M.Sc. Tezi Görüntüleme

Student: Taymaz RAHKAR FARSHİ
Supervisor: Assist. Prof. Dr. Orhan KESEMEN
Department: İstatistik
Institution: Graduate School of Natural and Applied Sciences
University: Karadeniz Technical University Turkey
Title of the Thesis: AUTOMATIC DETECTION DIFFRACTION HYPERBOLAS IN GROUND PENETRATING RADAR IMAGES WITH MULTIMODAL ARTIFICIAL BEE COLONY OPTIMIZATION ALGORITHM
Level: M.Sc.
Acceptance Date: 30/7/2013
Number of Pages: 48
Registration Number: i2696
Summary:

      In this study, ground penetrating radar tool used in the investigation of the shallow subsurface groundwater depth view of the data obtained from the measurements used in conversion is to obtain velocity information. Because of using Hough Transformation and using long time for that, practical solutions are impossible. For faster solution, a stochastic optimization method which is a ABC algorithm type is used. In order to adapt this method, Multimodal ABC Algorithm developed and the quality of that is tested on different apartments. In order to getting better performance in around the apartments blur filters are used. In developed multimodal ABC algorithm for improving the performance on radargram data, after pre-processing the Hilbert transform is applied to the radargram data. After applying the algorithm, in the obtained radargram image, the number of hyperbolas and also the parameters of them are determined automatically.

      

Key Words: Multimodal Optimization, Artificial Bee Colony Algorithm, Multi Circle Detection, Ground Penetrating Radar Hyperbolas.