结合MODIS植被指数时间序列的ＥＴＭ影像分类方法研究
Study of land cover classification based on MODIS Vegetation Index Time Serials and ETM image
MODIS数据由于其高时间分辨率和高光谱分辨率的特点，使其在全球及区域土地覆盖分类的研究中具有显著的优势。尤其MODIS植被指数产品（MOD13）能够反映植被生长的季相变化特征，并能够很好的区分植被区域的覆盖类型。本文利用NDVI时间序列可以有效区分植被区域这一特点，首先利用NDVI时序数列进行植被区与非植被区的分类，并制作掩膜进一步指导ETM影像的分类。并将所得的分类结果与传统的最大似然分类结果进行比较得出利用本文的分类方法可以提高对植被的分类精度。
MODIS data with high spectral and temporal resolution has more advantages in global and regional land cover classification. In particular, NDVI increased with the growth of the vegetation, and gradually decreased after reaching the maximum at some growth stage. This characteristic is similar to hyper-spectral data. This study relies on an annual time series of ten MODIS 16-days composited images (MOD13) monthly acquired during the year 2000. Support Vector Machine(SVM) are firstly employed to classify two classes, vegetated region and non-vegetated region. Then build mask separately to guide ETM image classification. Finally compares this result with classical classification by MLC.
蔡文婧、羌云娟
环境科学技术现状工程设计、工程测绘
MODISNDVI土地覆盖分类SVM最大似然法
MODISNDVIland cover classificationSVMmaximum likelihood classification
蔡文婧,羌云娟.结合MODIS植被指数时间序列的ＥＴＭ影像分类方法研究[EB/OL].(2010-05-06)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201005-110.点此复制
评论