95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 59-6 - A semi-automatic approach for characterizing and analyzing spatial heterogeneity in urban landscapes: Integrating visual interpretation and object-based classification

Wednesday, August 4, 2010: 9:50 AM
321, David L Lawrence Convention Center
Weiqi Zhou, Plant Sciences, University of California, Davis, Davis, CA and Mary L. Cadenasso, Department of Plant Sciences, University of California, Davis, Davis, CA
Background/Question/Methods

Urban areas are strikingly heterogeneous. To develop an ecological understanding of urban systems, it is critical to quantify the fine-scale heterogeneity of their built and natural components. Recent advances in object-based image analysis allow a cost-effective way to obtain highly accurate identification of land cover features that make up the urban landscape from high spatial resolution imagery. Visual interpretation, however, is better for delimiting ecologically realistic objects, or patches. This study presents an approach that combines visual interpretation and object-based classification, with high spatial resolution digital aerial imagery, to characterize and analyze the fine-scale heterogeneity in urban landscapes. We applied this approach to the Gwynns Falls watershed in Baltimore, Maryland, USA.

Results/Conclusions

Image objects were generated at two hierarchical scales – patches and features within patches.  Patches were generated through visual interpretation, based on the HERCULES (High Ecological Resolution Classification for Urban Landscapes and Environmental Systems) land cover classification scheme.  HERCULES classifies the biophysical structure of urban environments using six land cover features: (1) coarse-textured vegetation—trees and shrubs, (2) fine-textured vegetation—herbs and grasses, (3) bare soil, (4) pavement, (5) building, and (6) building typology. These patches served as pre-defined boundaries for finer-scale segmentation and classification of within-patch land cover features. Addition datasets including LIDAR data and building footprints were used to both facilitate the finer scale object segmentation and obtain greater classification accuracy. Patches were then classified based on the within-patch proportion cover of features. The object-based classification approach proved to be effective for classifying within-patch land cover features and the overall accuracy of the classification maps of 1999 and 2004 were 92.3% and 93.7%, respectively. This exercise demonstrates that by integrating visual interpretation with object-based classification, the fine-scale spatial heterogeneity in urban landscapes can be characterized and analyzed in a more efficient and ecologically meaningful way.