Automated Recognition of Post-Flare Loops in High-Resolution H$\upalpha$ Red-Wing Images

Abstract

The post-flare loop is a typical dynamic process of solar flares. In the past, a manual method was used to recognize the strand structure of loops. With higher spatial resolving power, the post-flare loop consists of more bundles of strands. For fast and reliable post-flare-loop recognition, an automated post-flare-loop detection method applied to high-resolution H$\upalpha$ red-wing images is proposed in this article. In this method, straight lines are detected from the edge-detection result, and then a curve-growing procedure is carried out on the basis of the detected straight lines. Finally, the grown curves are screened to eliminate isolated and intersected curves. To our knowledge, our research is the first trial to detect post-flare loops on the solar disk. Experimental results with the high-resolution H$\upalpha$ red-wing images from the Goode Solar Telescope (GST) have verified that the proposed method could detect post-flare loops effectively in all their developing phases. Based on the recognition results, the automatically computed loop widths (118±35 km) are consistent with the manually derived results (124±19 km). With the development of observing instrumentation, the proposed method may recognize each strand of the loops and the measured loop widths could be used to verify flare models.