Key points are not available for this paper at this time.
The use of multibeam MIMO-SONAR systems on marine vehicles (e.g. remotely operated vehicles, ROVs) enables the visual 3D reconstruction of the water-column by hydroacoustic ensonification of the surrounding environment in real-time. For underwater target detection and classification purposes, the processed SONAR data must be visualized for interpretation of the results by a human operator and to allow for a corresponding re-adjustment of the system parametrization during operation. Therefore, conventional 2D visualization approaches such as the plan position indicator (PPI) plot must be adapted to 3D. Challenges such as different signal-to-noise ratios, beamforming artifacts, and overlapping objects must be considered when choosing how to visualize the processed data to allow for the correct semantic interpretation of the scanned water-column. In this presentation, an approach for the 3D voxel- and mesh-based visualization of real-time processed multibeam SONAR data is shown. The focus will be on how to consider the 3D beamforming and signal correlation processing, combined with data interpolation and filtering techniques, to allow for a visual reconstruction of the water-column from the SONAR data. An implementation of this approach in the C++ programming language using the Qt visualization framework will be shown in the Kiel Real-time Application Toolkit (KiRAT) for a virtual ocean environment.
Kanarski et al. (Fri,) studied this question.