sar图像反演浅海地形的一种改进方法_电子测量技术2012.docVIP

SAR图像反演浅海地形的一种改进方法 李泽军1，2 王小青1 中国科学院电子学研究所微波成像技术重点实验室,北京 1001902 中国科学院研究生院,北京 100149hallow Sea Topography using SAR Image Li Zejun1，2 Wang Xiaoqing1 Yu Xiangzhen1，2 Liu Lei1，2 (1. Science and Technology on Microwave Imaging Laboratory, Institute of Electronics, Chinese Academy of Sciences, Beijing 100190,China;2. Graduate University of the Chinese Academy of Sciences, Beijing 100190,China ) Abstract Traditional Synthetic Aperture Radar (SAR) underwater topography detection is developed by optimizing the depth with iteration model only, the algorithm is too slow. In this passage, AH model is used to get the first guess shallow water depth as an improvement, and then the depth is optimized with iteration model, thus accelerates the iteration. Based on this algorithm, water depth of Taiwan Shoal is retrieved with 1 set of ERS-2 and 1 set of ENVISAT SAR image. The relative mean errors between the retrieved depths and measured ones are 8.9% and 10.8% respectively, and the iteration times reduced. It is indicated that the absolute error between the retrieved depth and measured depth is not a good standard to evaluate the accuracy of water depth retrieved from the SAR image. Thus we proposed to use the relative change of the topography as a standard of the retrieval accuracy, by taking off the base topography depth. Key words Synthetic Aperture Radar (SAR) Shallow Water Topography Retrieval 0 引 言 浅海水下地形信息不仅海上交通运输、浅海油气开发、海底管线埋设等工程，也是军事海洋环境动态监测管理的一项重要内容。水下地形。合成孔径雷达(Synthetic Aperture Radar, SAR)是浅海水下地形探测的重要手段之一。目前较为成熟的是Wensick和Vogelzang 等1993年建立的水深估算系统(Bathymetry Assessment SystemBAS)”[3]。然而，BAS需要密集实测水深数据(横向间隔不超过200m)作为输入，在此前提下系统获得了0.3m的水深反演精度。国内，等采用根据物理模型M4S(Model for Fully Two-dimensional Simulation)进行迭代反演，该方法需要提供并修正风和流参数。杨俊刚等建立最小代价函数迭代反演水深误差小于10%。Alpers和HenningsSAR浅海水下地形理论SAR浅海地形反演本文水深反演算法的流程如所示首先对SAR图像进行预处理，使用AH模型产生水深。将该水深带入模型。 图1 浅海水下地形反演流程图的预处理包括将斜距图像转化为地距图像，使得图像的位置与地理位置能准确对应，便于计算。然后对图像进行中值滤波和多次均值滤波，除去斑点噪声，并将图像变平滑。.2 初步水深反演 Alpers和Hennings最早于1984年提出基于Bragg