Relief-F筛选波段的植被伪装揭露研究

为了快速揭露植被伪装，基于Relief-F算法进行了高光谱波段选择，将高光谱研究问题转化为多光谱应用问题。首先以常见植物云杉模拟植被伪装目标，利用HH2地物光谱仪采集实验数据，然后引入Relief-F算法筛选特征波段子集，与其它两种常用算法得到的波段子集进行了分类实验。结果表明，使用Relief-F算法筛选特征波段子集分类精度达96.4%，高于其它两种算法。该研究对于揭露植被伪装问题是有帮助的。

Study on vegetation camouflage exposure in Relief-F screening band

In order to quickly expose vegetation camouflage and transform the hyperspectral research problem into the multi-spectral application problem,the hyperspectral band was selected based on the Relief-F algorithm was selected for the study of vegetation camouflage.First,the common plant spruce was used to simulate vegetation camouflage targets,and the HH2 ground-object spectrometer was used to collect experimental data.Then,the author introduced the Relief-F algorithm to screen the subset of feature bands,and conducted classification experiments with the subset of band obtained by other two common algorithms.The results show that the classification accuracy of using the Relief-F algorithm to choose the feature band subset is up to 96.4%,which is higher than the other two algorithms.This research is helpful for exposing the camouflage problem of vegetation.