International Journal

Ha-Yun Kim, Myung-Il Roh, Hye-Won Lee, In-Chang Yeo, Yeong-Min Jo, Jisang Ha, Nam-Sun Son, “A Hybrid Tracking Method for Maritime Obstacles Using Sensor Data”, Ocean Engineering, Vol. 312, Part 2, 2024.09.16

by SyDLab posted Sep 10, 2024
?

단축키

Prev이전 문서

Next다음 문서

ESC닫기

크게 작게 위로 아래로 댓글로 가기 인쇄
Extra Form
Abstract Safe navigation of ships depends on the accurate recognition and tracking of nearby maritime obstacles as detected by various sensors. Challenges arise in tracking maritime obstacles from sensor data because of sensor noise and incomplete sensor data. Traditionally, tracking algorithms such as the EKF (Extended Kalman Filter) have been applied to track the state of maritime obstacles, including the position, COG (Course Over Ground), and SOG (Speed Over Ground). This study implemented a combined EKF- and learning-based (hybrid) tracking method. In the EKF-based method, the parameters are related to the uncertainty of the system and the sensor data. These parameters are generally set manually by analyzing the noise of the sensor data and may not be optimal; we optimized the parameters to compensate for this. In the learning-based method, we trained a deep learning model using a DNN (Deep Neural Network) to predict obstacle states from sensor measurement data. We then propose a hybrid tracking method that combines the two tracking methods to compensate for the shortcomings of each method. We verified these three tracking methods using navigation data obtained through field tests. The verification results showed that the learning-based tracking method improved the SOG tracking accuracy by 11.47% compared with the traditional EKF-based tracking method. The tracking accuracy of the hybrid tracking method was reduced by 22.42% for the COG and 42.05% for the SOG. These results indicate that the hybrid tracking method effectively compensates for the limitations of the other methods, resulting in an enhanced tracking performance.
Publication Date 2024-09-16
Role Corresponding Author
Category SCI

Ha-Yun Kim, Myung-Il Roh, Hye-Won Lee, In-Chang Yeo, Yeong-Min Jo, Jisang Ha, Nam-Sun Son, “A Hybrid Tracking Method for Maritime Obstacles Using Sensor Data”, Ocean Engineering, Vol. 312, Part 2, 2024.09.16

https://doi.org/10.1016/j.oceaneng.2024.119242


Articles

4 5 6 7 8 9 10