PBTIN: Prismatic Buffered TIN un algoritmo di filtraggio per dati ALS

Very detailed high-resolution (3D) digital terrain models are required in many applications, both civilian and military, and can be obtained by using airborne laser scanner that can be considered one of the most important methods for large scale acquisition of height data. The extraction of points on bare Earth from point-clouds acquired by PBTIN: Prismatic Buffered TIN un algoritmo di filtraggio per dati ALS airborne laser scanning is the most time-consuming part of the production of digital elevation models with laser scanning. Over the last years various and valid algorithms for the classification of ground points have been proposed, which use statistical-geometrical procedures based on certain suppositions of bare-Earth morphology. This paper will discuss a new method supporting automatic construction of 3D digital terrain model, based on the iterative densification of TIN, that can be considered a variation of the Axelsson’s algorithm. Different filtering strategies propose that, in presence of data from zones with different structures, morphological characteristics and density, diversified treatments are allowed. This method and the different filtering strategies have been applied for the automatic generation of digital terrain models using data coming from different ALS surveys. In this paper results coming from the application to data used for 2003 ISPRS are presented. The comparison between the obtained results and the ones coming from the other algorithms used in the same test shows the satisfactory performances in the filtering errors reduction.