A Terrain Data-Assisted Tracking Method for Rolling Rock

The rolling rock's trajectory is variable and complex, which could easily cause false associations, and affect tracking accuracy. To address this issue, a rolling rock tracking method based on terrain data assistance is proposed. Firstly, three-dimensional (3D) terrain data is utilized to predict the rolling rock's motion direction and project it onto a two-dimensional(2D) radar coordinate system. Then, based on the deviation between the associated measurements and the predicted direction, false association are rejected, and the state and covariance matrix are corrected using the Extended Kalman Filter (EKF). Finally, the correction of rolling rock trajectory is achieved, enhancing tracking performance. Experimental data is used to validate the effectiveness of this method.