Yes / Solar images, along with other observational data, are very important for solar physicists and space weather researchers aiming to understand the way the Sun works and affects Earth. In this study a 3D modelling technique for visualizing solar magnetic field lines using solar images is presented. Photospheric magnetic field footpoints are detected from magnetogram images and using negative and positive magnetic footpoints, dipole pairs are associated according to their proximity. Then, 3D field line models are built using the calculated dipole coordinates, and mapped to detected pairs after coordinate transformations. Final 3D models are compared to extreme ultraviolet images and existing models and the results of visual comparisons are presented.

Based on a study of existing solar filament and tracking methods, a fully automated solar filament detection and tracking method is presented. An adaptive thresholding technique is used in a segmentation phase to identify candidate filament pixels. This phase is followed by retrieving the actual filament area from a region grown filament by using statistical parameters and morphological operations. This detection technique gives the opportunity to develop an accurate spine extraction algorithm. Features including separation distance, orientation and average intensities are extracted and fed to a Neural Network (NN) classifier to merge broken filament components. Finally, the results for two consecutive images are compared to detect filament disappearance events, taking advantage of the maps resulting from converting solar images to Heliographic Carrington co-ordinates. The study has demonstrated the novelty of the algorithms developed in terms of them now all being fully automated; significantly the algorithms do not require any empirical values to be used whatsoever unlike previous techniques. This combination of features gives the opportunity for these methods to work in real-time. Comparisons with other researchers shows that the present algorithms represent the filaments more accurately and evaluate computationally faster - which could lead to a more precise tracking practice in real-time. An additional development phase developed in this dissertation in the process of detecting solar filaments is the detection of filament disappearances. Some filaments and prominences end their life with eruptions. When this occurs, they disappear from the surface of the Sun within a few hours. Such events are known as disappearing filaments and it is thought that they are associated with coronal mass ejections (CMEs). Filament disappearances are generally monitored by observing and analysing successive solar H-alpha images. After filament regions are obtained from individual H-alpha images, a NN classifier is used to categorize the detected filaments as Disappeared Filaments (DFs) or Miss-Detected Filaments (MDFs). Features such as Area, Length, Mean, Standard Deviation, Skewness and Kurtosis are extracted and fed to this neural network which achieves a confidence level of at least 80%. Comparing the results with other researchers shows high divergence between the results. The NN method shows better convergence with the results of the National Geophysical Data Centre (NGDC) than the results of the others researchers.

Yes / In this study a method for 3D representation of active regions and sunspots that are detected from Solar and Heliospheric Observatory/Michelson Doppler Imager magnetogram and continuum images is provided. This is our first attempt to create a visual solar catalogue. Because of the difficulty of providing a full description of data in text based catalogues, it can be more accurate and effective for scientist to search 3D solar feature models and descriptions at the same time in such a visual solar catalogue. This catalogue would improve interpretation of solar images, since it would allow us to extract data embedded in various solar images and visualize it at the same time. In this work, active regions that are detected from magnetogram images and sunspots that are detected from continuum images are represented in 3D coordinates. Also their properties extracted from text based catalogues are represented at the same time in 3D environment. This is the first step for creating a 3D solar feature catalogue where automatically detected solar features will be presented visually together with their properties.

Based on a study of existing solar filament and tracking methods, a fully automated solar filament detection and tracking method is presented. An adaptive thresholding technique is used in a segmentation phase to identify candidate filament pixels. This phase is followed by retrieving the actual filament area from a region grown filament by using statistical parameters and morphological operations. This detection technique gives the opportunity to develop an accurate spine extraction algorithm. Features including separation distance, orientation and average intensities are extracted and fed to a Neural Network (NN) classifier to merge broken filament components. Finally, the results for two consecutive images are compared to detect filament disappearance events, taking advantage of the maps resulting from converting solar images to Heliographic Carrington co-ordinates. The study has demonstrated the novelty of the algorithms developed in terms of them now all being fully automated; significantly the algorithms do not require any empirical values to be used whatsoever unlike previous techniques. This combination of features gives the opportunity for these methods to work in real-time. Comparisons with other researchers shows that the present algorithms represent the filaments more accurately and evaluate computationally faster - which could lead to a more precise tracking practice in real-time. An additional development phase developed in this dissertation in the process of detecting solar filaments is the detection of filament disappearances. Some filaments and prominences end their life with eruptions. When this occurs, they disappear from the surface of the Sun within a few hours. Such events are known as disappearing filaments and it is thought that they are associated with coronal mass ejections (CMEs). Filament disappearances are generally monitored by observing and analysing successive solar H-alpha images. After filament regions are obtained from individual H-alpha images, a NN classifier is used to categorize the detected filaments as Disappeared Filaments (DFs) or Miss-Detected Filaments (MDFs). Features such as Area, Length, Mean, Standard Deviation, Skewness and Kurtosis are extracted and fed to this neural network which achieves a confidence level of at least 80%. Comparing the results with other researchers shows high divergence between the results. The NN method shows better convergence with the results of the National Geophysical Data Centre (NGDC) than the results of the others researchers.

Yes / Solar imaging is currently an active area of research. A fast hybrid system for the automated detection of filaments in solar images is presented in this paper. The system includes three major stages. The central solar region is detected in the first stage using integral projections. Intensity filtering and image enhancement techniques are implemented in the second stage to enhance the quality of detection in the central region. Local detection windows are implemented in the third stage to detect the positions of filaments and to define various sized arrays to contain them. The extracted arrays are fed later to a neural network for verification purposes.

