Research On Image Processing Method Based On The Fruit Fly Optimization Algorithm

The fruit fly optimization algorithm(FOA)is a new method for global optimization,which is based on the reasoning of fruit fly's foraging behavior.The algorithm can understand and implement quickly because of less parameters to be adjusted.Moreover,flies are more sensitive than other species in sense of smell and vision,so they can quickly explore the unique taste of food,even if the target source is far away.However,the traditional fruit fly optimization algorithm is not mature enough,such as: the overall search time is long,,Low convergence accuracy,and it is simple to fall in the part optimum.Because the step size of the traditional FOA is fixed,it leads to many problems.Therefore,this paper proposes a fruit fly optimization algorithm that automatically improves the step size through concentration changes and applies it to image processing.The important work is below:(1)A half-cursive Cauchy function,a mixed tangent function and a Cauchy operator's Tango Cauchy Operator Fruit Fly Optimization Algorithm(TCO-FOA),which increases or decreases the step size according to the iteration number.The variable of the tangent function is the ratio of the sum of the best concentration value and the worst concentration value in the current iteration to the average value of the best concentration value and the sum of the worst concentration value in the previous iteration of the fruit fly group.When the average change ratio of the concentration is greater than 1,using the characteristics of the half-century Cauchy function,the step length is uniform at the early stage and then increases in an S-shape.When the average change ratio of the concentration is less than or equal to 1,at this time,it enters the post-iteration stage,and the fruit fly is very close to the target source.As the ratio of the average concentration decreases,a local search is performed by reducing the step size to increase the success of finding the target source rate.The improvement of the Fruit Fly Optimization algorithm in this paper has effectively improved the convergence rate and optimization accuracy based on the original.(2)TCO-FOA is introduced into image thresholding to optimize segmentation: the maximum entropy thresholding is used to segment image.When improving image segmentation,most researchers will use image entropy segmentation for experiments.Otsu thresholding segmentation is a method that uses image binarization to solve image segmentation.It has a good effect in experiment,and most of them choose it in practice.However,no matter which method,the period division effect is not good,and the division time is long.The comparison with the traditional algorithm shows that,the TCO-FOA algorithm has made a breakthrough in the stability and effect of image segmentation.(3)TCO-FOA is introduced into the matching to combine the two.Image matching is a relationship of geometric space and gray intensity established between the original image and the template feature image.It has an irreplaceable position in the visual module of modern computer.It is mainly used in video tracking,target recognition,medicine and other fields.Image matching can be divided into two directions: one is based on unique features,the other is based on gray level.Although the two directions of the algorithm in the matching effect is good,but the convergence speed and matching rate still must be improved.By comparing the algorithm matching with several algorithms,it is proved that the adaptive change step size algorithm can achieve better matching algorithm.