摘要
在遥感影像分类应用中,不同分类器的分类精度是不同的,而同一分类器对不同类别的分类精度也是不相同的。多分类器结合的思想就是利用现有分类器之间的互补性,通过适当的方法将不同的分类器之间进行优势互补,往往可以得到比单个分类器更好的分类结果。本文研究了如何在Matlab下采用最短距离分类器、贝叶斯分类器、BP神经网络分类器对影像进行分类,并采用投票法进行多种分类器结合的遥感影像分类,最后进行分类后处理。实验结果表明多分类器结合的遥感影像分类比单一分类器分类的精度高。
In remote sensing image classification applications,the classification accuracies of different classifier are different and the classification accuracies of the same classifier of different categories are not the same.Combination of multiple classifiers is an idea which uses the complementary of the existing classification and find a appropriate way through which we can make use the advantage of different classifiers.Generally,we can get a better one than the classification result of a single classifier.This paper studies how can use the shortest distance classifier,Bayesian classifier,BP neural network classifier to image classification with Matlab.Then we use the voting method to do remote sensing image classification and the final post-processing.The experimental result shows that the accuracy of combination of multiple classifiers is higher than a single one for remote sensing image classification.
出处
《遥感信息》
CSCD
2009年第5期41-43,55,共4页
Remote Sensing Information
基金
基于异源遥感数据的三维重建模型研究
国家自然科学基金(40771159)
关键词
遥感影像分类
多种分类器结合
最短距离
贝叶斯
精度
隶属度
remote sensing image classification
multiple classifiers fusion
shortest distance
Bayesian
accuracy
membership