Kalman filtering is widely used in target tracking. However, conventional Kalman filtering may fail to track the target when there is acceleration, deceleration, and turn. In this paper, these maneuvers are characterized by two orthogonal components of the changed velocity in l 1 -norm. The adaptive factors to adjust the Kalman gain are then generated through a mapping function based on the characterization. Sea experimental results show that the proposed adaptive Kalman filtering is better at tracking the maneuvering target