中文摘要：

利用陕西省西安市阎良区Quickbird卫星遥感数据建立归一化植被指数模型提取城市绿地信息,这种基于光谱的信息提取模型受到＂同物异谱＂和＂同谱异物＂信息的干扰,影响了信息提取的精度和速度.通过地物判读发现＂同谱异物＂信息主要为蓝色房顶、白色建筑物,而＂同物异谱＂信息主要为城市道路两旁的行道树.对这几种地物类别同实际城市绿地的植被信息进行光谱采样对比分析,建立了避免蓝色房顶及白色建筑物被同时提取的植被提取光谱模型和消除＂同物异谱＂现象的植被提取光谱模型,经过实验验证,方法可行.

英文摘要：

The paper used the current commercial satellite Quickbird to establish NDVI model for extracting urban space information in Yanliang district of Xi＇an city in Shaanxi province.The model based on the spectrum was interfered by the information concerning ＂same object with different spectrums＂ and ＂same spectrum with different objects＂which affected the accuracy and speed of the information extraction.The interpretation found that the information concerning ＂same spectrum with different objects＂ was mainly the blue roofs and white architectures,while the information regarding ＂same object with different spectrums＂ was mainly avenue trees.Furthermore,these types of surface features were compared with the real vegetation information of the urban green space with spectrum sampling skill.The vegetation extraction spectrum model was established to avoid extracting the blue roofs and white architectures at the same time and to eliminate the phenomenon of ＂same spectrum with different objects＂.This method was verified to be feasible.