基于遥感指数的城市建城区界定与自动提取

Index-based definition and auto-extraction of the urban built-up region from remote sensing imagery

建筑用地是一类复杂的土地利用类型,难以用简单的方法将它们从遥感影像中准确提取出来.而城市建成区的界定,更是缺乏有效的方法,因此亟需一种科学的方法来准确提取城市建成区.新近提出的IBI建筑用地指数虽然可以将建筑用地较好地从影像中提取出来,但仍无法将城市建筑用地和农村建筑用地区别开来.为此,本文提出一种基于连通区域的新算法.在IBI指数所提取的建筑用地二值影像基础上,通过检测影像中连通区域的方法来鉴别各种建筑用地的聚合体,并根据其面积参数来判别城市和农村建筑用地,完成城市建成区的界定与自动提取,精度可达96%以上.

The extraction of built-up land information from remote sensing imagery is not an easy task due largely to the complicated composition of the built-up land class.The exact definition of the urban built-up area is even more difficult and still lack of an effective method.Although the recently proposed IBI（index-based built-up index） can effectively extract built-up land information from remote sensing imagery,the index still cannot distinguish between urban and rural built-up lands.Therefore,this paper proposed a new algorithm to work out this problem.It first detects connected areas from the IBI-extracted built-up land image and then recognizes the urban built-up region based on the size of the connected areas,because an urban built-up region is usually larger than a rural built-up region.The results show that the new method can effectively auto-extract the urban built-up lands and determine urban built-up region from remote sensing images with overall accuracy of over 96%.