Publication

Abstract : The manuscripts demonstrates the consequences on the way to advance in addition authenticate image processing techniques intended for deciding the finest physical models centered happening on solar image surveillances. This technique comprises of choosing the physical models centered proceeding on by means of coronal holes extracted from the images. Eventually, the objective is on the way to custom physical models in order to forecast geomagnetic storms. For coronal hole segmentation it is centered on physical restraints progressing a multi-modal scheme which customs segmentation maps commencing from three diversified methods in order to initialize a level-set technique which progresses the preliminary coronal hole segmentation towards the magnetic boundary. These coronal holes are clustered, mapped and validated. The suggested multi-modal segmentation technique vitally outstripped SegNet, U-net, Henney- Harvey, and FCN by giving precise boundary recognition. This suggested techniques gives 95.5% map classification precisely. An ontological representation IPCHSonto of the image processing method for Coronal Hole Segmentation have been built using protégé tool. This ontology is précised in formal languages by means of a well-defined semantics. Ontologies remain built upon a segmented understanding within a community.