基于极化似然比的极化SAR影像变化检测

Change detection based on polarimetric test statistic for multi-polarization SAR imagery

由于数据获取困难等问题,目前SAR影像变化检测方法多基于幅度,而较少引入极化信息。针对此方面的不足,以极化SAR数据为研究对象,在分析多极化SAR影像极化特征及其分布模型的基础上,构建极化似然比检验模型,以此进行不同时相的多极化SAR数据地表地物变化程度分析,通过设定恒虚警率确定变化区域,最后考虑地物空间信息剔出斑点噪声引起的孤立检测结果。利用多极化SAR数据进行算法的验证,并与图像比值法进行比较,实验表明:基于极化似然比方法可以有效区分地物的变化情况,且变化检测精度要优于图像比值法。

In this paper,a change detection approach using multi-polarization SAR imagery is proposed. For each pixel,its complex scattering vector follows a complex Gaussian distribution and its polarimetric convariance or coherency matrix follows a complex Wishart distribution. Based on this statistic distribution, a polarimetric test statistic is applied to evaluate the equality of two targets in two-pass polarimetric SAR images. And then,in order to find out the ＇real＇ changed area,a probability of false alarm is applied to get a threshold for change detection. In addition, a majority processing considering the context information of a given target is performed to improve the final accuracy of the change detection. A multi-polarization and multi-temporal SAR data are utilized to validate the proposed approach. Experiment results show the efficiency of the methods presented in the paper and a better change detection accuracy than SAR ratio method for change detection.