Geometric Feature Image Segmentation And Optimization Based On Nonlinear Programming Genetic Algorithm

Nowadays,with the rapid development of virtual reality technology,As an image preprocessing step,image segmentation becomes more important.How to better improve the quality of image preprocessing,improve image segmentation quality,and improve segmentation efficiency have become the key to image segmentation technology.However,due to the rapid development of image processing technology,the traditional segmentation techniques,such as threshold segmentation,region segmentation,cluster analysis,etc.,have been unable to meet the processing requirements of complex images.This paper combines the traditional image segmentation algorithm with the genetic algorithm to optimization processing.To solve the problem of local convergence and low partition efficiency in the segmentation process.The main work and innovation of this article are as follows:(1)First,on the basis of deep research on the genetic algorithm coding and decoding and its related steps,in order to verify the effectiveness of genetic algorithm in geometric feature image segmentation,the key techniques of geometric feature image segmentation based on genetic algorithm and largest interclass difference method are analyzed and implemented.The inter-class variance function as the fitness function for the genetic algorithm.By Selection,cross-cutting,variability operating of the genetic algorithm to preserve the species diversity,the local convergence can be effectively avoided,and the global-based optimal threshold can be obtained.(2)Based on the genetic algorithm,the fuzzy theory and the energy function are analyzed at the same time.In the geometric feature image segmentation algorithm based on the fuzzy energy function and the genetic algorithm,we first use the bi-convex fuzzy model to improve the traditional CV model.Then,The genetic algorithm is used for the optimization operation for the first time in the search of penalty length in the improved energy function.Finally,a global minimum energy function is obtained,so that a closed contour curve that can clearly segment the edges of the image is obtained.(3)Because of the weak local search ability of genetic algorithm,this article deeply studies the related properties of nonlinear programming,formed a genetic algorithm based on nonlinear programming,we combines with fuzzy C-means clustering algorithm at the some time,The objective function in the fuzzy C-means clustering is used as the fitness function in the nonlinear programming genetic algorithm.The nonlinear programming strengthens the local search,and the genetic algorithm itself guarantees the global optimum,so that a more ideal segmentation effect can be obtained.Experiments show that this paper analyzes the key techniques of geometric feature image segmentation based on genetic algorithm and verifies the effectiveness of genetic algorithm in geometric feature image segmentation.At the same time,the two optimized segmentation methods proposed in this paper have better segmentation results compared with the existing traditional segmentation algorithms.Secondly,the relevant numerical evaluation indexes are given,and the effectiveness and stability of the proposed method are proved through the related contrast experiments.