摘要
为了提高基于加速度传感器手指手势识别算法的性能,研究了一种双带通滤波优化随机森林算法模型,应用于手指手势识别。该方法利用MPU6050六轴传感器和HC—06蓝牙模块作为数据采集系统,采集食指的4个日常动作作为数据集。采用双带通滤波算法对数据集进行预处理并利用随机森林分类器对处理后的数据进行手势的分类预测。实验结果表明:该方法获得了98.5%的手指手势识别率,有效地识别了4种手指动作,具有良好的稳定性和准确性。
In order to improve the performance of finger gesture recognition algorithm based on acceleration sensor,a double band-pass filtering optimized random forest algorithm model is studied and applied to finger gesture recognition.The method uses MPU6050 six-axis sensor and HC—06 Bluetooth module as the data acquisition system,and collects four daily movements of the index finger as the dataset.The dataset is preprocessed by double band-pass filtering algorithm,and the gesture is classified and predicted by random forest classifier.Experimental results show that this method achieves 98.5%finger gesture recognition rate,effectively recognizes four finger movements,and has good stability and accuracy.
作者
程傲霜
王强
CHENG Aoshuang;WANG Qiang(School of Information Science Technology,Nantong University,Nantong 226000,China)
出处
《传感器与微系统》
CSCD
北大核心
2023年第8期165-168,共4页
Transducer and Microsystem Technologies
基金
江苏省高等学校自然科学研究重大项目(19KJ320004)。
关键词
手指手势识别
双带通滤波算法
随机森林算法
六轴传感器
蓝牙模块
finger gesture recognition
double band pass filtering algorithm
random forest algorithm
six-axis sensor
Bluetooth module
