3,747 Hits in 6.1 sec

Shape-aware surface reconstruction from sparse 3D point-clouds

Florian Bernard, Luis Salamanca, Johan Thunberg, Alexander Tack, Dennis Jentsch, Hans Lamecker, Stefan Zachow, Frank Hertel, Jorge Goncalves, Peter Gemmar
2017 Medical Image Analysis  
The reconstruction of an object's shape or surface from a set of 3D points plays an important role in medical image analysis, e.g. in anatomy reconstruction from tomographic measurements or in the process  ...  However, medical applications often provide contextual information about the 3D point data that allow to incorporate prior knowledge about the shape that is to be reconstructed.  ...  Conclusion and Outlook In this paper we have presented a methodology for a shape-aware surface reconstruction from sparse surface points.  ... 
doi:10.1016/ pmid:28282642 fatcat:3ouyknbrgbcjfawa3een5ecv74

RangeUDF: Semantic Surface Reconstruction from 3D Point Clouds [article]

Bing Wang, Zhengdi Yu, Bo Yang, Jie Qin, Toby Breckon, Ling Shao, Niki Trigoni, Andrew Markham
2022 arXiv   pre-print
We present RangeUDF, a new implicit representation based framework to recover the geometry and semantics of continuous 3D scene surfaces from point clouds.  ...  Extensive experiments show that RangeUDF clearly surpasses state-of-the-art approaches for surface reconstruction on four point cloud datasets.  ...  As shown in these two tables, we additionally explore the impact of color and surface point density on the performance of surface reconstruction and semantic segmentation.  ... 
arXiv:2204.09138v1 fatcat:ncjyo32qbzd33a6yw6zx4cxsje

DSP-SLAM: Object Oriented SLAM with Deep Shape Priors [article]

Jingwen Wang, Martin Rünz, Lourdes Agapito
2021 arXiv   pre-print
DSP-SLAM takes as input the 3D point cloud reconstructed by a feature-based SLAM system and equips it with the ability to enhance its sparse map with dense reconstructions of detected objects.  ...  We propose DSP-SLAM, an object-oriented SLAM system that builds a rich and accurate joint map of dense 3D models for foreground objects, and sparse landmark points to represent the background.  ...  The sparse SLAM backbone provides per-frame camera poses and a 3D point cloud.  ... 
arXiv:2108.09481v2 fatcat:2cxcoerz6vfjnkju5oskhxxuie

FroDO: From Detections to 3D Objects

Martin Runz, Kejie Li, Meng Tang, Lingni Ma, Chen Kong, Tanner Schmidt, Ian Reid, Lourdes Agapito, Julian Straub, Steven Lovegrove, Richard Newcombe
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Results on a real-world sequence from ScanNet [7].  ...  Figure 1: Given a localized input RGB sequence, FroDO dectects objects and infers their pose and a progressively fine grained and expressive object shape representation.  ...  First, we optimize the energy over a sparse set of surface points, using the point decoder G s (z) as a shape prior.  ... 
doi:10.1109/cvpr42600.2020.01473 dblp:conf/cvpr/RunzLTMKS0ASLN20 fatcat:rax2vea64nbdhlreuauajyqebi

Multiview Rectification of Folded Documents [article]

Shaodi You, Yasuyuki Matsushita, Sudipta Sinha, Yusuke Bou and Katsushi Ikeuchi
2016 arXiv   pre-print
Our main contribution is a new robust rectification method based on ridge-aware 3D reconstruction of a paper sheet and unwrapping the reconstructed surface using properties of developable surfaces via  ...  Prior methods either need expensive 3D scanners or model deformable surfaces using over-simplified parametric representations.  ...  Our technique recovers a ridge-aware 3D reconstruction of the document surface from a sparse 3D point cloud.  ... 
arXiv:1606.00166v1 fatcat:ulwkig5ug5ed7iknvtjevx75ee

DIFFER: Moving Beyond 3D Reconstruction with Differentiable Feature Rendering

Navaneet K. L., Priyanka Mandikal, Varun Jampani, R. Venkatesh Babu
2019 Computer Vision and Pattern Recognition  
In this work, we propose a deep learning system that can simultaneously reason about 3D shape as well as associated properties (such as color, semantic part segments) directly from a single 2D image.  ...  Perception of 3D object properties from 2D images form one of the core computer vision problems.  ...  The 3D perception of machines need to go beyond just the shape reconstruction from 2D images.  ... 
dblp:conf/cvpr/LMJB19 fatcat:djms5h3eavfcpk4qamxfr2le34

SSPU-Net: Self-Supervised Point Cloud Upsampling via Differentiable Rendering [article]

Yifan Zhao, Le Hui, Jin Xie
2021 arXiv   pre-print
Point clouds obtained from 3D sensors are usually sparse. Existing methods mainly focus on upsampling sparse point clouds in a supervised manner by using dense ground truth point clouds.  ...  To achieve this, we exploit the consistency between the input sparse point cloud and generated dense point cloud for the shapes and rendered images.  ...  Figure 5 : 5 Comparison results of point cloud upsampling (×4) and surface reconstruction results with different methods (b-d) from the input point clouds (a).  ... 
arXiv:2108.00454v2 fatcat:xswpljwwcbdzje5szfl5o5lgdm

Reconstruction of People in Loose Clothing [article]

