摘要
基于传统反向传播(back propagation,BP)网络,提出一种结合细菌觅食算法(bacterial fora-ging algorithm,BFA)的改进型BP网络(BFA-BP),并将其用于图像压缩.为克服传统BP网络容易陷入局部极小值的缺点,算法引入BFA特有的复制和驱散操作,以提高网络收敛速度,加强全局寻优能力.对标准测试图像进行仿真实验表明,该算法能有效提高重建图像质量.
An improved back propagation network （ BP network） combined with bacterial foraging algorithm （BFA） was applied to image compression. The particular operations, i.e., reproduction and elimination-dispersal in BFA, were introduced to the BFA-BP algorithm to prevent the premature saturation in traditional BP network. The convergence speed and global search ability are enhanced. The BFA-BP algorithm was evaluated on standard test images in comparison with traditional BP algorithm. The simulation results demonstrate the effectiveness of the proposed algorithm in improving the quality of reconstructed images.
出处
《深圳大学学报（理工版）》
EI
CAS
北大核心
2008年第2期153-157,共5页
Journal of Shenzhen University(Science and Engineering)
基金
国家自然科学基金资助项目(60572100)
国家自然科学基金委员会与英国皇家学会合作资助项目(60711130233)
深圳大学科研启动基金资助项目(200845)
关键词
人工神经网络
反向传播
细菌觅食算法
生物启发式计算
图像压缩
artificial neural network
back propagation network
bacterial foraging algorithm
biologically inspired computation
image compression
