Research On Image Processing Based On Sparrow Search Algorithm

With the development of society,video and image have become the main carriers for people to obtain information flow.Under the impact of diversified and complex image information flow,it is often difficult to achieve ideal results if the previous processing methods are still used when analyzing image features and semantics.The swarm intelligence optimization algorithm proposed according to the living and breeding behavior of biological communities in nature can help solve combinatorial optimization problems.This article mainly studies the use of sparrow search algorithm to optimize and solve the problem of insufficient accuracy in infrared image segmentation,as well as the problem of uneven enhancement in low illumination image enhancement,and has achieved good experimental results.The main research content of this paper includes the following two aspects:(1)Aiming at the problem of poor accuracy of traditional threshold segmentation in infrared image detection of power equipment,an image segmentation method based on variable spiral sparrow search algorithm(VSSSA)was proposed.Firstly,an optimization model combining two-dimensional dual threshold Otsu segmentation and automatic region growth method is constructed;Secondly,the initialization stage,the finder stage,and the follower stage of the sparrow search algorithm are improved by introducing variable spiral,Tent chaotic mapping,and Levi flight strategy to balance population diversity and algorithm convergence;Finally,the proposed VSSSA algorithm is used to optimize the evaluation function to improve the accuracy of segmented images.This method not only fully considers the neighborhood information of grayscale images,but also further extracts target regions from images with incomplete threshold segmentation.Experiments show that this method effectively improves the accuracy of infrared segmentation.(2)To solve the problem of uneven effect of multiscale Retinex in low illumination image enhancement and prone to excessive enhancement,a low illumination image enhancement method based on improved sparrow search algorithm is proposed.A multi scale Retinex image enhancement model based on adaptive weights is constructed,and the weights of scale parameters in different regions are determined according to the content characteristics of the image;At the same time,aiming at the problem that the standard sparrow search algorithm is difficult to jump out of the local extreme value in the late search period,an improved algorithm incorporating the idea of artificial bee colony(ABSSA)is proposed,and the proposed ABSSA algorithm is used to set appropriate scale parameters to ensure that each region achieves the best enhancement effect.Compared with other algorithms,this algorithm has clearer details and fuller colors.