基于改进射线矢量法的手部轮廓与手势实时跟踪方法

目前,在视频分析和处理过程中,运动物体的实时检测和与轮廓跟踪作为计算机视觉分析识别的基础,已变得越来越重要了.改进了传统的射线矢量法表示物体形状的方法,并结合拆分法和聚合法对单帧图像进行分割,以得到完整而准确的手部曲线,在相邻帧之间采用了Kalman滤波器估计帧间手部运动的轨迹来实现跟踪过程.该方法很好地克服了传统射线矢量法所无法表示的形状的缺陷,能够准确地跟踪手部的运动以及各种手势.

REAL-TIME HAND TRACKING BASED ON THE IMPROVED RADIUS-VECTOR FUNCTIONS

As the basis of recognition in machine vision field, real-time object tracking is becoming more and more important in the field of video analysis and process. A new approach to real-time hand contour tracking from color video sequences is presented. With the combination of splitting techniques and merging techniques, the shape is represented in a single frame by means of the improved radius-vector functions. Kalman filter is used to estimate the hand motion track between the neighbor frames. It is proved that the method can overcome the shortcoming of traditional radius-vector functions and track the hand motion or the transform of hand gesture robustly.