：针对目标跟踪过程中目标尺度伸缩和姿态形状的变化引起的目标丢失，以及使用单个模型跟踪

机动目标不够理想，提出一种基于SIFT特征的自适应滤波目标跟踪算法。仿真结果表明，该算法在

目标机动时，跟踪性能远优于其它特征匹配算法和多模型算法，而且计算量小，能保证跟踪的实时性。-Abstract：Aim at the lose of target due to scale。invarl’ant，position change and deformation，as welI as

dissatisfied single model tracking in the process of target tracking．an adaptive algorithm based on SIFT is

proposed．Simulation results show that tracking perform ance of method in this paper is far better than other

feature matching and multiple-model algorithms．Furtherm ore，the computational load of the proposed

method is less，and can ensure the real—time perform ance in tracking