Group Tracking of Flock Targets in Low-Altitude Airspace

Bird-strike avoidance system usually uses marine radar with limited resolution for surveillance of low-altitude airspace. When the target of interest is flock in high target density, the targets are so closely spaced so one cell on the radar display always represents several birds. Therefore, it is sometimes not necessary to detect and track individual target with accuracy. In this paper, a central group-target tracking algorithm is proposed to track the overall flock behavior, and we consider several criteria of splitting and merging based on clustering method. We experimentally compare the central group-target tracking algorithm with the multi-target tracking algorithm, which are complementary parts of the flying bird targets detecting and tracking scheme. It is shown that the proposed algorithm gives encouraging results with high efficiency, overcoming the data association problem with multi-target tracking for closely spaced targets. In addition, we include an application example on real radar data.