一种高分辨率遥感图像中居民区道路提取方法  

A New Algorithm of Road Extraction from High-Resolution RS Image of Residential Area

作　　者：满亢[1] 鲍远律[1] 马璐[1] 

机构地区：[1]中国科学技术大学自动化系,安徽合肥230027

出　　处：《计算机仿真》2009年第3期217-219,230,共4页Computer Simulation

基　　金：国家自然科学基金(50708107);国家高技术研究发展计划(863计划)(2007AA11Z222)

摘　　要：针对在遥感图像中提取居民区道路易受房屋等人工建筑干扰的问题,基于数学形态学提出了一种可以去除大量建筑物干扰,从而有效提取出居民区道路的算法。算法首先用顶帽变换和底帽变换对灰度图像进行了对比度争强,然后利用道路和建筑物之间的形态梯度以及道路特征对道路和建筑物进行分离,最后利用形态重建的方法得到了居民区道路网。仿真实验表明,区域中的大部分建筑物噪声被成功去除,算法是有效和可行的。The buildings such as house and square are the main noise sources of road extraction from High - Resolution RS（ Remotely Sensed） Image of Residential Area. To solve this problem, an algorithm based on mathematical morphology is proposed. Firstly, the top - hat and bottom - hat translation are used to enhance the contrast of the image. Secondly, the morphological grads between roads and buildings and the features of the road are used to separate the buildings from the road. Finally, the morphological reconstruction is used to get the road network. The simulation experiment shows that most of the noise is removed effectively and this algorithm is effective and feasible.

关 键 词：遥感图像 道路提取 数学形态学 道路特征 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]