Use of computer vision and machine learning for the analysis of solar rotation profile

• Main Author: Figueira, Francisco Correia de Berquó
• Publication Date: 2024
• Format: Master thesis
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10362/178096
• topic: Computer vision; Image processing; Match Template; Coronal Bright Points; Solar differential rotation; Solar cycle; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

With the unprecedented growth of solar images provided by recent missions, there is a growing necessity to harness the benefits they may offer. Using images from the Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA) instrument, this project employs state-of-the-art computer vision techniques to study solar differential rotation. Coronal Bright Points (CBPs), identified as small bright spots in the solar corona, serve as tracers across multiple images due to their distribution across all latitudes, independence from the solar cycle, and distinctiveness compared to other solar activities. Moreover, all CBPs share a similar shape, facilitating their detection and tracking. In the detection process, an image matching algorithm was employed, proving its reliability across various stages of the solar cycle. In the future, this tool could be improved by integrating Machine Learning, as post-processing method, to eliminate false positive CBP identifications. Computer vision techniques were also utilised to discern whether a CBP is inside or outside a Coronal Hole. Solar rotation values were calculated based on the movements of CBPs during their lifetime as a result of the performed tracking. Given the importance of this topic, it was created a website to display the information about the solar differential rotation, as well as data on the CBPs and Coronal Holes. This website is supported by a robust database where processed image data is stored, acting as a bridge between the detection and tracking tools and the user interface. This infrastructure enables the presentation of information to the user in near real-time. This tool’s performance results are also showcased, including its outcomes at various stages of the solar cycle. Specifically, these results are highlighted during periods of both low solar activity and high solar activity.