Filters








1,191 Hits in 6.1 sec

Single-Image Piece-wise Planar 3D Reconstruction via Associative Embedding [article]

Zehao Yu and Jia Zheng and Dongze Lian and Zihan Zhou and Shenghua Gao
2019 arXiv   pre-print
Single-image piece-wise planar 3D reconstruction aims to simultaneously segment plane instances and recover 3D plane parameters from an image.  ...  Then, the plane instances are obtained by grouping the embedding vectors in planar regions via an efficient mean shift clustering algorithm.  ...  Conclusion This paper proposes a novel two-stage method to singleimage piece-wise planar 3D reconstruction.  ... 
arXiv:1902.09777v3 fatcat:fpego7f2lvfmjnhr4jrigukhm4

PlaneTR: Structure-Guided Transformers for 3D Plane Recovery [article]

Bin Tan and Nan Xue and Song Bai and Tianfu Wu and Gui-Song Xia
2021 arXiv   pre-print
This paper presents a neural network built upon Transformers, namely PlaneTR, to simultaneously detect and reconstruct planes from a single image.  ...  Finally, the pixel-wise embeddings are computed to assign each pixel to one predicted plane instance which is nearest to it in embedding space.  ...  Related Work Single Image 3D Plane Recovery Geometric structures such as line segments and vanishing points are widely used in traditional methods [6, 2, 21, 15, 25, 14, 46, 9] for piece-wise 3D plane  ... 
arXiv:2107.13108v1 fatcat:abahcwyezjfctagkpq2ej3cuu4

PlaneSegNet: Fast and Robust Plane Estimation Using a Single-stage Instance Segmentation CNN [article]

Yaxu Xie, Jason Rambach, Fangwen Shu, Didier Stricker
2021 arXiv   pre-print
In this work, we propose a real-time deep neural architecture that estimates piece-wise planar regions from a single RGB image.  ...  We automatically label over 70,000 images as ground truth from the Stanford 2D-3D-Semantics dataset. Moreover, we incorporate our method with a state-of-the-art planar SLAM and validate its benefits.  ...  Piece-wise Plane Segmentation PlaneNet [19] is the first end-to-end neural network for piece-wise planar reconstruction from a single RGB image.  ... 
arXiv:2103.15428v1 fatcat:l6cczxtxajh3voxt3oknmqdzsm

Visual SLAM with Graph-Cut Optimized Multi-Plane Reconstruction [article]

Fangwen Shu, Yaxu Xie, Jason Rambach, Alain Pagani, Didier Stricker
2021 arXiv   pre-print
In the majority of existing work, geometric model estimation problems such as homography estimation and piece-wise planar reconstruction (PPR) are usually solved by standard (greedy) RANSAC separately  ...  This paper presents a semantic planar SLAM system that improves pose estimation and mapping using cues from an instance planar segmentation network.  ...  Initially PlaneNet [16] presented a end-to-end neural architecture for piece-wise planar reconstruction from a single RGB image.  ... 
arXiv:2108.04281v1 fatcat:yuqis7q27vabbempjef3pnb53u

Superpixel Soup: Monocular Dense 3D Reconstruction of a Complex Dynamic Scene [article]

Suryansh Kumar, Yuchao Dai, Hongdong Li
2019 arXiv   pre-print
This work addresses the task of dense 3D reconstruction of a complex dynamic scene from images.  ...  Hence, the task boils down to correctly assemble each rigid piece to construct a 3D shape that complies with the geometry of the scene under the ARAP assumption.  ...  of the scene between two frames is locally-rigid but globally as-rigid-as-possible. • Piece-wise planar model: Our method models a dynamic scene as a collection of piece-wise planar regions.  ... 
arXiv:1911.09092v1 fatcat:elfkvqnrmbbybcf7pcbymvdwwy

Piecewise Quadratic Reconstruction of Non-Rigid Surfaces from Monocular Sequences [chapter]

João Fayad, Lourdes Agapito, Alessio Del Bue
2010 Lecture Notes in Computer Science  
In this paper we present a new method for the 3D reconstruction of highly deforming surfaces (for instance a flag waving in the wind) viewed by a single orthographic camera.  ...  We assume that the surface is described by a set of feature points which are tracked along an image sequence.  ...  In this case we apply Isomap [21] to the reconstructed surface and recover its lower dimensional planar embedding.  ... 
doi:10.1007/978-3-642-15561-1_22 fatcat:7o3k75owbjd2ba33h6twfmx5vu

Indoor Panorama Planar 3D Reconstruction via Divide and Conquer [article]

Cheng Sun, Chi-Wei Hsiao, Ning-Hsu Wang, Min Sun, Hwann-Tzong Chen
2021 arXiv   pre-print
We create a benchmark for indoor panorama planar reconstruction by extending existing 360 depth datasets with ground truth H\&V-planes (referred to as PanoH&V dataset) and adopt state-of-the-art planar  ...  reconstruction methods to predict H\&V-planes as our baselines.  ...  Introduction Reconstructing planar surfaces from single-view images have many applications such as interior modeling, AR/VR, robot navigation, and scene understanding.  ... 
arXiv:2106.14166v2 fatcat:wctc6jcuuzdibktpqjc22ab6tm

Weakly Supported Plane Surface Reconstruction via Plane Segmentation Guided Point Cloud Enhancement