Yes / A fast hybrid system for the automated detection and verification of active regions (plages) and filaments in solar images is presented in this paper. The system combines automated image processing with machine learning. The imaging part consists of five major stages. The solar disk is detected in the first stage, using a morphological hit-miss transform, watershed transform and Filling algorithm. An image-enhancement technique is introduced to remove the limb-darkening effect and intensity filtering is implemented followed by a modified region-growing technique to detect the regions of interest (RoI). The algorithms are tested on H- and CA II K3-line solar images that are obtained from Meudon Observatory, covering the period from July 2, 2001 till August 4, 2001. The detection algorithm is fast and it achieves false acceptance rate (FAR) error rate of 67% and false rejection rate (FRR) error rate of 3% for active regions, and FAR error rate of 19% and FRR error rate of 14% for filaments, when compared with the manually detected filaments in the synoptic maps. The detection performance is enhanced further using a neural network (NN), which is trained on statistical features extracted from the RoI and non-RoI. With the use of this combination the FAR has dropped to 2% for active regions and 4% for filaments.© 2006 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 15, 199-210, 2005

Yes / The cadence and resolution of solar images have been increasing dramatically with the launch of new spacecraft such as STEREO and SDO. This increase in data volume provides new opportunities for solar researchers, but the efficient processing and analysis of these data create new challenges. We introduce a fuzzy-based solar feature-detection system in this article. The proposed system processes SDO/AIA images using fuzzy rules to detect coronal holes and active regions. This system is fast and it can handle different size images. It is tested on six months of solar data (1 October 2010 to 31 March 2011) to generate filling factors (ratio of area of solar feature to area of rest of the solar disc) for active regions and coronal holes. These filling factors are then compared to SDO/EVE/ESP irradiance measurements. The correlation between active-region filling factors and irradiance measurements is found to be very high, which has encouraged us to design a time-series prediction system using Radial Basis Function Networks to predict ESP irradiance measurements from our generated filling factors.

yes / The importance of real-time processing of solar data especially for space weather applications is increasing continuously. In this paper, we present an automated hybrid computer platform for the short-term prediction of significant solar flares using SOHO/Michelson Doppler Imager images. This platform is called the Automated Solar Activity Prediction tool (ASAP). This system integrates image processing and machine learning to deliver these predictions. A machine learning-based system is designed to analyze years of sunspot and flare data to create associations that can be represented using computer-based learning rules. An imaging-based real-time system that provides automated detection, grouping, and then classification of recent sunspots based on the McIntosh classification is also created and integrated within this system. The properties of the sunspot regions are extracted automatically by the imaging system and processed using the machine learning rules to generate the real-time predictions. Several performance measurement criteria are used and the results are provided in this paper. Also, quadratic score is used to compare the prediction results of ASAP with NOAA Space Weather Prediction Center (SWPC) between 1999 and 2002, and it is shown that ASAP generates more accurate predictions compared to SWPC. / EPSRC

Machine-learning algorithms are applied to explore the relation between significant flares and their associated CMEs. The NGDC flares catalogue and the SOHO/LASCO CME catalogue are processed to associate X and M-class flares with CMEs based on timing information. Automated systems are created to process and associate years of flare and CME data, which are later arranged in numerical-training vectors and fed to machine-learning algorithms to extract the embedded knowledge and provide learning rules that can be used for the automated prediction of CMEs. Properties representing the intensity, flare duration, and duration of decline and duration of growth are extracted from all the associated (A) and not-associated (NA) flares and converted to a numerical format that is suitable for machine-learning use. The machine-learning algorithms Cascade Correlation Neural Networks (CCNN) and Support Vector Machines (SVM) are used and compared in our work. The machine-learning systems predict, from the input of a flare¿s properties, if the flare is likely to initiate a CME. Intensive experiments using Jack-knife techniques are carried out and the relationships between flare properties and CMEs are investigated using the results. The predictive performance of SVM and CCNN is analysed and recommendations for enhancing the performance are provided. / EPSRC

Solar images are one of the most important sources of available information on the current state and behaviour of the sun, and the PICARD satellite is one of several ground and space-based observatories dedicated to the collection of that data. The PICARD satellite hosts the Solar Diameter Imager and Surface Mapper (SODISM), a telescope aimed at continuously monitoring the Sun. It has generated a huge cache of images and other data that can be analysed and interpreted to improve the monitoring of features, such as sunspots and the prediction and diagnosis of solar activity. In proportion to the available raw material, the little-published analysis of SODISM data has provided the impetus for this study, specifically a novel method of contributing to the development of a system to enhance, detect and segment sunspots using new hybrid methods. This research aims to yield an improved understanding of SODISM data by providing novel methods to tabulate a sunspot and filling factor (FF) catalogue, which will be useful for future forecasting activities. The developed technologies and the findings achieved in this research will work as a corner stone to enhance the accuracy of sunspot segmentation; create efficient filling factor catalogue systems, and enhance our understanding of SODISM image enhancement. The results achieved can be summarised as follows: i) Novel enhancement method for SODISM images. ii) New efficient methods to segment dark regions and detect sunspots. iii) Novel catalogue for filling factor including the number, size and sunspot location. v) Novel statistical method to summarise FFs catalogue. Image processing and partitioning techniques are used in this work; these methods have been applied to remove noise and detect sunspots and will provide more information such as sunspot numbers, size and filling factor. The performance of the model is compared to the fillers extracted from other satellites, such as SOHO. Also, the results were compared with the NOAA catalogue and achieved a precision of 98%. Performance measurement is also introduced and applied to verify results and evaluate proposal methods. Algorithms, implementation, results and future work have been explained in this thesis.