LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA
When viewed from distance, large parts of the topography of landmasses and the bathymetry of the sea and ocean floor can be regarded as a smooth background with local features. Consequently a digital elevation model combining a compact smooth representation of the background with locally added featu...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2015-08-01
|
| Series: | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-3-W3/565/2015/isprsarchives-XL-3-W3-565-2015.pdf |
| _version_ | 1818151323856535552 |
|---|---|
| author | T. Dokken V. Skytt O. Barrowclough |
| author_facet | T. Dokken V. Skytt O. Barrowclough |
| author_sort | T. Dokken |
| collection | DOAJ |
| description | When viewed from distance, large parts of the topography of landmasses and the bathymetry of the sea and ocean floor can be regarded as a smooth background with local features. Consequently a digital elevation model combining a compact smooth representation of the background with locally added features has the potential of providing a compact and accurate representation for topography and bathymetry. The recent introduction of Locally Refined B-Splines (LR B-splines) allows the granularity of spline representations to be locally adapted to the complexity of the smooth shape approximated. This allows few degrees of freedom to be used in areas with little variation, while adding extra degrees of freedom in areas in need of more modelling flexibility. In the EU fp7 Integrating Project IQmulus we exploit LR B-splines for approximating large point clouds representing bathymetry of the smooth sea and ocean floor. A drastic reduction is demonstrated in the bulk of the data representation compared to the size of input point clouds. The representation is very well suited for exploiting the power of GPUs for visualization as the spline format is transferred to the GPU and the triangulation needed for the visualization is generated on the GPU according to the viewing parameters. The LR B-splines are interoperable with other elevation model representations such as LIDAR data, raster representations and triangulated irregular networks as these can be used as input to the LR B-spline approximation algorithms. Output to these formats can be generated from the LR B-spline applications according to the resolution criteria required. The spline models are well suited for change detection as new sensor data can efficiently be compared to the compact LR B-spline representation. |
| first_indexed | 2024-12-11T13:37:00Z |
| format | Article |
| id | doaj.art-2166dad8ac2a49b5a2173fc4da981ca5 |
| institution | Directory Open Access Journal |
| issn | 1682-1750 2194-9034 |
| language | English |
| last_indexed | 2024-12-11T13:37:00Z |
| publishDate | 2015-08-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| spelling | doaj.art-2166dad8ac2a49b5a2173fc4da981ca52022-12-22T01:04:59ZengCopernicus PublicationsThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences1682-17502194-90342015-08-01XL-3/W356557010.5194/isprsarchives-XL-3-W3-565-2015LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATAT. Dokken0V. Skytt1O. Barrowclough2SINTEF ICT, Department of Applied Mathematics, PO. Box 124 Blindern, 0314 Oslo, NorwaySINTEF ICT, Department of Applied Mathematics, PO. Box 124 Blindern, 0314 Oslo, NorwaySINTEF ICT, Department of Applied Mathematics, PO. Box 124 Blindern, 0314 Oslo, NorwayWhen viewed from distance, large parts of the topography of landmasses and the bathymetry of the sea and ocean floor can be regarded as a smooth background with local features. Consequently a digital elevation model combining a compact smooth representation of the background with locally added features has the potential of providing a compact and accurate representation for topography and bathymetry. The recent introduction of Locally Refined B-Splines (LR B-splines) allows the granularity of spline representations to be locally adapted to the complexity of the smooth shape approximated. This allows few degrees of freedom to be used in areas with little variation, while adding extra degrees of freedom in areas in need of more modelling flexibility. In the EU fp7 Integrating Project IQmulus we exploit LR B-splines for approximating large point clouds representing bathymetry of the smooth sea and ocean floor. A drastic reduction is demonstrated in the bulk of the data representation compared to the size of input point clouds. The representation is very well suited for exploiting the power of GPUs for visualization as the spline format is transferred to the GPU and the triangulation needed for the visualization is generated on the GPU according to the viewing parameters. The LR B-splines are interoperable with other elevation model representations such as LIDAR data, raster representations and triangulated irregular networks as these can be used as input to the LR B-spline approximation algorithms. Output to these formats can be generated from the LR B-spline applications according to the resolution criteria required. The spline models are well suited for change detection as new sensor data can efficiently be compared to the compact LR B-spline representation.http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-3-W3/565/2015/isprsarchives-XL-3-W3-565-2015.pdf |
| spellingShingle | T. Dokken V. Skytt O. Barrowclough LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| title | LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA |
| title_full | LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA |
| title_fullStr | LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA |
| title_full_unstemmed | LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA |
| title_short | LOCALLY REFINED SPLINES REPRESENTATION FOR GEOSPATIAL BIG DATA |
| title_sort | locally refined splines representation for geospatial big data |
| url | http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-3-W3/565/2015/isprsarchives-XL-3-W3-565-2015.pdf |
| work_keys_str_mv | AT tdokken locallyrefinedsplinesrepresentationforgeospatialbigdata AT vskytt locallyrefinedsplinesrepresentationforgeospatialbigdata AT obarrowclough locallyrefinedsplinesrepresentationforgeospatialbigdata |