Razvoj sustava za prilagođavanje ponašanja robota temeljen na analizi ljudskih kretnji

undergraduate thesis

• Title (english): Development of a system for adapting robot behavior based on the analysis of human movements
• Author: Matija Pongračić
• Mentor: Marko Švaco (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2024-09-24, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Napretkom u robotici, tradicionalni industrijski roboti sve više ustupaju mjesto kolaborativnim robotima, tzv. „kobotima“, koji su namijenjeni da dijele radni prostor s ljudima i zajednički izvršavaju zadatke.
U završnom radu razvijen je sustav koji kombinira tehnologije prepoznavanja ljudskih gesti i procjene namjera kretanja čovjeka s ciljem prilagođavanja ponašanja robota kako bi se osigurala robusna interakcija i povećala sigurnost u radnom okruženju.
Provedena je analiza i evaluacija javno dostupnih biblioteka za procjenu ljudske poze i skeletonizaciju ljudskog tijela, nakon čega su koncipirani algoritmi za implementaciju u sustavu te je dodatno razvijen i evaluiran vlastiti model za prepoznavanje gesti. Konačno, sustav je implementiran integracijom na robotu u Regionalnom centru izvrsnosti za robotske tehnologije (CRTA-i).

Abstract (english)

With advancements in robotics, traditional industrial robots are increasingly being replaced by collaborative robots, also known as „cobots”, designed to share workspace with humans and perform tasks collaboratively.
In the Bachelor's thesis, a system has been developed that combines human gesture recognition technologies and movement intention estimation to adjust the robot’s behavior, ensuring robust interaction and enhanced safety in the working environment.
An analysis and evaluation of open-source libraries for human pose estimation and body skeletonization has been conducted, followed by the conceptualization of algorithms for implementation in the system and, additionally, the development and evaluation of a custom model for gesture recognition. Finally, the system has been implemented through integration with the robot at the Regional Center of Excellence for Robotic Technology (CRTA).

• Keywords: kobot; procjena ljudske poze; namjere kretanja; geste; algoritam
• Keywords (english): cobot; human pose estimation; movement intentions; gestures; algorithm
• Language: croatian
• URN:NBN: urn:nbn:hr:235:903849
• Study programme: Title: Mechatronics and Robotics; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka mehatronike i robotike (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka mehatronike i robotike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-09-19 13:09:30