A probabilistic framework for space carving

By Adrian Broadhurst, Tom Drummond, Roberto Cipolla. 2001

277 citations recorded by the source corpus.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Multi-view Supervision for Single-View Reconstruction via Differentiable Ray Consistency (2017, background): “Techniques based on this concept [2, 24] can obtain reconstructions of objects by space carving using multiple available views.”
3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction (2016, background): “In order to circumvent issues related to large baselines or non-Lambertian surfaces, 3D volumetric reconstruction methods such as space carving [12,13,14,15] and their probabilistic extensions [16] have become popular.”
Ray Markov Random Fields for image-based 3D modeling: Model and efficient inference (2010, background): “, occlusion) based on initial surface [19] or currently estimated surface [6, 12]; The other approach estimates the visibility based on the probabilistic opacity of the volume (usually for voxel based approach) [4, 15].”
Stereo for Image-Based Rendering using Image Over-Segmentation (2007, methodology): “Representative approaches include the original voxel coloring work (Seitz and Dyer, 1997), recovery of voxels with opacity (De Bonet and Viola, 1999; Reche et al., 2004), and treatment of voxels in a probabilistic framework (Broadhurst et al., 2001).”
MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction (2022, methodology): “Classic multi-view stereo (MVS) methods [2, 6–8, 35, 35, 62, 64, 65] consider either feature matching for depth estimation [6, 62] or represent shapes with voxels [2, 7, 8, 35, 51, 64, 72, 73].”
High Accuracy and Visibility-Consistent Dense Multiview Stereo (2012, unknown): “This includes space carving [49], [37], [8], [68], level sets [14], [32],[46], and volumetric graph cuts [65], [6], [28], [43], [61] (though [50], [27] propose regular volumetric grid adaptive to photo-consistency measures to push the resolution limit further).”
Robust Variational Segmentation of 3D Objects from Multiple Views (2006, background): “These limitations are partially alleviated by the probabilistic space carving method [1].”
Neural volumes (2019, methodology): “Volumetric reconstruction methods side-step this problem of explicit correspondence matching, beginning with Voxel Coloring [Prock and Dyer 1998; Seitz and Dyer 1997, 1999] and Space Carving [Bonet and Viola 1999; Broadhurst et al. 2001; Kutulakos and Seitz 2000].”
NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction (2021, methodology): “Traditional multi-view 3D reconstruction methods can be roughly classified into two categories: point- and surface-based reconstruction [2, 8, 9, 35] and volumetric reconstruction [6, 3, 36].”
Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for Multi-view Reconstruction (2022, background): “Explicit representation includes voxels [5, 29] and triangular mesh [3, 4, 14], which are limited by the resolution.”
3D Reconstruction by Shadow Carving: Theory and Practical Evaluation (2007, unknown): “An alternative approach is photo-consistency, as introduced by the work of Seitz and Dyer [33], Kutulakos and Seitz [14] and numerous extensions and/or alternative formulations [34, 35, 36, 37, 38, 39, 40].”
NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view Reconstruction (2022, methodology): “Recently, 3D reconstruction with neural implicit representations has become a promising alternative to traditional methods, because of its high spatial resolution and highly detailed reconstructions outperforming classical stereo algorithms [4, 12, 13, 50, 10, 5, 51].”
Volume Rendering of Neural Implicit Surfaces (2021, background): “Classical multi-view stereo approaches are generally either depth-based [1, 31, 8, 7] or voxel-based [5, 3, 32].”
Efficient Multi-View Reconstruction of Large-Scale Scenes using Interest Points, Delaunay Triangulation and Graph Cuts (2007, methodology): “Volumetric multi-view stereovision methods include space carving [9, 25, 34, 41, 43], level sets [13, 21, 32], and volumetric graph cuts [8, 20, 27, 36, 40, 42].”
Methods for Volumetric Reconstruction of Visual Scenes (2004, result): “We note that some of the content of this paper has appeared in previous workshop and conference papers [9, 42, 47].”
UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction (2021, background): “Before the era of deep learning, classic multi-view stereo (MVS) methods [2, 4, 5, 7, 21, 21, 48, 50, 51] focus on either matching features across views [4, 48] or representing shapes with a voxel grid [2, 5, 7, 21, 28, 42, 50, 54, 55].”
A complete statistical inverse ray tracing approach to multi-view stereo (2011, background): “…color I r weighted by the visibility, i.e., The value of each voxel color can be computed by accumulating the messages received from all the rays that passes through the voxel: where M cr → ℓ xr [1] and M cr → ℓ xr [2] denote respectively the ﬁrst and second value of the 2-d vector M cr → ℓ xr .”