摘要
提出一种基于10.525GHz微波传感器和支持向量机(Support Vector Machine,SVM)算法相结合的手势识别方法。微波传感器输出的多普勒信号经过放大和ADS1256采样后,发送给BCM2837B0SoC,再利用快速傅里叶变换(FFT)提取手势特征,最后借助SVM进行分类。实验结果表明,本方法能够分别以87%、84%和83%的平均准确率识别2种、4种和6种手势,具有较强的扩展性和较高准确率。
In the paper,agesture recognition method based on 10.525 GHz microwave sensor and Support Vector Machine(SVM)is proposed.The Doppler signal output from the microwave sensor is amplified and sampled by ADS1256,and then sent to BCM2837B0 SoC.The Fast Fourier Transform(FFT)is used to extract gesture features.Finally,the classification is performed by SVM.The experiment results show that the proposed method can identify 2,4 and 6 gestures with average accuracy of 87%,84% and 83%respectively,with strong scalability and high accuracy.
作者
王拥军
马维华
Wang Yongjun;Ma Weihua(College of Computer Science andTechnology,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《单片机与嵌入式系统应用》
2020年第2期57-60,共4页
Microcontrollers & Embedded Systems