Clustering method to extrct buildings from airborne laser data

We´ve proposed a clustering method based on wavelet analysis to extract buildings from airborne laser cloud data. In the case of 3 dimensional model of buildings are generated from the airborne laser data, usually laser data with GIS data are utulized. The reson is avoid the process to make polygon data from cloud data. Without spectral information provided by the airborne laser scanner, these extraction methods suffer tremendous difficulties when low-density laser cloud points are adopted. To mitigate the difficulties in extraction, the size of object should be taken into account. This study proposed a multi- resolution clustering approach based on wavelet to extract the buildings in a dense urban area from the low-density airborne laser scanner data with the assistance of extractable information from the aerial photographs. The proposed approach was tested in Shinjuku-ku, Tokyo, Japan, and showed its efficiency.