Unlike the radar target detection in a noise background, for the target detection in the background of the sea clutter, the increase of the transmitting power would not bring about a significant performance improvement and thus a refined modeling of the sea clutter and a full exploitation of the characteristics of the sea clutter become very important to improve the performance of the target detection in the background of the sea clutter. The compound-Gaussian model is a widely recognized model to characterize the sea clutter, which provides a powerful tool to implement a refined description of the sea clutter. Moreover, the related optimum detection theory and methods with regard to this model provide the technique for improving the target detection performance in the background of the sea clutter. This paper reviews three compound-Gaussian sea clutter models, including the K distribution, the generalized Pareto distribution and the inverse Gaussian texture and the existing optimum coherent detection and the near-optimum coherent detection with these models. The current development of the optimum coherent detection in the compound-Gaussian clutter plus noise is addressed and their ‘bottleneck’ in practical applications is analyzed. At last, we discuss several possible approaches to develop near-optimum and computationally implementable detection methods.

sea clutter  /  compound-Gaussian model  /  optimum coherent detection  /  near-optimal coherent detection  /  computationally implementable detection method