Improved urban land cover mapping using multi-temporal IKONOS images for local government planning

Abstract

This paper studies the potential role of multi-temporal (leaves-on ‐ spring and leaves-off ‐ fall) and multi-spectral IKONOS satellite images in the development of detailed and accurate urban land cover maps to assist planners and managers in the rapidly growing city of Columbia, Missouri, U.S.A. To improve the overall accuracy of the urban land cover map, multi-temporal images were classified separately and then outputs were combined. The result produced from the combined outputs was an improvement over the previous single season image. An up-to-date and detailed land cover map produced from these images can help planners in developing better urban land use plans. This paper also discusses in detail the problems encountered and lessons learned in using high-resolution satellite images in urban land cover classification so as to help planners proposing to use similar type images in future.