Matching Coronal Holes between Frames Algorithm Overview

In this project, we leverage machine learning techniques to match coronal holes between sequential frames. As time evolves, the coronal hole remains in proximity to its previous frame location. Therefore, in order to match CHs between sequential frames, we evaluate the CHs centroid (center of mass) and set of pixel location. Since computing the area overlap between two CHs can require intensive computational work, we first prune the list of possible classes by computing the centroid distance between the new CH and the ones identified in the previous frames.

The coronal hole classification algorithm is a two step process, where in step 1, we match CHs based on their centroid location using K-Nearest Neighbors (KNN), and in the step 2, we measure the area overlap between the coronal holes of which KNN resulted in a high probability of being associated.