20.500.12556/DKUM-84362 Zaznava in lociranje malin z uporabo YOLO algoritma magistrsko delo Detection and localization of raspberries using the YOLO algorithm V magistrskem delu smo raziskali delovanje LiDAR senzorjev ter uporabo umetne inteligence v strojnem vidu, vključno z nevronskimi mrežami, konvolucijskimi nevronskimi mrežami (CNN) in algoritmi YOLOv3, v4 in v4-tiny. V praktičnem delu smo testirali vse tri algoritme in nato izbrali najuspešnejšega, YOLOv4, ter ga dodatno analizirali. Preverili smo hitrost algoritmov ter razvili algoritem, ki je na podlagi oblakov točk in kamere sposoben določiti lokacijo malin. Ugotovili smo, da je uporaba LiDAR senzorjev v kombinaciji z umetno inteligenco učinkovita pri zaznavanju in lociranju malin v 3D-prostoru. Najuspešnejši algoritem YOLOv4 je bil sposoben razvrstiti zrele in nezrele maline z natančnostjo 84,13 %. Naš razviti algoritem je omogočil določanje lokacije malin s kombinirano uporabo oblakov točk in kamere ter tako skoraj v polovici izmerjenih primerov določil lokacijo z napako, manjšo od 2 cm. In this master's thesis, we explored the operation of LiDAR sensors and the use of artificial intelligence in machine vision, including neural networks, convolutional neural networks (CNN), and YOLOv3, v4, and v4-tiny algorithms. In the practical part, we tested all three algorithms and then selected the most successful one, YOLOv4, and further analyzed it. We also checked the speed of the algorithms and developed an algorithm capable of determining the location of raspberries based on point clouds and a camera. We found that the use of LiDAR sensors in combination with artificial intelligence is effective in detecting and locating raspberries in 3D space. The most successful YOLOv4 algorithm was able to detect ripe and raw raspberries with an accuracy of 84.13%. Our developed algorithm enabled the determination of the location of raspberries using a combination of point clouds and a camera and thus determined the location with an error of less than 2 cm in almost half of the measured cases. malina strojni vid YOLO nevronska mreža CNN oblak točk raspberry machine vision YOLO neural netwok CNN point cloud true true false [U. Kenda] Slovenski jezik Angleški jezik Magistrsko delo/naloga 2023-05-29 13:35:34 2023-06-15 13:55:54 2023-07-08 03:08:36 0000-00-00 00:00:00 2023 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (VII, 48 f.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.8.021:528.8.044.6(043.2) 158081283 MAG_Kenda_Urban_2023.pdf MAG_Kenda_Urban_2023.pdf 1 FCF35F0731C4B1D51224B12BA6A6555C d4f969584e21e7eceda0d89773f026d2cdcf090cc03ffee7aaf9efad2dab685e 80b61681-0b64-11ee-8d9c-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=168045 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0