基于OpenCV与MediaPipe的面部动作控制鼠标操作的技术
投稿时间:2024-02-04  修订日期:2024-03-25  点此下载全文
引用本文:
摘要点击次数: 37
全文下载次数: 0
作者单位邮编
王骏祥* 南开大学软件学院 300457
马健为 南开大学软件学院 
宋笛秋 南开大学软件学院 
朱思哲 南开大学软件学院 
中文摘要:为丰富人机交互模式,本文基于OpenCV和Mediapipe在面部识别中的应用,在个人电脑上实现通过面部动作进行常用的鼠标控制。具体的方法是通过捕捉摄像头传入视频帧中用户脸部的关键点,并根据关键点的位置变化来执行对应鼠标操作。经实际测试结果表明:该程序仅通过简单的面部动作就可以较好地实现诸如移动光标,单击鼠标左右键以及拖动滚动条等常用鼠标操作,且识别准确率高,是一种可为特殊用户和特定环境条件下替代鼠标操作的低成本实用技术。
中文关键词:MediaPipe  OpenCV  面部识别  鼠标操作
 
Research on Facial Action Control Mouse Operation Based on OpenCV and MediaPipe
Abstract:In order to enrich the human-computer interaction methods, this work realizes common mouse control through facial movements on PC based on the application of OpenCV and MediaPipe in facial recognition. The specific method is to capture the key points of user’s face in the video frame transmitted by the camera, and perform corresponding operations according to the position changes of the key points. The actual test results show that this program can effectively achieve common mouse operations such as moving the cursor, clicking the left and right mouse buttons, and dragging the scroll bar through simple facial movements, and has high recognition accuracy. It is a low-cost practical technology that can replace mouse operations for special users and specific environmental conditions.
keywords:MedaiPipe  OpenCV  face recognition  mouse-control
查看全文   查看/发表评论   下载pdf阅读器