【摘要】 针对单一分类器的不足,文中采用最短距离分类器、马氏距分类器、K-均值分类器对多光谱遥感影像进行分类,并在测量级的融合方法下进行多种分类器融合分类实验,最后采用混淆矩阵进行分类结果精度评价。实验结果表明:多分类器融合的遥感影像分类方法在精度上高于单一分类器分类。更多还原

【Abstract】 Because of the limitations of the single classifier,the paper classifies the multi- spectral remote sensing image with the minimum distance classifier,Mahalanobis distance classifier and K- means classifier and makes the multiple classifier fusion experiment in the measurement level,based on which the paper makes precision accuracy of the experimental results with Confusion Matrix. It shows that the remote sensing classification method of multi- classifier is more accurate than the single classifier.更多还原