Research On Neutrosophic Image Segmentation Based On Improved Fuzzy C-Means Algorithm

Image segmentation is a key step of image processing and image recognition.How to segment image accurately and efficiently has always been the key to computer vision and image processing.The purpose of image segmentation is to divide the image into non overlapping and characteristic regions.However,in the actual segmentation process,in addition to the noise of the image itself,there are a lot of other uncertain factors,resulting in the unsatisfactory segmentation results.In order to obtain a good segmentation effect,in addition to improving the existing image segmentation algorithm,scholars put forward a method that combines other intelligent algorithms with specific theories,such as image segmentation based on feature space clustering,which combines pixels in image space.According to their aggregation in the feature space,the feature space is segmented,and then they are mapped back to the original image space to get the segmentation results.The fuzzy C-means(FCM)algorithm based on cluster analysis belongs to this type of algorithm,which achieves the clustering purpose by optimizing the corresponding fuzzy objective function.The fuzzy C-means(FCM)algorithm based on clustering analysis achieves the goal of clustering by optimizing the corresponding fuzzy objective function.Because of its simple design and less human intervention,it is suitable for dealing with the uncertainty in the real object,so it has attracted more and more attention and has been widely used in various fields.In order to better deal with the uncertainty information in the image and improve the accuracy of image segmentation,this paper introduces neutrosophic theory into the image segmentation algorithm.Neutrosophy belongs to the category of philosophy,which can better quantify the uncertainty of the data.In this paper,three image segmentation methods are proposed by combining FCM algorithm with the theory of neutrosophic,and the proposed method is further applied to the field of medical image,which has a certain practical application value.The main innovative work of this paper can be summarized as follows:(1)Analyze and research the shortcomings of FCM algorithm in the current image segmentation process,introduce the neutrosophic theory into the FCM algorithm,and propose the FCM algorithm based on the neutrosophic fuzzy set.This method handles the fuzzy edges more reasonably,and applies it to the neutrosophic image segmentation process.The experimental results show that the image segmentation algorithm introduced by neutrosophic theory can obtain better segmentation results than FCM algorithm.(2)In order to improve the FCM image segmentation process,the FCM algorithm has a strong dependence on the initial value parameters and the segmentation accuracy is not high.Particle Swarm Optimization(PSO)is introduced into the FCM image segmentation process,and the rapid convergence of the PSO algorithm is used to improve the the ability of anti-noise interference,thereby improving the image segmentation accuracy.The experimental results show that in the image segmentation process of neutrosophology,the PSO-FCM proposed in this paper can obtain better results than FCM.(3)When PSO algorithm is applied to FCM image segmentation,the accuracy of image segmentation is significantly improved,but the number of iterations of the algorithm is increased by ten times.In order to reduce the number of iterations,the velocity and position update mode of the PSO algorithm is improved.In the novel method,the velocity updating method of PSO algorithm is divided into two kinds,and the objective function obtained by comparing the two velocities is compared.The corresponding velocity of the optimal objective function is used as the current generation velocity and the position of the PSO particles is updated.Then the IPSO-FCM is used to segment the neutrosophic image.The experimental results show that the method in this paper can converge to obtain the optimal solution in a faster time,and get a better result.(4)The essence of FCM algorithm is to use gradient descent method to find the global optimal solution,so it is easy to fall into the local optimum;The convergence speed of PSO algorithm is too fast in the early stage of the iterative optimization process,which is easy to cause the diversity of the population to decrease.And the convergence speed is slowed in the later stage of the iterative process,and it is easier to fall into the local optimal.In order to improve the diversity of the convergence process and balance the global search ability and local development ability of the algorithm,this paper introduces a particle swarm algorithm that can guarantee global convergence,namely the Quantum-behaved Particle Swarm Optimization(QPSO).The QPSO-FCM algorithm is proposed and applied to the process of neutrosophic image segmentation.The experimental results show that this method can find the global optimal solution,and the boundary segmentation of the test image is clear and accurate,and it can also obtain satisfactory results for the medical image that requires high-precision segmentation.