The Use Of High-Resolution Imagery For Identification Of Urban Climax Forest Species Using Traditional And Rule-Based Classification Approach
Classification, Classification and regression tree (CART), Multitemporal satellite imagery, Plant identification
IEEE Transactions on Geoscience and Remote Sensing
9 PART I
Columbia, MO, is one of the fastest growing cities in the United States. The rapid urbanization is jeopardizing the city's urban climax forests, which are defined as any woodland community of more than 7.53 has dominated by plant species such as oak and hickory. The increasing urban pressure places heavy demands on city planners to seek better management approaches to ensure that these plant species, which are native to Missouri, are protected even when threatened by new developments. Traditionally, planners are identifying these areas by surveys, which are costly, time consuming, and conducted on an as-needed basis. The current study is to test the feasibility of high-resolution satellite and airborne imagery for the identification of these forests and to assist planners in preserving them. In order to identify these climax forests, 4-m multispectral IKONOS images for April and August 2000 and 25-cm and 1-m multispectral airborne photographs for September and November 2001 were acquired and used in this study. These images were classified using a traditional classifier [maximum likelihood (ML)] and a rule-based classifier [classification and regression tree (CART)]. Results show that the images taken in September using the ML classifier are more useful in identifying the tree species in the study area. Among the spatial resolutions used, 1-m images proved to be optimal in recognizing trees and at the same time minimizing shadows.
Original Publication Date
DOI of published version
Sugumaran, Ramanathan; Pavuluri, Manoj Kumar; and Zerr, Daniel, "The Use Of High-Resolution Imagery For Identification Of Urban Climax Forest Species Using Traditional And Rule-Based Classification Approach" (2003). Faculty Publications. 3241.