One Approach to the Recognition of Bottom Objects Using Monitoring Systems Installed on an Autonomous Underwater Vehicle

One of the tasks assigned to autonomous underwater vehicles (AUVs) is the search for designated bottom objects. To solve it, AUVs are equipped with hydroacoustic, electromagnetic and optical systems for bottom monitoring. Considering that each of these systems has limitations, their complex use is supposed. The most difficult task in the search for bottom objects is their recognition. To solve it, the paper proposes the use of deep learning technology for a problem-oriented artificial neural network with the peculiarity that the training material is generated programmatically in the form of digital images of the desired bottom object at the output of hydroacoustic, optical and electromagnetic bottom monitoring systems in various conditions of their observation. The application of this technology is shown on the example of recognition of aircraft from its images obtained using a side-scan sonar.