Research On Coronal Jet Detection Method Based On Fast RPCA

Coronal jets are a common solar activity phenomenon in the solar atmosphere.The study of coronal jets is helpful to reveal the mechanism of energy transfer and release in the solar atmosphere.It tends to occur in relatively small areas and is closely related to magnetic levitation and magnetic cancellation.If jets in the sun’s atmosphere can be well identified,especially small,weak jets,it will be clear how much of a role coronal jets play in solar wind acceleration and coronal heating.In previous studies,jet eruptions were accompanied by flares or coronal mass ejections,so the detection of coronal jets is also beneficial to the study of other solar activity phenomena.With the development of science,detection technology for moving objects of computer is constantly improved,and many solar activity phenomena(such as:sunspot,flare,prominence)automatic recognition and detection technology has been widely studied in recent years.In this paper,two sets of coronal image data obtained by the Atomospheric Imaging Assembly(AIA)on the Solar Dynamic Observatory(SDO)satellite in channels 171(?) and 193(?) were used as research objects to detect the coronal jet.Coronal jet detection consists of the following five steps: First,image alignment in the solar software package,solar autobiographic correction method and image processing technology are used to preprocess the experimental image data;Secondly,Fast RPCA method is used to detect the coronal jet in the coronal image data.The basic idea of detection is to combine the idea of low rank and sparse decomposition in Fast RPCA with the characteristics of the coronal jet with small scale and large scale in coronal sequence images,so as to realize the separation between the random and complex dynamic background and sparse moving objects.The coronal jet as the foreground change is detected;Again: as the corona also contains in addition to the coronal jet of solar activity,so detected candidate outlook matrix may contain other sports other than jet change,here we combine the corona jet movement has the directivity,the characteristics of the form shall be straight shape,using connected domain analysis operations,excluding geometric parameters and coronal jet characteristics is not in conformity with the mixed and disorderly change area,in order to achieve the purpose of detecting corona jets;Finally,the ratio of length to width and the characteristic parameters of transverse velocity of the coronal jets detected were estimated.In this paper,the subjective and objective analysis method is used to verify the applicability,feasibility and accuracy of the algorithm.Experimental results show that when the experimental data of the corona jets in line with the large scale change,pixel intensity increased more obviously,at the same time there are some brightness enhancement of other coronal activity change,but the change range of the other coronal activity is far less than the change range of corona jets,the algorithm is compared with running difference method,the proposed method can detect the intensity of the weaker corona jet and improve the accuracy of the corona jet detection.Compared with the Gaussian mixture model,the proposed method is more robust.