基于kinect系统的手势识别与应用

Gesture Recognition & Application Based on Kinect System

文章基于kinect系统,运用人体骨骼跟踪技术,研究了人体手势的计算机识别问题。根据东方人的手势习惯和特点,设定了8种特定的手势操控计算机模式,通过kinect系统采集操作者手势的三维深度图像,分析识别手势的对应模式,执行相应的计算机操控程序。以PPT演播的操控为例,设计并实现了基于手势的人机交互系统。实际应用结果表明:基于手势的人机交互系统识别精度较高、稳定可行,有效地提升了计算机操控的简便性和用户使用计算机的愉悦性。

This thesis uses human skeleton tracking technology to study the computer recognition with the help of human hand posture based on kinect system.According to the habits and characteristics in Asians of hand posture expression,eight kinds of specific hand postures are set in this paper to control computer mode,three- dimensional images through hand posture analysis to identify corresponding modes then to perform appropriate computer handling procedures.Taking PPT studio manipulation as an example,interactive system of hand posture is designed and implemented.Practical application results have shown high identification accuracy,stable and feasible operation with more simplicity and user's pleasantness.