面向林地信息提取的 Pleiades-1影像融合算法评价分析

以Pleiades-1影像为对象,以探究其面向林地信息提取的融合算法选择为视角,通过基于目视和定量特征分析、辅助面向对象分类分析的方法,为林业部门遥感影像大规模融合应用提供参考。研究认为Pansharp和GramSchmidt算法融合结果目视效果良好,各波段与原多光谱相关系数均高达0.8以上,清晰度与纹理增强明显。两种融合算法影像在不同林地层次信息提取能力各有优势,Pansharp融合结果在林地层次分类总精度可达86.55%,Gram-Schmidt融合结果则在森林类型层次具有最高的分类总精度78.76%。具体融合算法的选取需根据其应用的信息提取层次而定。

Evaluation analysis of fusion algorithms for Pleiades-1 data oriented to extraction of forestland information

Base on the Pleiades-1 data ,the optimal fusion algorithm is selected in the perspective of the extraction of forestland information .The fusion images are compared and analyzed by using both visual judgments and statistical method .Furthermore ,the object-oriented classification method is used to evaluate the performance of the fusion methods at information level .This paper is supposed to provide the reference for large-scale image fusion application in forestry industry .The result indicates that Pansharp and Gram-Schmidt fusion images have good visual effect ,strong ability of spectral preservation with the correlation coefficient up to 0.8 high spectral fiedelity and clear texture .In term of the classification accuracy ,each fusion algorithm has an advantage in extraction of forestland information of different levels .Pansharp fusion image has highest overall accuracy(86.55% ) at forestland level ,and Gram-Schmidt transformation is superior to that of other fusion algorithms at forest type level(78.76% ) .Specific fusion algorithm shall be chosen differently from its application levels .