Semi-automatic Building Extraction from Quickbird Imagery

Abstract

Extraction of geospatial data from digital images is one of the most complex and challenging tasks faced by computer vision and photogrammetry communities. Geospatial data and buildings in particular are required for varieties of applications such as urban planning, creation of GIS databases, and urban city models. For decades, extraction of geospatial data in urban areas has been by conventional photogrammetry using aerial photos. This method is expensive, manually operated and require well-trained personnel. In recent years we have experienced the emergence of high-resolution space borne images, which have disclosed a large number of new opportunities for medium and large-scale topographic mapping. In this paper, we have developed a semi-automatic method to extract buildings in structured and unstructured urban settlements areas from high-spatial resolution panchromatic imagery. The proposed method uses radial casting algorithm to initialize snakes contours, and the fine measurements of building outlines is carried out using snakes model. The results are satisfactory with an extraction rate of 91 percent as demonstrated by examples over a variety of selected test areas. The potential, limitations and future work of our approach is discussed.