Research On Key Technology Of Submarine Pipeline Detection With Side Scan Sonar

As an important carrier for transporting oil and natural gas,submarine pipelines are laid on the bottom of the sea for a long time.Under the influence of the complex environment of the sea,there will be many phenomena such as hanging and bare with a huge risk of security.At present,the detection of submarine pipeline state mostly relies on acoustic devices such as side scan sonar and subbottom profiler.We estimate the state of submarine pipelines by the detected acoustic images,existing low efficiency and poor accuracy.This paper did a systematic study on the detection of submarine pipelines based on side scan sonar and get the initial results.(1)According to all kinds of information for the side scan sonar imaging principle,recombing the correct imaging mechanism,the form and applicability of the emission signal are analyzed,The concepts of several performance detection indexes of side scan sonar system are distinguished and its influence factors are analyzed.(2)Designing the data acquisition process and method,collecting the data by side scan sonar system,several states of the submarine pipeline in sonar image are analyzed,according to the measured side scan sonar images and the depth of water,a rough interpretation of the states for the submarine pipeline is made.(3)In order to evaluate the existence of the submarine pipeline more accurately,the original XTF format side scan sonar data are decoded.Analyzing the each step from data parsing to the geographically coded map of the side scan sonar,a joint single beam seabed detection method is proposed,finally we get corrected geographically coded map of the side scan sonar,Finally,each algorithm of preprocessing is encapsulated and integrated,and a set of submarine pipelines side scan sonar data visualization system is formed,all information of the raw data can be quickly viewed,extracted and process.(4)According to the obtained side scan sonar image,we discussed the noise types of sonar image,studying the denoising algorithms and segmentation algorithms of side scan sonar image,through the test we found wiener filtering is best for Ping filtering and bilateral filtering is best for image filtering.The side scan sonar image segmentation mainly focuses on the research of fuzzy clustering algorithm and threshold segmentation algorithm,the experimental results show that the effect of clustering algorithm for experimental data is poor,this paper presents a dual threshold method based on histogram double dip,and compared with other threshold segmentation algorithm.The results show that the double threshold Ostu algorithm made the optimal result,the pipeline shape recovery method by its shadow is proposed,extracting the shadow and target boundary coordinates by the Canny operator,judging the distribution state of the submarine pipeline and calculating its hanging height,Compared with the measured data of the single beam cross section data,the validity of this method is verified,Finally,each algorithm of image processing is integrated to develop a set of seabed pipeline inspection processing software.