Multiview Stereo via Volumetric Graph-Cuts and Occlusion Robust Photo-Consistency

By George Vogiatzis, Carlos Hernández Esteban, Philip H. S. Torr, Roberto Cipolla. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007

329 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/14d079a4f3655034083fc749ba2f8f370f28a81a

How researchers see this paper

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

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Continuous depth estimation for multi-view stereo (2009, methodology): “The volumetric graph cuts is implemented based on [ 27 ], and we use the PMVS software [3] to obtain the patch based MVS results.”
Continuous Global Optimization in Multiview 3D Reconstruction (2007, result): “It is similar to the approach of (Vogiatzis et al., 2007 A great limitation of both previously presented energy models is that they use silhouettebased regional terms, which do not capture surface indentations, since concavities do not affect the observed silhouettes.”
Accurate, Dense, and Robust Multi-View Stereopsis (2007, background): “11), while a box function is chosen in [31] with discretized voxels, which ends up simply casting votes to voxels.”
BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis (2023, methodology): “Most recent approaches to view synthesis sidestep the difficulty of high-quality mesh reconstruction by using volumetric representations of geometry and appearance, such as voxel grids [Lombardi et al. 2019; Penner and Zhang 2017; Szeliski and Golland 1999; Vogiatzis et al. 2007] or multiplane image”
Minimizing the Multi-view Stereo Reprojection Error for Triangular Surface Meshes (2008, background): “This makes the use of additional energy terms like ballooning [18], visual hull [7, 14] or contours unnecessary and completely removes the minimal surface bias present in many other problem formulations.”
Multi-View Stereo: A Tutorial (2015, background): “MVS algorithms went through two major developments that took them to their current state: They left the laboratory setting to a small-scale outdoor scenes [174, 102, 85, 169, 190], e.”
Using Multiple Hypotheses to Improve Depth-Maps for Multi-View Stereo (2008, methodology): “In order to evaluate the improvement to multi-view stereo we combined our depth map estimation with a modified implementation of the volumetric regularisation framework of [6].”
Accurate, Dense, and Robust Multiview Stereopsis (2010, methodology): “Although the fundamental idea has not changed from [ 31 ], there are a couple of differences worth mentioning.”
Point-Based Multi-View Stereo Network (2019, background): “A number of 3D representations are adopted, including volumes [26, 9], deformation models [3, 31], and patches [5], which are iteratively updated through multi-view photometric consistency and regularization optimization.”
Ray Markov Random Fields for image-based 3D modeling: Model and efficient inference (2010, methodology): “, 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].”
Large-Scale Data for Multiple-View Stereopsis (2016, methodology): “…can be divided into two main categories: point-cloud-based methods (e.g. Campbell et al. 2008; Tola et al. 2012; Goesele et al. 2006; Hiep et al. 2009; Furukawa and Ponce 2010; Vogiatzis et al. 2007) and volume-based methods (e.g. Liu and Cooper 2011; Kolev et al. 2012; Hernández et al. 2007).”
State of the art in high density image matching (2014, methodology): “The most common multi-view algorithms rely on silhouette and fusion (Hern (cid:2) andez Esteban and Schmitt, 2004), volumetric graph-cut (Vogiatzis et al., 2007), patch-based methods (Furukawa and Ponce, 2010) and global optimisation (Vu et al., 2012).”
High Accuracy and Visibility-Consistent Dense Multiview Stereo (2012, unknown): “Many dense multiview methods rely on this information either as an initial guess for further optimization [26], [18], [29], [28], [55], [61], [64], [69], as a soft constraint [26], [35], or even as a hard constraint [51], [18] to be fulfilled by the reconstructed shape.”
Large Scale Multi-view Stereopsis Evaluation (2014, methodology): “Feature expansion [3] and depthmap fusion [2, 5, 7, 19, 20].”
A complete statistical inverse ray tracing approach to multi-view stereo (2011, methodology): “Based on how to handle the visibility, these methods can be categorized into two groups: one relies on an initial surface to compute the visibility, and is represented by the graph cuts based methods [15, 17]; the other does not assume any prior visibility knowledge, and is represented by the work o”