The project will apply advanced signal and image processing techniques to acquire the information required to adaptively drive the platform. Major contributions in this field are the definition of statistical processing techniques that use the macroscopic morphology of the observed field to correct for artefacts produced by disturbance factors (such as currents), the automatic adjustment of performance to the strict needs of platform guidance (since perception is not understood as a goal on its own, but rather a tool for guidance), and the use of video-mosaicing techniques to provide ground-truthing for validating adaptive surveying.