基于图像区域搜索法的彩色球目标识别与跟踪

Color-ball Object Recognizing & Tracking Using the Image Region Searching

本文着重探讨如何实时地进行彩色球目标或者具有一定刚性目标的跟踪。结合先验颜色和形状特征 ,进行目标识别 ,可以增强系统的鲁棒性。将目标搜索限制在图像局部小区域内 ,增强实时性。给出的方法首先利用颜色信息确定辨识的区域 ,然后根据我们给出的椭圆分析方法进行鲁棒的目标识别。

In this paper, we introduced the novelty method of tracking the color ball or rigid objects. In the method, we used the characters of the color and shape to recognizing the object. Using these characters, we can achieve the better robust of the system. Limiting the researching space of object in the small part of the image, we can get the high real time of the system. Firstly, the method used the information of the color to determining the discriminating regions, and processed the object recognition using the ellipse analyzing.