The automated detection algorithm of coronal loop and statistical analysis of its width

Abstract

Coronal loop is an important component of solar corona, and it is useful for us to study its size, morphological structure and evolution as a result of that the studies will help us to better understand the internal structure of coronal loop. In the meantime, the studies can further provide us a clue of comprehending the construction of solar corona, the evolution of solar active region, coronal heating process, coronal magnetic field and the properties of plasma of solar corona. Currently, the majority of the scientific studies associated with coronal loop just focus on small sample analyses, which is mainly due to that the advances of automated detection algorithms of coronal loop and their scientific applications lag behind the theoretical studies of coronal loop. Although, some automated detection algorithms of coronal loop have been developed, just the algorithm named as Oriented Coronal CUrved Loop Tracing (OCCULT), which has a satisfactory performance in comparison with other existing algorithms, is applied to the scientific studies associated with coronal loop. However, the algorithm, OCCULT, is not sensitive to the weak and fine coronal loops, resulting in that its detection results lack of the coronal loops with weak intensities and fine structures. Clearly, the disadvantage of OCCULT mentioned above makes its results of large sample statistical analyses incomplete. For example, in the large sample statistical analysis of the widths of cross-section profiles of coronal loops, the statistical results of OCCULT are incomplete because the coronal loops with weak intensities and fine structures cannot be detected by it. Therefore, in this paper, a more effective detection algorithm of coronal loop is developed and further applied to the large sample statistical analysis of the widths of coronal loops to obtain more complete statistical distribution of the widths of coronal loops. First of all, in this paper, the developed algorithm of automatically detecting coronal loops based on matched filtering technique and unsharp-masking enhancement is described. Secondly, the coronal loops contained in the images of Atmospheric Imaging Assembly (AIA) of Solar Dynamics Observatory (SDO) and High-resolution Coronal Imager (Hi-C) are detected by the developed algorithm, and then the detection results are compared with those of the algorithm, OCCULT, so as to indicate that the developed algorithm can detect more coronal loops, especially the coronal loops with weak intensities and fine structures. Thirdly, the local directions of the main axes of the detected coronal loops are extracted, and corresponding cross-section profiles of the coronal loops are obtained, resulting in that the widths of the coronal loops are automatically calculated and statistically analyzed. Finally, the conclusions are obtained. (1) In comparison with OCCULT, the performance of the developed algorithm is more excellent, indicating that in addition to calculating and statistically analyzing the widths of coronal loops, the developed algorithm can also be used to other scientific applications related to coronal loops. (2) In comparison with OCCULT, the large sample statistical results of the widths of coronal loops of the developed algorithm are more complete. (3) In comparison with the high resolution of Hi-C, the resolution of AIA/SDO cannot distinguish most of coronal loops with fine structures.