基于SonarWiz的多波束声纳图像智能底质分类技术研究

Research on intelligent seabed sediment classification technology of multibeam sonar image based on SonarWiz

多波束回波强度信息与海底底质类型具有较强的相关性,通过海底声纳图像能够实现底质类型的划分。为提高海底底质分类质量,依托SonarWiz的智能底质分类优势,在海底声纳图像纹理特征自动分类基础上,引入地形属性信息修正分类结果。以三亚崖州湾附近海域为例,基于实测海底地形数据和海底表面声纳图像,利用数据处理技术和图像分类方法,完成对该区域海底声纳图像处理分析和海底表层沉积物的分类研究,获得试验区海底底质分类成果。该方法对未知海底沉积物类型的大面积底质分类研究,以及港口、抛锚地等重点区域的底质分类工作具有借鉴意义。

The intensity information of multibeam echoes have a strong correlation with the properties of seabed sediment.According to the surface texture structure of sonar image,seabed sediment types can be classified.Relying on the advantages of intelligent sediment classification of SonarWiz,this study introduces terrain attribute information to modify the classification results on the basis of automatic classification of texture features of seabed sonar images for improving the quality of seabed sediment classification.The sea area near Yazhou Bay in Sanya is selected as the test area.Based on the seabed terrain data and surficial seabed sonar image,the data processing technology and image classification method are used to complete the processing and analysis of seabed sonar image and classification of surficial seabed sediment in this area.The classification result of the seabed sediment in the test area is obtained.This method can be used as a reference for the analysis of large-area sediment classification in unknown seabed sediment types and the classification of sediments in key areas such as ports and anchorages.