Multilevel-thresholding is an efficient method used in image segmentation. This paper presents a hybrid meta-heuristic approach for multi-level thresholding image segmentation by integrating both the artificial bee colony (ABC) algorithm and the sine-cosine algorithm (SCA). The proposed algorithm, called ABCSCA, is applied to segment images and it utilizes Otsu's function as the objective function. The proposed ABCSCA uses ABC to optimize the threshold and to reduce the search region. Thereafter, the SCA algorithm uses the output of ABC to determine the global optimal solution, which represents the thresholding values. To evaluate the performance of the proposed ABCSCA, a set of experimental series is performed using nineteen images. In the first experimental series, the proposed ABCSCA is assessed at the low threshold levels and compared with the ABC and SCA as traditional methods. Moreover, the second experimental series aims to evaluate the ABCSCA at high threshold levels and it is compared with six algorithms in addition to the SCA and ABC. Besides, the proposed method is evaluated using the fuzzy entropy. The results demonstrate the effectiveness of the proposed algorithm and showed that it outperforms other algorithms in terms of performance measures, such as Peak Signal-to-Noise Ratio (PSNR) and the Structural Similarity Index (SSIM).

Improved Artificial Bee Colony Using Sine-Cosine Algorithm for Multi-Level Thresholding Image Segmentation

The shortest path (SP) problem constitutes one of the most prominent topics in graph theory and has practical applications in many research areas such as transportation, network communications, emergency services, and fire stations services, to name just a few. In most real-world applications, the arc weights of the corresponding SP problems are represented by fuzzy numbers. The current paper presents a fuzzy-based Ant Colony Optimization (ACO) algorithm for solving shortest path problems with different types of fuzzy weights. The weights of the fuzzy paths involving different kinds of fuzzy arcs are approximated using the α -cut method. In addition, a signed distance function is used to compare the fuzzy weights of paths. The proposed algorithm is implemented on three increasingly complex numerical examples and the results obtained compared with those derived from a genetic algorithm (GA), a particle swarm optimization (PSO) algorithm and an artificial bee colony (ABC) algorithm. The results confirm that the fuzzy-based enhanced ACO algorithm could converge in about 50% less time than the alternative metaheuristic algorithms.

A novel ant colony algorithm for solving shortest path problems with fuzzy arc weights

Modified hybrid bat algorithm with genetic crossover operation and smart inertia weight for multilevel image segmentation

A Grunwald–Letnikov based Manta ray foraging optimizer for global optimization and image segmentation

Multi-level image thresholding is the most direct and effective method for image segmentation, which is a key step for image analysis and computer vision, however, as the number of threshold values increases, exhaustive search does not work efficiently and effectively and evolutionary algorithms often fall into a local optimal solution. In the paper, a meta-heuristics algorithm based on the breeding mechanism of Chinese hybrid rice is proposed to seek the optimal multi-level thresholds for image segmentation and Renyi’s entropy is utilized as the fitness function. Experiments have been run on four scanning electron microscope images of cement and four standard images, moreover, it is compared with other six classical and novel evolutionary algorithms: genetic algorithm, particle swarm optimization algorithm, differential evolution algorithm, ant lion optimization algorithm, whale optimization algorithm, and salp swarm algorithm. Meanwhile, some indicators, including the average fitness values, standard deviation, peak signal to noise ratio, and structural similarity index are used as evaluation criteria in the experiments. The experimental results show that the proposed method prevails over the other algorithms involved in the paper on most indicators and it can segment cement scanning electron microscope image effectively.

Renyi’s Entropy Based Multilevel Thresholding Using a Novel Meta-Heuristics Algorithm

For the traditional multi-thresholds segmentation algorithms, usually it would take too much time in finding the optimal solution. As one of the widely used swarm-intelligence optimization algorithms, ant colony optimization (ACO) algorithm has been introduced to optimize the thresholding search process. The traditional ACO is improved in this paper to get a faster convergence speed and applied in Otsu multi-thresholds segmentation algorithms. When the ant colony is initialized, each member of the ant colony is distributed evenly in the solution space, so that it could search the entire solution space as fast as possible. In the search process, the random step length of ants moving is generated by the Levy flight pattern, but the global transition probability of the traditional ACO is used to control the search range of the ant colony. The experimental results show that the proposed algorithm could obtain the optimal thresholds faster and more effectively than the traditional Otsu algorithm and the Otsu based on traditional ACO.

Zheng Fang

Papers

An Otsu multi-thresholds segmentation algorithm based on improved ACO