Target Detection and Correction Method for Unmanned Surface Vehicles Marine Radar Based on Image Contour Features

Marine radar has the advantages of a wide detection range and all-weather operation, and it is the main method used by unmanned surface vehicles(USV) to detect obstacles on the sea. However, for large targets such as islands and freighters, marine radar cannot accurately describe the area in which they are located and may misjudge them as multiple scattered targets. For this reason, a method for correcting the target detection result of marine radar based on image contour features was proposed with the use of the radar image characteristics. First, the radar image is preprocessed to simplify the image data and enhance the detectability of related information. Second, the target contour is extracted according to the corresponding pixel coordinates of the target. If different targets correspond to the same contour, a misjudgment is considered to have occurred, and multiple scattered targets generated by the misjudgment are combined into one target. Finally, the area where the target was located was described using the distance and orientation characteristic parameters contained in the target contour. The results of the experimental tests demonstrate that this method can effectively solve the problem of misjudgment of large targets and reduce the distance error in the detection results of marine radar by more than 79%, and the azimuth error by up to 60%.