Uporaba projekcijske geometrije i aktivnog vida u tumačenju scena

scientific master's theses

• Title (english): Interpreting real scenes by means of projective geometry and active vision
• Author: Siniša Šegvić
• Mentor: Slobodan Ribarić (mentor)
• Committee member: Leo Budin (predsjednik povjerenstva)
• Committee member: Slobodan Ribarić (član povjerenstva)
• Committee member: Nikola Pavešić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Electronics, Microelectronics, Computer and Intelligent Systems); Zagreb
• Defense date and country: 2000-05-03, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering; Electronics
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Samostalno određivanje položaja u okolišu (lokalizacija) je preduvjet za postizanje samostalne navigacije. Predložen je pristup lokalizaciji samostalnog robota u sustavu hodnika, koji se temelji na projekcijskoj geometriji i jednookom aktivnom računarskom vidu. Količina korisne informacije za snalaženje koja je sadržana u slici razmatranog okoliša jako ovisi o položaju i orijentaciji kamere kojom je slika pribavljena. Predložena lokalizacijska procedura stoga, pored postupaka za obradu slike, propisuje i postupke za njihovo pribavljanje. Tako se smjerovi pružanja hodnika odnosno mogući smjerovi napredovanja određuju obradom niza slika pribavljenih iz zadanog položaja robota, za različite smjerove gledanja kamere. Obrada pribavljenih slika se sastoji od traženja hipotetskih točaka nedogleda izlučenih pravocrtnih segmenata, u skladu sa principima projekcijske geometrije. Konačno, smjerovi pružanja hodnika se određuju grupiranjem pronađenih nedogleda u koordinatnom sustavu robota. Pojedinačni postupci i cjelokupna lokalizacijska procedura su isprobani na stvarnim scenama.

Abstract (english)

The capability of a mobile robot to determine its position in the environment (self-localisa\-ti\-on) is a prerequisite for achieving autonomous navigation. The proposed approach to self-localisation in a system of corridors is based on the projective geometry and active computer vision using a single camera. The quantity of navigationally useful information contained in an image of the considered environment strongly depends on position and orientation of the camera used to acquire the image. The presented self-localisation procedure therefore specifies both image processing algorythms and methods for controlling the image acquisition process. Thus, the directions of longitudinal corridor axes in the environment (or, equivalently, the viable directions of advancement) are determined by processing the sequence of images acquired from the given position, with different orientations of the camera. The processing of acquired images consists of finding hypothetical vanishing points of extracted line segments, in accordance with principles of the projective geometry. Finally, the viable advancement directions in the corridor are determined by grouping of obtained vanishing points in the coordinate system of the robot. Both isolated techniques and the whole self-localisation procedure are tested on real scenes.

• Keywords: Računarski vid; tumačenje trodimenzionalnih scena; projekcijska geometrija; aktivna percepcija; rubna segmentacija; Houghova transformacija; pravocrtni segmenti; nedogledi; ortogonalna triedarska struktura
• Keywords (english): Computer vision; 3D scene analysis; projective geometry; active perception; edge segmentation; Hough transform; straight line segments; vanishing points; orthogonal trihedral structure
• Language: croatian
• URN:NBN: urn:nbn:hr:168:291475
• Study programme: Title: Postgraduate master programme in electrical engineering; Study programme type: university; Study level: postgraduate; Academic / professional title: Magistar znanosti elektrotehnike (Magistar znanosti elektrotehnike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-04-02 13:42:23