Određivanje prostorne dubine pomoću pseudoslučajnih uzoraka

master's thesis

• Title (english): Depth estimation using pseudorandom patterns
• Author: Igor Ključariček
• Mentor: Tomislav Pribanić (mentor)
• Committee member: Tomislav Pribanić (predsjednik povjerenstva)
• Committee member: Igor Sunday Pandžić (član povjerenstva)
• Committee member: Mario Cifrek (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2014-07-09, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Dan je opis i pretpostavljeni princip rada Kinect uređaja. Implementirana je metoda normalizirane križne korelacije koja je potrebna za izračun matrice dispariteta. Opisana je integralna slika, algoritam kojim se optimira NCC-a. Prikazan je kratak uvod u pisanje MEX funkcija koje služe za ubrzanje postojećeg koda u Matlabu. Izvedena su mjerenja i napravljena je statistika implementiranog algoritma normalizirane križne korelacije. Predstavljen je sustav za projiciranje pseudoslučajnog uzorka kao i sam uzorak koji smo projicirali na ravnu površinu zida i lutku. Napravljena je statistika točnosti pseudoslučajnog uzorka kao i sama 3D rekonstrukcija.

Abstract (english)

In this thesis it's presented assumed working principle of the Kinect. For calculation of disparity matrix it's implemented and described normalized cross-correlation method. To optimize the NCC it is implemented an integral image algorithm. There is also a introduction to writing MEX functions which serve to speed up existing code in Matlab. Measurements are made as well as statistics of the implemented normalized cross-correlation algorithm. System for projecting a random speckle pattern is presented. Experiments projecting a pattern on the flat surface and on a doll are demonstrated. Finally, after statistics of pseudorandom pattern is done it's shown that 3D reconstruction is possible.

• Keywords: Kinect; prostorna dubina; normalizirana križna korelacija; NCC; matrica dispariteta; integralna slika; MEX funkcija; pseudoslučajni uzorak; 3D rekonstrukcija
• Keywords (english): Kinect; spatial depth; normalized cross correlation; NCC; disparity matrix; integral image; summed area table; MEX function; pseudorandom pattern; 3D reconstruction
• Language: croatian
• URN:NBN: urn:nbn:hr:168:427537
• Study programme: Title: Information and Communication Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka informacijske i komunikacijske tehnologije (magistar/magistra inženjer/inženjerka informacijske i komunikacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2018-02-06 09:41:10