Markerless technology makes it possible to acquire kinematic and dynamic parameters without the use of stereophotogrammetric systems and without the need for dedicated laboratory environments. The use of commercial video cameras, combined with computer vision and artificial intelligence algorithms, allows the estimation of the subject’s three-dimensional pose and the calculation of fundamental biomechanical quantities such as joint angles, forces, and moments. While offering advantages related to ease of use, low cost, and the possibility of application in real-life contexts, markerless techniques still present critical issues in terms of precision and reliability, particularly when employed to estimate inverse dynamics. The aim of this thesis was to assess the validity and robustness of different markerless approaches. To this end, two case studies were conducted: three subjects performing repeated squats, used as a standardized movement for method comparison; two athletes from the Italian National Rugby Academy, assessed at BBSoF for the analysis of critical flexion angles of the knee and ankle. The collected data were processed through dedicated software and custom Matlab scripts, applying statistical methods to compare the different approaches. The analysis not only highlighted the strengths but also the main limitations of markerless technologies in the contexts considered, thus providing a clear picture of their actual potential in both clinical and sports applications.
La tecnologia markerless consente di acquisire parametri cinematici e dinamici senza l’impiego di sistemi stereofotogrammetrici e senza la necessità di ambienti di laboratorio dedicati. L’utilizzo di videocamere commerciali e di algoritmi di visione artificiale e intelligenza artificiale permette infatti di stimare la posa tridimensionale del soggetto e di calcolare grandezze biomeccaniche fondamentali come angoli articolari, forze e momenti. Pur offrendo vantaggi legati alla semplicità d’uso, ai costi contenuti e alla possibilità di applicarle anche in contesti di vita reale, l’applicazione delle tecniche markerless presenta ancora delle criticità in termini di precisione e affidabilità, in particolare quando vengono impiegate per stimare la dinamica inversa. Il presente lavoro di tesi ha avuto come obiettivo quello di verificare la validità e la robustezza di diversi approcci markerless. A tal fine sono stati analizzati: - tre soggetti impegnati nell’esecuzione di squat ripetuti, utilizzati come gesto standardizzato per il confronto tra metodi; - due atleti dell’Academy della Nazionale Italiana di Rugby, valutati presso l’azienda BBSoF per lo studio degli angoli critici di flessione di ginocchio e caviglia. I dati ottenuti sono stati elaborati attraverso software dedicati e script realizzati in Matlab, applicando metodi statistici per confrontare i diversi approcci. L’analisi, non solo ha permesso di evidenziare i punti di forza, ma anche le principali criticità delle tecnologie markerless nei contesti considerati, fornendo così un quadro chiaro delle loro effettive potenzialità sia in ambito clinico che sportivo.
Stima della dinamica inversa da metodologie di Motion Capture Markeless: Validazione e Valutazione dell’Affidabilità
MARIUZ, FEDERICO
2024/2025
Abstract
Markerless technology makes it possible to acquire kinematic and dynamic parameters without the use of stereophotogrammetric systems and without the need for dedicated laboratory environments. The use of commercial video cameras, combined with computer vision and artificial intelligence algorithms, allows the estimation of the subject’s three-dimensional pose and the calculation of fundamental biomechanical quantities such as joint angles, forces, and moments. While offering advantages related to ease of use, low cost, and the possibility of application in real-life contexts, markerless techniques still present critical issues in terms of precision and reliability, particularly when employed to estimate inverse dynamics. The aim of this thesis was to assess the validity and robustness of different markerless approaches. To this end, two case studies were conducted: three subjects performing repeated squats, used as a standardized movement for method comparison; two athletes from the Italian National Rugby Academy, assessed at BBSoF for the analysis of critical flexion angles of the knee and ankle. The collected data were processed through dedicated software and custom Matlab scripts, applying statistical methods to compare the different approaches. The analysis not only highlighted the strengths but also the main limitations of markerless technologies in the contexts considered, thus providing a clear picture of their actual potential in both clinical and sports applications.| File | Dimensione | Formato | |
|---|---|---|---|
|
Mariuz_Federico.pdf
accesso aperto
Dimensione
1 MB
Formato
Adobe PDF
|
1 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/91732