Automatic Detection of Prominence Eruption

In this paper, we present a method to automatically detect prominence eruption on the solar limb, based on analysis of full-disk Halpha images. As the part of the detection program, we extract the properties of prominences and their eruptions, such as location, height, and eruption speed. In the first step, we normalize the images to the same contrast and remove noises. In the second step, we detect the solar limb and isolate the limb and disk components. In the third step, we identify the prominence features. Finally, based on the extracted parameters of detected features, we characterize prominence eruption as a function of time. The method will be used to detect prominence eruptions in real time.