Multi-View Stereo for Community Photo Collections

By Michael Goesele, Noah Snavely, Brian Curless, Hugues Hoppe, S. M. Seitz. IEEE International Conference on Computer Vision, 2007

883 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b59964ff729bbde324af83743cd3cf424ce69758

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Multi-View Stereo: A Tutorial (2015, methodology): “In the first phase, visibility is estimated coarsely by clustering the initial set of images and reducing the large-scale MVS problem into a sequence of small sub-problems [93, 79, 74, 69].”
High Accuracy and Visibility-Consistent Dense Multiview Stereo (2012, neutral): “For 3D reconstruction, Goesele et al. [24] compute depth maps that form a point clouds and water-tight mesh using Poisson surface reconstruction.”
Towards Internet-scale multi-view stereo (2010, methodology): “[12] proposed the first MVS method applied to Internet photo collections, which handles variation in image sampling resolutions by selecting images with the most compatible resolution.”
State of the art in high density image matching (2014, background): “The image-matching problem is nowadays solved using stereopairs ( stereomatching ) (Hirschm € uller, 2008; Gehrig et al., 2009; Tola et al., 2010; Haala and Rothermel, 2012; Hermann and Klette, 2013) or via identi ﬁ cation of correspondences in multiple images ( multi-view stereo – MVS) (Collins, 19”
Building Rome on a Cloudless Day (2010, methodology): “There also exist techniques for accurate reconstruction of dense 3D models from community photo collections [4,5], but they are currently much slower and more computationally intensive than the SfM approaches.”
Solving the multi-way matching problem by permutation synchronization (2013, background): “Our primary motivation and running example is the classic problem of matching landmarks (feature points) across many images of the same object in computer vision, which is a key ingredient of image registration [2], recognition [3, 4], stereo [5], shape matching [6, 7], and structure from motion (SF”
Accurate, Dense, and Robust Multiview Stereopsis (2010, background): “Goesele et al. [21] have also proposed an algorithm to handle Internet photo collections containing obstacles and produce impressive results with a clever view selection scheme.”
DeepMVS: Learning Multi-view Stereopsis (2018, background): “Recent advances include robust neighbor view selection [24], incorporation of visibility consistency [10], and clustering-based techniques for efﬁcient reconstruction [13, 8].”
Evaluation of image-based modeling and laser scanning accuracy for emerging automated performance monitoring techniques (2011, methodology): “Ongoing research is working on automating this task [4] and on optimizing the SfM point cloud models for higher quality reconstructions [37].”
Consistent Depth Maps Recovery from a Video Sequence (2009, background): “STEREO reconstruction of dense depth maps from naturalvideo sequences is a fundamentally important and challenging problem in computer vision.”
Using Multiple Hypotheses to Improve Depth-Maps for Multi-View Stereo (2008, background): “region growing and filtering to build up a final dense surface [13, 15, 16].”
SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering (2023, methodology): “We then introduce a method that exploits this alignment to extract a mesh from the Gaussians using Poisson reconstruction, which is fast, scalable, and preserves details, in contrast to the Marching Cubes algorithm usually applied to extract meshes from Neural SDFs. Finally, we introduce an…”
Straightforward reconstruction of 3D surfaces and topography with a camera: Accuracy and geoscience application (2012, methodology): “Compact cameras and camera phones could also be used, but they may not achieve the precisions given here [e.g., Goesele et al. , 2007].”
Tour the world: Building a web-scale landmark recognition engine (2009, methodology): “[5] employed the geometric constraints to construct 3D visualization of landmarks, based on a set of relatively clean landmark photos.”
Global, Dense Multiscale Reconstruction for a Billion Points (2015, methodology): “We use the data set and the sparse reconstruction from Wilson and Snavely (2014) and Goesele et al. (2007) to compute the point cloud.”
Large scale mining and retrieval of visual data in a multimodal context (2008, background): “, 2007], automatic 3D reconstruction of landmarks [Vergauwen and Van Gool, 2006; Goesele et al., 2007], or for tourist guide applications on mobile devices [Paletta et al.”
Autonomous MAV flight in indoor environments using single image perspective cues (2011, background): “model of the environment takes considerable computation power [3].”
Neural volumes (2019, background): “2016] or 3D reconstruction [Goesele et al. 2007; Nießner et al. 2013], but not both.”
Neural RGB-D Surface Reconstruction (2021, background): “Reconstructing objects and scenes can be done using passive stereo systems that rely on stereo matching from two or multiple color views [29, 62], Structure-from-Motion [63], or SLAM-based [20, 21, 23] methods.”