Measuring Detailed Urban Vegetation with Multisource High-Resolution Remote Sensing Imagery for Environmental Design and Planning

Abstract

The availability of high-resolution remote sensing imagery brings both opportunity and challenge to environmental designers and planners in obtaining high-quality landscape information for better design and planning decision making. To meet the challenge, in this paper we introduce a comprehensive approach to measuring urban vegetation data detailed to single patches of trees or shrubs and single patches of lawn or grass with multisource remote sensing imageries. Methodologically, the approach integrates advanced geospatial technologies to achieve the research objective. First, an automatic registration algorithm is applied to align an unorthorectified QuickBird satellite multispectral imagery to a high-resolution United States Geographical Survey orthoimage. Next, an image segmentation process extracts landscape objects from such multisource data for further object-based image classification. Third, the approach takes advantage of the strong power of a group of prioritized spectral, geometric, topological, and thematic image object features to produce satisfactory classification results. The approach was tested in the Oakland Metropolitan Area in California, USA and was assessed with both groundtruthing and imagetruthing data. The paper concludes with a discussion on the potential applications of both the approach and the generated data in environmental design and planning.