一种融合MBI的多特征建筑物变化检测方法  

A Method for Multi-feature Building Change Detection Based on MBI

作　　者：俞晓飞 YU Xiaofei(Ningbo Civil Building Design ＆ Research Co.Ltd.,Hangzhou 315000,China)

机构地区：[1]宁波市民用建筑设计研究有限公司,浙江杭州315000

出　　处：《测绘与空间地理信息》2018年第11期144-146,150,154,共5页Geomatics & Spatial Information Technology

摘　　要：多特征遥感影像变化检测是目前变化检测研究的趋势与方向,本文基于形态建筑物指数(MBI)改进模型与特征融合方法,提出了基于证据融合的城市建筑物变化检测方法。该法首先利用搜索窗口与匹配窗口计算两时相影像的结构相似度,然后利用Dempster-Shafer理论构造基本概率赋值函数,由函数值界定的阈值决定像素变化,进而提取变化信息。实验发现,融合MBI的多特征建筑物检测方法能有效提高变化检测精度,且对影像辐射校正要求较低,明显优于单特征变化检测。It is a trend and direction that multi-feature change detection are used in a remote sensing imagery currently, based on the improved MBI model and feature fusion, a city building change detection method is proposed in this paper. This method firstly used the search window and matching window to calculate the structural similarity of two imageries, then constructed basic probability assignment function by using Dempster-Shafer theory, thus pixels change was decided by the threshold that made by the function, and the change information was extracted. The experimental results show that the multi-feature building detection method based on MBI can effectively improve the detection accuracy and reduce the requirement of image radiometric correction, obviously better than single feature change detection.

关 键 词：变化检测 基本概率赋值函数 结构相似度 形态建筑物指数 

分 类 号：P237[天文地球—摄影测量与遥感]