The Research Of Extracting Building's Feature From Vehicle-Borne Laser-Scanning Data

Laser scanning is an emerging and attractive technology in the field of remote sensing that is capable of rapidly capturing three dimensional geo-spatial information in multi-scales, and also gives us a new technology for obtaining short time resolution and high spatial resolution data. With more then ten years' development, airborne laser scanning systems available on the market are presently in a fairly mature state of art, the research of ground-based laser scanning systems also has some development, while the processing of laser scanning data is comparatively behindhand. The processing of laser scanning data is one of the focuses of current research of worldwide countries.The research presented in this paper is focused on extracting building's feature from laser scanning data, especially captured by vehicle-borne laser scanning systems. Firstly, laser scanning technology is introduced, the method of image segmentation and feature extraction, the method of classification and feature extraction of laser scanning data, are also analyzed. Then a method is proposed for classification and building's feature extraction from laser scanning data captured by vehicle-borne laser scanning. This method is also feasible to multi-view ground-based laser scanning data after registration. The spatial distribution characteristics of the points in the range image and the geometric characteristics of various objects in urban environment are adequately utilized in this method. First, all the data points in the range image are projected on the horizontal grids; second, for each grid the max value of the height before being projected of all the data points is calculated; then this max value is used to classify the data points. Some experimental examples using the proposed method are presented and the result of the experiment is analyzed in this paper. Finally, the focus of future research is presented.