基于Bagged CART和Boosted CART的高光谱影像分类技术研究

Research on Hyperspectral Image Classification Based on Bagged CART and Boosted CART

本文针对高光谱影像数据光谱分辨率高,数据量大的特点,采用以CART决策树为弱分类器的Bagging和Boosting集成学习算法对该影像进行分类,通过实验分析比较,体现出了Bagged CART和Boosted CART算法用于分类时的有效性和准确性。

This paper classed the hyperspectral image using Bagging and Boosting that makes CART be weak classification , which aims at the characteristic of high spectral resolution and massive data. The experiment represented the validity and veracity of Bagged CART and Boosted CART when they classed.