Detekcija izvođenja rehabilitacijskih vježbi ramena pomoću robota Pepper

undergraduate thesis

• Title (english): Detection of performing shoulder rehabilitation exercises using the Pepper robot
• Author: Lovro Furač
• Mentor: Zdenko Kovačić (mentor)
• Committee member: Zdenko Kovačić (član povjerenstva)
• Committee member: Dean Martinović (član povjerenstva)
• Committee member: Ivo Vatavuk (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-07, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Korištenje humanoidnih robota pri interakciji s pacijentima dodaje dimenziju društvenosti i omogućava lakše i ugodnije obavljanje vježbi rehabilitacije. Prilikom detekcije i klasifikacije položaja tijela čovjeka, koristi se algoritam linearne regresije strojnog učenja, koji se pokazao optimalnim za ovu primjenu. Moguće je pratiti broj ponavljanja određene vježbe i vrijeme zadržavanja u određenom položaju, što je omogućava provedbu treninga i istezanja. Analiziran je pristup dohvata videozapisa Pepper-ove kamere na dva načina i navedene su prednosti i mane u obje situacije. Za povezivanje s ostatkom mreže koristi se ROS. Demonstrirano je dodavanje novih vježbi u program i uspoređeni su rezultati detekcije u različitim uvjetima.

Abstract (english)

Using humanoid robots in interaction with patients adds a dimension of sociability and ensures unchallenging and pleasant experience doing rehabilitation workouts. A machine learning algorythm of linear regression is used during detection and body pose classification, which has shown as an optimal algorythm for this use. It is possible to track repetitions of a certain exercise and time spent in a certain pose, which enables implementation of working out and stretching.
Two means of accessing Pepper robot`s video feed are analyzed, with positive and negative sides stated in both situations. Connecting with the rest of the network is implemented in ROS.
Adding new exercises into the program is demonstrated while comparing results of detection in different conditions.

• Keywords: strojno učenje; robot Pepper; ROS; računalni vid; klasifikacija položaja tijela
• Keywords (english): machine learning; Pepper robot; ROS; computer vision; body pose classification
• Language: croatian
• URN:NBN: urn:nbn:hr:168:836909
• 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: 2023-01-15 10:08:59