3D Diver Positioning in Water Column Using Mills Cross Sonar

In typical configurations, 2D forward-looking sonars (FLS) use a wide vertical beam that compresses elevation into a single plane, providing only a horizontal slice of the scene. This limits 3D localization accuracy, especially for stealthy underwater targets moving in depth, such as divers and Autonomous Underwater Vehicles (AUVs). This paper presents a novel framework for 3D localization of underwater targets using the Seapix-L sonar system configured in a dual-view from a Mills Cross antennas arrangement. The proposed system incorporates ping-to-ping time-averaging to improve the signal-to-noise ratio (SNR), a Constant False Alarm Rate (CFAR) detector to ensure consistent detection thresholds, and DBSCAN clustering to aggregate detections into coherent targets. A 3D Kalman filter fuses information from both sonar views, yielding robust and precise target localization. Experimental validation in real-world underwater scenarios confirms the effectiveness of the proposed method, demonstrating its potential to enhance situational awareness and security in maritime environments. The implications of this research extend to applications such as maritime infrastructure protection and environmental monitoring.