prikaz prve stranice dokumenta Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu
  Download
PDF 9.6 MB
master's thesis
Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu
2021. urn:nbn:hr:200:098658
Radičević, Valentin
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Communications
Chair of Multimedia Systems and Digital Television

Cite this document

Radičević, V. (2021). Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:098658

Radičević, Valentin. "Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2021. https://urn.nsk.hr/urn:nbn:hr:200:098658

Radičević, Valentin. "Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2021. https://urn.nsk.hr/urn:nbn:hr:200:098658

Radičević, V. (2021). 'Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 20 April 2024, https://urn.nsk.hr/urn:nbn:hr:200:098658

Radičević V. Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2021 [cited 2024 April 20] Available at: https://urn.nsk.hr/urn:nbn:hr:200:098658

V. Radičević, "Proračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2021. Available at: https://urn.nsk.hr/urn:nbn:hr:200:098658

Please login to the repository to save this object to your list.
Metadata
TitleProračun vektora pokreta na razini blokova i njihovo grupiranje prema sličnosti uz implementaciju na realnu ADAS razvojnu platformu
Title (english)Calculation of motion vectors at the block level and their grouping according to similarity with implementation on a real ADAS development platform
AuthorValentin Radičević
MentorMario Vranješ (mentor)
Committee memberMario Vranješ (predsjednik povjerenstva)
Committee memberMarijan Herceg (član povjerenstva)
Committee memberRatko Grbić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Communications)
(Chair of Multimedia Systems and Digital Television)
Osijek
Defense date and country2021-09-16, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Electrical Engineering
Telecommunications and Informatics
Abstract
Ovaj rad se bavi metodama za pronalazak vektora pokreta i njihovim grupiranjem s ciljem detekcije broja i smjera kretanja pokretnih objekata u sceni te implementacijom metoda na ugradbenu ADAS Alpha ploču. Nakon implementacije i optimizacije rješenja u C programskom jeziku na osobnom računalu, odabrane metode su implementirane na ADAS ploču, također korištenjem C programskog jezika. S dovršetkom prve implementacije na Alpha ploču, pažnja je posvećena na raspodjelu poslova na dostupne procesore Alpha ploče. Nakon prikladne raspodjele posla, uspoređene su performanse triju metoda za pretragu vektora pokreta (metoda iscrpne pretrage, metoda tri koraka s mogućnošću povećanja broja koraka te metoda tri koraka s uključenim predviđanjem i prilagodbom na odabir oblika pretrage), prilikom izvođenja na osobnom računalu i na ugradbenoj ADAS Alpha ploči. Rezultati su pokazali da je predložena metoda iscrpne pretrage najtočnija metoda za pronalazak VP, a ostale metode imaju prihvatljivu točnost. Testiranjem rješenja na video signalima iz prirode (s prometnica) uočeno je da je predloženo rješenje za grupiranje VP primjenjivo u slučajevima gdje kamera miruje, dok je za slučaj video signala u kojima se kamera kreće dalo neprihvatljive rezultate točnosti. Rezultati testiranja brzine izvođenja optimiziranog rješenja pokazali su da je moguća obrada uz više od 50 okvira u sekundi za ulazne slike rezolucije 1280x720 elemenata slike.
Abstract (english)
This paper presents methods for the discovery of motion vectors (MV) and their grouping with the goal of detection of number and direction of movable objects contained in the scene as well as the implementation of those methods onto an ADAS Alpha board. After implementing and optimizing the solutions in C programming language on a personal computer, chosen methods were implemented onto an ADAS board using the same programming language. With the completion of the first implementation on the Alpha board, attention was given to the allocation of tasks to the available Alpha board processors. After appropriate task allocation, the performance of three motion vector search methods (exhaustive search method, three-step search method with the possibility of increasing the number of steps, and three-step method with prediction and adaptation of search form selection) was compared both on a personal computer as well as on the ADAS Alpha board. The results have shown that the exhaustive search method is the most accurate method for finding MV with other methods having acceptable accuracy. Testing the solutions on real-life (taken from roads) video signals have shown that that the proposed solution for grouping MV is applicable in cases where the camera is stationary while it gave unacceptable accuracy results in cases of video signals where the camera is moving. The results of testing the execution speed of an optimized solution have shown that processing is possible with more than 50 frames per second for input images with a resolution of 1280x720 pixels.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:098658
Study programmeTitle: Graduate University Study Programme in Electrical Engineering; specializations in: Communications and Informatics, Power Engineering
Course: Communications and Informatics
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike (magistar/magistra inženjer/inženjerka elektrotehnike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2021-10-05 15:14:35