Background: Unmanned Surface Vehicles (USV) can undertake risks or special tasks in
marine independently and will be widely used in the future. In the autonomous navigation of USV
equipped with vision camera, the water boundary line needs to be detected in real time and it is one
of these key intelligent environment perception methods for USV.
Methods: An efficient water boundary line detection method based on Gray Level Co-occurrence
Matrix (GLCM) texture entropy is proposed. In image preprocessing, the high-brightness areas are
eliminated to avoid the effects of water boundary line detection.
Results: GLCM entropy is employed to segment water, land and air for water line regression. The
proposed method is efficient for the images with high-brightness areas.
Conclusion: The experimental results demonstrate that the proposed method is not only more accurate
than the existing water boundary line detection method, but also has good real-time performance
and is suitable for the application in USV.