Ugotavljanje ujemanja lokalnih lastnosti oblaka točk

magistrsko delo

Denis Kolednik (Author), Borut Žalik (Mentor), Domen Mongus (Co-mentor)

Abstract

V magistrskem delu predstavimo dva postopka za iskanje sprememb nad oblakoma točk, zajetima s tehnologijo LiDAR (ang. Light Detection And Ranging). S prvim postopkom ugotavljamo količino zemeljskega površja, ki ga je naneslo ali odneslo na določenem področju. Ta metoda temelji na iskanju najbližjih sosedov točk z uporabo drevesa k-d. Metoda je zmožna 90-odstotne natančnosti. Z uporabo mreže višinskih vrednosti in optičnega toka smo v drugem postopku cenili morebitne premike na zemeljskem površju. V nasprotju s prvo metodo ta omogoča zaznavo premikov glede na ravnino xy. Natančnost zaznave premikov je 78-odstotna. Kombinacija obeh metod zagotavlja učinkovito zaznavo ujemanja lokalnih lastnosti oblakov točk ter opisa razlik.

Keywords

računalniška geometrija;razpoznava vzorcev;LiDAR;zaznava sprememb;optični tok;drevo k-d;

Data

Language:	Slovenian
Year of publishing:	2013
Typology:	2.09 - Master's Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[D. Kolednik]
UDC:	004.94(043)
COBISS:	17534486
Views:	1583
Downloads:	149
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	DETERMINATION OF POINT CLOUD LOCAL PROPERTIES
Secondary abstract:	This masters thesis presents two methods for difference detection of point clouds captured by LiDAR (Light Detection And Ranging) technology. The first method searches for the quantity of Earth's surface erosion and coating on a particular area. The main procedure that is used, is k-d tree nearest neighbour search. This method is capable of 90% accurate measurements. In the second method, height grids and optical flow are used to determine movement of the surface. In contrast to the first method, this one detects movement in the xy plane. A combination of both proposed methods provides an effective determination of local point cloud properties and a description of determined differences.
Secondary keywords:	pattern recognition;computational geometry;LidAR;different detection;optical flow;k-d tree;
URN:	URN:SI:UM:
Type (COBISS):	Master's thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko
Pages:	XII, 36 f.
ID:	8727386