| Abstrakt: |
Image segmentation is one of the most significant and required procedures in pre-processing and analyzing images. Metaheuristic optimization algorithms are used to solve a wide range of different problems because they can solve problems with different dimensions in an acceptable time and with quality results. It can show different functions in solving various problems. So, a metaheuristic algorithm should be adapted to solve the target problem with different mechanisms to find the best performance. In this paper, we have used the improved African Vultures Optimization Algorithm (AVOA) that uses the three binary thresholds (Kapur's entropy, Tsallis entropy, and Ostu's entropy) in multi-threshold image segmentation. The Quantum Rotation Gate (QRG) mechanism has increased population diversity in optimization stages, and optimal local trap escapes to improve AVOA performance. The Association Strategy (AS) mechanism is used to obtain and faster search for optimal solutions. These two mechanisms increase the diversity of production solutions in all optimization stages because the AVOA algorithm focuses on the exploration phase almost in the first half of the iterations. So, in this approach, it is possible to guarantee a wide variety of solutions and avoid falling into the local optimum trap. Standard criteria and datasets were used to evaluate the performance of the proposed algorithm and then compared with other optimization algorithms. Eight images with large dimensions have been used to evaluate the proposed algorithm so that the ability of the proposed algorithm and other compared algorithms can be accurately checked. A better solution to large-scale problems requires good performance of the algorithm in both the exploitation and exploration phases, and a balance must be created between these two phases. According to the experimental results from the proposed algorithm, it is determined that it has a good and significant performance. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink