【摘要】 提出一种从空间日冕观测图像中自动识别日冕物质抛射事件(CME)并分析计算事件特征参数的新方法,该方法主要包括以下3个步骤:通过预处理去除原始观测图像中的各种噪声;针对CME的视觉统计特性,利用频谱突变分析从预处理后的观测图像中准确分离CME像区域;基于区域协方差分析提取出角宽度、速度等CME特征参数.由于采用了频谱突变分析技术而非亮度增强检测技术来检测CME,该方法在识别多重和晦暗CME时具有更好的效果.更多还原

【Abstract】 An automatic method was presented to detect coronal mass ejections(CME) in Coronagraph images and analysis the characteristic parameters of the events.The method consisted of the following three steps: preprocessing to remove the noise in the original observation image;according to the visual statistical characteristics of CMEs,separated CME region from the observed image after preprocessing using spectrum sudden change analysis;calculated the CME feature parameters such as angular width and speed based on region covariance analysis.Different from brightness comparison which was widely used in currently methods,CMEs were detected by analyzing the sudden change of frequency spectrum in coronagraph.This method should be more effective while detecting multiple and faint CME.更多还原