基于ENVI的遥感影像监督分类方法的研究

Research on Supervised Classification Method of Remote Sensing Images Based on ENVI

随着遥感技术的发展和遥感图像采集方式的多样性,对遥感图像处理技术的要求更高.文章介绍了三种常见的监督分类算法:支持向量机、最大似然法、BP神经网络;并利用上述三种算法对南泥湾地区同一幅SuperView-1遥感影像进行了分类,获得分类解果并评估准确性;然后对三种算法进行了比较,分析三种算法的优缺点,得出支持向量机分类精度最低,最大似然法次之,神经网络最高;最后得出结论,BP神经网络是一种较为优良的遥感影像分类算法.

With the development of remote sensing technology and the diversity of remote sensing image acquisition methods, there is higher demand for remote sensing image processing technology. This paper introduces three common supervised classification algorithms: support vector machine, maximum likelihood method and BP neural network. The same SuperView-1 remote sensing image in Nanniwan area is classified by using the above three algorithms, and the classification results are obtained and the accuracy is evaluated. Then, the three algorithms are compared, and the advantages and disadvantages of the three algorithms are analyzed. It is concluded that the classification accuracy of support vector machine is the lowest, the maximum likelihood method is the second, and the neural network is the highest. Finally, it is concluded that BP neural network is an excellent remote sensing image classification algorithm.