20.500.12556/DKUM-95983 Spremljanje gostote in hitrosti prometa z uporabo kamere na platformi RPI magistrsko delo Monitoring traffic density and speed using a camera on the RPI platform Magistrsko delo obravnava razvoj sistema za spremljanje gostote in hitrosti prometa z uporabo kamere na platformi Raspberry Pi in pospeševalnika Coral TPU. Namen raziskave je zasnova cenovno dostopnega in energetsko učinkovitega sistema, ki lokalno v realnem času obdeluje videoposnetke brez zunanje infrastrukture. Sistem temelji na metodah računalniškega vida in umetne inteligence ter omogoča zaznavanje, sledenje, razvrščanje in ocenjevanje hitrosti vozil. Za zaznavanje je uporabljen prilagojen obstoječ model, za sledenje pa algoritem SORT, medtem ko se hitrost ocenjuje z analizo zaporednih slik prek več virtualnih črt brez dodatnih senzorjev. Delo prikazuje celoten proces načrtovanja, implementacije in delovanja sistema na robni napravi The master's thesis focuses on the development of a system for monitoring traffic density and vehicle speed using a camera on the Raspberry Pi platform with a Coral TPU accelerator. The aim of the research is to design a cost-effective and energy-efficient system capable of processing video streams locally and in real time without external infrastructure. The system is based on computer vision and artificial intelligence methods, enabling vehicle detection, tracking, classification, and speed estimation. A customized detection model is used together with the SORT algorithm for object tracking, while vehicle speed is estimated by analyzing consecutive frames using multiple virtual lines without additional sensors. The thesis presents the complete process of system design, implementation, and operation on an edge device. Računalniški vid Umetna inteligenca Zaznavanje vozil Sledenje objektov Ocena hitrosti Robne naprave Raspberry Pi Coral TPU Computer vision Artificial intelligence Vehicle detection Object tracking Speed estimation Edge devices Raspberry Pi Coral TPU true true false [J. Kozole] Slovenski jezik Angleški jezik Magistrsko delo/naloga 2025-11-18 09:29:14 2025-12-22 11:56:43 2026-01-31 03:15:48 0000-00-00 00:00:00 2025 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (65 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 [004.8:004.93]:656.021(043.2) 266816515 MAG_Kozole_Jure_2025.pdf MAG_Kozole_Jure_2025.pdf 1 D57344305AECD08BBCC57458891E5D47 bfbe6d8c9bd57d8d457ca28b43bc3f6475750adebf18f2ff0a02e6b9f80f5576 aad665da-ca08-11f0-8fc7-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=200885 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0