Autonomous underwater vehicle for rugosity and obstacle detection

In this paper, we present a path planning frame- work for underwater exploration and rugosity estimation using Autonomous Underwater Vehicles (AUVs). Rugosity, a measure of surface variation in height, is often used to characterize seafloor habitat. This project's purpose is to optimise the survey of an unknown area. The proposed method makes use of a diverse set of sensors, including multibeam sonar and stereo cameras, whose varying resolution and range make them complementary for this task. Sonar is used for broad-scale exploration to generate a coarse understanding of the area's structure. The stereo cameras are used for fine-scale investigation to refine the high- resolution estimation of rugosity. The results show the simulation of two situations based on real structural data collected with an AUV and a diver-held sensor. The first scenario considers a circumstance in which there is no broad-scale information and the robot must examine the landscape optically. The second simulates the two-pass example and exhibits our method's ability to estimate rugosity with high accuracy faster than other survey planning approaches.