Extraction of buildings in informal settlement areas from high-resolution data

S.D. Mayunga, Y. Zhang, D.J. Coleman

Research output: Chapter in Book/Report/Conference proceedingConference contribution


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 buildings in urban areas has been by conventional photogrammetry using aerial photos. This method is considered accurate but very slow, expensive, manually operated and require well-trained personnel. In recent years we have experienced the emergence of high-resolution space borne images, which have shown potential for medium and large-scale topographic mapping. In this paper, we have developed a semi-automatic method to extract buildings in informal settlements areas. The proposed method uses radial casting algorithm to initialize snakes control points. The fine measurements of building outlines from high-spatial resolution panchromatic imagery is carried out using snakes model. The preliminary results are encouraging and demonstrated by examples over a variety of selected test areas. Our approach is capable of extracting buildings consisting of rectilinear and non-rectilinear lines; limitations and future work of our approach is discussed.
Original languageEnglish
Title of host publicationAmerican Society for Photogrammetry and Remote Sensing - Annual Conference 2005 - Geospatial Goes Global
Subtitle of host publicationFrom Your Neighborhood to the Whole Planet
Number of pages12
Publication statusPublished - 2005
EventAnnual Conference 2005 - Geospatial Goes Global:: From Your Neighborhood to the Whole Planet -
Duration: Mar 7 2005Mar 11 2005


ConferenceAnnual Conference 2005 - Geospatial Goes Global:


Dive into the research topics of 'Extraction of buildings in informal settlement areas from high-resolution data'. Together they form a unique fingerprint.

Cite this