Moving Dim-Target Tracking Algorithm using Template Matching

This paper proposes the method for tracking a dim target included in images acquired by an infrared (IR) sensor. By using target information obtained in the first frame for target detection in the next frame, execution time was shortened, and unnecessary computations were removed. Also by suppressing IR sensor noise and emphasizing the target, this enabled more effective tracking. In order to verify the performance of proposed method, error between the center of real target and the center which was obtained by algorithm was compared. Moreover, the result of the experiment investigates the proposed method's outstanding performance for tracking dim targets.