基于GIS辅助数据的样本像斑变化检测方法研究

Research of change detection using GIS auxiliary data and samples

土地利用变化检测是国内外全球化进程研究的重要内容。虽然利用遥感影像进行变化检测的研究已取得了巨大进展,但由于对象获取方式单一、算法参数需要人为设定/调试等问题使现有变化检测方法在实际应用中的效果大打折扣。提出了一种以GIS辅助数据获取像斑建立样本特征数据库的方法进行遥感影像变化检测。该方法可充分利用多源数据,并提高变化检测方法的自动化程度。实验证明了方法的有效性。

Land use change detection is important in the study of the globalization in the world.The research of change detection using remote sensing images have made great progress.But the auto-detection method is still not practical because of the method of the image segment,the low degree of automation.This paper proposes a new method,which analyzes the images change in segments obtained from the GIS auxiliary data.The sample-auto-selection method and sample-auto-update method can reduce the human involvement.The experiment shows the effectiveness of the method.