CME Classification Based on Wavelet Spectra

Abstract

We study the internal structure of coronal mass ejections (CMEs) using wavelet analysis. We derive wavelet spectra, spatially integrated over regions of interest within LASCO C2 white-light coronographic images. These spectra show an inflection point, which we use to characterize each spectrum. In a diagram of flux vs. spatial scale of the inflection point, we find that the analyzed structures fall into two, distinct groups: a lowflux, small-spatial-scale group (which we call the "homogeneous" type), and a high-flux, larger-spatial-scale group (the "collimated" type). Interestingly, if we study different regions of a given image, all of the structures fall into one of the two groups described above. From a qualitative comparison with the images, it is clear that the two groups identified by the wavelet analysis correspond to two types of morphologies, which are seen as either more-homogeneous or more-collimated structures.

An end-to-end method of Coronal Mass Ejections detection

“…(2) 基 于 频 域 变 换 的 图 像 处 理 方 法 . Gonzalez-Gomez等人 [10] 采用小波分析的方法对CME进行 了分类研究, 分析的结果是从形态上把CME分成两 类: 一类是低流量、小尺度的活动现象, 但未涉及 CME检测问题; 另一类则是高流量大尺度的活动现 象. Gallagher等人 [11] 认为CME是一种多尺度特征, 而 小波分析在增强典型CME前端曲线的可观性时效果 欠佳, 他们采用更高阶的脊小波和curvelet来进行分 类 和 监 测 . Goussies 等 人 [12] 提 出 了 " 事 件 的 纹 理 (texture of the event)"这一概念, 利用灰度共生矩阵 来描述纹理, 并利用有监督的分割算法来提取感兴 趣区域, 进行CME的检测, 并提出了CME跟踪的概 念.…”