Računalni model za procjenu položaja tijela promatrane osobe

undergraduate thesis

• Title (english): Computational model for estimating the position of the body of the observed person
• Author: Korina Belokleić
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2022-09-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Razni alati stvaraju se kako bi se pojednostavili i olakšali pojedini aspekti života. Umjetna inteligencija se već koristi za procjenu položaja preko slike i videa u raznim područjima kao što je prepoznavanje sigurnosnim kamerama pa sve do praćenja pokreta u npr. video igrama. To je ostvareno procjenom položaja zglobova i dijelova/točaka ljudskog tijela dane slike koja je prethodno implementirana kroz MediaPipe biblioteku. Na taj način je moguće povećati brzinu i točnost svakodnevnih radnji što može imati primjenu u poboljšanju npr. sportskih performansi. Ovdje se procjena položaja ostvaruje na primjeru računanja kuta u laktu ruke što je implementirano kroz programski jezik Python gdje su uvedene, uz MediaPipe, biblioteke kao što su NumPy i OpenCV.

Abstract (english)

Various technologies are being developed to streamline and simplify certain aspects of life. As a result, artificial intelligence is already being used to assess the position via photos and videos in a variety of contexts, from the identification of security cameras to the tracking of movements, for instance, in video games. In this study, a computational model was developed for estimating the postion of the body an observed person by a computer camera in real time. This was accomplished by using the MediaPipe library to estimate the position of joints and human body landmarks in a given image. By doing this, routine tasks can be performed faster and with greater accuarcy, which can help, for instance, with athletic and sporting performance. Here, pose estimation is achieved on the basis of calculating the angle at the elbow, which is implemented through the Python programming language, which also includes the NumPy, OpenCV and MediaPipe libraries.

• Keywords: umjetna inteligencija; procjena položaja tijela; računalni model; Python; MediaPipe
• Keywords (english): artificial intelligence; pose estimation; computational model; Python; MediaPipe
• Language: croatian
• URN:NBN: urn:nbn:hr:235:057537
• 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: 2022-09-21 20:32:35