提出一种用于实现手位手形跟踪的颜色滤波方法的计算模型。它可以取代虚拟现实中的数据手套 ,成为人机交互的工具。本方法是基于计算机视觉技术 ,模仿人眼的颜色视觉的特性 ,通过提取手的像素的特征颜色作为参考 ,对整个图像进行颜色滤波处理 ,从而将背景和干扰物体的像滤除 ,只保留手的像。然后经过平滑和轮廓提取处理 ,得到只含有手的轮廓的图像。颜色滤波算法能大大提高跟踪和识别的可靠性和实时性 A computational model of color filtering method for hand tracking is proposed. It can replace the data glove in virtual reality and become a tool of human-computer interaction. The method is based on the computer vision technology and imitates the characteristics of the human eye’s color vision. By extracting the characteristic color of the pixel of the hand as a reference, the whole image is color-filtered to filter out the background and the interfering object, leaving only the hand Like. After smoothing and contour extraction, an image containing only the outline of the hand is obtained. Color filtering algorithm can greatly improve the tracking and identification of the reliability and real-time