AUV based flushed and buried object detection

Within the framework development program of the SeaOtter MK2 Autonomous Underwater Vehicles (AUVs), in 2013, ATLAS ELEKTRONIK GmbH designed a combined multi sensor system capable of deep sea operations also known as DDAUV (Deep Diving AUV). The DDAUV is equipped with mid-and high-frequency sonar systems, magnetometers and optical based sensors (image and video). The purpose of the DDAUV is to perform independent detection, localization, tracking and categorization of specified objects of interest (OOI) that are lying on the seafloor or buried within the upper sediment layer [1]. In this paper we present the detection improvements achieved by fusing information from the following DDAUV sensors: video camera; side scan sonar (SSS); multi beam echo-sounder (MBES); and buried object scanning sonar (BOSS). The last is capable for both: time delay beamforming (the result of this beamforming is referred to as Voxel Layer VL) and synthetic aperture sonar (SAS) processing (referred to as tomographic imaging). In the OOI detection, automatic real-time processing combined with high-resolution data enables an immediate detection, localization and categorization to take place; thus allowing for autonomous decision-making based on the up-to-date measurements and observations. Results from sea trials are also demonstrated.