位置:成果数据库 > 期刊 > 期刊详情页
Hα全日面太阳图像云污染的去除
  • ISSN号:1672-7673
  • 期刊名称:天文研究与技术
  • 时间:2014
  • 页码:134-139
  • 分类:P182[天文地球—天文学]
  • 作者机构:[1]昆明理工大学云南省计算机技术应用重点实验室,云南昆明,650500
  • 相关基金:国家自然科学基金(11263004,11163004)资助.
  • 相关项目:红外太阳塔观测图像质量评价方法研究
中文摘要:

Hα全日面太阳图像对太阳物理研究有着重要的意义,但其观测过程中可能受天气的影响,导致观测图像含有云层覆盖的污染.主要通过对这种Hα全日面太阳图像的云污染进行处理,从而得到较为清晰的太阳活动细节.具体算法为:首先将含云污染的图像扣除模板图像的标准临边昏暗轮廓,然后对其进行中值滤波获得云层图像,最后在原始图像上扣除云层图像得到修复后的全日面图像.实验证明此方法可以有效地去除Hα全日面太阳图像上的云污染,并对太阳活动区域影响较小.

英文摘要:

Hα full-disk solar images are important in research of solar science. Observations leading to such images may be affected by weather so that they show appreciable cloud extinction. This paper focuses on an approach to remove cloud extinction in Hα full-disk solar images. The approach can make details of solar activities easily recognizable in result images. We first evaluate and select a template image of Hα solar surface brightness. The template is constructed as an image modeling the Hα solar full disk with an isotropic surface brightness distribution following the direction-averaged Hα solar radial profile of a ( cloud-extinction free) real image. We next subtract the template from a raw Hα full-disk solar image affected by cloud extinction, and perform median filtering on the difference image to obtain an image of cloud extinction. We finally subtract the raw image by the image of cloud extinction. The subtraction result is an Hα full-disk solar image with cloud extinction effectively removed. We have conducted some experiments, which validate our approach and show that it has little effect on image details of solar activities. Particularly, our experiments demonstrate that the performance of the approach is stable against patterns of cloud extinction. We also discuss shortcomings of our approach.

同期刊论文项目
同项目期刊论文
期刊信息
  • 《天文研究与技术》
  • 主管单位:中国科学院
  • 主办单位:中国科学院国家天文台
  • 主编:钱声帮
  • 地址:昆明110信箱
  • 邮编:650011
  • 邮箱:ynaotk@ynao.ac.cn;
  • 电话:0871-63920157
  • 国际标准刊号:ISSN:1672-7673
  • 国内统一刊号:ISSN:53-1189/P
  • 邮发代号:
  • 获奖情况:
  • 国内外数据库收录:
  • 被引量:800