【摘要】 遥感影像分类后处理是为了提高分类的精度,优化分类结果。参考传统的分类后处理方法,本文提出了基于数学形态学的遥感影像分类后处理方法,利用数学形态学的基本概念和算法,在遥感软件ERDAS的平台上,对结构元灵活的组合﹑分解,应用形态变换算法达到了消除噪声﹑填补孔穴和光滑边界的效果,最大程度的保留了影像的信息,同时优化了分类后的影像;并实际验证了本文方法较传统方法的优越性与可靠性。更多还原

【Abstract】 In order to improve precision of image classify, we can make use of some traditional ways to optimize results of post classification image. In this paper, I propose post classification with the method of Mathematical morphology. We can make good use of its basic conception and algorithm, such as structural element combining and decomposition and so on. After dealing with the morphological sequence ,we can eliminate hole, noise in the image and smooth bound. At last, we obtain a better result .更多还原