为了准确快速地获取高分辨率影像中橡胶林的分布信息,设计了一种基于纹理特征和多光谱特征的信息提取方法。方法选取合适的植被指数,将多光谱和植被指数的影像进行地统计半方差分析,获得最佳纹理提取窗口并实现各种纹理信息的提取,将纹理信息和光谱信息一起作为参考特征构建地物的分类规则并用C5决策树分类算法实现。选取某高分辨率遥感影像区域对该方法进行验证,橡胶树林提取的生产者精度为81.00%,提取用户精度为82.65%,总精度为83.50%,Kappa系数为0.78。与其他方法分类结果对比表明,本文方法是一种有效的橡胶林提取方法。 In order to acquire the distribution information of rubber plantations in high-resolution images accurately and quickly, an information extraction method based on texture features and multi-spectral features was designed. Methods The appropriate vegetation index was selected, and the semi-variance analysis of geostatistics was performed on the images of multi-spectral and vegetation indices to obtain the best texture extraction window and to extract various texture information. Texture information and spectral information were used as reference features to construct features Classification rules and use C5 decision tree classification algorithm. The method was validated by selecting a high-resolution remote sensing image. The accuracy of the extracted tree was 81.00%, the precision of extraction was 82.65%, the overall accuracy was 83.50%, and the Kappa coefficient was 0.78. Comparison with the results of other methods shows that this method is an effective method for rubber plantation.