Improvement Artificial Bee Colony Algorithm And Its Application In Image Segmentation

Artificial bee colony algorithm(ABC)is an intelligent optimization method based on swarm intelligence.It is an algorithm produced by simulating the social behavior of bees in nature.It has good search performance in a variety of optimization problems.All along,artificial bee colony algorithm has become popular because of its good performance and simple structure.However,the performance of ABC algorithm is often affected by the development ability and exploration ability,and how to balance the development and exploration in the algorithm is particularly important.For some complex problems,although the solution search equation of ABC algorithm has strong exploration ability,its development ability is poor.In addition,image segmentation is one of the most difficult problems in digital image processing,which is one of the classic problems in the field of computer vision.But image segmentation is always a kind of complex optimization problem.In this thesis,the improved artificial bee colony algorithm is used to deal with the gray threshold segmentation problem of image.,and the current more efficient intelligent optimization algorithm is used to solve the actual application problems,which can realize the automatic image processing.The main research work of this thesis is summarized as follows:(1)ABC algorithm has a simple structure and is difficult to solve complex optimization problems.A Group-guided artificial bee colony algorithm with elastic adjustment strategy(EGABC)is proposed.In EGABC,we propose a group boot strategy with elastic adjustment capability.EGABC proposes two improved strategies and a grouping guidance model.To solve the problem of slow convergence of standard ABC algorithm,a grouping guidance strategy guided by elite individuals is designed to divide the whole population into several groups,and the individuals in each group are guided by elite individuals or global optimal individuals.The current known information can be used effectively to accelerate the convergence rate of the algorithm.Secondly,in order to balance the exploration and development capabilities of the algorithm,the adaptive parameters of dynamic elastic adjustment are added,which can effectively adjust the search efficiency of the algorithm and balance the exploration and development performance of the algorithm.In order to verify the validity of EGABC,we conducted experiments on 13 benchmark functions and cec2013 test functions.Compared with other ABC variants,the results show that the strategy proposed by EGABC has strong competitiveness and can effectively improve the performance of ABC algorithm.(2)the improved artificial bee colony algorithm EGABC is introduced to solve the image segmentation problem in image processing.The improved algorithm is applied to the traditional gray threshold segmentation problem of Kapur entropy method and Otsu method,and the improved artificial bee colony algorithm is used to find the optimal threshold.In the gray threshold segmentation experiment,the authoritative standard Berkeley segmentation data set is used,which are three classical evaluation functions: peak signal-to-noise ratio(PSNR),structural similarity index SSIM and feature similarity FSIM.By combining the two improved strategies,the improved algorithm obtained has good performance and strong adaptability in the experiment,which can enhance the quality and effect of image segmentation.In summary,I have applied an improved artificial bee colony algorithm to grayscale image threshold segmentation,and using this algorithm to find the optimal threshold can achieve better results.