In movement analysis, kinematic data are of fundamental importance. Although measurements taken using markers and cameras are considered the gold standard, their use outside operational environments is almost impossible. IMU sensors based on MEMS technology offer a solution to these issues due to their high precision and compactness. This thesis presents the internship experience carried out at the company 221e, with the aim of developing a program for the recognition of bodyweight exercises, identifying the optimal sensor placement. The MEMS sensors used in this study include an accelerometer, gyroscope, and magnetometer, which will be described in detail in the first chapter. The exercises chosen for the study are divided into static exercises, such as lunges and squats, and dynamic exercises, such as jumping jacks, burpees, and jumps. Depending on the exercise, the sensor was placed in different areas of the body: knee, ankle, wrist, and pelvis. The data collected were processed using Matlab software to improve measurement quality and used to create an exercise recognition program. The results were very positive, with a recognition efficiency greater than 90%. Additionally, it was found that the best sensor placement is the pelvis, which allows for more precise data acquisition and is less affected by undesired micro-movements.
Nell'analisi del movimento, i dati cinematici sono di fondamentale importanza. Sebbene le misurazioni effettuate attraverso marcatori e telecamere siano considerate il gold standard, il loro utilizzo al di fuori degli ambienti operativi è quasi impossibile. I sensori IMU basati sulla tecnologia MEMS offrono una soluzione a queste problematiche grazie alla loro elevata precisione e compattezza. In questo elaborato verrà presentata l’esperienza di tirocinio svolta presso l'azienda 221e, con l'obiettivo di sviluppare un programma per il riconoscimento di esercizi a corpo libero, identificando il miglior posizionamento del sensore. I sensori MEMS utilizzati in questo studio includono accelerometro, giroscopio e magnetometro, che verranno descritti dettagliatamente nel primo capitolo. Gli esercizi scelti per lo studio si suddividono in statici, come affondi e squat, e dinamici, come jumping jack, burpees e balzi. In base all’esercizio, il sensore è stato posizionato in diverse aree del corpo: ginocchio, caviglia, polso e bacino. I dati raccolti sono stati elaborati con il software Matlab per migliorare la qualità delle misurazioni e utilizzati per creare un programma di riconoscimento degli esercizi. I risultati sono stati molto positivi, con un’efficienza di riconoscimento superiore al 90%. Inoltre, si è riscontrato che il miglior posizionamento del sensore è il bacino, che consente un’acquisizione dei dati più precisa e meno influenzata da micro movimenti indesiderati.
Utilizzo di sensori indossabili IMU per il riconoscimento e caratterizzazione di esercizi a corpo libero
CALORE, MARCO
2023/2024
Abstract
In movement analysis, kinematic data are of fundamental importance. Although measurements taken using markers and cameras are considered the gold standard, their use outside operational environments is almost impossible. IMU sensors based on MEMS technology offer a solution to these issues due to their high precision and compactness. This thesis presents the internship experience carried out at the company 221e, with the aim of developing a program for the recognition of bodyweight exercises, identifying the optimal sensor placement. The MEMS sensors used in this study include an accelerometer, gyroscope, and magnetometer, which will be described in detail in the first chapter. The exercises chosen for the study are divided into static exercises, such as lunges and squats, and dynamic exercises, such as jumping jacks, burpees, and jumps. Depending on the exercise, the sensor was placed in different areas of the body: knee, ankle, wrist, and pelvis. The data collected were processed using Matlab software to improve measurement quality and used to create an exercise recognition program. The results were very positive, with a recognition efficiency greater than 90%. Additionally, it was found that the best sensor placement is the pelvis, which allows for more precise data acquisition and is less affected by undesired micro-movements.File | Dimensione | Formato | |
---|---|---|---|
Calore_Marco.pdf
accesso aperto
Dimensione
1.24 MB
Formato
Adobe PDF
|
1.24 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/20.500.12608/67358