红外目标图像跟踪系统中应用单一特征跟踪算法,易造成跟踪点漂移和目标丢失。提出目标图像的复合特征跟踪算法,综合利用目标图像的灰度特征和点特征,联合灰度特征的易用性和点特征的空间不变性,分别克服单一特征易受光照、噪声干扰,以及实时性较差等方面的不足。提出了复合特征联合匹配的算法流程和权值分配策略,通过仿真试验对比三种方法的目标跟踪性能,证明了该算法的有效性。

The problems of tracking point drift and target losing are prone to occur to the single feature tracking algorithm of infra-red imaging guidance system. A new tracking algorithm based on composite features was proposed. The algorithm utilizes gray feature and point feature of the infrared target image, and overcomes the default of single matching method. The flow and power assign strategy of the algorithm were proposed, by comparing the characters of three tracking algorithms, the validity of the algorithm was proved.