期刊文献+

基于SVM的多光谱影像与SAR图像融合地物分类研究 被引量:6

Classification of Terrain Surfaces Using Fused Images of Multi-Spectral and SAR Image Based on Support Vector Machine
下载PDF
导出
摘要 多光谱遥感图像反映了不同地物的光谱信息,而SAR图像则反映了地表不同地物的后向散射强度信息。通过二者结合,可以实现优势信息互补,提高遥感影像分类的精度。多光谱影像与单波段单极化SAR图像融合分类有2种策略:一种是将SAR图像作为一个波段加入多光谱影像中进行分类;另一种先把多光谱影像与SAR图像融合,然后对融合后的图像进行分类。以成都市使用支持向量机分类方法对2种分类策略下的分类精度进行验证。结果表明,后者分类精度要高于前者,同时2种分类方法的分类精度都明显高于单独使用多光谱影像的分类精度。 The multi-spectral remote sensing images reflected the spectral features of diverse surface features,although the synthetic aperture radar(SAR)images reflected the backscatter information.So the accuracy of the image classification could be effectively improved by using fused data of multi-spectral images and SAR images.Two methods were used for the fused data classification.One method was putting SAR image as a band into multi-spectral image to classify.The other was fuse SAR image and multi-spectral image and then classify with fused image.An example of the Chengdu City using both of the two methods had been taken.Experimental results showed that the first method was better than the second method,and both of the two methods'classification accuracy was higher than that using only multi-spectral images.
出处 《安徽农业科学》 CAS 北大核心 2010年第20期10662-10664,共3页 Journal of Anhui Agricultural Sciences
关键词 地物分类 图像融合 支持向量机 SAR图像 Surface classification Multi-spectral Support vector machine SAR image
  • 相关文献

参考文献14

二级参考文献79

共引文献86

同被引文献73

引证文献6

二级引证文献22

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部