Robotsko izuzimanje objekata temeljem boje i oblika

master's thesis

• Title (english): Robotic separation of objects based on color and shape
• Author: Luka Sinjeri
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2022-12-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Razvojem umjetne inteligencije uz sve veću dostupnost potrebne računalne opreme omogućeno je automatizirano prepoznavanje objekata korištenjem vizijskih senzora smještenih u stvarnu okolinu. Zadatak ovog diplomskog rada je razviti programsku podršku koja će omogućiti robotsko izuzimanje temeljeno na značajkama boje i oblika predmeta korištenjem određenih metoda iz područja strojnog vida. Korišteni vizijski senzor za snimanje predmeta iz okoline u stvarnom vremenu je web kamera. Prepoznavanje predmeta omogućeno je putem programa napravljenog korištenjem algoritama OpenCV knjižnice implementirane kroz programski jezik Python. Koordinate pronađenog predmeta šalju se na kolaborativnog robota Universal Robots UR5 koji vrši izuzimanje predmeta traženih karakteristika. Komunikacija između računala i robota ostvaruje se TCP protokolom. Robotska hvataljka potrebna za rukovanje predmetom oblikovana je i izrađena korištenjem aditivne tehnologije. Na početku rada dan je uvod u navedeno područje nakon čega se opisuju koraci izrade programskog rješenja. Na kraju rada analiziraju se dobiveni rezultati, nakon čega se izvodi odgovarajući zaključak.

Abstract (english)

Continuous development of artificial intelligence alongside increased accessibility of required computer equipment provides automated object detection by using visual sensors located in the real environment. Goal of this master’s thesis is to develop program support which would enable robotic separation based on attributes of color and shape of the object by utilizing certain methods of machine vision. Visual sensor used to record objects in real time located in environment is web camera. Object detection is achieved by program based on the algorithms that are part of OpenCV library implemented in Python programming language. Coordinates of detected object are sent to collaborative robot Universal Robots UR5 in order to perform robotic separation on object of required attributes. Communication between computer and robot is achieved by using TCP protocol. Robotic gripper needed for manipulation with objects is modeled and constructed by using additive manufacturing. Introduction into mentioned field is given at the beginning of thesis which is followed by describing steps of creating program support. After analyzing results, valid conclusion is given at the end of thesis.

• Keywords: prepoznavanje objekata; robotsko izuzimanje; OpenCV; UR5
• Keywords (english): object detection; robotic separation; OpenCV; UR5
• Language: croatian
• URN:NBN: urn:nbn:hr:235:669909
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka strojarstva (magistar/magistra inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2022-12-01 14:04:37