Airborne laser scanning: System evaluation and building extraction

Airborne laser scanning technology is impressive in its capability of collecting a large number of points in a very short time and providing a reasonable depiction of complex objects in the scanned areas. It has been used in a wide range of applications with promising results. In this research, a detailed evaluation of the system is presented and the quality of the data used is quantified. In the second part of the research, a new technique to detect and reconstruct buildings with detailed roof delineation in urban areas is presented, using airborne laser scanning data exclusively. The quality of any final product naturally depends on the original data and methods of generating it. Thus the quality of the data should be evaluated before assessing any of its products. The work described in this research is aimed at a quantitative accuracy evaluation of the laser data itself. This is an area that has been under-emphasized in much published work on the applications of airborne laser scanning data. The results will address both the planimetric as well as the height accuracy of the laser data. This research also shows that dense laser data is very suitable for 3D extraction of urban features such as buildings, by presenting a building detection and reconstruction technique. This concept is based on local statistical inferences. The consistency of the data with surfaces determines how they will be modeled. In order to evaluate the building extraction method that is presented in this thesis, samples of the buildings that were extracted were also measured using conventional digital photogrammetry. Building features extracted from actual laser data collected over the Purdue campus are presented in the thesis