izpis_h1_title_alt

Večkamerni sistem za lokalizacijo objekta v prostoru
ID Beličič, Ernest (Author), ID Skočaj, Danijel (Mentor) More about this mentor... This link opens in a new window

.pdfPDF - Presentation file, Download (23,58 MB)
MD5: 31994F6C1DA8DC0D9E37744FECE54401

Abstract
Lokalizacija igra pomembno vlogo v številnih aplikacijah, na primer filmska in igračarska industrija uporabljata lokalizacijske sisteme za zajem gibanja telesa. V diplomskem delu implementiramo postopek lokalizacije v prostoru s pomočjo kamer. Objekt, ki ga lokaliziramo, je okrogel svetleči barvni označevalnik. Kameram določimo notranje in zunanje parametre ter jih tako umestimo v prostor, pri čemer notranje parametre določimo z MATLAB-ovim orodjem za kalibracijo kamer, zunanje pa izračunamo s pomočjo metode DLT in SVD. V realnem času zaznavamo označevalnik na slikah in s postopkom triangulacije določimo njegov položaj v prostoru. Sistem tudi ustrezno evalviramo in ocenimo napake lokalizacije. Izkaže se, da so napake v sistemu večinoma sistematične in se jih lahko popravi.

Language:Slovenian
Keywords:lokalizacija, triangulacija, večkamerni sistem, računalniški vid, kalibracija kamer, DLT, SVD
Work type:Bachelor thesis/paper
Typology:2.11 - Undergraduate Thesis
Organization:FRI - Faculty of Computer and Information Science
Publisher:[E. Beličič]
Year:2015
Number of pages:57 str.
PID:20.500.12556/RUL-72605 This link opens in a new window
COBISS.SI-ID:1536518595 This link opens in a new window
Publication date in RUL:29.09.2015
Views:1637
Downloads:271
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:English
Title:Multi-camera system for object localization in space
Abstract:
Localization plays a big role in many applications for example film and game industry are using localization systems for capturing of body movement. In this thesis we implement localization procedure using cameras. Localization object is a light emitting spherical color marker. We compute cameras intrinsic and extrinsic parameters effectively fitting camera in space. Intrinsic parameters are determined using MATLAB camera calibration tool. Extrinsic parameters are calculated using DLT and SVD methods. From images we detect marker in real time and use triangulation to determine its position in space. We properly evaluate the system and measure localization accuracy. It turns out that system errors are mostly systematic and can be corrected.

Keywords:localization, triangulation, multi-camera system, computer vision, camera calibration, DLT, SVD

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back