Poor Quality Hα Full Day Sun Image Repair

Ha full-disk solar image observation is an important part of solar observation, it’s observation image has been a basic data of solar physics,Sun-Earth relationship and related disciplines studies, these data for monitoring solar activity and study the Sun-Earth relationship is significant, therefore a large number of domestic and international ground-based observatories are regular full-day observation. However, the qualities of these images are different due to the change of observation condition, such as seeing, cloud and so on. Images with bad quality will necessarily bring difficult to the subsequent research. Standard full-disk Ha image after deducting limb darkening background will be "flat" in the sun with a round face, so that the surface of the sun round event details can well distinguish it from the sky in the background, to facilitate research on solar activity. And in this way the whole day face poor image quality processing, can not be "flat" in the sun with a round face, it must be repaired for the poor quality full-face Ha images. In this paper, for the cloud of pollution and poor optical distortion both quality Ha full-face sun image processing.In this paper, we use Ha full-disk image of GONG plans observation,mainly to do the following three aspects. First, we propose two evaluation algorithm selected from a large number of full-disk images of poor quality full-disk image. Second, we have these poor quality full-disk Ha images were restored with the specific process is:first full-disk images were of poor quality limb darkening background subtraction, and then deducting the background image after filtering to filter out the details of solar activity on the image, the final image filtering after deducting all day to get the repaired surface image with the original image, this paper used the median filter, a single morphological filtering and morphological filtering three full days of poor quality surface images were processed. Third, we also adopted SSIM operator fixes these three filtering algorithms are compared. Meanwhile, we also wavelet domain and frequency domain for poor quality full-disk Ha image repair attempt.In each filter repair algorithm, we have an image of the original image and the results were compared to verify the effectiveness of the filtering algorithm repair; but also through different observer, at different times and in different quality vary greatly throughout the day face image processing, proved that the filtering algorithm has the versatility to repair. By comparing the three filtering algorithms to repair, we found that three morphological filtering algorithm is better than the other two algorithms to repair. Finally, we also present in our experiment at the shortcomings discussed.