THE COMPUTER SYSTEM OF MEDICAL IMAGE SEGMENTATION BY ANT COLONY OPTIMIZATION

The image segmentation is one of the most important and complex low-level image analysis tasks. Because it is one of the first stages of image recognition, the next steps, such as the allocation of entities, classification and recognition, largely depend on its results. Therefore, the image segmentation is the subject of intense research. There are a lot of segmentation methods, but each of them has its own advantages and disadvantages. New segmentation methods based on swarm intelligence look are promising for researching. They are ant colony optimization algorithm, swarm optimization, fish and bacteria fouraging algorithms etc. These algorithms are based on the behavior modeling of set of agents and inspired by the nature, especially by biological systems. The mixed segmentation algorithm of K-means and ant colony optimization was implemented and analyzed in the presented paper. The software system for visualization and approbation of the developed algorithm was implemented too. The algorithm was tested on public benchmark Berkley. We have obtained the output processed images, as well as the values of heuristic coefficients of the algorithm. The results are compared with output data obtained by Osiriss system.