Accurate and Efficient Stereo Processing by Semi-Global Matching and Mutual Information

By Heiko Hirschmüller. Computer Vision and Pattern Recognition, 2005

979 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/c38fe172d1c96650ee918b4d731e341859504bb7

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

Exploiting the Power of Stereo Confidences (2013, methodology): “For all considered metrics, the cost term from the local method is replaced with the accumlated costs of SGM.”
Stereo Matching with Color-Weighted Correlation, Hierachical Belief Propagation and Occlusion Handling (2006, methodology): “Many methods [9, 3, 15, 16, 8, 2] have been proposed to solve this ambiguity problem.”
Review of Stereo Vision Algorithms: From Software to Hardware (2008, methodology): “In (Hirschmuller 2005) an algorithm based on a hierarchical calculation of a mutual information based matching cost is proposed.”
Pyramid Stereo Matching Network (2018, positive): “Their method also exploits typical stereo matching procedures, including cost aggregation, SGM, and other disparity map reﬁne-ments to improve matching results.”
BA-Net: Dense Bundle Adjustment Network (2018, methodology): “Here, Oracle uses ground truth camera poses to solve the multi-view stereo by SGM (Hirschmuller, 2005), while SIFT , FF , and Matlab further use sparse features, optical ﬂow, and KLT tracking respectively for feature correspondence to solve camera poses by the 8-pt algorithm (Hartley, 1997).”
Making Bertha See (2013, methodology): “Just like SGM, the Stixel computation is performed on an FPGA platform.”
Assessing the Accuracy of Georeferenced Point Clouds Produced via Multi-View Stereopsis from Unmanned Aerial Vehicle (UAV) Imagery (2012, methodology): “A number of alternative MVS approaches have been developed such as Semi-Global Matching (SGM) [34,35], Plane-sweep strategies [36], and the MVS pipeline developed by Vu et al.”
Semi-global Stereo Matching with Surface Orientation Priors (2017, methodology): “Semi-Global Matching (SGM) is a widely-used stereo matching technique introduced by Hirschmüller [24, 26].”
Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching (2019, methodology): “More specifically, works in [16, 35] use belief propagation and semi-global matching [10] to approximate the global optimization with dynamic programming.”
The Fast Bilateral Solver (2015, neutral): “89 SGM[16] 24.”
Multi-task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics (2017, methodology): “The disparity (inverse depth) metrics were computed against the CityScapes depth maps, which are sparse and computed using SGM stereo [21].”
Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlation (2022, methodology): “Zbontar and LeCun [54] proposed to train a CNN to initialize the matching cost between patches, which is refined by cross-based aggregation and semiglobal optimization as in SGM [14].”
Unmanned aerial systems for photogrammetry and remote sensing: A review (2014, methodology): “Particularly, the introduction of Semi-Global Matching (SGM) (Hirschmüller, 2005) has been key in using optical cameras as a stand-alone solution for dense DSM production, and this is also reflected in UAS literature (Bulatov et al., 2011; Küng et al., 2011; Haala et al., 2013).”
Self-Supervised Monocular Depth Hints (2019, methodology): “[30] also exploits proxy ground truth labels generated with a traditional stereo matching method [13].”