中文摘要：

针对红外成像末导引阶段飞行器姿态调整及高速运动导致的目标尺度和姿态迅速变化的问题，提出了一种基于加速鲁棒特征（SURF）的红外成像目标跟踪算法。为了实现在末导引阶段对目标进行精确跟踪，采取了点跟踪的策略。首先根据跟踪点在上一帧的位置，在当前帧选取以相同位置为中心的图像子块并求其SURF特征，通过SURF特征匹配得到当前帧图像子块和模板的匹配点集，采用随机抽样一致性（RANSAC）算法剔除误匹配点对，进一步用最小二乘算法（LSA）精确地估计出对应的单应性矩阵；然后通过单应性矩阵把跟踪点映射到当前帧获取跟踪点在当前帧的位置，从而实现精确跟踪。试验结果表明，本文算法有较高的跟踪精度和较好的实时性。

英文摘要：

For the rapid changes of target scale and posture caused by aircrafts attitude adjustment and high-speed movement in the infrared imaging terminal guidance stage, speeded-up robust features （SURF） based infrared imaging target tracking algorithm is proposed. In order to precisely track target in the terminal guidance stage, the point tracking strategy is suggested. Firstly, according to the tracking point position in the previous frame, the image sutyblock centered on the same location in the current frame is selected; the matching point set between the image sub-block and model is attained by the SURF matching, and the random sample consensus （RANSAC） is employed to eliminate false matching points;the corresponding homography matrix is exactly estimated through the least squares algorithm （LSA）. Then,the exact location of tracking point in the current frame is attained through the homography matrix mapping. In order to validate the efficiency of the algorithm,the simulation analysis is carried out, and the simulation results show that the proposed algorithm can obtain high tracking accuracy and meet real time demand.