الفهرس 
Galaxy classificationOnly 14 pages are availabe for public view
 |  | from | 119 |  |  |
| | | from | 119 | | |
Abstract
Clues and traces of the universe origin and developmental process are deeply buried in galaxies shapes and formations so that galaxies images are considered one of the most important astronomical images. The problem of automating galaxies classification from their images is complicated by galaxies images faintness, conflicting bright background stars, and images noise. The current technique preprocesses the images to rectify brightness and noise problems, and then employs two different algorithms.
The first algorithm is a novel classification algorithm that differentiates between edge-on galaxies and face-on galaxies using a novel slimness weighting factor. The developed algorithm has been tested on 1800 galaxies from the EFIGI catalog. The achieved overall classification accuracy was 98.6%.
The second algorithm is a novel face-on classification algorithm that analyses galaxies morphological raw data and automatically detects the galaxies visual centers, regions, and classifications. This technique has been tested by using a collection of 1000 galaxies from the EFIGI catalog. Results demonstrated a success rate of 97.2 % in galaxies classification.
Finally, the algorithms were combined to create an automated system to differentiate between edge-on galaxies and face-on galaxies then classify face-on galaxies. The automated system has been tested by using the full set of 1800 galaxies (edge-on and face-on galaxies) from the EFIGI catalog. Results demonstrated a success rate of 97.5 % in galaxies classification with an average processing time of 0.26 seconds per galaxy on an average laptop. The high success rates and the low processing time proved the efficiency of the current work.
The thesis aims to propose a technique that has superior performance in detection and classification of galaxies images in the shortest possible processing time.
The thesis comprises five chapters, which are organized as follows:
Chapter one: presents a brief introduction of the research study which includes an overview, objectives and thesis organization.
Chapter two: Presents the related works of galaxies detection and classification techniques.
Chapter three: presents the proposed technique includes detection and classification technique for edge-on galaxies based on mathematical treatment of galaxies brightness data. Secondly, detection and classification technique for face-on galaxies based on galaxies brightness variation patterns.
Chapter four: presents results and discussion. The result of proposed techniques implemented for detection and classification on edge-on galaxies images and face-on galaxies images. Also, a comparison between the proposed techniques with other exiting techniques.
Chapter five: presents the conclusions and suggests headlines for future