为了探索和实现和谐自然的人机交互方式,提出了一种新的三维人机交互方法,该方法利用单个摄像头捕捉人的手指及其阴影的运动,实时进行图像分析处理,检测指尖的位置.并利用手指的指尖和阴影的指尖的距离计算指尖的高度,从而还原指尖的在空间中的移动.最后使用Kalman滤波来使指尖的运动更平滑,使三维交互更自然.该方法对软硬件要求低,易于实现.实验结果表明:该方法在手指跟踪和指尖检测上可达到较高的准确率,可实现自然友好的三维人机交互,并可应用于虚拟现实或三维游戏中.
This paper presents a novel approach for three dimensional human-computer interaction (3D HCI). The approach uses one single camera for tracking the motion of the finger and the fiflger's shadow, and.deteets the finger's position by analyzing the image in real time. The distance between the finger's tip and its shadow's tip is calculated to compute the 3D motion of the finger tip. Finally, to make the 3D interaction more natural, the Kalman Filter is used to smooth the motion of the finger tip. This approach has a low demand of hardware and software, and is implemented easily. Experiments show that the proposed system has high accuracy in finger tracking and fingertip detection. The approach achieves natural and friendly 3D human-computer interaction, and can be used in virtual reality applications or 3D games.