By mapping the spatial distribution of specific cultural and natural features from high-resolution airborne and satellite imagery, users can cost-effectively expand traditional base mapping products to incorporate extract other thematic layers such as land-use and land-cover, impervious surfaces, soils, tree species, fire fuels, and benthic habitats. Once overlaid on maps delineating topography or political divisions, the data can be viewed or queriedand analyzed in a GIS to reveal patterns and trends that support vast numbers of analyses for multiple user populations.
- Object-Oriented Classification. Fugro EarthData employs automated object-orientation which, unlike traditional methods that rely on individual pixels, examines groups of pixels to incorporate spatial characteristics like texture and context within the landscape along with the normal spectral data to segment the imagery into polygons. These polygons are then classified using specific classification schemes and data correlation techniques such as Classification and Regression Tree (CART) analysis. This advanced process results in faster, more accurate, and more consistent classification, especially with regard to high-resolution airborne and satellite imagery.
- In-field data collection. Analysts visit specific sites to collect information based on pre-determined classification scheme rules for field verification and accuracy assessment purposes. In-field data collection tools include GPS tracking, customized digital field forms, and hardcopy image maps.
- Data Correlation. Variations in the imagery are correlated with variations on the ground using the signature-recognition sites. These contain field information of ground conditions, information from the imagery, and sometimes information from ancillary data sources. Rules for classifying objects based on these correlations can be made using the object-oriented classification and/or CART analysis software. Accuracy assessment is used to determine where errors are present and where to focus efforts for error correction.
Image classification expands GIS capability by providing new and specialized thematic input that can be used in any number of GIS applications to better manage and sustain resources for public good. Because these processes are largely automated, customers already involved in orthophoto updates, for example, can acquire value-added datasets depicting land-use/land cover, impervious surfaces, or change-detection, among others, at minimal costs.
- Orthoimagery land use / land cover. Available at scales from 1:500 to 1:24,000 and pixel resolutions 5 cm and higher. orthoimagery is offered in two formats—ground orthoimagery and true orthoimagery.
- Impervious surfaces. Autocorrelated digital surface model (DSM) data serves to orthorectify each image and also is available for 3D visualization and line-of-site applications. A derived bare-earth digital elevation model (DEM) enhanced with 3D breaklines supports topographic contour generation.
- Vegetation (crops, tree species, etc). With supplemented breaklines, Fugro EarthData provides contours at 30 cm intervals and higher.
- Benthic habitats. Stereo pairs of the ADS40-SH52 imagery are used in the compilation of 3D planimetric and topographic data, including building footprints, transportation networks, hydrography, and breaklines of spot heights.
- Wetlands delineation. Using photointerpretation and/or automated classification methods—including pixel-based and object-oriented classification—land-use/land-cover classification and impervious surface data can be derived from the ADS40-SH52 imagery.
- Fire fuels.
- Change detection.