• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Fifth international workshop on the analysis of multi-temporal remote sensing images >SPECTRAL PRE-SORTING TO ISOLATE POTENTIAL PSEUDO-INVARIANT FEATURES (PIFs)
【24h】

SPECTRAL PRE-SORTING TO ISOLATE POTENTIAL PSEUDO-INVARIANT FEATURES (PIFs)

机译：频谱预排序以隔离潜在的伪不变特征（PIF）

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The Ridge Method provides a convenient method for atmospheric normalization of multitemporal imagery without extensive reference to atmospheric observations or required field observations. The method does, however, require that Pseudo-Invariant Features (PIFs) be identified in the image, and this usually requires either a detailed prior knowledge of the scene or a labor-intensive selection process. Several procedures have been proposed for eliminating outliers in order define a selection of PIFs in which one can have reasonable confidence. These processes all begin with a well-behaved set of data that limits the utility of the method. A spectral pre-sorting of the data is proposed to isolate potential PIFs for use with the Ridge Method. The pre-sorting is based on a pixel-by-pixel spectral comparison and is demonstrated using hyperspectral data from the Hyperion sensor, but the technique is not specific to the Hyperion sensor, nor even to hyperspectral data. Most importantly, the ranking of pixels by spectral similarity is an automated process not requiring expert intervention.
机译：Ridge方法为多时相影像的大气标准化提供了一种便捷的方法，而无需广泛参考大气观测或所需的现场观测。然而，该方法确实需要在图像中标识伪不变特征（PIF），并且这通常需要对场景进行详细的先验知识或劳动密集型的选择过程。为了消除离群值，已经提出了几种程序，以便定义一组可以具有合理置信度的PIF。所有这些过程都从一组行为良好的数据开始，从而限制了该方法的实用性。建议对数据进行频谱预分类，以分离潜在的PIF，以供Ridge方法使用。预排序基于逐个像素的光谱比较，并使用来自Hyperion传感器的高光谱数据进行了演示，但该技术并非特定于Hyperion传感器，甚至不特定于高光谱数据。最重要的是，通过光谱相似性对像素进行排名是一个自动化过程，不需要专家干预。

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