基于MediaPipe的手势识别算法研究与应用  

Research and application of gesture recognition algorithm based on MediaPipe

作　　者：高加瑞 杜洪波[1] 蔡锦 李佳翰 GAO Jiarui;DU Hongbo;CAI Jin;LI Jiahan(School of Science,Shenyang University of Technology,Shenyang 110870,China)

机构地区：[1]沈阳工业大学理学院,沈阳110870

出　　处：《智能计算机与应用》2024年第4期157-161,共5页Intelligent Computer and Applications

基　　金：辽宁省教育厅高等学校基本科研项目(LJKZ0157);辽宁省大学生创新训练计划项目(S202210142057)。

摘　　要：随着人机交互的飞速发展,手势识别技术逐渐引起研究人员的关注。然而复杂的环境会对手势识别的准确性产生很大的影响,本文旨在探索一种基于MediaPipe的高精度手势识别算法,算法采用MediaPipe的手部关键点检测模块,对手部姿态进行检测和识别,利用本文的手势识别算法对手势进行分类识别。通过实验证明,该算法可以有效地克服光照变化、背景干扰和手部遮挡等问题,并可同时对多只手进行识别,提高了手势识别算法的准确性和鲁棒性。With the rapid development of human-computer interaction,gesture recognition technology has gradually attracted the attention of researchers.However,complex environments can significantly affect the accuracy of gesture recognition.This paper aims to explore a high-precision gesture recognition algorithm based on MediaPipe.The hand pose is detected and recognized using MediaPipe′s hand keypoint detection module.The gesture recognition algorithm proposed in this paper is employed for gesture classification and identification.Through experiments,when compared with traditional gesture recognition algorithms,this algorithm effectively overcomes issues such as lighting variations,background interference,and hand occlusions.Additionally,it can simultaneously recognize multiple hands,thereby enhancing the accuracy and robustness of the gesture recognition algorithm.

关 键 词：MediaPipe 手势识别 OPENCV 手势跟踪 人机交互 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]