Inclusion of a Second-Order Prior into Semi-Global Matching
作者： Simon Hermann , Reinhard Klette , Eduardo Destefanis
DOI: 10.1007/978-3-540-92957-4_55
关键词: Artificial intelligence 、 Robustness (computer science) 、 Computer vision 、 Stereopsis 、 Belief propagation 、 Computer science 、 Stereo pair 、 Advanced driver assistance systems 、 Preprocessor 、 Semi global
摘要: Today's stereo vision algorithms and computing technology allow real-time 3D data analysis, for example driver assistance systems. A recently developed Semi-Global Matching (SGM) approach by H. Hirschmuller became a popular choice due to performance robustness. This paper evaluates different parameter settings SGM, its main contribution consists in suggesting include second order prior into the smoothness term of energy function. It also proposes tests new cost function SGM. Furthermore, some preprocessing (edge images) proved be great value improving SGM results on real-world sequences, as previously already shown S. Guan R. Klette belief propagation. There is gain engineered (e.g.) currently used Middlebury website. However, fact that are not impressive .enpeda.. sequences indicates optimizing does neccessarily improve real world analysis.
core.ac.uk
springer.com
acm.org 
sci-hub.st 参考文章(7)
Shushi Guan, Reinhard Klette, Belief-Propagation on Edge Images for Stereo Analysis of Image Sequences Robot Vision. pp. 291- 302 ,(2008) , 10.1007/978-3-540-78157-8_22
Hernán Badino, A robust approach for ego-motion estimation using a mobile stereo platform IWCM'04 Proceedings of the 1st international conference on Complex motion. pp. 198- 208 ,(2004) , 10.1007/978-3-540-69866-1_15
H. Hirschmuller, Stereo Vision in Structured Environments by Consistent Semi-Global Matching computer vision and pattern recognition. ,vol. 2, pp. 2386- 2393 ,(2006) , 10.1109/CVPR.2006.294
Stan Birchfield, Carlo Tomasi, Depth Discontinuities by Pixel-to-Pixel Stereo International Journal of Computer Vision. ,vol. 35, pp. 269- 293 ,(1999) , 10.1023/A:1008160311296
Stefan K. Gehrig, Uwe Franke, Improving Stereo Sub-Pixel Accuracy for Long Range Stereo international conference on computer vision. pp. 1- 7 ,(2007) , 10.1109/ICCV.2007.4409212
H. Hirschmuller, Accurate and efficient stereo processing by semi-global matching and mutual information computer vision and pattern recognition. ,vol. 2, pp. 807- 814 ,(2005) , 10.1109/CVPR.2005.56
David St. George Vernon, Robot Vision ,(1986)