Entropy Map Based Active Seabed Scanning Method Using Underwater Image Sonar for AUV

In underwater exploration, where high turbidity and signal attenuation are prevalent, generating 3-dimensional (3D) maps using autonomous underwater vehicles (AUVs) equipped with image sonar is essential for successfully completing underwater missions. To achieve autonomous underwater mapping for AUV, this paper addresses two-stage seabed scanning method using an autonomous underwater vehicle's image sonar. In the first stage, the AUV generates a 3D volumetric map over a wide area. In the second stage, the entropy-based seabed scanning, is carried out on a interested area using the previous occupancy voxel map. This proposed method was validated through underwater environment simulation experiments. By comparing the residual map entropy and the reconstructed object shape with those obtained using the traditional lawnmower scanning method, the results demonstrate that the proposed method yields better results compared to the conventional approach.