Combining imaging sonar and photogrammetry for underwater object localization and mapping

Abstract. During the flooding event in September 2024 in Austria, an unknown number of industrial barrels was washed into a freshwater pond near Loosdorf, leading to this study aimed at locating and measuring these barrels. Due to significant turbidity, with a Secchi depth of 1.1 m, traditional airborne inspection methods failed to locate the barrels. Consequently, we employed a Remotely Operated Vehicle (ROV), equipped with various sensors including an imaging sonar Oculus M3000d, a 4K camera, and a laser scaler. Using the imaging sonar, we successfully identified twelve barrels within a study area exceeding one hectare. After identifying the barrels, our goal was to determine the dimensions (height and diameter) of the barrels. While visibility limitations restricted photogrammetric evaluations, sonar data allowed accurate height measurements. However, the diameter could not be determined using the imaging sonar, so that the camera and laser scaler were employed for this task. Despite challenges posed by turbidity, the combination of these sensors proved effective for this case of underwater inspections. The objects were found, and the dimensions could be determined. This study demonstrates the potential of combining imaging sonar, cameras, and laser scaling techniques in underwater environments, particularly under low visibility conditions.