Fall Detection and Protection System Based on Characteristic Areas Algorithm

• Published in: Intelligent Robotics and Applications Vol. 13013; pp. 170 - 178
• Main Authors: Du, Jun, Shi, Jingyi, Wei, Xiaodong, Xu, Ying, Chen, Diansheng
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Airbag devices; Characteristic areas; Fall detection; Fall protection
• ISBN: 9783030890940; 3030890945
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-89095-7_17

Hip fracture caused by falls and its complications is one of the greatest threats to disability and death of the elderly. To reduce physical damage from falls in the elderly, the current solution to achieve effective protection is detecting fall trends and turning on protective devices. However, the existing products have the problems of low accuracy and poor real-time. In this paper, a high accuracy and high real-time human fall detection and protection system based on characteristic areas algorithm is designed, which can detect the trend of falls within 400 ms after the human body begins to fall and is filled with the airbag in the 400 ms later, realizing effective protection of the human hip. The system got 95.33% accuracy, with an average airbag opening time of 70 ms.