测试技术学报2025，Vol.39Issue(5)：540-547,8.DOI:10.62756/csjs.1671-7449.2025088
基于GO-SWCNT@PDMS压力传感器的手势识别
Gesture Recognition Based on GO-SWCNT@PDMS Pressure Sensors
摘要
Abstract
Gestures,as the earliest and still widely used communication method among humans,occupy an important position in posture language.To solve the problems of environmental impact and high cost in traditional gesture recognition based on computer vision and radio frequency radar,a flexible pressure sensor was encapsulated at the finger joint,and gesture recognition was performed by monitoring the resistance changes caused by finger bending,eliminating the influence of the surrounding environment on gesture recognition and solving the limitation of high cost in traditional radio frequency radar.The sensitivity of the prepared flexible pressure sensor can reach up to 13.57 kPa-1,with fast response time and recovery time of 63 ms and 84 ms respectively,and excellent stability(>4 000 cycles).The gesture recognition system used the TCN-FCN parallel deep learning algorithm to train and test 9 gestures with strong spatiotemporal coupling characteristics,and the final recognition average accuracy remains stable at over 95%.This study provides a convenient and efficient gesture recognition method for fields such as augmented reality and human-computer interaction.关键词
压力传感器/褶皱结构/深度学习/手势识别Key words
pressure sensor/wrinkled structure/deep learning/gesture recognition分类
信息技术与安全科学引用本文复制引用
樊磊,闫天昊,王兆欣,杨涛,谭秋林..基于GO-SWCNT@PDMS压力传感器的手势识别[J].测试技术学报,2025,39(5):540-547,8.基金项目
国家重点研发计划资助项目(2023YFB3209100) （2023YFB3209100）
山西省基础研究计划青年科学研究项目(202303021222087) （202303021222087）
国家自然科学基金区域创新发展联合基金(U24A20136) （U24A20136）
