摘要
Camshift算法主要利用物体的颜色信息进行跟踪,在复杂背景条件下容易造成目标的跟丢,且在目标被遮挡时,也容易造成跟踪失效。本文提出了一种改进的Camshift目标跟踪算法。首先将目标图像的HSV模型的三个分量进行加权建立一种新的目标颜色模型,然后由对整帧图像计算反向投影改为比搜索窗口稍大的区域计算反向投影,减少了相似背景的干扰。同时为了解决遮挡问题,结合了Kalman滤波器,有效地预测了目标的位置。实验表明,本算法能够避免背景颜色干扰和解决遮挡问题,实现了对运动目标准确跟踪。
Camshift algorithm mainly utilizes the color information of object to track object, thus it is easy to cause goal lost under the condition of complicated background. And the object is also subject to be lost when being sheltered. This paper puts forward an improved Camshifl object tracking algorithm. Firstly, the three components of the target image in the HSV model are weighted to create a new kind of target color model, and then it calculates the back projection in a slightly larger area than the search window only instead of calculating the back projection of the whole image, thereby reducing the interference of similar background. Meanwhile in order to solve the sheltered problem, the Kalman filter is combined with improved Camshift algorithm to predict the position of the target effectively. Experimental results show that the proposed algorithm can avoid the interference of the background color and solve the sheltered problem of the object, so that achieving a precise tracking of moving objects.
出处
《电子技术（上海）》
2014年第1期11-13,共3页
Electronic Technology
