双群交换微粒群算法及在Shearlet图像去噪中的应用.pdfVIP

ComputerEngineeringandApplications计算机工程与应用 2011，47(19) 207 双群交换微粒群算法及在Shearlet图像去噪中的应用 赵 嘉’，曹寒问，孙 辉’，李文力 ‘ ZHAo Jia’，CAO Hanwen0，SUN Hui。，LIWenli‘ 1．南昌工程学院 信息工程学院，南昌 330099 2．南昌工程学院 理学系，南昌 330099 1．SchoolofInformation Engineering，Nanchang InstituteofTechnology，Nanchang 330099，China 2DepartmentofScience，NanchanglnsitituteofTechnology，Nanchang 330099，China ZHAO Jia。CAO Hanwen，SUN Hui，eta1．Two sub-swarmsexchangeparticle swarm optimization and itsapplication in Shear!~timagedenoising．ComputerEngineeringandApplications。2011，47(19)：207—210． Abstract：By analyzing the deficienciesoftwo sub--swarmsexchange particle swarm optimization basedon differentevolu-- tionary models，an improved two sub·swarms exchangeparticle swarm optimization algorithm isproposed．Thealgorithm di— videsthe particles into two sub-swamr softhe same size，the firstsub—swarm using the standard PSO evolutionary model and the second sub-swarm using the Cognition Only evolutionary mode1．When theparticlesevolve to a stable state，part of theparticlesfrom the first sub-swaltn arerandomly selected to exchange with particleswith theworstfitnessvalues from the second sub．swarm．The above．mentioned operations are repeated untilthe optimalsolution is found．Experimentresults show thattheproposed algorithm hasbeaer globalsearch capability and optimalconvergence rate．To verify thepracticality of the algorithm ，the improved algorithm isapplied to Shearletimage denoising．According to the distribution characteristics ofSheariettransfomr domain coefficientsofdifferentscalesand directions，themethod usesthe improvedalgorithm to adap- tively determine theoptimalthresholdsof differentscalesand directions，to achieve image content-based adaptive denoising． Experimentresultsshow thatthemethod can effectively filteroutimagenoiseandbetterretain edge information，and thede— noisedimageshavehigherPeakSignaltoNoiseRatio(PSNR)． Keywords：particleswarm optimization；exchan