Estimation Of ViBe Coronal Mass Projectile Based On Otsu Method

Coronal Mass Ejection (CME) is a process of the largest scale and most vigorous energy release. The release of these energies may have a significant impact on the earth. So detection for coronal mass ejections is very important.With the development of science and technology, the method of automatic detection of CME has been studied extensively in recent years. So I also work on the detection of coronal mass ejections. The main contents are as follows:(1) This thsis introduces the several existing coronal mass ejections detection technology, such as the Computer Aided CME Tracking software package(CACTus) method based on Hough transform to detect the CME, the Solar Eruptive Event Detection System (SEEDS) method based on luminance enhancement detection identification technology and artificial detection method; then I have analyzed the testing results of three methods with the artificial detection as a foundation, I carried on the multiple perspectives analysis of the CACTus Method and artificial detection method. And I find that artificial detection is accurate but takes longer, CACTus method is prone to leakage inspection problems.(2) In order to solve the above problems, this thsis proposes a new method for coronal mass ejections detection, this method uses the detection foreground method of ViBe (Visual Background extractor) algorithm for CME detection, the advantage of the algorithm can adapt to the changing of the coronal mass ejection scenario and has high real-time performance and robustness; but it is prone to "ghost" in the process of CME detection. In order to solve these problems, this thsis introduces the Otsu method to have a quadratic discriminant of the pixels which are classified as foreground pixels. Then according to the judgment of CME conditions, I determine the number of the CME and some physical properties, then formulate the CME catalog after transforming preliminary CME test results to polar coordinates. Finally I compare my experiment with CACTus method and artificial detection from three aspects, and find that the method in this thsis has a higher rate in successful detection than CACTus method, and a lower rate in in error detection than CACTus method. And the thsis could detect faint CME moving targets that CACTus method did not detect but artificial detected.