M.Sc. Tezi Görüntüleme

Student:	Büşranur KÜÇÜKUĞURLU
Supervisor:	Assist. Prof. Dr. Eyüp GEDİKLİ
Department:	Bilgisayar Mühendisliği
Institution:	Graduate School of Natural and Applied Sciences
University:	Karadeniz Technical University Turkey

Title of the Thesis:	IMAGE SEGMENTATION BASED ON META-HEURISTIC APPROACHES
Level:	M.Sc.
Acceptance Date:	19/6/2019
Number of Pages:	81
Registration Number:	i3602

Summary:

Numerous methods for image segmentation have been proposed and discussed in the literature for many years. Thresholding is the most commonly used technique in image segmentation, which is easy to implement and produces favorable results. Kapur Entropy and Otsu method are proven methods, which frequently used in the literature.

In this thesis, segmentation of several images has been performed by using Kapur’s Entropy and Otsu thresholding methods with existing meta-heuristic algorithms and a new meta-heuristic algorithm, Symbiotic Organisms Search, together. The computational complexity of the processes was reduced by these methods and more favorable results were obtained. In addition, comparison and experimental results of all methods were reported.

Keywords: Segmentation, Thresholding, Meta-Heuristic Algorithm, Otsu, Kapur’s Entropy, Particle Swarm Optimization, Firefly Algorithm, Artificial Bee Colony, Genetic Algorithm, Symbiotic Organisms Search.