Wavelet-based system for classification of airborne laser scanner data

A new semi-automatic processing system for classification of airborne laser scanner cloud points is developed. To mitigate the difficulty caused by the complex distribution of objects on Earth's surface, wavelet was adopted in size-based clustering of laser points. A hybrid method of processing laser scanner data in both grid and raw formats was also adopted to speed up the processing time and adjust the smoothing effect of interpolation. The processing focused on processing the data acquired over urban area. This paper presents and explains the components of the system using the test data acquired over Shinjuku area, Tokyo, Japan.