Visual features and Deep Maxout Network‐based moving target detection using radar signals under sea clutter

Abstract

Radar echoes have emerged as a rapid technology that is utilized in a wide range of applications especially in target identification under sea-clutter background. The detection of moving targets under a sea-clutter background is still a complex task because of the entering and evacuating of targets, which is highly unpredictable and unexpectable. Because of the lack of essential information, target detection and parameter estimations have become key constraints. Besides, it is very difficult to differentiate directly the small targets from the clutter background. However, the accurate location of the target cannot be determined exactly by the existing approaches.To address such limitations, an efficient strategy is proposed in this study for recognizing the moving targets under sea-clutter background using the proposed Magnetic Ant Lion Optimization algorithm, which is derived by the integration of Magnetic Optimization and Ant Lion Optimization. However, the proposed method achieves effective performance in terms of detection rate, mean-squared error, and the missed target rate with the values of 0.988, 0.012, and 0.224 for vertical-horizontal polarity.