Detection and Tracking of Underwater Object Based on Forward-Scan Sonar

Abstract

Underwater object detection is critical in a lot of applications in maintenance, repair of undersea structures, marine sciences, and homeland security. However, because optics camera is subject to the influence of light and turbidity, its visibility is very poor in underwater environment. Therefore, forward-scan sonar is widely applied to the underwater object detection in recent years. But there are still some problems such as: Forward-scan sonar imaging, which is different from optics imaging, processes echo information from acoustic signal in the water. Generally, sonar images are with high noise and low contrast. It is difficult for the operator to identify underwater objects from the images. In addition, the surveillance of underwater objects is tedious and time consuming, and it is easy to make mistakes due to the fatigue and distraction of the operator. To solve the above problems, an image processing strategy to detect and track the underwater object automatically is presented. Firstly, the sonar images are enhanced by the Gabor filter. And then underwater objects are extracted. Finally the tracking method based on Kalman filter is adopted. The experimental results validate that the presented methods are valid.