摘要
研究了基于深度学习算法的可穿戴设备手势识别系统设计。以智能手表为例,利用可穿戴设备的加速度传感器进行数据采集,通过在执行不同手势时三轴加速度数值的不同以及执行时长的差异,使用TensorFlow以及CNN神经网络实现手势识别。实验结果表明该系统对交互手势的识别有着良好稳定的识别率,准确率达到97%以上,并且系统的各项性能指标都较好。所提出的手势识别系统还可以应用到数字签名、个人安全或用户标识等其他领域。
This paper studies the design of wearable device gesture recognition system based on deep learning algorithm.The wearable device′s acceleration sensor is used for data acquisition,and the system using TensorFlow and CNN to recognize the gesture models based on the difference of the execution time and the acceleration values.The experimental results show that the system has a good and stable recognition for interactive gestures,the accuracy rate is 97% and the performance of the system is satisfactory.The gesture recognition system proposed in this paper can also be applied to other fields such as digital signature,personal security and user identification.
作者
惠丹
Hui Dan(Xi′an University of Technology,Xi′an 710048,China)
出处
《信息技术与网络安全》
2019年第9期30-33,共4页
Information Technology and Network Security
基金
陕西省教育厅科研计划项目(17JK0521)
西安理工大学人文社会科学研究专项项目(2018RY003)
关键词
深度学习算法
手势识别
系统设计
deep learning algorithm
gesture recognition
system design
