20.500.12556/DKUM-44193 GRUČENJE PODATKOV LiDAR CLUSTERING OF LiDAR DATA Cilje diplomske naloge je raziskati možnosti uporabe algoritmov gručenja za obdelavo podatkov LiDAR. Prvi del diplomske naloge predstavlja podatke LiDAR in algoritme gručenja. S senzorji LiDAR, pritrjenimi na letala ali helikopterje, je omogočeno hitro in natančno modeliranje površja. Metode gručenje predstavljajo enega izmed pristopov za detekcijo objektov na površju. Gručenje je vrsta nenadzorovane klasifikacije podatkov. Za učinkovito implementacijo algoritmov gručenja so potrebne posebne podatkovne strukture. Pri diplomski nalogi uporabljamo drevesa KD. V drugem delu diplomske naloge predstavimo aplikacijo. Pripravi podatkov s predobdelavo sledi uporaba algoritma DBSCAN za iskanje zgradb in cest. Za iskanje dreves smo uporabili algoritem k-means. V zaključku podamo rezultate in uporabljene parametre. The aim of this thesis is to explore clustering algorithms of point clouds collected by aerial LiDAR technology. The first part consideres LiDAR data and clustering algorithms. With LiDAR sensors that are mounted on air-planes or helicopters, fast and accurate surface modelling is possible. Clustering methods are one of the approaches for the detection of above-ground objects. Clustering is unsupervised classification of observed data into groups. For efficient implementation of clustering algorithms spatial data structures are required. KD-trees were used in this work. In the second part of diploma work, the developed application is presented. At first, the data was adequaetly preprocessed, then DBSCAN algorithm was used for buildings detection and road extraction. For tree crown delineation, the k-means algorithm was considered. The work is concluded with the results of clustering on real LiDAR data. algoritmi gručenja k-means DBSCAN iskanje zgradb iskanje cest iskanje drevesnih krošenj podatki LiDAR clustering algorithms k-means DBSCAN building detection road extraction tree crown delineation LiDAR data true true false [B. Založnik] Slovenski jezik Angleški jezik Diplomsko delo 2014-04-28 02:41:03 2014-05-20 07:58:20 2022-05-12 03:07:09 2021-05-01 03:09:37 2014 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.92.021(043.2) 17872150 URN:SI:UM:DK:DPQZBTSQ UNI_Zaloznik_Bostjan_2014.pdf UNI_Zaloznik_Bostjan_2014.pdf 1 263658EBF569C5B5D3E46E4FF469E8FE 5d223ad3a7a5fc6d35a624d6e8b05a0076b5c817d93bb9904c60d6a47c1a0de6 cd923476-7c0a-11eb-bb7a-00155d0001ca https://dk.um.si/Dokument.php?lang=slv&id=63350 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0