基于多波束水体点云信息的海底气体羽状流快速识别方法研究

Submarine gas seepage is a natural phenomenon which is widely distributed in global marine, and it may affect the global climate change. Meanwhile, the submarine gas seepage is an important sign for the identification of submarine gas hydrates. Modern multi-beam echosounder systems with high resolution and wide coverage, show its great value in seabed backscatter and water column data application. Integrating seabed topography, seabed backscatter and water column data detected in seepage area, combining with the spatial analysis of seabed topography and backscatter intensity, along with the information processing method of water column point cloud model, this study aim to establish a fast recognition model for gas seepage area and gas plumes, develop software packages for recognition of underwater gas plumes based on multi-beam probe information, it can provide technical support for rapid detection of seabed gas plumes.

海底气体渗漏是在全球海洋环境中广泛分布的一种自然现象,渗漏出的气体可能影响全球的气候变化。同时，海底气体渗漏也是天然气水合物成藏区识别的重要标志之一。新一代多波束测深系统分辨率高，覆盖宽，其底质及水体的后向散射数据具备较大应用价值。该项目综合多波束探测的海底气体渗漏区地形地貌特点、底质后向散射强度特征以及含气水体声学特性，结合基于地形特征、底质后向散射的空间分析和水体点云模型的信息处理方法，建立海底渗漏区气体羽状流的快速识别模式，研发基于多波束探测信息的海底气体羽状流识别软件包，可为多波束测深系统快速探测海底气体羽状流提供技术支撑。

本项目基于多波束探测的原始声呐数据，分离提取了海底地形、海底后向散射及水体数据信息，针对多波束海底后向散射数据进行声学底质分类方法研究，针对多波束水体点云数据，实现了水体三维点云信息数据模型构建，重点研究了水体数据噪声压制与点云滤波算法优化，建立海底气体渗漏区羽状流的识别模式，实现了基于全局阈值分割法和基于图像分割及开操作的海底渗漏区气体羽状流的快速识别方法，以及基于数据挖掘方法的精细化海底渗漏区气体羽状流的识别方法，研发了基于多波束水体探测信息的海底气体羽状流识别和提取软件包，可实现基于多波束测深系统的海底气体羽状流目标快速识别与提取。