Automated Detection and Classification in High-Resolution Sonar Imagery for Autonomous Underwater Vehicle Operations

摘要

Autonomous Underwater Vehicles (AUVs) are increasingly being used by military forces to acquire high-resolution sonar imagery, in order to detect mines and other objects of interest on the seabed. Automatic detection and classification techniques are being developed for several reasons: to provide reliable and consistent detection of objects on the seabed; to free human analysts from time-consuming and tedious detection tasks; and to enable autonomous in-field decision-making based on observations of mines and other objects. This document reviews progress in the development of automated detection and classification techniques for side-looking sonars mounted on AUVs. Whilst the techniques have not yet reached maturity, considerable progress has been made in both unsupervised and supervised (trained) algorithms for feature detection and classification. In some cases, the performance and reliability of automated detection systems exceed those of human operators.