Analiza i vrednovanje odabranih izvedbenih detalja postupaka optičkog toka

master's thesis

• Title (english): Analysis and evaluation of optical flow implementation details
• Author: Igor Smolkovič
• Mentor: Siniša Šegvić (mentor)
• Committee member: Siniša Šegvić (predsjednik povjerenstva)
• Committee member: Zoran Kalafatić (član povjerenstva)
• Committee member: Marin Golub (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2014-07-17, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U radu se razmatra problem određivanja gustog optičkog toka u videu. U okviru rada proučeni su postupci procjene optičkog toka varijacijskim računom, gdje je cilj minimizacija energijskog funkcionala. Izvedene su diferencijalne jednadžbe odabranih postupaka te je prikazan postupak njihova rješavanja. Razvijen je programski sustav koji omogućava evaluaciju i vizualizaciju rezultata dobivenih procijenom optičkog toka. Na kraju su prezentirani eksperimentalni rezultati dobiveni provođenjem evaluacije postupaka na ispitnoj bazi Middlebury te je dana njihova ocjena.

Abstract (english)

In this paper we study estimation of dense optical flow in the video. Selected methods based on calculus of variations are analyzed. Variational methods aim to minimize an energy functional for the estimation of optical flow field. Differential equations of selected methods are derived and procedures for their solving are shown. A system for evaluation and visualisation of results obtained by optical flow estimation procedures is developed. Finally, experimental results obtained by conducting of evaluation procedures on Middlebury database are presented and graded.

• Keywords: analiza pokreta; gusti optički tok; varijacijski račun; energijski funkcional; norma ukupne varijacije; konveksne optimizacije; Middlebury; računalni vid
• Keywords (english): motion analysis; dense optical flow; calculus of variations; energy functional; total variation norm; convex optimizations; Middlebury; computer vision
• Language: croatian
• URN:NBN: urn:nbn:hr:168:083343
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2018-02-06 11:50:18