Coronal mass ejection detection method based on Faster R-CNN

Abstract

The invention discloses a coronal mass ejection detection method based on a Faster R-CNN. The method comprises the following steps: firstly, carrying out preprocessing, alignment, median filtering noise reduction, polar coordinate conversion and data enhancement on an intercepted image; secondly, constructing a Faster R-CNN target detection network, designing a multi-channel region suggestion neural network, and adopting two RPNs for generating strong coronal mass ejection target candidate boxes and weak coronal mass ejection target candidate boxes respectively; and then pre-training the RPN,training the FEN by utilizing a suggestion box generated by the RPN, and training the RPN again on the basis of the fixed FEN. During RPN1 and RPN2 training, an alternate training strategy is adopted,and weak CME target training is assisted by means of strong CME target characteristics. And finally, coronal sequence image substance ejection identification is carried out. According to the method,the overall performance of a traditional coronal mass ejection identification method is improved to a great extent, and the detection precision of CME of different scales is improved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a detection method for coronal mass ejection based on Faster R-CNN. Background technique [0002] A Coronal Mass Ejection (CME) is the apparent event that accompanies the release of material from the Sun by the corona. CME originates from the solar corona and is ejected from the lower layers of the sun to the heliosphere. It is a large-scale eruption phenomenon composed of plasma and magnetic field. Coronal mass ejections are closely related to the solar magnetic field and have a great impact on the space environment and human activities. They have always been considered as the most important driving source of space disaster weather. Strong CMEs can cause geomagnetic disturbances such as geomagnetic storms, ionospheric storms, and auroras. When this disturbance is severe enough, it will also have a catastrophic impact on high-tech activities that humans rely on for survival, suc...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the present invention provides a detection method for coronal mass ejection based on Faster R-CNN, which is used to solve the problem that the highlight features of the early and late stages of the CME phenomenon in the existing detection method for coronal mass ejection are very weak and appear The problem of missed and misjudged images and the problem of low detection accuracy

Images