Gray scale image segmentation using artificial bee colony optimization algorithm

Digitally evaluation of the gray scale and color images in an important step. Image segmentation process is partitioning a given image according to various features. Image thresholding is known as the basic image segmentation method but it is not enough good for all image segmentation applications. In this thesis, a new method is proposed based on wavelet transformation, entropy function and artificial bee colony algorithm for gray texture segmentation. For feature extraction, the wavelet transformation is used. The entropy function is used to find the optimum threshold on the normalized wavelet coefficients. Fort his purpose, an optimization procedure called the artificial bee colony algorithm is used. Artificial bee colony algorithm, which is based on the swarm intelligence is an optimization algorithm and used for finding the optimum automatic threshold fort he segmentation of the input image. Based on the experimental results, the proposed method yields better segmentation results. The proposed method yields worse results for several images but the running time of the algorithm is considerable short. Key Words: Gray texture image segmentation, Swarm intelligence, Artificial bee colony algorithm, Wavelet transform, Entropy function.