Integrated tracking with features matching and adaptive model updating

Algorithms for correlation-based IR(infrared) target tracking can be implemented easily and have high precision, so they are used widely. There are some problems to be solved for IR target track, The problems include how to realize steady and fast tracking and how to assure the validity of model updating. In this paper we propose a novel high performance matching algorithm including integrated features selection and accelerating phase, and a model updating scheme based on confidence integration and multiple frame cumulation, the proposed IR target tracking algorithm is real time and accurate. From the experiment results we can conclude the proposed solution is validate and practical, and is very useful.