摘要
本文利用多时相Landsat TM/ETM+影像分析了兖州市1998年和2002年的土地利用/覆盖变化。综合考虑波段间相关系数和OIF指数,选择最佳波段组合进行图像解译,并在此基础上运用最大似然分类器(MLC)和支持向量机(SVM)的分类方法对遥感影像进行分类。进而利用SVM分类结果进行土地利用遥感动态监测,获取兖州市土地利用/覆盖变化信息,并与社会经济统计资料的统计结果进行比较。最后提取TM/ETM+影像的RDVI,基于线性混合像元分解模型分析了植被覆盖的变化。结果表明,基于多时相TM/ETM+影像分析的土地利用/覆盖变化与实际统计数据较吻合,适合动态监测土地利用变化,且精度较高。
The Land Use/Cover Change （LUCC） of Yanzhou city was analyzed based on multi-temporal Landsat TM/ETM ＋ data in 1998 and 2002 in this paper. Considering the correlation matrix of bands and OIF index, optimal bands combination was selected to interpret images, and then the images were categorized using Maximum Likelihood Classifier （MLC） and Support Vector Machines （SVM） method. By comparing, the dynamic monitoring information obtained from SVM classification results was closed to the regional social and economic statistic data. Lastly, the vegetation cover changes were discussed based on RDVI data and linear mixed spectral model as special interests. The results indicated that dynamic monitoring LUCC using Landsat TM/ETM ＋ data has high accuracy.
出处
《测绘科学》
CSCD
北大核心
2009年第3期171-174,240,共5页
Science of Surveying and Mapping
