首页|基于MediaPipe的手势识别研究

基于MediaPipe的手势识别研究

扫码查看
文章研究一种基于MediaPipe的快速手势识别方法,阐述了手势识别过程数据特征提取方法,通过构建手部检测模型、关键点定位和数据采集等步骤,建立了基于MediaPipe的数字手势识别模型.实验证明:该模型具有良好的实时性和鲁棒性,能够准确地识别手势并实现互动操作.本研究为基于MediaPipe手势识别的互动机器手研制提供了有效的方法和技术基础.
Research on gesture recognition based on MediaPipe
This paper studies a fast gesture recognition method based on MediaPipe,and expounds the data feature extraction method in the gesture recognition process.By constructing hand detection model,key point location and data acquisition and so on,a digital gesture recognition model based on MediaPipe is established.Experiments show that the model has good real-time perfor-mance and robustness,and can accurately recognize gestures and realize interactive operation.This research provides an effective method and technical basis for interactive robot hand based on MediaPipe gesture recognition.

MediaPipegesture recognitionfeature extractKNN algorithm

李达良

展开 >

福建洁博利厨卫科技有限公司,福建 福州 3500118

MediaPipe 手势识别 特征提取 KNN算法

福建省科技厅创新资金项目

2022C0022

2024

木工机床
福州木工机床研究所,中国机床工具工业协会,木工机床分会

木工机床

影响因子:0.105
ISSN:1005-1937
年,卷(期):2024.(1)
  • 8