Generating Land Cover and Land Use Database to Support GFC research activity.

Principle Investigator
William Wu, Quantitative Engineering Design(QED)
Geographic area of interest
All SIIL countries
Collection of ground-truth samples remains one of the most challenging tasks for extracting information from remote sensing imagery. Visual interpretation of very high resolution image (spatial resolution less than 1m) is becoming a cost-effective alternative to field visits. Geosurvey platform is specifically designed to collect land use land cover information by interpreting the imagery available in Google Earth and Bing Maps. The current project will deploy single-pass Geosurveys (web-based surveys of satellite imagery) using CSV files of latitudes and longitudes to answer the following three questions based on a visual inspection of a 250m-by-250m square of very high resolution satellite imagery centered at each point to detect croplands, woody cover and human settlements, where croplands will include tree crops such as oil palms, rubber, and tea plantations.
Detect croplands, woody cover and human settlements for the 500000 survey points provided by GFC
Key Achievements (last update: Sep 2016)
  • The survey results have been delivered, which will be hosted in the GFC data website
  • Developed a comprehensive training material/guide for interpreting aerial photos to identify different land use classes
Land use land cover interpretation guide