opencv基于视频的手势识别
这篇文章实现了一个使用视频手势获取进行人机交互的原型系统,大致可分为四个部分:视频搜集和分离、图画预管理、手势获取和识别、人机交互。其间的重要部分是手势的获取和识别,手势获取需求选用图画管理的方法,而手势势识别选用的则是决议方案树的方法,通过对手势几何特征的分析和决议方案来断定手势类型。选用以上方法对一段视频进行实时管理和分析,进而实现人机交互。其间首要的交互功能可以实现,实时性也较好,准确性也较高,不过该方法的鲁棒性不是很好,对光照强度较灵敏。选定一:基于视频的人机交互技术研究论文.doc
(This paper implements a prototype system for human-computer interaction using video gesture acquisition, which can be roughly divided into four parts: video collection and separation, picture pre-management, gesture acquisition and recognition, and human-computer interaction. The most important part is the acquisition and recognition of gestures, the method of image management is used to acquire gestures, and the method of resolution tree is used for gesture recognition. The above methods are used to manage and analyze a video in real time, and then realize human-computer interaction. The main interactive function can be realized, the real-time performance is also good, and the accuracy is also high, but the robustness of this method is not very good, and it is more sensitive to the light intensity.
Option 1: Video-based Human-Computer Interaction Technology Research Paper.doc)
页:
[1]