Shen Yan, Yang Peng, Guangyue Wang, Shiming Lai, Maojun Zhang
2019 IEEE Access  
INDEX TERMS 3D reconstruction, weakly supported surface, real image reconstruction.  ...  state-of-the-art 3D reconstruction systems.  ...  For a certain registered image I i containing a weakly supported plane, after piece-wise planar instance segmentation with CNNs, an inaccurate weakly supported plane mask M i pre is obtained.  ... 
doi:10.1109/access.2019.2946456 fatcat:vhuxpls6sjfftnf75oea2mbqxe

Factored Facade Acquisition using Symmetric Line Arrangements

Duygu Ceylan, Niloy J. Mitra, Hao Li, Thibaut Weise, Mark Pauly
2012 Computer graphics forum (Print)  
We introduce a novel framework for image-based 3D reconstruction of urban buildings based on symmetry priors.  ...  Operating both on 2D image data and intermediate 3D feature representations, we perform iterative feature consolidation and effective outlier pruning, thus eliminating reconstruction artifacts arising  ...  [JTC09] perform camera calibration from a single image by exploiting symmetry, and allow the user to interactively annotate architectural components using the reconstructed 3D points as anchors, producing  ... 
doi:10.1111/j.1467-8659.2012.03046.x fatcat:hn6buxh4fvdptdx2jtejgl5g6i

Monocular template-based 3D surface reconstruction: Convex inextensible and nonconvex isometric methods

F. Brunet, A. Bartoli, R.I. Hartley
2014 Computer Vision and Image Understanding  
We study the 3D reconstruction of an isometric surface from point correspondences between a template and a single input image. The template shows the surface flat and fronto-parallel.  ...  Our experimental results on simulated and real data show that our convex point-wise method and our nonconvex method outperform respectively current initialization and refinement methods in 3D reconstructed  ...  The second and third methods are applicable to two images.  ... 
doi:10.1016/j.cviu.2014.04.003 fatcat:dy2w6jro6ne7hiy7uc4djdujse

Stereo reconstruction using top-down cues

Simon Hadfield, Karel Lebeda, Richard Bowden
2017 Computer Vision and Image Understanding  
We further evaluate the technique using the recently proposed HCI stereo metrics, finding significant improvements in the quality of depth discontinuities, planar surfaces and thin structures.  ...  To demonstrate the importance and generality of scene understanding during stereo-reconstruction, the proposed approach is integrated with 3 different state-of-the-art techniques for bottom-up stereo reconstruction  ...  We are greatly indebted to Katrin Honauer for performing a significant portion of the evaluation for this work via the HCI stereo metrics [8] .  ... 
doi:10.1016/j.cviu.2016.08.001 fatcat:3ky3xv5ujzahxll6oav2dchmja

A Comprehensive Survey on Pose-Invariant Face Recognition [article]

Changxing Ding, Dacheng Tao
2016 arXiv   pre-print
(a) Piece-wise warping; (b) Patchwise warping; (c) Pixel-wise displacement.  ...  The 3D methods can be classified into three subcategories, i.e., 3D pose normalization from single image, 3D pose normalization from multiple images, and 3D modeling by image reconstruction. 3D Pose  ...  "Feature" stands for the pose-robust feature extraction category. "2D Synthesis" stands for the 2D-based face synthesis category. "3D Synthesis" stands for the 3D-based face synthesis category.  ... 
arXiv:1502.04383v3 fatcat:opz5onz775cj5hxvoaggvmp52q

A Comprehensive Survey on Pose-Invariant Face Recognition

Changxing Ding, Dacheng Tao
2016 ACM Transactions on Intelligent Systems and Technology  
(a) Piece-wise warping; (b) Patchwise warping; (c) Pixel-wise displacement.  ...  The 3D methods can be classified into three subcategories, i.e., 3D pose normalization from single image, 3D pose normalization from multiple images, and 3D modeling by image reconstruction. 3D Pose  ...  "Feature" stands for the pose-robust feature extraction category. "2D Synthesis" stands for the 2D-based face synthesis category. "3D Synthesis" stands for the 3D-based face synthesis category.  ... 
doi:10.1145/2845089 fatcat:dbzheowl5vha7h3rglrpkphsla

Learning and Meshing from Deep Implicit Surface Networks Using an Efficient Implementation of Analytic Marching [article]

Jiabao Lei, Kui Jia, Yi Ma
2021 arXiv   pre-print
Reconstruction of object or scene surfaces has tremendous applications in computer vision, computer graphics, and robotics.  ...  We prove that under mild conditions, the identified analytic faces are guaranteed to connect and form a closed, piecewise planar surface.  ...  Deep implicit surface networks are also used in [20] for surface reconstruction from as few as a single image.  ... 
arXiv:2106.10031v1 fatcat:7btqis7nkjcfzkbytg6725ipjm

A LOW-COST PANORAMIC CAMERA FOR THE 3D DOCUMENTATION OF CONTAMINATED CRIME SCENES

D. Abate, I. Toschi, C. Sturdy-Colls, F. Remondino
2017 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
All the algorithms adopted for panoramas pre-processing, photogrammetric 3D reconstruction, 3D geometry registration and analysis, are presented and currently available in open-source or low-cost software  ...  A 3D reconstruction of the environment is first generated by processing panoramas acquired with the low-cost Ricoh Theta 360 camera, and further analysed to highlight potentials and limits of this emerging  ...  The single images are stitched together using embedded software and there is no possibility to access the raw prestitching shots in the current version of camera.  ... 
doi:10.5194/isprs-archives-xlii-2-w8-1-2017 fatcat:2hkfwunskvalrc3txzqhvuj4sm
« Previous Showing results 1 — 15 out of 1,191 results