International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Download
Citation
Volume XLII-3/W2
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-3/W2, 209–212, 2017
https://doi.org/10.5194/isprs-archives-XLII-3-W2-209-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-3/W2, 209–212, 2017
https://doi.org/10.5194/isprs-archives-XLII-3-W2-209-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.

  16 Nov 2017

16 Nov 2017

PROVING THE CAPABILITY FOR LARGE SCALE REGIONAL LAND-COVER DATA PRODUCTION BY SELF-FUNDED COMMERCIAL OPERATORS

M. W. Thompson, J. Hiestermann, and L. Moyo M. W. Thompson et al.
  • GeoTerraImage Pty Ltd, Pretoria, South Africa

Keywords: Land-cover, big data, automated mapping, cloud-based processing, disruptive technologies, southern Africa

Abstract. For service providers developing commercial value-added data content based on remote sensing technologies, the focus is to typically create commercially appropriate geospatial information which has downstream business value. The primary aim being to link locational intelligence with business intelligence in order to better make informed decisions. From a geospatial perspective this locational information must be relevant, informative, and most importantly current; with the ability to maintain the information timeously into the future for change detection purposes. Aligned with this, GeoTerraImage has successfully embarked on the production of land-cover/land-use content over southern Africa. The ability for a private company to successfully implement and complete such an exercise has been the capability to leverage the combined advantages of cutting edge data processing technologies and methodologies, with emphasis on processing repeatability and speed, and the use of a wide range of readily available imagery. These production workflows utilise a wide range of integrated procedures including machine learning algorithms, innovative use of non-specialists for sourcing of reference data, and conventional pixel and object-based image classification routines, and experienced/expert landscape interpretation. This multi-faceted approach to data produce development demonstrates the capability for SMME level commercial entities such as GeoTerraImage to generate industry applicable large data content, in this case, wide area coverage land-cover and land-use data across the sub-continent. Within this development, the emphasis has been placed on the key land-use information, such as mining, human settlements, and agriculture, given the importance of this geo-spatial land-use information in business and socio-economic applications and decision making.