Edge detection in range images based on scan line approximation
Computer Vision and Image Understanding
Multidimensional binary search trees used for associative searching
Communications of the ACM
The Ball-Pivoting Algorithm for Surface Reconstruction
IEEE Transactions on Visualization and Computer Graphics
Geometry and texture recovery of scenes of large scale
Computer Vision and Image Understanding
Context-based surface completion
ACM SIGGRAPH 2004 Papers
Robust moving least-squares fitting with sharp features
ACM SIGGRAPH 2005 Papers
Segmentation of architecture shape information from 3D point cloud
Proceedings of the 8th International Conference on Virtual Reality Continuum and its Applications in Industry
Edge detecting for range data using laplacian operators
IEEE Transactions on Image Processing
MM '11 Proceedings of the 19th ACM international conference on Multimedia
Creating Large-Scale City Models from 3D-Point Clouds: A Robust Approach with Hybrid Representation
International Journal of Computer Vision
O-snap: Optimization-based snapping for modeling architecture
ACM Transactions on Graphics (TOG)
Technical Section: Reconstructing shape boundaries with multimodal constraints
Computers and Graphics
| Hi-index | 0.04 |
We present a pipeline to reconstruct complete geometry of architectural buildings from point clouds obtained by sparse range laser scanning. Due to limited accessibility of outdoor environments, complete and sufficient scanning of every face of an architectural building is often impossible. Our pipeline deals with architectures that are made of planar faces and faithfully constructs a polyhedron of low complexity based on the incomplete scans. The pipeline first recognizes planar regions based on point clouds, then proceeds to compute plane intersections and corners (in this paper, we use the informal terms corner or vertex corner to stand for a polyhedron vertex. See the Overview section for notation declarations), and finally produces a complete polyhedron. Within the pipeline, several algorithms based on the polyhedron geometry assumption are designed to perform data clustering, boundary detection, and face extraction. Our system offers a convenient user interface but minimizes the necessity of user intervention. We demonstrate the capability and advantage of our system by modeling real-life buildings.