Combining Multiple Data Sources for Urban Data Acquisition

摘要: One of the most promising approaches to solve challenging task thematic mapping in urban environments is integration multiple data sources benefit from complementary types available information. Within paper combination Digital Surface Models (DSM) resulting airborne laser scanning and colour aerial imagery for landuse classification described. The DSM used obtain information on local height above terrain surface each pixel. This can be applied order separate objects higher than like buildings trees at level streets grass-covered areas. basic idea proposed algorithm combine geometric with additional multispectral provided imagery. Both are integrated a pixel-based classification, where as an channel together spectral channels. It demonstrated that by this approach scenes improved considerably compared scene labelling, only one type applied.