一种基于卡尔曼滤波器的彩色目标跟踪方法

提出了一种基于卡尔曼滤波器的彩色目标跟踪方法。该方法首先在YCbCr颜色空间使用高斯模型构建目标颜色模型，根据该颜色模型在图像中进行目标检测，获取目标中心，然后利用卡尔曼滤波预测目标中心在下一帧图像中可能出现的位置，以该位置为中心建立搜索区域，在搜索区域内进行目标检测，计算目标中心位置，同时以该位置作为观测值，进行下一帧预测。实验结果表明，该方法具有较好的实时性和鲁棒性。

Kalman Filter for Color Object Tracking

An algorithm of color object tracking based on kalman filter was proposed. The color model of object was built by Gauss model in YCbCr color space. This color model was used to detect object, so the position of the object center can be computed. The possible position of the object center in the next frame image was predicted by kalman filter, and a searching window of object was generated around the center position. After detecting the object in the searching window, the position of the object center can be computed. This position was used as an observed valued for next prediction. The experimental results show that the method is effective and robust.