摘要
提出基于小波神经网络的图像去噪方法,该方法兼有小波分析的良好时频域特性和神经网络的自适应能力。实验结果表明,该方法在去除噪声上优于中值滤波等传统去噪声方法,其散斑指数较小,峰值信噪比较大,在有效去除噪声同时,又能很好地保护图像的细节信息。
An image noise-fihering method based on wavelet neural network was put forward, which has good time-frequency features of the wavelet analysis and the adaptive ability of neural network. Experimental results showed that the method excels traditional noise filtering method such as median filtering method in noise-filtering; it has lower speckle index and high PSNR; it is effective in noise removal and good at protection of detail information of images.
出处
《包装工程》
CAS
CSCD
北大核心
2009年第8期28-30,共3页
Packaging Engineering
关键词
散斑图像
小波变换
去噪
神经网络
speckle image
wavelet transforrn
noise removal
neural network
