极地研究

极地研究 ›› 2005, Vol. 17 ›› Issue (2): 93-98.

• 目次 •    下一篇

基于粗阈值和梯度算法的南极多光谱影像云区的提取

鄂栋臣;陈刚;   

  • 出版日期:2005-06-30 发布日期:2005-06-30

CLOUD DETECTION OF MULTISPECTRAL IMAGES IN ANTARCTICA BASED ON WIDE THRESHOLD AND GRADIENT ALGORITHM

  • Online:2005-06-30 Published:2005-06-30

PDF (PC)

1203

被引次数

1

摘要：

利用Landsat-7的ETM+多光谱影像,提出使用粗阈值以及梯度算法提取极地上空的云。试验结果表明,粗阈值增加了阈值算法的灵活性,并结合梯度算法,由粗到精,达到了较好的极地目标识别效果。

关键词: 云区提取;粗阈值;梯度算法;ETM+

Abstract:

This work presents a new method for cloud detection in polar area based on the wide threshold and gradient algorithm using ETM + multispectral images. The approach identifies the cloud, shadow, ice sheet, glacial and bare rock through visible, IR and thermal images. More dynamic adaptation is given by the wide threshold and the gradient algorithm is combined to confine the ranges of cloud, shadow and ice sheet. The experiment result shows us the good efficiency of the algorithm.