Prepoznavanje pokreta ljudskog tijela fuzijom podataka ultraširokopojasnih lokalizacijskih senzora i inercijskih senzora.

doctoral thesis

• Title (english): Human body motion recognition using ultra-wideband localization sensors and inertial sensors data fusion.
• Author: Hrvoje Mihadlinec; MBZ: -330578
• Mentor: Hrvoje Džapo (mentor)
• Committee member: Hrvoje Džapo (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Electronic Systems and Information Processing); Zagreb
• Defense date and country: 2020, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering
• Universal decimal classification (UDC): 621.3; APPLIED SCIENCES. MEDICINE. TECHNOLOGY; Electrical engineering

Abstract

Praćenje pokreta ljudskog tijela (engl. Human Motion Tracking, HMT) nalazi primjenu u brojnim područjima, poput medicine, rehabilitacije, sporta, interaktivnih sustava itd. HMT sustavi prikupljaju kinetičke i kinematičke informacije o kretanju tijela ili njegovih dijelova u svrhu kvantitativne analize parametara pokreta. Tradicionalno korišteni HMT sustavi temelje se na računalnom vidu i inercijskim (IMU) senzorima, a u posljednjih nekoliko godina pojavom tržišno dostupnih ultraširokopojasnih (UWB) tehnologija otvorile su se mogućnosti za realizaciju HMT sustava boljih značajki. Istraživanje su motivirali uočeni nedostaci postojećih HMT metoda temeljenih na fuziji UWB tehnologije i IMU senzora. Razvijene su odgovarajuće metode kompenzacije mjernih pogrešaka i pokazano je da se predloženim pristupom postižu bolji rezultati u odnosu na postojeće metode. Razvijeni su specijalizirani nosivi bežični umreženi senzorski čvorovi te je razrađen model i algoritam za određivanje položaja i pokreta ljudskog tijela primjenom fuzije senzorskih podataka iz UWB i IMU senzora. Postignuti su sljedeći znanstveni doprinosi: • model za određivanje položaja i pokreta ljudskog tijela iz podataka dobivenih nosivim ultraširokopojasnim lokalizacijskim senzorima i inercijskim senzorima, • arhitektura sklopovske i programske izvedbe mjernog postava niske potrošnje za određivanje položaja i pokreta ljudskog tijela • algoritam za određivanje položaja i prepoznavanje pokreta ljudskog tijela primjenom fuzije senzorskih podataka i metoda strojnog učenja.

Abstract (english)

Human Motion Tracking (HMT) is used in various fields, such as medicine, rehabilitation, sports, interactive systems, etc. HMT systems collect kinetic and kinematic data about movement of the body or its individual parts for purpose of quantitative analysis of movement parameters. Traditionally, HMT systems were based on computer vision techniques and inertial (IMU) sensors, while in recent years commercial availability of ultra-wideband (UWB) technologies opened new opportunities for realization of HMT systems with better characteristics. The research was motivated by the perceived shortcomings of existing HMT approaches based on UWB and IMU sensor fusion. Methods for measurement error compensation were developed and it was shown that the proposed approach can achieve better performance compared to the existing methods. Specialized wearable wireless sensor nodes were developed, and the model and algorithm for determination of the position and motion of the human body, based on data fusion between UWB and inertial sensors, were developed. The following scientific contributions were achieved: • model for determination of the human body position and motion from the data obtained by wearable UWB localization sensors and inertial sensors • hardware and software architecture of low-power measurement system for determination of the human body position and motion • algorithm for determination of the human body position and motion recognition by using sensor data fusion and machine learning methods

• Keywords: praćenje pokreta ljudskog tijela; ultraširokopojasna (UWB) tehnologija; inercijski senzori; senzorska fuzija; prepoznavanje pokreta; strojno učenje
• Keywords (english): human motion tracking; ultra-wideband (UWB) technology; inertial sensors; sensor fusion; motion recognition; machine learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:573490
• Promotion: 2020
• Study programme: Title: Doctoral study programme "Electrical Engineering and Computing"; Study programme type: university; Study level: postgraduate; Academic / professional title: doktor/doktorica znanosti, po-dručje tehničkih znanosti (doktor/doktorica znanosti, po-dručje tehničkih znanosti)
• Type of resource: Text
• Extent: 178 str.
• File origin: Born digital
• Access conditions: Open access
• Created on: 2020-09-11 11:43:44