基于视觉的动态手势识别综述  

A Review of Vision Based Dynamic Hand Gestures Recognition

作　　者：沙洁 麻建 牟海军 侯建华[1] 

机构地区：[1]中南民族大学电子信息工程学院,湖北 武汉 [2]武汉杰目科技有限公司,湖北 武汉

出　　处：《计算机科学与应用》2020年第5期990-1001,共12页Computer Science and Application

基　　金：企业委托项目资助(编号HZY18010).

摘　　要：随着计算机视觉研究的迅速发展,动态手势识别在人机交互、智能驾驶、虚拟现实等领域得到了广泛的应用。为了解动态手势识别的发展现状,通过对近年来动态手势识别算法的研究,从多模态输入、手部检测、手势建模三个方面进行梳理总结,介绍当前动态手势识别算法研究进展、面临的难题和应用场景。With the rapid advancement of computer vision, dynamic gesture recognition has been applied in a range of applications, such as human-computer interaction, intelligent driving, virtual reality, etc. In order to understand the current status of dynamic gesture recognition, the paper provides a comprehensive survey from three aspects: multi-mode input, hand detection and dynamic gesture modeling. The challenges faced by dynamic gesture recognition and application scenarios are also introduced.

关 键 词：动态手势识别 多模态输入 手部检测 手势建模 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]