一种基于图像序列的脑电极点的跟踪定位方法  

A Head Electrode Points Tracking Location Approach Based on Image-sequence

作　　者：梁刚[1] 朱虹[1] 张军[1] 李晗[1] 

机构地区：[1]西安理工大学信息与自动控制工程学院,西安710048

出　　处：《计算机工程与应用》2005年第10期230-232,共3页Computer Engineering and Applications

摘　　要：为了对脑电检测诊断中的各脑电极点进行自动定位跟踪,该文提出了一种基于图像序列的自动定位方法,首先在单帧图像中,根据控制点的标识,确定首个基准点,之后根据基准点与局部周围电极之间的相互关系,采用动态迭代搜索方法,获得后续的置信基准点及周围电极点的标识名。实验结果表明,该方法对旋转拍摄时角度的随机变动,以及镜头的仰俯变动均具有很好的适应性,可以准确地给出同一个电极在不同帧中的位置匹配关系。In order to track the electrode points in head electrode diagnoses automatically,a tracking location algorithm is proposed in this paper.Firstly,a reference point is confirmed based on the information of control points in single frame.Next,for gaining the sequent believing reference points and its neighboring electrode points,a dynamic iterative search is used according to the relation between the reference point and electrode points around it in local region.Experimental results show that this algorithm can overcome random alteration of angle and elevation of rotary camera.At the same time,a matching relationship of the same electrode in different frames can be given out by our method.

关 键 词：脑电极诊断 图像序列 自动定位跟踪 动态迭代搜索 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]