基于广义高斯模型的SAR幅度图像震害检测

本文以2008年5月12日汶川地震前后的星载ALOS 合成孔径雷达图像作为数据源, 利用广义高斯分布模型作为差异图像中变化类和未变化类的分布模型, 在假设变化和未变化像元服从广义高斯分布的条件下, 估计变化和未变化像元的概率密度参数, 采用KI算法计算最佳分割阈值并提取变化区域。 以都江堰地区为例, 自动检测这次地震导致的地表变化, 快速圈定一些变化差异较大的区域。 研究表明, 该变化检测算法能较准确地描述差异影像的分布, 能够在地震灾害提取中发挥作用。

Earthquake Disaster Assessment Based on Generalized Gaussian Model Using SAR Amplitude Data

In this paper we use ALOS data before and after the 2008 Wenchuan earthquake to evaluate the seismic damages caused by the earthquake. Surface change caused by the earthquake has been automatically detected based on the generalized Gaussian model and KI optimal threshold value change detection method. Under the assumption that the changed pixels and unchanged pixels follow generalized Gaussian distribution , probability densities of the two classes of pixels are estimated. The KI threshold selection criterion is derived under the generalized Gaussian assumption for modeling the distributions of changed and unchanged classes. Applying this method to the city of Dujiangyan, some serious changed regions can be quickly identified. The results show that the changed map can be extracted based on the optimal threshold. SAR technique is an effective means to monitor natural disasters due to its all-weather characteristics.