Plots-Centroid Method for USV-Borne Millimeter-Wave Radar Based on Euclidean Distance

To solve the problems of false alarm and target-plot splitting of millimeter-wave(MMW) radar used in the detection of obstacles in unmanned surface vessels(USVs) on the sea, most recent studies have reduced false alarms by processing radar echo based on waveform characteristics. These studies have also compared multi-dimensional information such as the distance and angle of plots to determine whether the plots belong to the same target, this studies are cumbersome. And the centroid methods are not sufficiently accurate to represent the target state. Therefore, this study proposes a plots-centroid method based on Euclidean distance that solves the problems of false alarm and target-plot splitting by processing radar plot data. First, using prior knowledge related to radar echo intensity and effective detection range, a threshold filtering method is used to remove invalid plots. Then, based on the rule stating that plots belonging to the same target have the same speed and close distance, the Euclidean distance is used to measure the similarity of information between plots. Finally, target-plot clustering is realized. Finally, the position and section width of the obstacle target are calculated, where the range of the obstacle target is denoted by a rectangular dangerous area to enable accurate detection of the obstacle target in front of the USV. The effectiveness of this method is verified by an actual ship test conducted.