汶川地震震灾图像处理中的融合方法及其比较——以北川地区遥感影像为例

Image Processing and Comparison of the Wenchuan Earthquake Using Fusion methods --Case Study of the Beichuan Area

图像融合是一种对多源图像信息进行提取与综合处理的技术，其任务是根据多源图像信息进行提取与综合，以获得对同一场景、目标的更准确、更全面、更可靠的图像描述。本文以北川地区的遥感图像为例介绍了汶川地震中遥感图像处理使用的图像融合方法，包括IHS变换，主成分变换（PCA），高通滤波和小波变换等，肯定了图像融合在震害信息提取中的作用。然后对各种方法进行了定性与定量地对比、分析及评价，根据各种方法的特点，分析了不同应用目的所适用的遥感图像融合方法。

Image fusion is a kind of technology dealing with information extraction and comprehensive treatment of multi-source images. The tasks of image fusion are to fuse the original images of the same target obtained by multiple sensors and to improve the description of image. Many kinds of methods based on pixel-level, such as IHS transform, principal component analysis（PCA）, high pass filter and wavelet transform are introduced in this paper. As a case study, these methods are applied to the Beichuan area and proved to be valuable in image process of the disaster area after the Ms8.0 Wenchuan earthquake. Comparison and analysis in both quality and quantity are given in several aspects. The applications of image fusion methods for different purposes are discussed according to the characteristics of the methods.