Research On Seabed Sediment Classification With Multibeam Sonar And Software Designing

With rapid development of multibeam sonar technology, single water depth measurement already can't satisfy people's needs. No matter in the excavation industry,fishery, and environmental science research, or in military defense aspect, people want to bo able to get information on sea bottom sediment efficiently and quickly. However, traditional mechanical sampling method and visualization method have many disadvantages such as high cost, low efficiencies and big intensity of work, which are not suitable for classification and recognition of large area substrate. In this case, doing research on sediment classification, marine surveying and mapping and ocean development with acoustic method has important significance.Multibeam sonar system works in the form of stripes, in which each stripe width can reach 5-9 times of the depth of the water. With high resolution water depth data, we can also get underwater acoustic image map like a sidescan sonar. We can use backscatter intensity data of multibeam sonar system and sediment classification method to classify measuring area of the sediment. Therefore, using the data got by multibeam sonar system to classify sediment has a broad application prospect. This paper mainly studies the sediment classification method based on the backscatter data of multibeam sonar system, which includes three parts:Firstly, the working principle of multibeam sonar system is introduced to clear the way how the error of the backscatter intensity data causes. Then, in order to get pure data which can reflect the seabed sediment characteristics, the paper studies the method of data preprocessing method, including the transmission loss, Lambert law revision and beam irradiation area correction. The paper deals with some experiment data with the preprocessing method to verify the correctness of it.Secondly, the feature vector and bottom sediment classification commonly used on sediment classification are studied in this paper. Then, the paper selects the mean and standard deviation of the backscatter intensity as a feature vector, and selects the Bayes maximum likehood classification algorithm and applies it on sediment classification. Then, the paper verifies the feasibility of sediment classification by using Bayes classification algorithm.Finally, the sediment classification software is designed in this paper. The paper introduces the working process and the specific function of the software through actual operation pictures , and it shows the stability and practicability of the software through data processing.