高维遥感图像的快速分类算法  

A fast classification algorithm for high-dimensional remote sensing images

作　　者：孙华生[1] 李晓轩[1] 

机构地区：[1]辽宁工程技术大学测绘与地理科学学院,辽宁阜新123000

出　　处：《测绘科学》2016年第8期19-23,37,共6页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41201454);辽宁省大学生创新创业训练计划项目(201410147047)

摘　　要：为了实现对高维遥感图像的快速准确分类,提出了一种基于k均值二叉树支持向量机(SVM)的分类方法。该方法通过对选取的训练样本进行k均值聚类,生成支持向量机分类二叉树,作为确定最佳分类顺序的依据,以降低分类过程中的误差累积并提高整体分类精度,而且可缓解由样本数量不均衡导致的分类误差。该方法可在不进行降维处理的情况下,对高维遥感图像进行快速准确分类。测试结果表明,其分类速度和分类精度都优于传统的支持向量机分类结果。To achieve fast and accurate classification for high-dimensional remote sensing images,a classification method based on k-means binary tree was developed.The k-means clustering method was carried out to generate binary tree for SVM classification according to the selected samples,and it was the basis of the determination of optimal classification sequence in order to reduce the error accumulation in the classification process and to improve the overall classification accuracy.Furthermore,to relieve the errors caused by sample unbalance.The method could realize image classification quickly and accurately for highdimensional remote sensing images without reducing the dimension.Test results indicated that both of the speed and the accuracy were better than the traditional SVM classification result.

关 键 词：支持向量机SVM k均值二叉树 图像分类 高维数据 

分 类 号：TP701[自动化与计算机技术—检测技术与自动化装置]