| Autor: |
Zhao, Haochen, Esfahani, Mohammad Iman Mokhlespour, Krishnan, Suresh, Abari, Omid |
| Zdroj: |
IEEE Sensors Journal; November 2024, Vol. 24 Issue: 21 p35610-35620, 11p |
| Abstrakt: |
Continuous back posture monitoring and correction can help to prevent back pains associated with improper back postures. However, existing solutions are expensive, use wearable sensors which usually require regular maintenance, or use cameras which have privacy issues. We introduce Di-Angle, a low-cost, battery-free, and reusable sensor capable of monitoring harmful back angles with high accuracy. Our novel design is using two off-the-shelf passive backscattering dipole radio frequency identification (RFID) tags, placed in a novel crossed configuration. This placement prevents interference between tags in both near field and far field while allowing differential reading which helps mitigate wireless environment variance. We have evaluated our system in both simulation and real-world experiments. Our simulation results validate the robustness with crossed configuration against self-interference and wireless channel variation. Our real-world bad back posture monitoring case study showed that our system achieves an accuracy of 97.0% in identifying harmful back postures of excessive torso angle with a variety of subjects and activities. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
