高分遥感影像与矢量数据结合的变化检测方法

Research of change detection using high-resolution remote sensing images and vector data oriented to geographic national conditions monitoring

针对高分辨率遥感影像与矢量数据配准套合之后存在的不一致性问题,该文探讨了多尺度分割算法获取同质像斑的方法。在此基础上,发展了一种基于最近邻分类算法的分类后处理变化检测方法。结果表明,通过多尺度分割算法能够获取"类内光谱相同"和"类间光谱相异"的像斑;分类后处理变化检测方法能够正确检测出80%以上的变化区域,在获取变化检测结果的同时能够获取变化像斑的类别,而且检测精度较高,具有较高的应用价值。

With the combination of two phases high-resolution remote sensing images and vector data, this paper analyzed the inconsistencies after matching vector data and remote sensing images and then proposed a multi-scale segmentation method to obtain homogeneous image segments in the two phase images. According to the result, a post-classification comparison change detection method based on the Nea rest Neighbor algorithm was developed. The results showed that the image segments obtained by the multi scale segmentation algorithm matched the goals of ＂same spectrum within classes＂ and ＂different spectrum among classes＂ . And post-classification comparison change detection method could detect over 80 percent of the changed regions. The method could obtain the image segments categories and had higher precision and higher application value.