Fine Scale Land Use-Land Cover Mapping
Project lead:
Analyst: Claudine Tobalske
In this contract for Missoula’s , we segmented and classified 2009 NAIP imagery using , an object-based image analysis software that classifies whole “objects” instead of individual pixels, making it possible to use textural and contextual information in addition to spectral information.
Using 1m and multi-resolution segmentation and classification within eCognition, we were able to map five land cover classes (Barren, Impervious, Grass, Tree and Shrubland, and Water) with an accuracy >92%.
Source imagery (left) and mapped land cover classes (right) for the Butte-Silver Bow urban area.