论文部分内容阅读
讨论了面向对象的高空间分辨率遥感影像地质灾害体信息提取方法.该方法首先利用像元的色调和形状特征判别所要生成的影像对象的异质性,采取自下而上区域合并的方法将影像划分为互不相交的影像对象,然后根据这些对象的影像特征,如光谱、纹理、形状和上下文等,进行分类,提取出地质灾害体.结果表明,该方法能有效地提取地质灾害信息,为震灾调查和快速预评估提供决策支持数据.
This paper discusses the object-oriented method of extracting geological hazard information from high spatial resolution remote sensing images by using the hue and shape features of pixels to discriminate the heterogeneity of the image objects to be generated, taking the method of bottom-up regional combination The images are divided into disjoint image objects, and then the geological disasters are extracted according to the image features of the objects such as spectrum, texture, shape and context, etc. The results show that the method can effectively extract geological hazard information, Provide decision support data for earthquake disaster investigations and rapid pre-assessment.