Underwater Gas Target Detection Based On Sonar Image

With the development of Marine industry,people pay more and more attention to the research of underwater gas target.Underwater gas target is often closely related to seabed mineral resources,and underwater gas is generally released in the process of offshore oil and gas development.Therefore,it is of great significance to study underwater gas target detection technology for exploiting Marine resources and protecting Marine environment.Based on the national major science and technology project,this paper studies the underwater gas target detection technology based on sonar image according to the actual engineering requirements.Firstly,this paper summarizes various methods of underwater gas target detection at home and abroad,and highlights the advantages of using multi-beam sounding sonar to detect underwater gas targets through comparative analysis.Secondly,the beamforming algorithm based on multi-beam depth sounding sonar is studied.In view of the features of CBF algorithm,such as side lobe height and resolution difference,MVDR algorithm and DCBF algorithm are respectively studied,and the performance of the algorithm is investigated and verified through simulation test.In order to detect the underwater gas target,the image transformation algorithm,image preprocessing algorithm and image segmentation algorithm based on sonar image are further studied.The image transformation algorithm converts the sonar image in polar coordinate system into the sonar image in Cartesian coordinate system through interpolation processing.The image preprocessing algorithm can suppress the circuit noise and seabed reverberation interference and improve the image quality.The image segmentation algorithm is to segment the pre-processed image and successfully extract the underwater gas target.Finally,in view of the spatio-temporal transformation of underwater gas targets,Hu invariant moment feature and inter-frame feature are combined to extract the feature of segmated underwater gas targets,and the identification of underwater gas targets is completed based on the support vector machine platform.By processing the actual test data of Songhua Lake,South China Sea and noise-suppressing pool,this paper verifies the feasibility and effectiveness of each algorithm,and realizes the underwater gas target detection based on sonar image.