国产高分一号数据估算草地植被覆盖度方法研究——以呼伦贝尔草原露天煤矿区为例

发布时间：2018-03-12 17:22

本文选题：GF-卫星　切入点：SPOT卫星　出处：《地理与地理信息科学》2017年02期 　论文类型：期刊论文

【摘要】：为了探讨与分析国产高分一号(GF-1)数据在北方露天煤矿区草地植被覆盖度估测中的精度及适用性,该文基于GF-1与SPOT6多光谱影像数据,以多个植被指数为自变量,利用像元二分模型、偏最小二乘(PLS)回归、支持向量机(SVM)回归3种模型对区内植被覆盖度进行估算,结合野外同步实地植被样方数据,对比分析不同估算模型的精度及适宜性,并通过蒙特卡洛模拟多尺度交叉建模的误差传播,分析空间分辨率不同对植被覆盖度估测的精度影响。结果表明:GF-1数据基于增强型植被指数的SVM回归模型(R~2=0.8149,RPD=2.336,RMSE=8.694%)与SPOT6数据基于归一化植被指数的SVM回归模型(R~2=0.8755,RPD=2.870,RMSE=7.032%)估算效果较好。不同分辨率数据交叉传递过程中SVM回归模型的精度高于PLS回归模型。因此,基于GF-1数据构建的SVM回归模型可以高精度地估算区域草地植被覆盖度。
[Abstract]:In order to discuss and analyze the accuracy and applicability of GF-1) data in the estimation of grassland vegetation coverage in northern opencast coal mine area, this paper takes multiple vegetation indices as independent variables based on GF-1 and SPOT6 multispectral image data. Using pixel dichotomy model, partial least squares regression and support vector machine SVM regression, the vegetation coverage in the area was estimated. The accuracy and suitability of different estimation models were compared and analyzed based on field synchronous vegetation sample data. The error propagation of multi-scale cross-modeling is simulated by Monte Carlo. The effect of spatial resolution on the accuracy of vegetation coverage estimation was analyzed. The results show that the SVM regression model based on enhanced vegetation index (SVM regression model) and the SVM regression model based on normalized vegetation index (R20.8755RPD2.870RMSE 7.032) and SPOT6 data are used to estimate the accuracy of vegetation coverage estimation. The accuracy of SVM regression model is higher than that of PLS regression model. The SVM regression model based on GF-1 data can accurately estimate the vegetation coverage of grassland.
【作者单位】：东北大学测绘遥感与数字矿山研究所;
【基金】：国家自然科学基金青年基金项目(41401233)
【分类号】：TP79

【相似文献】