WAVELET AND SCALE-SPACE THEORY IN SEGMENTATION OF AIRBORNE LASER SCANNER DATA

摘要

An important property of any object in the world is that it only appears over a certain ranges of scales. Therefore, the scales should be considered in describing or analyzing of an object. Scale-space theory has been well developed in computer vision. Recent years, wavelet, a new and impressive mathematical tool, which can be said as an evolution of scale-space theory, has raised the consideration of the scientists in many fields including geoinfomatics. Based on the idea well developed in pattern recognition, this paper is aimed to introduce the application of wavelet in segmentation algorithms of airborne laser scanner. In this paper, a segmentation wavelet based framework was set up. Afterwards, several experiments applied in testing site were done to prove the feasibility of proposed method in segmentation of airborne laser scanner data. This wo rk is only a part of author's doctoral dissertation. As a consequence, this paper only presents the preliminary results of this work and the further developments will be pointed out.