Repozytorium publikacji - Politechnika Gdańska

Airborne laser scanning (ALS) provides survey results in the form of a point cloud. The ALS point cloud is a source of data used primarily for constructing a digital terrain model (DTM). To generate a DTM, the set of ALS observations must be first subjected to the point cloud processing methodology. A standard methodology is composed of the following stages: acquisition of the ALS data, initial processing (including filtration), and the main processing (including DTM generation). Problems primarily concerning the time span necessary for processing a very high number of observations occur at each of the stages mentioned above. In previous studies the authors proposed modification of the ALS point cloud methodology. The modification introduced an optimization algorithm to reduce the size of the survey dataset at the stage of initial processing. Those studies analyzed the “optimization–filtration” and “filtration–optimization” variants, applying methods based on the multi (M) estimation principle at the filtration stage. This study presents a modified process of the initial data processing stage with the application of filtration using the adaptive triangulated irregular network (TIN) model and an optimization algorithm. The algorithm used the Visvalingam–Whyatt generalization method.

Autorzy

• dr inż. Wioleta Błaszczak-bąk,
• Artur Janowski,
• prof. dr hab. inż. Waldemar Kamiński link otwiera się w nowej karcie ,
• Jacek Rapiński