Detekcija i očitanje registarskih oznaka vozila u otežanim uvjetima

undergraduate thesis

• Title (english): Detection and Reading of Vehicle Registrations in Difficult Conditions
• Author: Fran Hančić
• Mentor: Marko Subašić (mentor)
• Committee member: Marko Subašić (predsjednik povjerenstva)
• Committee member: Ivana Zadro Matovinović (član povjerenstva)
• Committee member: Karlo Koščević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad bavi se izmjenom Faster R-CNN modela kako bi bolje odgovarao problemu detekcije registarskih oznaka u otežanim uvjetima. U radu su detaljno objašnjeni hiperparametri i njihov utjecaj te je kroz iteracije razvoja modela razvijen vrlo dobar model koji rješava zadan problem. Također je opsežno objašnjena metodologija u radu i koncepti kojima su se racionalizirale odluke donesene prilikom razvoja. Rad se također dotiče optičkog očitanja registarskih oznaka, gdje opisuje korake pred-procesiranja i samog očitanja tablica.

Abstract (english)

This paper deals with the modification of the Faster R-CNN model to better address the problem of detection of registration labels in difficult conditions. The paper explains in detail the hyperparameters and their influence, and through the development iterations of the model, a very good model has been developed to solve the given problem. The work methodology is also extensively explained along with concepts that rationalize decisions made during development. The work also touches upon the optical reading of the registration labels, describing the steps of pre-processing and reading vehicle registrations themselves.

• Keywords: računalni vid; OCR; detekcija objekata; registarske oznake; Faster R-CNN; konvolucijske neuralne mreže; duboko učenje
• Keywords (english): computer vision; OCR; object detection; registration plates; Faster R-CNN; convolutional neural network; deep learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:846636
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 18:22:21