Target Recognition and Tracking of Sonar Images with Ring-Like Background Noise

As the main tool for underwater detection, sonar has a longer underwater propagation distance and stronger penetration than visible light and infrared rays, and is widely used in oceanographic surveys and hydraulic engineering. However, sonar images are characterized by poor contrast, low resolution, and severe noise. When sonar is used for motion detection, the traditional static background construction method has poor recognition accuracy and slow speed for high-speed moving targets. This paper proposes a method for target detection in underwater sonar images in a ring-shaped rapidly changing noise environment. The median filtering method is firstly used to eliminate the interference of salt and pepper noise, and then the gray integration is carried out on the polar coordinate system to separate the foreground target. The proposed method can locate the target, and offer an accurate target trajectory in which the Kalman filter is employed. The simulation results show that the proposed method has a high recognition rate and can predict the target trajectory even when the target disappears for a while of time.