This paper proposes a sonar-based underwater object classification method for autonomous underwater vehicles (AUVs) by reconstructing an object’s three-dimensional (3D) geometry. The point cloud of underwater objects can be generated from sonar images captured while the AUV passes over the object. Then, a neural network can predict the class given the generated point cloud. By reconstructing the 3D shape of the object, the proposed method can classify the object accurately through a straightforward training process. We verified the proposed method by performing simulations and field experiments.

This paper proposes a method to recognize underwater target objects and estimate their yaw angle using an imaging sonar. First, a light sonar simulator generated template images of the target objects from various viewing angles. Next, a generative adversarial network predicted a semantic map by segmenting the real sonar image for reliable recognition. Then, matching the template images and semantic map identifies the target object and its yaw angle. We verified the proposed method by installing objects in the indoor water tank. The proposed method can provide relative pose information of sensing platforms which is useful for pose control and navigation.