Prepoznavanje šahovske ploče i figura temeljeno na metodama dubokog učenja

undergraduate thesis

• Title (english): Chessboard and figure recognition based on deep learning methods
• Author: Karlo Filek
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2023-09-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Zadnjih 2 desetljeća metode računalnog vida i strojnog učenja su se drastično razvile i modernom arhitekturom računala omogućili real-time izvedbu zadanih algoritama prepoznavanja i detekcije objekata. Tema ovog rada je izrada računalnog programa za detekciju i prepoznavanje šahovske ploče i figura te usporedba u učinkovitosti modela s obzirom o kutu gledanja na šahovsku ploču. Rješenje je bazirano na implementaciji duboke neuronske mreže doučene na naš set podataka te je izrađeno u programskom jeziku Python uz korištenje prograskih biblioteka otvorenog koda OpenCV za područje obrade slika i računalnog vida i TensorFlow za područje učenje neuronske mreže i obradu podataka. Učenje mreža je ubrzano primjenom GoogleColab kolaborativnog okruženja gdje se omogućava pristub Google-ovim radnim jedinicama.

Abstract (english)

For the past 2 decades the methods for computer vision and object recognition have been rapidly developing. With the evolution of modern computer processing units real-time object detection and recognition has been made a possibility. The goal of this thesis is to create a computer program for chess board detection and figure recognition based on deep neural network recognition. Operating program is built in Python 3.10 and uses open source libraries such as OpenCV for image manipulation and TensorFlow for neural network fitting and image evaluation. The training of the neural network was sped up by implementing Googles server resources through GoogleColab platform.

• Keywords: Python; OpenCV; TensorFlow; duboke neuronske mreže; prepoznavanje objekata; VGG16
• Keywords (english): Python; OpenCV; TensorFlow; deep neural network; object recognition; VGG16
• Language: croatian
• URN:NBN: urn:nbn:hr:235:329600
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2023-09-29 14:42:26