Multilevel image threshold segmentation using an improved Bloch quantum artificial bee colony algorithm

Multilevel threshold segmentation is one of the most broadly used image segmentation methods. The key problem is how to obtain the optimal threshold as soon as possible. So a novel method based on the analysis of artificial bee colony algorithm, quantum Bloch sphere and Kapur's entropy is put forward, and it is applied to the multilevel thresholds of typical images efficiently. In the first place, in order to improve the performance of artificial bee colony (ABC) algorithm, the updating strategy is improved by combining the Bloch spherical coordinates of qubit with ABC algorithm. Then an improved Bloch quantum artificial bee colony (IBQABC) is proposed. There is one more point that IBQABC is applied to the optimization of multidimensional benchmark function, and it is proved that the algorithm has quick convergence speed compared with other algorithms. Finally, IBQABC combined with Kapur's entropy segments the benchmark gray images with different characteristics. After comparing the results of threshold segmentation of different images by using GA, PSO, ABC and IBQABC algorithms, it is verified that the IBQABC algorithm is superior to other conventional algorithms in the overall performance of gray image multilevel threshold segmentation, and it is determined that the improved algorithm has superior segmentation effect and strong generalization ability in gray image multilevel threshold segmentation.