Sustav za automatsko optičko očitanje digitalnih prikaznika

master's thesis

• Title (english): Automatic Optical Readout System for Digital Displays
• Author: Luka Škorić
• Mentor: Davor Petrinović (mentor)
• Committee member: Davor Petrinović (predsjednik povjerenstva)
• Committee member: Marko Subašić (član povjerenstva)
• Committee member: Nikola Banić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2018-07-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

U okviru ovog rada razvijen je programski sustav namjenjen automatskom
optičkom očitavanju sadržaja digitalnih prikaznika. Razvijeni sustav objedinjuje
algoritam lokalizacije prikaznika unutar učitane slike, predobradu sadržaja
prikaznika i segmentaciju promatranog zapisa lokalizacijom znakova. Značenje
pojedinog lokaliziranog znaka prepoznaje se obradom od strane tri, u tu svrhu
implementirana algoritma. Algoritmom usporedbe slike sa zadanim predloškom,
direktno se iz prikaza lokaliziranog znaka saznaje tražena informacija, dok
algoritmi neuronske mreže i K najbližeg susjeda to ostvaruju primjenom metoda
strojnog učenja. Za potrebe učenja i testiranja navedenih algoritama strojnog
učenja kreirana je baza podataka sa slikama 7 – segmentnog formata. Razvijeni
programski sustav potpunu funkcionalnost ostvaruje implementacijom na
ugradbenom računalu s odgovarajućim ulazno - izlaznim periferijama. Takvi
sustavi primjenu pronalaze u polju mjeriteljstva, nadzora ili automatiziranog
upravljanja.

Abstract (english)

This thesis describes the course of development of an automatized system
for optical readout of the content from digital displays. The developed system
implements various algorithms with tasks of localizing a display within the input
picture, pre - processing of the localized content and segmentation of the
observed sequence by character localization methods. Value of the observed
localized character is recognized by three different algorithms implemented for this
purpose. Template comparison algorithm has the capability of direct extraction of
sought after information form the content of the input image. Neural network and K
nearest neighbour algorithms accomplish that task by applying the methods of
machine learning. Special database of 7 – segment character images has been
created for the purpose of training and testing earlier mentioned algorithms. The
developed application accomplishes its full functionality when implemented within
the embedded computer with appropriate input - output peripherals. Such systems
can be used for purposes of metrology, surveillance and automated control.

• Keywords: računalni vid; 7 – segmentni format; OCR; binarizacija slike; prepoznavanje predloškom; neuronska mreža; Knn algoritam; Unix; ugradbeni sustavi
• Keywords (english): computer vision; 7 – segment format; OCR; image thresholding; template recognition; neural network; Knn algorithm; Unix; embedded systems
• Language: croatian
• URN:NBN: urn:nbn:hr:168:338880
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-03-14 01:19:34