基于人手的非接触式虚拟鼠标

Non-Contact Virtual Mouse Based on Hand Gesture

摘要: 手势识别技术，是一种高效、直接、自然的人机交互方式，属于计算机视觉领域中较为活跃的研究主题。同时，手的高自由度导致手势复杂、变化多样、移动高速等特性，使手势识别技术成为极具挑战的课题之一。本文在HSV和YCgCr颜色空间对不同的手势进行识别，借助OPENCV2.3.1视觉库和Microsoft Visual Studio 2010实现了基于C++语言的检测和特征提取。本文对基于计算机视觉的自然手势识别的相关技术：手势图像采集和预处理、手势图像的分割和特征提取，以及手势建模等，并且提出了适合虚拟鼠标、键盘系统的解决方案。

Abstract: Gesture Recognition, which is efficient, direct and natural, has become one of the key techniques of human-compute interaction. Meanwhile, the high degree of freedom of hand results in the complexity of gesture, diversity of variation and rapidity of movement, which makes gesture recognition full of challenges. The author uses the HSV and YCgCr color space to recognize different gestures, and depending on the OPENCV2.3.1 visual library and Microsoft Visual Studio 2010, the detection and feature extraction based on C++ language was realized. The main works are listed as following: Acquisition and preprocessing of the hand gesture image, segmentation and feature extraction of the image, and hand gesture modeling, etc. Finally based on the research on the technologies of the vision-based hand gesture recognition, the author gives an optimal solution of virtual mouse and keyboard system.