Elder Fall Detection System Based on Two-level SVM Algorithm
Shaolong Peng, Bokang Yang, Qiang Peng · 2022 37th Youth Academic Annual Conference of Chinese Association of Automation (YAC) · 2022
With the acceleration of population aging, the health and life safety of the elderly has gradually become a common problem facing the whole society. In this paper, an elderly fall detection system based on embedded microcontroller system and SVM algorithm is studied. In terms of hardware design, acceleration sensor, communication module and positioning module are integrated. In the algorithm part, THE SVM algorithm and threshold algorithm are integrated. The threshold algorithm is used for preliminary judgment of the action, and the SVM algorithm is used for accurate judgment, which can effectively distinguish the fall and daily behavior, reduce misjudgment and improve the accuracy of the fall judgment. Can effectively protect the life safety of the elderly