利用决策级融合进行遥感影像分类

Classification for Remote Sensing Data with Decision Level Fusion

提出了一种基于决策级融合的遥感影像分类方法。该方法对遥感影像特征以最大似然分类器进行预分类,应用Adaboost算法将分类的结果进行决策级融合,实现影像的分类。实验结果表明,该方法的分类精度较传统分类方法有明显的提高。

With the development of remote sensing technology, dealing with high-dimension features with traditional classification methods is difficult. Multiple classifiers fusion technology not only deals with high-dimension features but also improves the classification accuracies. We focuses on classifier fusion in decision level, and proposes a new classification method for remote sensing data based on Adaboost. Experiments show that this method is more effective than traditional classification algorithms.