Geometrical DCC-Algorithm for Merging Polygonal Geospatial Data - Silvija Stankute and Hartmut Asche
0 likes525 views
The document outlines a geometrical DCC-algorithm for merging polygonal geospatial data, which addresses the challenges of manually acquiring missing information and creating a combined dataset that meets specific task requirements. The algorithm involves pre-processing, object assignment, and ensuring uniform data formats to facilitate the integration of different datasets. It concludes with the potential for improving geospatial data quality and accuracy while emphasizing further developments in data fusion techniques.
Geometrical DCC-Algorithm for Merging Polygonal Geospatial Data - Silvija Stankute and Hartmut Asche
- 1. lv-DCC 1/?? Geometrical DCC-Algorithm for Merging Polygonal Geospatial Data ICCSA 2010 Fukuoka / Japan 23-26 March, 2010 Silvija Stankute and Hartmut Asche | University of Potsdam | Geoinformation Research | Germany
- 2. data-fusion 2/12 Problem information needed information available input dataset 1 input dataset 2 input dataset 3 © stankute·asche·ifg·uni·potsdam 2009
- 3. data-fusion 3/12 Problem and Objective manual-acquisition of missing information - time-consuming and costly a combination of two or more different datasets allows for an output dataset which fulfills the demands of the particular task an incorporation of suitable features required for specific task © stankute·asche·ifg·uni·potsdam 2009
- 4. data-fusion 4/12 Workflow dataset 1 pre-processing object assignment dataset 2 pre-processing data sources new dataset the DCC-algorithm is based on Direct Coordinate Comparison between two datasets © stankute·asche·ifg·uni·potsdam 2009
- 5. data-fusion 5/12 Workflow dataset 1 pre-processing object assignment dataset 2 pre-processing data sources 1. uniform data format 2. transfer to the same new dataset coordinate system 3. verification and 4. geometrical correction © stankute·asche·ifg·uni·potsdam 2009
- 6. data-fusion 6/12 Workflow dataset 1 pre-processing object assignment dataset 2 pre-processing data sources 1. uniform data format 2. transfer to the same new dataset coordinate system 3. verification and 1. new geometrical 4. geometrical correction information only 2. new semantical information only 3. new geometrical and semantical information © stankute·asche·ifg·uni·potsdam 2009
- 7. data-fusion 7/12 Relation between two corresponding objects source dataset target dataset © stankute·asche·ifg·uni·potsdam 2009
- 8. data-fusion 8/12 Relation between two corresponding objects source dataset target dataset © stankute·asche·ifg·uni·potsdam 2009
- 9. data-fusion 9/12 Relation between geometrical objects mean centre (MC), minimum bounding rectangle centre (MBR) and centroid (C) the choice of centre depends on the type of polygon © stankute·asche·ifg·uni·potsdam 2009
- 10. data-fusion 10/12 Relation between geometrical objects | MBR © stankute·asche·ifg·uni·potsdam 2009
- 11. data-fusion 11/12 Object Assignment source dataset target dataset SDS TDS © stankute·asche·ifg·uni·potsdam 2009
- 12. data-fusion 12/12 Object Assignment enhanced dataset source dataset target dataset SDS TDS © stankute·asche·ifg·uni·potsdam 2009
- 13. data-fusion 13/12 Object Assignment dp <= dmax , where dmax user-defined Z1 – source polygon centre Z1´- target polygon centre © stankute·asche·ifg·uni·potsdam 2009
- 14. data-fusion 14/12 Object Assignment dp source dataset SDS target dataset TDS © stankute·asche·ifg·uni·potsdam 2009
- 15. data-fusion 15/12 Object Assignment dp <= dmax , where dmax user-defined Z1 – source polygon centre Z1´- target polygon centre © stankute·asche·ifg·uni·potsdam 2009
- 16. data-fusion 16/12 Object Assignment dp <= dmax , where dmax user-defined Z1 – source polygon centre Z1´- target polygon centre © stankute·asche·ifg·uni·potsdam 2009
- 17. data-fusion 17/12 Object Assignment dp <= dmax , where dmax user-defined Z1 – source polygon centre Z1´- target polygon centre to compare: perimeter area polygon extent © stankute·asche·ifg·uni·potsdam 2009
- 18. data-fusion 18/12 Transfer of Semantical and Geometrical Information 6 5 3 4 2 1 source dataset ID use 1 library 2 university 3 apartment 4 apartment 5 apartment 6 apartment © stankute·asche·ifg·uni·potsdam 2009
- 19. data-fusion 19/12 Transfer of Semantical and Geometrical Information 6 6 7 5 5 3 4 3 4 2 2 1 1 source dataset target dataset ID use ID level 1 library 1 4 2 university 2 4 3 apartment 3 5 4 apartment 4 2 5 apartment 5 6 6 apartment 6 6 7 6 © stankute·asche·ifg·uni·potsdam 2009
- 20. data-fusion 20/12 Transfer of Semantical and Geometrical Information 6 6 7 6 7 5 5 5 3 4 3 4 3 4 2 2 2 1 1 1 source dataset target dataset output dataset ID use ID level ID use level 1 library 1 4 1 library 4 2 university 2 4 2 university 4 3 apartment 3 5 3 apartment 5 4 apartment 5 6 4 apartment 99999 5 apartment 7 6 5 apartment 6 6 apartment 6 apartment 99999 7 99999 6 © stankute·asche·ifg·uni·potsdam 2009
- 21. data-fusion 21/12 Transfer of Semantical and Geometrical Information 6 6 7 6 7 5 5 5 3 4 3 4 3 4 2 2 2 1 1 1 source dataset target dataset output dataset ID use ID level ID use level 1 library 1 4 1 library 4 2 university 2 4 2 university 4 3 apartment 3 5 3 apartment 5 4 apartment 5 6 4 apartment 99999 5 apartment 7 6 5 apartment 6 6 apartment 6 apartment 99999 7 99999 6 © stankute·asche·ifg·uni·potsdam 2009
- 22. data-fusion 22/12 Transfer of Semantical and Geometrical Information 6 6 7 6 7 5 5 5 3 4 3 4 3 4 2 2 2 1 1 1 source dataset target dataset output dataset ID use ID level ID use level 1 library 1 4 1 library 4 2 university 2 4 2 university 4 3 apartment 3 5 3 apartment 5 4 apartment 5 6 4 apartment 99999 5 apartment 7 6 5 apartment 6 6 apartment 6 apartment 99999 7 99999 6 © stankute·asche·ifg·uni·potsdam 2009
- 23. data-fusion 23/12 Results source dataset SDS target dataset TDS output dataset 58 shapes 306 shapes 362 shapes 3 attributes 12 attributes 15 attributes About 96% of geometrical information is transferred! © stankute·asche·ifg·uni·potsdam 2009
- 24. data-fusion 24/12 Conclusion and Future Work datafusion - updating and adding new geospatial features increasing the quality and accuracy of geospatial information datafusion of more complex polygon types (i.e. landuse) comprehensive algorithm that combines results of linear datafusion and polygonal datafusion © stankute·asche·ifg·uni·potsdam 2009
- 25. data-fusion 25/12 Thank you for your attention! Autor: Silvija Stankutė IfG 2010 Kontakt: silvija.stankute@uni-potsdam.de data-fusion © stankute·asche·ifg·uni·potsdam 2009