SAR变化检测技术发展综述

The Art in SAR Change Detection-A Systematic Review

合成孔径雷达(SAR)具有全天候、全天时的特点,是很好的变化检测信息源,研究SAR图像变化检测技术有着非常广阔的应用前景。系统介绍了SAR变化检测的整个流程——数据预处理、变化检测算法、检测后处理及变化检测算法评价。为了方便算法设计者,分门别类地阐述了变化检测算法的主要方法及核心决策规则,包括差值法、统计假设检验法、预测模型、相干模型;并对每种方法的特征进行了分析和比较。针对SAR技术的迅猛发展,我们对将来SAR变化检测技术的发展方向给予了预测。

SAR system has important properties at the operational level, such as its capability of acquiring data in all weather and time conditions. SAR imagery is an important source data for change detection and its application is so promising. In this paper, we systematically describe the whole process of change detection using SAR images. Firstly, we present the pre-process of SAR images, which comprise of calibration and geometric correction. In the second part, we present the art in SAR change detection, including the methods of difference image, hypothesis test, predictive models and coherence models. As the development of SAR technique and computer science, more and more new approaches for change detection using SAR images are proposed; in this paper, we illustrate some representative works. Then, the principles for performance evaluation and the post-process of change detection are discussed. Lastly, the oriented development of SAR change detection is discussed at our viewpoint. We hope that our classification of algorithm using SAR images into a relatively small number of categories will provide useful guidance to the algorithm designers in SAR researcher field.