ABSTRACT
Real-time water boundary extraction with high accuracy is still a challenging problem in remote-sensing field. This article presents a shipborne water surface boundary extraction technique by integrating civilian marine radar and GPS compass. Meanwhile, a framework is developed to automatically extract water boundary from radar image. In this framework, direct transformation is adopted to rectify the geometric distortion of the radar image, morphological operation is used to fill holes and filter speckles in image, an edge tracing algorithm is design to extract image boundaries, visibility analysis and optimizations are conducted to recognize the real water surface boundary and to produce a vector map. A case study in Nanjing reach of Yangtze River are intended to validate performances of the method, and the final output of this method is verified using synchronous and detailed GPS-RTK (real-time kinematics) survey points. The experimental results demonstrate that the proposed water boundary extraction method can reach a positioning accuracy within ±1 m, which is significantly better than that of satellite-based optical remote-sensing and synthetic aperture radar (SAR) techniques, and the algorithm is computationally efficient and has the capability of real-time, full-automatic water boundary generation for various applications.
Acknowledgements
The authors greatly appreciate the editors and anonymous reviewers for their efforts and time which they have spent on this article.
Disclosure statement
No potential conflict of interest was reported by the authors.