Tracking Omnidirectional Surfaces Using a Low-Cost Autonomous Underwater Vehicle

Tracking and observing underwater structures at close range are important for gathering high-resolution information. This article proposes a method to track omnidirectional surfaces using an autonomous underwater vehicle (AUV). This method is suitable for surveying various kinds of underwater structures, such as rugged seafloors, ship hulls, pools, dams, and quay walls. It is based on stochastic means, and has two important features. First, it depends on only commercial off-the-shelf sensors (mechanical scanning imaging sonar, a water current sensor, and an attitude sensor). Second, it is applicable to fast moving vehicles. This method is implemented in the AUV highly agile terrain tracker for ocean research and investigation, and verification tests in the sea and in a tank are carried out to evaluate its performance. In the sea, the AUV succeeded in tracking the rugged seafloor at high speed. In the tank, the AUV succeeded in tracking vertical walls while keeping its roll angle at 90$^\circ$.