A Multisensor Data Fusion Method for Seabed Topography Reconstruction Based on Image Enhancement and Intensity Matching

In seabed topography reconstruction, side-scan sonar (SSS) provides wide-area seafloor imaging as the detection equipment moves, but it cannot capture the region directly beneath the equipment, leaving gaps in coverage. On the other hand, Forward-looking sonar (FLS), typically mounted at the front of the equipment, offers real-time imaging of the seafloor ahead, though its detection range is limited and it cannot cover the entire area. Therefore, combining FLS with SSS, along with navigation data, enables comprehensive seafloor mapping by filling the gaps in SSS coverage and improving the accuracy of seabed topography reconstruction. Existing methods insert FLS images into the SSS coverage gaps by combining navigation data, but FLS images suffer from low resolution, blurred target details, and visible seams between the filled regions and the surrounding SSS images, leading to a lack of overall image coherence. This paper proposes a multi-sensor data fusion method that integrates data from SSS, FLS, altimeter, GPS and inertial navigation systems (INS). The method employs FLS image enhancement, intensity matching, and a weight-adjusted fusion strategy to improve the clarity of FLS images and significantly enhance the overall quality of the fused imagery. Experimental results show that our method greatly improves the visual coherence of the fused FLS and SSS images, achieves smooth edge transitions, eliminates visible seams, and enhances the precision of seabed topography reconstruction.