for 3D Object Reconstruction [article]

Dominic Jack, Jhony K. Pontes, Sridha Sridharan, Clinton Fookes, Sareh Shirazi, Frederic Maire, Anders Eriksson
2018 arXiv   pre-print
By learning to deform points sampled from a high-quality mesh, our trained model can be used to produce arbitrarily dense point clouds or meshes with fine-grained geometry.  ...  Although point clouds are more efficient than voxel representations as they only cover surfaces rather than volumes, they do not encode detailed geometric information about relationships between points  ...  Computational resources and services used in this work were provided by the HPC and Research Support Group, Queensland University of Technology, Brisbane, Australia.  ... 
arXiv:1803.10932v1 fatcat:c5cn2sjcfjfddeazctcyltxrqm

GraphX-Convolution for Point Cloud Deformation in 2D-to-3D Conversion [article]

Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sanghoon Lee
2019 arXiv   pre-print
In this paper, we present a novel deep method to reconstruct a point cloud of an object from a single still image.  ...  We propose to overcome these by deforming a random point cloud to the object shape through two steps: feature blending and deformation.  ...  The per-point and global features are processed by a deformation network to produce a point cloud for the given object.  ... 
arXiv:1911.06600v1 fatcat:abjucudg2zeopjproq72vsvtra

Pano2Scene: 3D Indoor Semantic Scene Reconstruction from a Single Indoor Panorama Image

Wei Zeng, Sezer Karaoglu, Theo Gevers
2020 British Machine Vision Conference  
the full scene point cloud.  ...  3D indoor semantic scene reconstruction from 2D images is challenging as it requires both scene understanding and object reconstruction.  ...  requires clean and dense meshes to supervise object mesh reconstruction.  ... 
dblp:conf/bmvc/ZengKG20 fatcat:ecvl7wgh2fh43j5wamtxg6htte

3D Correspondence and Point Projection Method for Structures Deformation Analysis

Aurelio G. Melo, Milena F. Pinto, Leonardo M. Honorio, Felipe M. Dias, Juliano E. N. Masson
2020 IEEE Access  
Inspected Object K K P a ir 1 After alignment a dense set of points p t i , a point is projected in all visible images.  ...  However, due to the lack of regular structures and smooth objects, the deformation analysis of natural scenes is particularly demanding for point cloud since two points cannot be readily identified in  ... 
doi:10.1109/access.2020.3027205 fatcat:liyg2smqevbc7ga6oiprlxk7na

Joint Reconstruction and Registration of a Deformable Planar Surface Observed by a 3D Sensor

Umberto Castellani, Vincent Gay-Bellile, Adrien Bartoli
2007 3-D Digital Imaging and Modeling  
We address the problem of reconstruction and registration of a deforming 3D surface observed by some 3D sensor giving a cloud of 3D points at each time instant.  ...  Since the Hessian-matrix H = J J must be inverted at each LM iteration, the problem is not tractable if the number of model points is too high (if the deformation field is too dense).  ...  Acknowledgments We would like to thank Jhonny Park from Purdue University for providing us 3D scans of the paper sheet, and eVS ( for the cover sequence.  ... 
doi:10.1109/3dim.2007.31 dblp:conf/3dim/CastellaniGB07 fatcat:bhfv7hjw7ravvc5fxqp4yohamq


W. Sun, J. Wang, F. Jin, Z. Liang, Y. Yang
2018 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
datum feature and analysing deformation based on normal vector of point cloud was proposed.  ...  Firstly, the kd-tree is used to establish the topological relation. Datum points are detected by tracking the normal vector of point cloud determined by the normal vector of local planar.  ...  Especially in the field of deformation monitoring, the surface dense "point cloud" on the deformable body makes up for the shortcomings of traditional surveying method that could hardly reflect the overall  ... 
doi:10.5194/isprs-archives-xlii-3-1601-2018 fatcat:nz5c2wmibbfgpf7zot5yg222xa
« Previous Showing results 1 — 15 out of 5,556 results