Accurate estimation of a body’s position is crucial in various fields and often involves using Motion Capture systems. Recently, there has been significant development in this technology, particularly in camera-based systems, thanks to advancements in computer vision. However, although these techniques provide accurate pose estimation results, they are expensive and require significant computational resources. As a cost-effective alternative, fiducial markers like ArUco have gained popularity due to their simplicity, low computational requirements, and adaptability. Nevertheless, there is limited experimental data in the literature regarding the accuracy of Motion Capture systems based on these markers. This study therefore aims to fill the identified gap by conducting an experimental analysis of the performance of a Motion Capture system based on ArUco markers, comparing it with a modern commercial system developed by Optitrack. The ArUco-based setup in this study is initially implemented using an ELP webcam and later a ZED 2i stereo vision camera. The research evaluates pose estimation errors in both static and dynamic scenarios, analyzing different regions within the camera’s working area and also exploring the relationship between pose estimation error and the use of an increasing number of ArUco markers.
L’accurata stima della posizione di un corpo é cruciale in vari campi e spesso coinvolge l’uso di sistemi di Motion Capture. Recentemente, si é assistito ad un forte sviluppo di questa tecnologia, in particolare nei sistemi basati su telecamere, grazie ai progressi nella visione artificiale, ma, sebbene queste tecniche forniscano risultati accurati nella stima della posa, comportano costi elevati e richiedono significative risorse computazionali. Come alternativa economica, i fiducial-marker come ArUco hanno guadagnato popolaritá per la loro semplicitá, le basse esigenze computazionali e l’adattabilitá, tuttavia, in letteratura sono presenti dati sperimentali limitati sulla precisione dei sistemi di Motion Capture basati su di essi. Questo studio mira quindi a colmare la lacuna evidenziata, conducendo un’analisi sperimentale delle prestazioni di un sistema di Motion Capture basato su marker ArUco, confrontandolo con un moderno sistema commerciale sviluppato da Optitrack. L’apparato basato su ArUco in questo studio é realizzato utilizzando inizialmente una webcam ELP e successivamente una telecamera con visione stereo ZED 2i. La ricerca valuta gli errori di stima della posa in scenari sia statici che dinamici, analizzando diverse regioni della zona di lavoro della telecamera, esplorando anche la relazione tra l’errore di stima della posa e l’utilizzo di un numero crescente di marker ArUco.
Analisi dei metodi di stima della posa basati su marker ArUco e sistema di visione OptiTrack
CORRATA, DAVIDE
2023/2024
Abstract
Accurate estimation of a body’s position is crucial in various fields and often involves using Motion Capture systems. Recently, there has been significant development in this technology, particularly in camera-based systems, thanks to advancements in computer vision. However, although these techniques provide accurate pose estimation results, they are expensive and require significant computational resources. As a cost-effective alternative, fiducial markers like ArUco have gained popularity due to their simplicity, low computational requirements, and adaptability. Nevertheless, there is limited experimental data in the literature regarding the accuracy of Motion Capture systems based on these markers. This study therefore aims to fill the identified gap by conducting an experimental analysis of the performance of a Motion Capture system based on ArUco markers, comparing it with a modern commercial system developed by Optitrack. The ArUco-based setup in this study is initially implemented using an ELP webcam and later a ZED 2i stereo vision camera. The research evaluates pose estimation errors in both static and dynamic scenarios, analyzing different regions within the camera’s working area and also exploring the relationship between pose estimation error and the use of an increasing number of ArUco markers.File | Dimensione | Formato | |
---|---|---|---|
Corrata_Davide.pdf
accesso aperto
Dimensione
6.6 MB
Formato
Adobe PDF
|
6.6 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/20.500.12608/74731