A Topological Descriptor of Acoustic Images for Navigation and Mapping

The use of robots in underwater exploration is increasing in the last years. The automation of the monitoring, inspection and underwater maintenance tasks require a good mapping and localization system. One of the key issues of these systems is how to summarize the sensory information in order to recognize an area that has already been visited. This paper proposes a description method of acoustic images acquired by a forward looking sonar (FLS) using a graph of Gaussian probability density function. This structure represents both shape and topological relation. Furthermore, we also presented a method to match the descriptors in a efficient way. We evaluated the method in a real dataset acquired by a underwater vehicle performing autonomous navigation and mapping tasks in a marine environment.