基于TM遥感影像的分类方法研究与探讨

为了能够更加方便、快捷和准确地采用分类方法对遥感影像进行分类，对lsodata分类法、最大似然分类法以及决策树分类法这三种遥感图像分类方法进行了机理分析，并以阜新TM遥感影像作为实验数据，对这三种方法进行实验数据比较分析，实验数据表明决策树分类法精度最高，不仅能有效利用地物的光谱信息，而且增强了地物非光谱信息对分类结果的作用。

Remote Sensing Image Classification Method and Discussion Based on the TM

In order to enable the people to classify the remote sensing image more convenient a more, faster and more accurate with the classification of the use of remote sensing image,the paper adopts three classification methods to make the mechanism analysis lsodata classification, maximum likelihood classification and decision tree classification and takes Fuxin TM remote sensing image as the experimental data to compare with this three methods. The experiment shows that the accuracy of decision tree classification is the highest which not only make full use of spectral information of the features and improves the effect of the results that the non-spectral information of the features works on classification.