一种基于运动估计的红外目标跟踪方法  

Infrared target tracking algorithm based on motion estimation

作　　者：修彬 李成龙[2] 汤进[1,2] 罗斌[1,2] 

机构地区：[1]安徽省工业图像处理与分析重点实验室,合肥230039 [2]安徽大学计算机科学与技术学院,合肥230601

出　　处：《计算机工程与应用》2014年第12期125-128,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61073116;No.61003038);安徽省教育厅自然科学重点基金(No.KJ2010A006)

摘　　要：为解决红外目标跟踪中目标的交错、遮挡等问题,提出了一种新的基于运动估计的目标跟踪方法。建立目标的方向梯度-灰度直方图特征模型,该模型能较准确地刻画目标特征。使用最大后验概率指标在搜索区域进行目标匹配,该指标能很好地突出目标、抑制背景,并容易得到全局最优解。提出一种新的运动估计方法,即轨迹预测算法,对目标的运动进行较准确的估计。实验结果证明,该方法不仅计算复杂度低,而且能够较好地解决目标交错、遮挡等问题。A novel object tracking method based on motion estimation is proposed in this paper to solve stagger and occlu-sion problems in infrared object tracking. The gradient direction-joint gray histogram model of an object is established, which can depict characteristics of the object accurately. Then maximum posterior probability is used to match the objects in the search area, which is efficient to enhance the target, inhibit the background and easy to find the global optimal solution. A novel method based on motion estimation, trajectory prediction, is proposed to evaluate motion trajectory of object accurately. Experimental results illustrate that the proposed method not only can solve the stagger and occlusion prob-lems better, but also has lower computation complexity.

关 键 词：红外目标跟踪 方向梯度-灰度直方图特征 最大后验概率 运动估计 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]