Monitoring urban tree cover using object-based image analysis and public domain remotely sensed data

The heterogeneity of tree canopies makes them difficult to classify and monitor using traditional per-pixel classification methods with Landsat imagery, which lacks the resolution to map small urban features. In this study, object-based image assessment (OBIA) was used in combination with publicly available, free aerial imagery to identify nine land use/land cover classes in a Seattle, WA, neighborhood: buildings, grass, developed, impervious, shrub, tree, ground, water, and other. Some of the algorithms are described. The OBIA method was found to result in good and repeatable classifications that were appropriate for tree cover assessments in cities and detailed enough for parcel-based analyses.