一种面向对象的居民地变化检测方法  

A Method of Object-Oriented Residential Area Change Detection

作　　者：陈鑫镖[1] 赵仁亮[2] 周晓光[1] 

机构地区：[1]中南大学地球科学与信息物理学院,湖南长沙410083 [2]国家基础地理信息中心,北京100830

出　　处：《地理信息世界》2012年第6期50-53,共4页Geomatics World

基　　金：国家"863计划"项目(2009AA122004)资助

摘　　要：文章提出了一种新的面向对象的居民地变化检测方法。首先对同一地区的新旧影像进行影像分割,然后选取训练样地,根据居民地的各种特征得出决策规则,进而提取居民地,最后通过叠加分析的方法检测出居民地4种变化形态。实验证明本方法能够有效地检测出居民地的变化信息,总体精度达到了86.15%。This paper presents a novel method of object-oriented residential area change detection. Firstly, image segmentation is applied to the two images of the same region at different time. Then we summarize the decision rule according to features of residential area obtained by training samples in the images, and extract the residential area. Finally, by using overlay analysis, four change types of residential area are detected. Experimental results show that this method can deteel change information of residential area effectively. The overall accuracy is 86.15%.

关 键 词：居民地 面向对象 变化检测 叠加分析 

分 类 号：TP311.132.4[自动化与计算机技术—计算机软件与理论]