计算机技术与发展Issue(6)：149-152,4.DOI:10.3969/j.issn.1673-629X.2014.06.037
基于粒子群优化的BP神经网络图像复原方法
A Method of Image Restoration Based on Particle Swarm Optimization for BP Neural Network
摘要
Abstract
Aiming at the problem of local minimum,slow convergence of the BP neural network,the learning algorithm based on particle swarm optimization is designed and analyzed,which is applied to image restoration. Firstly,noiseless images are processed by Gaussian noise. Then,the result image and the noiseless images are made training pair,which is used in training the optimized BP neural network. Lastly,use the BP neural network to restore test images for the purpose of removing noise. The simulation results show that the effect of PSO-BP algorithm to recover the image have fast convergence rate and less iterations,is better than the BP neural network both in Nor-malized Mean Square Error ( NMSE) and the Peak Signal to Noise Ratio ( PSNR) .关键词
BP神经网络/粒子群优化/图像复原Key words
BP neural network/particle swarm optimization/image restoration分类
信息技术与安全科学引用本文复制引用
宋发兴,高留洋,刘东升,米兰,刘力维..基于粒子群优化的BP神经网络图像复原方法[J].计算机技术与发展,2014,(6):149-152,4.基金项目
国防“十二五”预研基金(40405070102) （40405070102）
