Integration of Radar Sequential Images and AIS for Ship Speed and Heading Estimation Under Uncertainty

High-quality ship kinematic information (i.e., speed and heading) is important for Vessel Traffic Service (VTS) to assess traffic situation. Presently, the ship speed and heading are derived from the Automatic Identification System (AIS) and radar signal transmission in VTS. However, there are uncertainties on ship speed and heading estimation caused by the false echo in radar images and packet loss rate in AIS data transmission, which affects the supervision of ship navigation and its safety. To address the uncertainty on ship speed and heading estimation and enhance the supervision of the marine traffic system, this paper proposes a ship speed and heading estimation approach by using radar sequential images and AIS data fused. The proposed approach consists of three steps: 1) ship detection using radar sequential images, 2) ship tracking with inter-frame information, and 3) ship speed and heading fusion. The experimental results in selected radar sequential images of the Yangtze River show that the variance of ship speed and heading distribution at 10s sampling frequency reduce from 2.55 to 0.38, and 28.35 to 23.48, respectively. It is indicated that the proposed approach can reduce the uncertainty on ship speed and heading estimation and quantitatively describe the ship navigation status.