机场区域变化检测研究

区域变化检测在遥感信息处理中有着越来越重要的作用。但现有的变化检测方法存在两个缺点,一是需要很多的图像数据,二是难以取得足够多的定量信息。利用单波段机场遥感图像,采用Hough变换检测出机场跑道,通过区域生长划定感兴趣区域。在简单几何校正的基础上,用边缘关联的方法判断有无变化。最后用弹性模板匹配的方法对变化作出解释。试验取得了令人满意的结果。

Study on Airport Change Detection

Change detection is playing more and more important role in remote sensing information processing. However, there are two aspects of shortcoming in existing change detection algorithms. One is that too large amount of raw data is usually needed to implement the change detection. The other is that it is sometimes difficult to acquire enough amount of quantitative information. In this paper, the runway in airport is detected by taking Hough transform in single-band images. The region of interest is then located by region-growing. After a coarse step of calibration, a judgement is made on whether there is change occurring within ROI based on the edge-correlation. Finally, the elastic mask matching procedure is adopted to make an explanation to the changes detected. Satisfactory results are obtained in the experiments.