手势
可穿戴计算机
计算机科学
手势识别
压力传感器
人工智能
材料科学
信号(编程语言)
声学
计算机视觉
工程类
机械工程
嵌入式系统
物理
程序设计语言
作者
Feilu Wang,Wangyong Zhang,Yang Song,Xiuli Jiang,Niuping Sun
出处
期刊:ACS applied electronic materials
[American Chemical Society]
日期:2023-11-28
卷期号:5 (12): 6704-6715
被引量:7
标识
DOI:10.1021/acsaelm.3c01199
摘要
Flexible pressure sensors are important for various fields including human–machine interaction, motion detection, and gesture recognition. In this study, a piezoresistive pressure sensor is developed using a composite material of carbon nanotubes and a polyurethane sponge. The sensor is fabricated through a dipping-drying method, which enables the carbon nanotubes (CNTs) to adhere to the skeleton of the polyurethane sponge (PUS). The sensor obtained displays superior features: excellent sensitivity (2.7% kPa–1), prompt response (response/recovery time of 60/100 ms), and remarkable long-term stability demonstrated by a consistent response signal during loading/unloading cycles with the range of 0–100 kPa at 0.1 Hz for a period of 18,000 s. In addition, the sensor was placed on different parts of the human body to detect human motion signals. It has been demonstrated that the sensor can effectively capture these diverse signals to distinguish between different motion states. Additionally, the sensor can accurately convey the Morse code of the 26 letters of the alphabet and the 10 Arabic numerals through regular pressing. Finally, a sensory glove was created using the sensors, which is used to express the gestures of Arabic numerals 0–9. A deep-learning algorithm based on the Inception Network has achieved a high-accuracy (99.5%) gesture recognition for 10 gestures. This work offers a cost-effective and simple way to produce flexible pressure sensors that can be employed in various applications, including human motion detection, wearable devices, gesture recognition, and human-machine interaction.
科研通智能强力驱动
Strongly Powered by AbleSci AI