Klasifikacija tipova vozila korištenjem dubokih neuronskih mreža

master's thesis

• Title (english): Classification of vehicle types using deep neural networks
• Author: Dominik Bakarić
• Mentor: Vedran Mornar (mentor)
• Committee member: Vedran Mornar (predsjednik povjerenstva)
• Committee member: Gordan Gledec (član povjerenstva)
• Committee member: Marko Horvat (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-12, Croatia

Abstract

U današnje vrijeme, napretkom u području prepoznavanja objekata i računalnom vidu, jasno se vidi mogućnost primjene tih tehnologija u prometu i transportu. Te tehnologije mogu se primjenjivati na razne aspekte prometa kao što su regulacija prometa, sigurnost u prometu te razvoj autonomnih vozila. U sklopu ovog rada treba ostvariti model koji je će na slici prepoznavati vozila te ih klasificirati u kategorije kao što su automobil, motocikl, autobus, kamion ili kombi. Za realizaciju tog zadatka koristiti biblioteku Tensorflow 2 programskog jezika Python s ciljem izgradnje dubokog modela koji je baziran na konvolucijskim neuronskim mrežama te treniran korištenjem metoda prenesenog učenja.

Abstract (english)

Nowadays, with progress in the field of object recognition and computer vision, the possibility of applying these technologies in traffic and transport is clearly visible. These technologies can be applied to various aspects of traffic such as traffic regulation, traffic safety and the development of autonomous vehicles. As part of this work, a model should be created that will recognize vehicles in the image and classify them into categories such as car, motorcycle, bus, truck or van. To realize this task, we use the Tensorflow 2 library of the Python programming language with the aim of building a deep model based on convolutional neural networks and trained using transfer learning methods.

• Keywords: Duboke neuronske mreže; konvolucija; Tensorflow; Prepoznavanje objekata; Python
• Keywords (english): Deep neural networks; convolution; Tensorflow; Object recognition; Python
• Language: croatian
• URN:NBN: urn:nbn:hr:168:200483
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 20:00:50