The parameters of the modified mixed Ant Colony Optimization (ACO) - k-means image segmentation algorithm are considered. There have been investigated such parameters as n - the number of ants; heuristic coefficients of ACO algorithm and their dependence on the image scale and number of iterations before and after parameters correction. The proposed...
Purpose. Development of new methods for medical images segmentation, since modern methods of medical diagnostics are largely based on image processing MRI, CT scan etc. Materials and Methods. A hybrid particle swarm algorithm for medical image segmentation is proposed, which includes a modified and elite exponential particle swarm segmentation algorithm in combination with the k-means method. The time complexity of the developed algorithm is investigated on the basis of the drift analysis. It is shown that the developed algorithm for segmentation of MRI images has a polynomial time complexity. Results. To test the developed algorithms, images of the Ossirix test set and real medical images were used. When comparing the operating time of the proposed segmentation methods, it was found that the hyper-heuristic swarm method shows the operating time on average 2 times less than when using the hybrid ant method, and the segmentation results for good quality images and blurry images are comparable. Conclusions. Several modifications of swarm image segmentation algorithms have been proposed for various types of medical images, including contrasting, noisy and blurred images, which showed good results during testing and low time complexity.
Image segmentation is the process of dividing an image into homogeneous regions according to certain features and is widely used in image processing. Complexly structured images usually contain complex and essential objects. These images are non-linear structural images and they contain a large number of elements with required specifications. The main goal of the proposed EPSO (Exponential Particle Swarm Optimization) algorithm is to prevent local solutions and find the exact global optimal solutions for the task of segmenting medical images. The execution time is compared with well-known segmentation algorithms. The EPSO method is superior to the segmentation methods studied, including the graph algorithm. Comparisons were made with existing segmentation algorithms (Grow cut, Random Walker, DPSO, K-means PSO, and hybrid-K-means ant colony optimization algorithm) in tabular form.