自适应区域生长耦合神经P系统的图像分割方法

打开文本图片集

关键词：图像分割；金豺优化算法；耦合神经P系统；区域生长

中图分类号：TP391.41 文献标志码：A 文章编号：1001-3695（2026）01-034-0280-11

doi：10. 19734/j.issn. 1001-3695. 2025.06.0163

Image segmentation method for adaptive region growing coupled neural P systems

XuJiachang†，YeXuxiu

Abstract：Toaddress theissue that thecoupledneuralPsystemreliesontheselectionof initialseed points whenusing the spiking mechanism for region growing，this paper proposed an image segmentation method based on adaptive region growing coupled neuralPsystems（ARGCNP）.This method employed theglobal search capabilityof the golden jackal optimization （GJO）algorithmandintroducedfourstrategies toenhancethe globaloptimizationperformanceofGJO.Theaimwastodentify theoptimalthresholdpointsintheimagetooptimizetheselectionofseedpointsinregiongrowing.Theexperiments firstly evaluatedtheperformanceof theimprovedGJOontheCEC2O17benchmark functions，andshowed thatitachieved hebest overallperformance.ThenapliedARGCNPtothesegmentationofbothcolorandmedical images，ndquantitativelysseed thesegmentationperformanceusingthreeevaluationmetrics，includingpeaksignal-to-noiseratio.Theresultsdemonstratethat the proposed method improvessegmentationaccuracyand enhances the stabilityof theresults，confirming thatARGCNPoffers clear advantages in practical applications and efectively meets the requirements of image segmentation tasks.

Key words： image segmentation；golden jackal optimization；coupled neural P systems；region growing

0 引言

图像分割[12]是一项基本的计算机视觉任务，旨在将图像划分为多个具有相似特征的区域，分割后图像的精确性直接影响后续任务的效果。（剩余23885字）