Automatska kalibracija vrha alata robota korištenjem tehnika strojnog vida

master's thesis

• Title (english): Automatic robot tool tip calibration using machine vision techniques
• Author: Petar Grabar
• Mentor: Filip Šuligoj (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2025-02-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

U ovom radu razvijeno je korištenje vizijskog sustava pri kalibraciji alata robota. Ključni dijelovi razvijenog vizijskog sustava su kamera, hardver za procesuiranje slika i softver koji sadrži funkcije za obradu slike. U uvodu je dano osnovno o vizijskim sustavima i kameri te su prikazane vrste vizijskih sustava i vrste konfiguracija robota i kamera. Nadalje je pokazana oprema koja se koristi za procesuiranje slike i dohvaćanje scene. Za procesiranje i obradu slike korištena je Jetson Nano platforma za razvijanje s instaliranim OpenCV-om za obradu slike, a za dohvaćanje scene korištene su dvije IMX219-120 MIPI kamere. Kako bi se na sustavu moglo raditi potrebno je na njega priključiti periferiju što omogućavaju priključci Jetson platforme. Da bi se oprema zaštitila napravljeno je kućište za Jetson platformu i napravljena je kalibracijska kutija koja je odvojena od procesne jedinice radi sigurnosti i robusnosti segmentacije. Dalje je proveden eksperiment odabira boje pozadine scene kako bi se odredila najprikladnija boja za dobar kontrast u odnosu na alat. Odabirom optimalne boje pozadine omogućuju se robusniju rezultati segmentacije, koja je ključna za dobru kalibraciju. Kako bi se sa kamera dobila dobra informacija potrebno ju je kalibrirati pa je kratko objašnjena te provedena kalibracije kamere. Pokazan je način segmentacije alata od podloge sa uključenim svim radnjama obrade slike. Definiran je način detekcije vrha alata robota i izračuni gibanja robota kako bi se alat nalazio u sjecištu optičkih osi kamera. Za izračunavanje kalibracije alata odnosno transformacije koordinatnog sustava vrha alata robota u odnosu na prirubnicu robota, što je i cilj kalibracije, prikazani su korišteni algoritmi. Pokazan je i način komunikacije robota i vizijskog sustava gdje je robot klijent, a vizijski sustav server u socket arhitekturi komunikacije. Dani je programski kod na robotu za proces kalibracije alata korištenjem vizijskog sustava. Na kraju su prikazani rezultati kalibracije korištenjem vizijskog sustava i dana je usporedba takvog načina kalibracije alata robota naspram klasične metode kalibracije robotskog alata.

Abstract (english)

In this paper, the use of a vision system during the calibration of robot tools was developed. The key parts of the developed vision system are the camera, the image processing hardware, and the software that contains the image processing functions. In the introduction, the basics about vision systems and cameras are given, and the types of vision systems and the types of configurations of robots and cameras are presented. Furthermore, the equipment used for image processing and scene retrieval is presented. A Jetson Nano development platform with OpenCV installed for image processing is used for processing of images, and two IMX219-120 MIPI cameras are used for scene acquisition. In order to be able to work on the system, it is necessary to connect the peripherals that enable the connections of the Jetson platform. In order to protect the equipment, a case was made for the Jetson platform and a calibration box was made that was separated from the processing unit for the safety and robustness of the segmentation. Next, an experiment was conducted to select the background color of the scene to determine the most suitable color for good contrast with the tool. Choosing the optimal background color enables robust segmentation of the results, which is essential for good calibration. In order to get good information from the camera, it needs to be calibrated, so the camera calibration was briefly explained and carried out. The method of segmentation of the tool from the background with all the actions of image processing included is shown. The method of detecting the tooltip of the robot and calculating the motion of the robot so that the tool is located at the intersection of the optical axes of the cameras are defined. Algorithms used to calculate the tool calibration, i.e. the transformation of the coordinate system of the tip of the robot tool in relation to the robot flange, which is the goal of the calibration, are presented. The method of communication between the robot and the vision system is also shown, where the robot is the client and the vision system is the server in the socket communication architecture. The programming code on the robot is given for the tool calibration process using the vision system. At the end, the results of the calibration using the vision system are presented and a comparison of this method of robot tool calibration versus the classic method of robot tool calibration is given.

• Keywords: kalibracija alata robota; vizijski sustav; kamere; komunikacija
• Keywords (english): calibration of robot tools; vision system; cameras; communication
• Language: croatian
• URN:NBN: urn:nbn:hr:235:699806
• Study programme: Title: Mechanical Engineering; specializations in: Design, Process and Energy Engineering, Production Engineering, Engineering Modelling and Computer Simulation, Marine Engineering, Industrial Engineering and Management, Materials Engineering, Mechanics and Robotics, Autonomous Systems, Mechatronics and Robotics; Course: Mechatronics and Robotics; Study programme type: university; Study level: graduate; Academic / professional title: sveučilišni/a magistar/magistra inženjer/inženjerka strojarstva (sveučilišni/a magistar/magistra inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2025-02-03 10:31:48