Sai Sagar Jinka, Rohan Chacko, Astitva Srivastava, Avinash Sharma, P.J. Narayanan
2021 arXiv   pre-print
3D human body reconstruction from monocular images is an interesting and ill-posed problem in computer vision with wider applications in multiple domains.  ...  In this paper, we propose SHARP, a novel end-to-end trainable network that accurately recovers the detailed geometry and appearance of 3D people in loose clothing from a monocular image.  ...  Effect of Upsampling on the Predicted Point-cloud The predicted point-cloud P, generated from the fused depth maps, can sometimes be sparse and contain holes.  ... 
arXiv:2106.04778v3 fatcat:5as3byi76ve5tez6qzio3mwb7y

ANISE: Assembly-based Neural Implicit Surface rEconstruction [article]

Dmitry Petrov, Matheus Gadelha, Radomir Mech, Evangelos Kalogerakis
2022 arXiv   pre-print
We present ANISE, a method that reconstructs a 3D shape from partial observations (images or sparse point clouds) using a part-aware neural implicit shape representation.  ...  The part implicit functions can then be combined into a single, coherent shape, enabling part-aware shape reconstructions from images and point clouds.  ...  We present ANISE, a method that reconstructs a 3D shape from partial observations (images or sparse point clouds) using a part-aware neural implicit shape representation.  ... 
arXiv:2205.13682v1 fatcat:73xrvfyeeffd3ijkpo3vckttii

Point Scene Understanding via Disentangled Instance Mesh Reconstruction [article]

Jiaxiang Tang, Xiaokang Chen, Jingbo Wang, Gang Zeng
2022 arXiv   pre-print
Semantic scene reconstruction from point cloud is an essential and challenging task for 3D scene understanding.  ...  Based on the accurate proposals, we leverage a mesh-aware latent code space to disentangle the processes of shape completion and mesh generation, relieving the ambiguity caused by the incomplete point  ...  -We design a disentangled instance mesh reconstruction strategy to mitigate the ambiguity of learning complete shapes from incomplete point cloud observations, by leveraging a mesh-aware latent code space  ... 
arXiv:2203.16832v2 fatcat:umhqlzcierh6pclnaf77avvqm4

FroDO: From Detections to 3D Objects [article]

Kejie Li, Martin Rünz, Meng Tang, Lingni Ma, Chen Kong, Tanner Schmidt, Ian Reid, Lourdes Agapito, Julian Straub, Steven Lovegrove, Richard Newcombe
2020 arXiv   pre-print
Key to FroDO is to embed object shapes in a novel learnt space that allows seamless switching between sparse point cloud and dense DeepSDF decoding.  ...  We introduce FroDO, a method for accurate 3D reconstruction of object instances from RGB video that infers object location, pose and shape in a coarse-to-fine manner.  ...  First, we optimize the energy over a sparse set of surface points, using the point decoder G s (z) as a shape prior.  ... 
arXiv:2005.05125v1 fatcat:ein67ktbl5bcla5l3kizbstjee

Deep Point Cloud Simplification for High-quality Surface Reconstruction [article]

Yuanqi Li, Jianwei Guo, Xinran Yang, Shun Liu, Jie Guo, Xiaopeng Zhang, Yanwen Guo
2022 arXiv   pre-print
In this paper, we propose a novel point cloud simplification network (PCS-Net) dedicated to high-quality surface mesh reconstruction while maintaining geometric fidelity.  ...  Therefore, simplifying point clouds for achieving compact, clean, and uniform points is becoming increasingly important for 3D vision and graphics tasks.  ...  We propose a new point cloud simplification network (PCS-Net) to produce sparse and uniform point clouds to take advantage of compact storage and reconstruct surfaces faithfully.  ... 
arXiv:2203.09088v1 fatcat:nmtgbd73kvfzjh4qz7ary7ixfq

Learning to Infer Implicit Surfaces without 3D Supervision [article]

Shichen Liu, Shunsuke Saito, Weikai Chen, Hao Li
2019 arXiv   pre-print
While explicit representations, such as point clouds and voxels, can span a wide range of shape variations, their resolutions are often limited.  ...  Recent advances in 3D deep learning have shown that it is possible to train highly effective deep models for 3D shape generation, directly from 2D images.  ...  For point cloud representation, we also visualize the meshes reconstructed from the output point cloud. Implementation details.  ... 
arXiv:1911.00767v1 fatcat:vq33mtgfvvc7llzkigbvpljo4i

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis [article]

Ben Fei, Weidong Yang, Wenming Chen, Zhijun Li, Yikang Li, Tao Ma, Xing Hu, Lipeng Ma
2022 arXiv   pre-print
Point cloud completion is a generation and estimation issue derived from the partial point clouds, which plays a vital role in the applications in 3D computer vision.  ...  The progress of deep learning (DL) has impressively improved the capability and robustness of point cloud completion.  ...  This method could be reiterated many times to reconstruct a 3D shape from a combination of numerous surface elements.  ... 
arXiv:2203.03311v2 fatcat:e2kvryolufearetp4ujlw2gwwy

SLAM-based dense surface reconstruction in monocular Minimally Invasive Surgery and its application to Augmented Reality

Long Chen, Wen Tang, Nigel W. John, Tao Ruan Wan, Jian Jun Zhang
2018 Computer Methods and Programs in Biomedicine  
A robust global 3D surface reconstruction framework has been developed for building a dense surface using only unorganized sparse point clouds extracted from the SLAM.  ...  Both camera tracking and surface reconstruction based on a sparse point cloud are effective and operated in real-time.  ...  method to generate the smooth surface from an unstructured sparse point cloud.  ... 
doi:10.1016/j.cmpb.2018.02.006 pmid:29544779 fatcat:v4bhbs7wzzej7pifadeffhwzqm
« Previous Showing results 1 — 15 out of 3,747 results