Praćenje položaja ljudskog tijela u pokretu korištenjem OpenPose biblioteke

undergraduate thesis

• Title (english): Tracking of a Human Body in Motion using OpenPose Library
• Author: Daniel Rey Šparemblek
• Mentor: Tomislav Petković (mentor)
• Committee member: Tomislav Petković (predsjednik povjerenstva)
• Committee member: Kristijan Bartol (član povjerenstva)
• Committee member: David Bojanić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-09-02, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Problem precizne digitalizacije kretanja ljudskog tijela nije popularan samo u industriji vizualnih efekata, već i u biomehanici, robotici i za općenito razumijevanje podsvjesnog ljudskog ponašanja. Ovaj se rad fokusira na uspoređivanju dva različita sustava: Motivea - profesionalnog 3D kinematičkog sustava za analizu ljudskog pokreta koji je razvila tvrtka OptiTrack i OpenPose-a - sustava koji u stvarnom vremenu za više osoba nalazi ključne točke ljudskog tijela, lica, ruku i stopala. Usporedba je napravljena istodobnim snimanjem snimaka za Motive i za OpenPose te, nakon obrade, usporedbom točnosti s kojom svaki sustav nalazi ključne točke (lakta i ručnog zgloba) na ljudskom tijelu. Dobiveni rezultati pokazuju da trenutna implementacija OpenPose-a nije dovoljno dobra da zamjeni profesionalne sustave za 3D kinematičku analizu ljudskog pokreta. U slučajevima kada je potrebna izrazita točnost i preciznost preporučuje se koristiti OptiTrackov sustav. Međutim, zbog jednostavnosti korištenja, cijene i dostupnosti, OpenPose može biti prihvatljiva alternativa u primjenama gdje 3D rekonstrukcija ne treba biti savršena i gdje točnost nije ključna.

Abstract (english)

The problem of accurately digitizing human body movement is not only popular in the visual effects industry but also in biomechanics, robotics, and in general for the understanding of subconscious human behavior. This bachelor thesis focuses on comparing two different systems: Motive – a professional motion capture system developed by OptiTrack and OpenPose – a real-time multi-person pose tracking system that jointly detects the human body, hand, facial, and foot keypoints in images. The comparison was done by simultaneously recording footage for both Motive and OpenPose and, after processing, comparing the key point tracking accuracy of each system for elbow and wrist. The obtained results show that the current implementation of OpenPose is not good enough to replace professional motion capture systems. In cases where extreme accuracy and precision are required, it is recommended to use OptiTrack's system. Although, due to OpenPose’s ease of use, cost, and availability, it may be an acceptable alternative in applications where 3D reconstruction does not need to be perfect and where accuracy is not critical.

• Keywords: OpenPose; digitalizacija kretanja; OptiTrack Motive; BODY_25 model
• Keywords (english): OpenPose; motion capture; OptiTrack Motive; BODY_25 model
• Language: croatian
• URN:NBN: urn:nbn:hr:168:017746
• 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: 2023-01-14 10:04:48