中文摘要：

遥感技术作为信息获取的有效手段,广泛应用于气象观测、资源考察、地图测绘和军事侦察等重要领域.遥感影像的归一化植被指数（NDVI）是土地利用、覆盖变化研究的重要指标.本文以Landsat 8获取的郑汴都市区遥感影像为例,分别计算了DN值、辐射定标、大气纠正三个不同处理阶段的NDVI指数,并对三者的关系进行了分析.对比实验研究表明：（1）NDVIDN、NDVIL（λ）、NDVIFLAASH三个指数取值范围逐渐增大,数据相应拉伸.（2）三个指数的均值NDVIL（λ）最小,NDVIFLAASH最大,标准差依次增大.（3）均值—标准差分类显示,相同类别3个指数图各类型面积相差不大,但NDVIFLAASH在高值区域的面积相对减少,低值区域空间上更加分散.（4）样点统计结果表明NDVIDN、NDVIL（λ）、NDVIFLAASH三个指数两两之间的相关系数都在0.95以上,具有极强的正相关关系.（5）总体上,三个NDVI指数既关系密切,又差距很大,本文根据具体情况,给出了不同指数的选择建议.

英文摘要：

Remote sensing technology,as an effective means of access to information,is widely used in meteorological observation,resource investigation,mapping,military reconnaissance and other important fields.The normalized difference vegetation index（NDVI）of remote sensing image is an important index in the field of LUCC（Land-Use and Land-Cover Change）.Based on Landsat 8 remote sensing images of Zhengzhou-Kaifeng Metropolitan Area,the NDVIindexes of three different processing stages including DN value,radiation calibration and atmospheric correction were computed.Contrast experiment research shows that：（1）The range of values of NDVIDN,NDVIL（λ）and NDVIFLAASH increases gradually,and they are stretched with the changes.（2）Three averages of indexes drop before they rise,and the standard deviations increase gradually.（3）The classification according to the averages and standard deviations shows the raster category maps are similar,and the area of the peak value of NDVIFLAASH is relatively reduced,and spatially,the distribution of the low value region is more dispersed.（4）The sample statistic results show that the correlation coefficient between NDVIDN,NDVIL（λ）and NDVIFLAASHis greater than 0.95,with a strong positive correlation.（5）On the whole,the relationship between NDVIDN,NDVIL（λ）and NDVIFLAASH is very close,but the difference between them is obvious.The selection methods are given according to the specific conditions in this paper.