计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (12): 133-136.

• 图形图像处理 • 上一篇    下一篇

一种基于肤色和深度的第一人称人手识别方法

郭训力,俞  扬   

  1. 南京大学 计算机软件新技术国家重点实验室,南京 210023
  • 出版日期:2014-06-15 发布日期:2015-05-08

Hand recognition method based on color and depth information from first-person view

GUO Xunli, YU Yang   

  1. National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China
  • Online:2014-06-15 Published:2015-05-08

摘要: 手是人类与外界交互的主要工具,因此在可穿戴增强现实系统中,引入手势操作将会为人机交互过程提供非常自然的操作体验。以往的手势识别,一方面并不是考虑应用在可穿戴增强现实的场景中,有着不同的视角差,另一方面往往只基于二维信息,而忽视三维深度信息。在传统的肤色模型基础上,融合了三维深度信息,构建了满足实时性要求的手势操作系统。

关键词: 增强现实, 人手识别, 颜色模型, 深度, 实时

Abstract: Since human beings interact with the world mainly by their hands, introducing hands operations will provide them with very natural operation experience in wearable augmented reality systems. Traditional hands recognition techniques, on one hand, are not applied in wearable augmented reality settings, and have different parallaxes in different environments. On the other hand, they usually focus on only the 2D information, neglecting the depth information. This paper implements a real-time hands recognition system by considering the depth, in addition to the traditional color model.

Key words: augmented reality, hand recognition, color model, depth, real-time