首页> 外文会议>Symposium on multispectral image processing and pattern recognition >Change detection in high resolution SAR images based on multi-scale texture features
【24h】

Change detection in high resolution SAR images based on multi-scale texture features

机译:基于多尺度纹理特征的高分辨率SAR图像中的变化检测

获取原文
获取外文期刊封面目录资料

摘要

This paper studied on change detection algorithm of high resolution (HR) Synthetic Aperture Radar (SAR) images based on multi-scale texture features. Firstly, preprocessed multi-temporal Terra-SAR images were decomposed by 2-D dual tree complex wavelet transform (DT-CWT), and multi-scale texture features were extracted from those images. Then, log-ratio operation was utilized to get difference images, and the Bayes minimum error theory was used to extract change information from difference images. Lastly, precision assessment was done. Meanwhile, we compared with the result of method based on texture features extracted from gray-level cooccurrence matrix (GLCM). We had a conclusion that, change detection algorithm based on multi-scale texture features has a great more improvement, which proves an effective method to change detect of high spatial resolution SAR images.
机译:本文研究了基于多尺度纹理特征的高分辨率(HR)合成孔径雷达(SAR)图像变化检测算法。首先,利用二维双树复小波变换(DT-CWT)对经过预处理的多时相Terra-SAR图像进行分解,并从这些图像中提取出多尺度纹理特征。然后,利用对数比运算获得差异图像,并利用贝叶斯最小误差理论从差异图像中提取变化信息。最后,进行了精度评估。同时,我们与基于灰度共生矩阵(GLCM)提取的纹理特征的方法的结果进行了比较。我们得出的结论是,基于多尺度纹理特征的变化检测算法有了较大的改进,证明了一种有效的高分辨率空间SAR图像变化检测方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号