prikaz prve stranice dokumenta Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila
  Download
PDF 3.8 MB
master's thesis
Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila
Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. urn:nbn:hr:200:670680
Ovžetski, Marija
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

Ovžetski, M. (2023). Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila (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:200:670680

Ovžetski, Marija. "Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. https://urn.nsk.hr/urn:nbn:hr:200:670680

Ovžetski, Marija. "Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. https://urn.nsk.hr/urn:nbn:hr:200:670680

Ovžetski, M. (2023). 'Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 14 March 2025, https://urn.nsk.hr/urn:nbn:hr:200:670680

Ovžetski M. Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2023 [cited 2025 March 14] Available at: https://urn.nsk.hr/urn:nbn:hr:200:670680

M. Ovžetski, "Algoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:200:670680

Please login to the repository to save this object to your list.
Metadata
TitleAlgoritam za uključivanje alarmnog sustava prilikom izlaska iz vozila
Title (english)Algorithm for turning on the alarm system when leaving the vehicle
AuthorMarija Ovžetski
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 country2023-09-12, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Electrical Engineering
Telecommunications and Informatics
Abstract
U ovom diplomskom radu opisano je rješenje za uključivanje alarmnog sustava prilikom izlaska vozača iz vozila s ciljem povećanja sigurnosti vozača i sudionika prometa koji mu dolaze odostraga. Rješenje se sastoji od YOLOv7 algoritma za detekciju sudionika prometa u okviru video sekvence, algoritma za praćenje sudionika prometa te algoritma za procjenu udaljenosti koji je potreban za procjenu brzine i vremena do potencijalne kolizije s vozačevim vratima na temelju kojeg se uključuje... More alarmni sustav. Rješenje je implementirano na osobnom računalu i na Raspberry Pi 4 platformi. Osobno računalo ima jači procesor i grafičku karticu, što omogućuje brže izvršavanje rješenja. S druge strane, Raspberry Pi 4 platforma ima ograničenje u računalnoj snazi, što utječe na vrijeme izvršavanja rješenja i točnost rezultata. Na Raspberry Pi 4 platformi zbog ograničenih resursa detekcija je napravljena pomoću YOLOv7 Tiny modela. Rješenje je testirano na vlastitoj bazi od 34 video sekvence. Detekcijom je određen položaj sudionika prometa u video okviru pomoću istreniranog YOLOv7 modela koji klasificira pet klasa: automobil, kamion/autobus, pješak, biciklist i motociklist. Brža obrada omogućena je praćenjem detektiranog sudionika prometa između okvira pomoću DeepSort algoritma. Procjena brzine sudionika prometa se izvršava na temelju prijeđene udaljenosti između dvaju okvira video sekvenci. Na 34 testne video sekvence rezultati su pokazali da se postotna pogreška za udaljenost na osobnom računalu kreće se od -24% do +27%, dok na Raspberry Pi 4 platformi varira od -42% do +28%. Za brzinu, postotna pogreška na osobnom računalu kreće se od -39% do +52%, dok na Raspberry Pi 4 platformi iznosi od -48% do +69%. Važno je napomenuti da se postotne pogreške razlikuju ovisno o klasi sudionika prometa, udaljenosti i brzini na kojoj su testirane. Međutim, unatoč navedenim rezultatima, rješenje kao takvo ima prostora za doradu kako bi se povećala preciznost u procjenu udaljenosti, brzine i TTC-a. Less
Abstract (english)
This thesis describes a solution for turning on the alarm system when the driver exits the vehicle, with the aim of enhancing the safety of both the driver and the traffic participants approaching from the rear. The solution consists of the YOLOv7 algorithm for detecting traffic participants in video sequences, a tracking algorithm for tracking traffic participants, and a distance estimation algorithm needed for calculating speed and time-to-collision with the driver's door, which... More triggers the alarm system. The solution is implemented on a personal computer and on the Raspberry Pi 4 platform. The personal computer has a more powerful processor and graphics card, enabling faster execution of the solution. On the other hand, the Raspberry Pi 4 platform has limited computational resources, affecting the execution time and accuracy of the results. On the Raspberry Pi 4 platform, detection is performed using the YOLOv7 Tiny model due to a lack of resources. The solution is tested on a custom dataset consisting of 34 video sequences. Detection determines the position of traffic participants in the video frame using the trained YOLOv7 model that classifies five classes: car, truck/bus, pedestrian, cyclist, and motorcyclist. Faster processing is achieved by tracking the detected traffic participant between frames using the DeepSort algorithm. Speed estimation of traffic participants is performed based on the distance traveled between two frames of the video sequence. Over the 34 test video sequences, the results showed that the percentage error for distance on the personal computer ranges from -24% to +27%, while on the Raspberry Pi 4 platform, it varies from -42% to +28%. For speed, the percentage error on the personal computer ranges from -39% to +52%, while on the Raspberry Pi 4 platform, it ranges from -48% to +69%. It is important to note that the error percentage varies depending on the class of traffic participant, distance, and speed at which they were tested. However, despite these results, the solution has room for improvement to increase the accuracy of distance, speed, and TTC (Time-to-Collision) estimation. Less
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:670680
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
RepositoryFaculty of Electrical Engineering, Computer Science and Information Technology Osijek
Created on2023-09-22 09:07:20
accessibility

closeAccessibilityrefresh

If you wish to permanently save changes, click on Save, if not - your settings will be reset when you restart the browser.