High-Precision Underwater Image Sonar Mapping in Aquatic Structural Environment using LiDAR Odometry and Mapping for Safety Inspection

This study proposes a method for underwater sonar image mapping for aquatic structural environments. Sonar is commonly used in autonomous maritime vehicles (AMVs) owing to the limitations of optical sensors; however, its low resolution and noise make pose estimation difficult, requiring support from global positioning system (GPS) or Doppler velocity log (DVL)-based navigation. However, GPS is not appropriate for all cases owing to jamming and interference in structural environments such as floating infrastructures, dams, and power plants, and underwater navigation sensors are vulnerable to several situations such as shallow water, interference with underwater structures, and unstable signal processing. Consequently, for the aquatic structural areas, we used the results of light detection and ranging (LiDAR) odometry and mapping in water surface environments to improve the accuracy of underwater sonar mapping. When compared to using only sonar data, the proposed method could obtain navigation data for AMVs easily and accurately. We verified the proposed method by conducting field tests, and the results showed high-precision sonar image mapping of the underwater environment in a turbid and low-depth underwater environment. Therefore, the proposed LiDAR-based alignment enabled mosaic mapping with better continuity and fewer distortions, as well as localization trends in field tests.