Research On Otsu Image Segmentation Algorithm Based On Firefly Algorithm Optimized By Fractional-order

In computer vision,image segmentation is an important step in image processing.The Otsu segmentation method is one of the classic image segmentation methods,which can effectively segment most images and has a wide range of applications.Based on the in-depth study of traditional two-dimensional Otsu threshold segmentation and traditional firefly algorithm,this paper eombines the fractional differential with the firefly algorithm according to the shortcoming of its inherent convergence speed,and optimizes the traditional Otsu algorithm and firefly algorithm.The new algorithm is applied to image segmentation.The main research contents of this paper are as follows:(1)The status quo of image segmentation algorithm,firefly algorithm and fractional calculus theory at home and abroad are studied.The classification of image segmentation algorithm and the principle of one-dimensional and two-dimensional Otsu algorithm are studied.The basic theory,mathematical model and algorithm flow of firefly algorithm are studied.And the parameter setting and performance have been deeply studied and analyzed.(2)Proposed fractional firefly improvement algorithm:The traditional firefly algorithm has the advantages of simple implementation and strong searching ability.However,in the search process,each firefly uses only the last memory,and there is a problem of slow convergence.Considering that the fractional differential can fully describe the memory and genetic characteristics of the event,it can improve the memory ability of the firefly to control its convergence.Therefore,an improved fractional firefly algorithm is proposed to introduce the fractional differential into the standard firefly position update formula.Find a reasonable solution through the more historical memory of each firefly.The improved fractional firefly algorithm has a faster convergence speed than the standard firefly algorithm,which can effectively improve the traditional firefly early maturity and fall into local optimum problems.(3)Improved fractional firefly algorithm applied to image segmentation:Validation of the proposed Otsu image segmentation algorithm based on improved fractional firefly,applied to various types of classic images for simulation experiments,and traditional The firefly-optimized Otsu algorithm,the fractional-order particle swarm optimization-based Otsu algorithm and the proposed algorithm respectively perform subjective and objective comparison analysis on the segmentation results.The algorithm is validated by three kinds of inconsistent images,namely,person image,scene image and medical image.The experimental results show that the proposed algorithm can achieve accurate segmentation of the target image.Compared with the traditional firefly-optimized Otsu algorithm and the Otsu algorithm based on fractional-order particle swarm optimization,the convergence speed is faster under the premise that the segmentation accuracy is basically the same.