Ship Detection in Complex Environment Using SAR Time Series

Ship detection in complex environment is a challenging task due to strong background inferences, for which various deep-learning-based methods have been proposed. However, they poor performance on detecting nearshore ships medium-resolution synthetic aperture radar (SAR) images the loss of typical features and confusion with land scatterers. The availability multitemporal SAR gives opportunity separate scatterers by using temporal characteristics. In this article, we propose ship method based time series. First, investigate statistical stability series preclassification identify potential changed pixel clusters. Then, discriminate between candidates combining rotating object detector transition algorithm generate corresponding frozen reference (FBR) image. addition, dynamic framework proposed FBR image two-stage outlier approach. experiments show that enables monitoring high accuracy low false alarm rate targets.