Submarine pipeline tracking technology based on AUVs with forward looking sonar

Abstract

Submarine pipelines have provided an important approach to transport marine oil and gas resources because of their simplicity and efficiency. However, natural or man-made causes can easily lead to pipeline damage and thus result in waste of resources and environmental pollution. At present, the manual routine inspection of submarine pipelines is mainly completed by the survey vessels, which requires a large workload and high cost. The use of autonomous underwater vehicles with multibeam sonar system for real-time surveys is a good solution to this problem. Yet, the accumulation of errors in the integrated navigation system of the vehicle over a long period of operation can cause relatively large position deviations, since the signals of global positioning system cannot be received in the deep sea. To deal with it, the forward looking sonar is adopted in this study to obtain the precise position of the submarine pipeline and a tracking algorithm is proposed to track it. Firstly, the software development kit of the forward looking sonar is used to continuously obtain the pseudo-color images of the submarine pipeline, which are converted into 8-channel gray-scale images for image segmentation and binarization. Secondly, the optimized Rosenfeld algorithm is applied to thin the binary images, after which the submarine pipeline is identified and the corresponding function is fitted to determine the location of the target point to be tracked. Finally, the geodetic coordinates of the target point are determined through coordinate conversion to position the submarine pipeline. A sea trial is carried out and the results prove that accurate and stable pipeline tracking can be achieved by combining the integrated navigation system and the proposed algorithm.