| Popis: |
This paper proposes a simulation environment for creating simulated ranging data of high resolution sonar systems. It enables the assessment of underwater object reconstruction techniques and the verification of various methods for automated target detection. As an input, the proposed environment uses the AUV's trajectory, a model of the sonar system attached to the AUV and 3D CAD models of underwater objects. As an output one can get a point cloud of the whole underwater object or 2D images, i.e. range data, which are smaller parts of a large object's 3D model. Noise is modeled and used as additional input in order to generate outputs that are closer to the real- world. The simulated data is processed and evaluated, where well known sonar signal processing methods are used to process the simulated acoustic signals. Automated target detection is achieved by processing the simulated 2D range data with advanced image processing techniques. |
