Statistical autofocus of synthetic aperture sonar images using image contrast optimisation

In this paper we present preliminary results of anew technique for autofocus of synthetic aperture sonar (SAS) imagery. This employs global optimisation methods based on image quality measures and motion statistics of the towfish. From each reconstructed image a quality measure is calculated and balanced with a cost measure calculated from the known statistics of the towfish motion. This is used to perturb the estimated towfish path and a new image is reconstructed. Results show a high degree of accuracy is possible but is dependent on the image quality measure used. A number of simple image contrast measures are compared. When combined with a bulk-error removal method, or INU measurements, this technique shows great promise for high-quality autofocus of regions of interest