Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3, 197-204, 2014
https://doi.org/10.5194/isprsarchives-XL-3-197-2014
© Author(s) 2014. This work is distributed under
the Creative Commons Attribution 3.0 License.
 
11 Aug 2014
Reconstruction and simplification of urban scene models based on oblique images
J. Liu1 and B. Guo2 1Wuhan University, School of Remote Sensing and Information Engineering, Wuhan, China
2Wuhan University, State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan, China
Keywords: Urban scene models, oblique images, Self-Adaptive Patch, 2-manifold surface, simplification Abstract. We describe a multi-view stereo reconstruction and simplification algorithms for urban scene models based on oblique images. The complexity, diversity, and density within the urban scene, it increases the difficulty to build the city models using the oblique images. But there are a lot of flat surfaces existing in the urban scene. One of our key contributions is that a dense matching algorithm based on Self-Adaptive Patch in view of the urban scene is proposed. The basic idea of matching propagating based on Self-Adaptive Patch is to build patches centred by seed points which are already matched. The extent and shape of the patches can adapt to the objects of urban scene automatically: when the surface is flat, the extent of the patch would become bigger; while the surface is very rough, the extent of the patch would become smaller. The other contribution is that the mesh generated by Graph Cuts is 2-manifold surface satisfied the half edge data structure. It is solved by clustering and re-marking tetrahedrons in s-t graph. The purpose of getting 2- manifold surface is to simply the mesh by edge collapse algorithm which can preserve and stand out the features of buildings.
Conference paper (PDF, 1297 KB)


Citation: Liu, J. and Guo, B.: Reconstruction and simplification of urban scene models based on oblique images, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3, 197-204, https://doi.org/10.5194/isprsarchives-XL-3-197-2014, 2014.

BibTeX EndNote Reference Manager XML