摘要
遥感影像上的建筑物是一种非常重要的地物。利用高分辨率遥感影像上建筑物的光谱、空间特征及上下文特征等,并通过贝叶斯理论计算对象的归属概率属性,提出了一种面向对象的高分辨率影像建筑物提取方法。实验证明该方法有较好的提取效果。
Remote sensing image building is a very important surface features.High-resolution remote sensing spectral images of buildings,spatial characteristics and context characteristics,and calculated by Bayesian probability of property ownership of the object,a object-oriented high-resolution imaging Building Extraction.Experiment shows that the better extraction.
出处
《城市勘测》
2011年第1期84-86,共3页
Urban Geotechnical Investigation & Surveying
关键词
面向对象
对象归属概率
贝叶斯理论
规则
Object-oriented
The probability of an object belonging
Bayesian theory
Rule
