Automatic detection of sunspots from solar images using fractional-order derivatives and extraction of their attributes

Abstract

A huge collection of solar images to visualize sunspot are acquired by various solar observatories spread across the globe. This necessitates efficient tools for detecting and analyzing the sunspots encompassing diverse solar features. Particularly sunspot characterization requires precise intensity refinement to distinguish the intricate sunspot structures the umbra and penumbra. Accordingly, this work delivers a sunspot detection module by forging Fractional-Order Derivative Mask (FODM) in segmentation for embracing the diverse solar regions present in solar images. The mechanism initially localizes the statistical intensity variations and bounds them followed by FODM generalization for effectively outlining the heterogenous regions concerned with diverse solar regions. Rigorous analysis of the developed mechanism in terms of the diverse Receiver Operating Characteristics (ROC) parameters was performed on images obtained from diverse benchmarked solar repositories. The proposed model’s nature of acutely packing high-frequency image features using the localized statistical intensity characterization supersedes the conventional Integer Order Differential Mask (IODM) investigated using the ROC parameters. Furthermore, relative ROC analysis with trending models reveals the consistent superiority of this scheme. Additionally, comparative analysis in terms of the diverse extracted solar features with relevant solar bulletins affirms the model’s efficacy.