Mosaic construction from side-scan sonar: A comparison of two approaches for beam interpolation

This paper presents a unified methodology for the mosaic construction of the seafloor from side-scan sonar (SSS) imagery. This methodology applies techniques in SSS data correction to compensate the sonar beam pattern and the motion of the vehicle transporting the sonar that is doing the data acquisition. The vehicle motion is estimated through an Extended Kalman Filter (EKF). The EKF reduces the error propagation in the mosaic alignment by using the information on the past trajectory for predicting the new position of the vehicle. This process not only improves the navigation accuracy but also allows producing georeferenced mosaics. These georeferenced mosaics are enhanced using Nearest-Neighbor and Bilinear interpolation classic methods. The experimental data were acquired in the Atlantic Ocean near South Portugal utilizing an L-3 Klein System 3000 Digital SSS.