Target Classification Using YOLOv2 in Land-Based Marine Surveillance Radar

In this study, it is aimed to identify the measured profile data obtained from a land-based marine surveillance radar by using YOLOv2 (You Only Look Once version 2) object detection and classification method. Accordingly, a total of 2401 profile data were collected for 12 different target classes from a marine surveillance radar operating in the X frequency band. The data to which two different labeling forms were applied were subjected to the identification process by YOLOv2 method. For the identification of profile data of marine targets, it was found that YOLOv2 method which was applied for the first time with this study provided high performance.