广东省主要人工林树种光谱分类

为了探索不同树种叶片光谱区分的最佳特征波段以及不同树种光谱分类的 性能，利用地物光谱仪对广东省10个主要人工林树种的叶片光谱数进行了采集。采用遗 传算法(Genetic Algorithm, GA)和连续投影变换算法(Successive Projections Algorithm, SPA) 进行了高光谱数据降维处理，然后结合支持向量机(Support Vector Machine, SVM)和随机森林(Random Forest, RF) 两种方法进行了树种分类。研究发现，通过两种变量选择方式筛选的用于树种分类的光谱范围 主要位于近红外波段。其中，经GA算法筛选的变量建模和预测精度与基于全波段光谱数据的分类精 度较为接近，且比经SPA算法变量筛选的分类结果好。通过比较可知，RF算法所建模型的性能比SVM算 法更稳定，且GA-RF算法在几种分类处理中的性能最好。结果表明，GA-RF算法 可用于基于光谱数据的树种分类研究。

Classification of Main Tree Species of Plantation in Guangdong Province by Leaf Spectra

To explore the optimal feature wavebands for spectral discrimination of different tree species and the ability of classifying different tree species by spectroscopy, the leaf spectra of 10 tree species in the plantation in Guangdong Province are collected by using a spectroradiometer. The Genetic Algorithm (GA) and Successive Projection Algorithm (SPA) are used to reduce the data dimension. Then, the Support Vector Machine (SVM) and Random Forest Algorithm are used to classify the tree species. The result shows that the wavebands selected by two data dimension reduction methods are mainly located in the near infrared region. GA performs better than SPA in variable selection. The model established by RF is more stable than that established by SVM in performance. The GA-RF algorithms can be used for the tree species classification based on spectral data.