Sustav za raspoznavanje znakovnog jezika

undergraduate thesis

• Title (english): System for Sign Language Recognition
• Author: Nikola Tomažin
• Mentor: Tomislav Hrkać (mentor)
• Committee member: Tomislav Hrkać (predsjednik povjerenstva)
• Committee member: Marin Oršić (član povjerenstva)
• Committee member: Ivan Grubišić (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

Broj ljudi koji koristi znakovni jezik je svakako značajan te bi se čitanje istog računalnim vidom moglo olakšati ljudima koji koriste taj jezik za komunikaciju ili sasvim suprotno, onima koji nisu upoznati s njime, a žele biti dio komunikacije. Prevođenje abecede korak je k tome. U ovom radu prikazan je način rada neuronskih mreža kod problema raspoznavanja jednoručnih i dvoručnih znakova hrvatske znakovne abecede. Objašnjena su dva pristupa problemu (histogram boje ruke, oduzimanje pozadine) te postupak pripreme i nadopunjavanja skupa podataka. Na kraju su analizirana ponašanja implementiranog sustava, prikazani problemi kod prepoznavanja te opisana uputstva za korištenje programa.

Abstract (english)

The number of people who use sign langauge is certantly significant and reading it using computer vision would make it easier from people who use it to communicate or, contrarily, those who are unfamiliar with it and want to be part of communication. Translating the alphabet is a step to it. This paper presents the way neural networks work in problems of recognizing letters of the Croatian sign language alphabet. Two approaches(hand histogram, background substraction) are explained as well as the procedure for preparing and upgrading the dataset. Finally, the behaviors of the implemented system were analyzed, recognition problems were described and the instructions for use of the program were shown.

• Keywords: strojno učenje; neuronske mreže; konvolucijske neuronske mreže; klasifikacija; znakovni jezik; abeceda; oduzimanje pozadine
• Keywords (english): machine learning; neural networks; convolutional neural networks; classification; sign langauge; alphabet; background substraction
• Language: croatian
• URN:NBN: urn:nbn:hr:168:723803
• 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 22:39:23