Ispitivanje učinkovitosti procjene položaja tijela korištenjem biblioteke OpenPose

undergraduate thesis

• Title (english): Testing of body position estimation efficiency using the OpenPose library
• Author: Lucija Marinčić
• Mentor: Marko Horvat (mentor)
• Committee member: Marko Horvat (predsjednik povjerenstva)
• Committee member: Marina Bagić Babac (član povjerenstva)
• Committee member: Mihael Kovač (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-05, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Jedan od postupaka računalnog vida je detekcija položaja tijela. OpenPose je biblioteka koja
položaj tijela procjenjuje identificiranjem ključnih točaka na tijelu. Brojni su izazovi s kojima se OpenPose susreće pa je korisno odrediti u kojim uvjetima on daje dovoljno točne
rezultate kako bi se znalo kada procjenu položaja tijela dobivenu pomoću biblioteke OpenPose treba uzeti u obzir.
U ovom radu korišten je model koji detektira 18 ključnih točaka. Izrađeno je 5 skupova
fotografija. Oni opisuju različite uvjete koji bi mogli predstavljati izazov za OpenPose. Korištenjem OpenPose-a provedena je procjena položaja tijela na fotografijama iz navedenih
skupova. Za dobivene procjene određuju se matrice konfuzije te računaju mjere deskriptivne
statistike.

Abstract (english)

One of the computer vision techniques is body pose estimation. OpenPose is a library that
estimates body position by identifying key points on the body. There are many challenges
that OpenPose faces, so it is useful to determine under which conditions the results it gives
are accurate enough, in order to know when the body position estimation obtained using the
Open-Pose library should be considered.
In this paper, an 18 keypoints model was used. 5 sets of photos were taken. They describe
various conditions that could pose a challenge for OpenPose. Using OpenPose, an assessment of the body position in the photographs from the mentioned sets was performed. Confusion matrices have been determined and the descriptive statistics measures have been calculated

• Keywords: biblioteka OpenPose; prepoznavanje položaja tijela; deskriptivna statistika; ključne točke
• Keywords (english): OpenPose library; body pose detection; descriptive statistics; keypoints
• Language: croatian
• URN:NBN: urn:nbn:hr:168:811298
• 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: 2025-02-25 13:36:43