Side-looking sonar modeling using inverse imaging techniques

Developing a new sonar for side-scan imaging can be a complicated and expensive proposition. Sonar models and simulations can be used to alleviate some of the expense and time required for development by providing some early simulated imagery of proposed designs. This paper describes a modeling approach where inverse imaging techniques are applied in order to generate simulated stave-level sonar data for analysis. Using this approach, any image can be used as a reflectance map and converted to stave-level data. Once generated, the data can be down sampled to match the parameters for a proposed system (real or synthetic aperture) and re-beamformed to produce simulated imagery. This process has the benefit that it is fast, efficient, and can be performed on typical desktop or laptop computers. However, this approach is intended for imaging systems and may not produce reasonable results when comparing simulated and field-collected data in frequency ranges where sediment penetration or other non-specular phenomena occur. The implications of the underlying assumptions of this modeling approach are discussed and the model is used to convert imagery from one synthetic aperture sonar (SAS) system to the specifications of another SAS system. Sample imagery including the simulated comparison with actual imagery from the two systems is shown and the limitations of the comparison are discussed.