摘要
神经网络在智能化信息处理和非线性滤波方面的优异表现,使得越来越多的人将神经网络应用到图像处理中。提出一种将非线性滤波器和神经网络结合所构成的滤波器,实验证明这种滤波器在消除图像的混合噪声、保护图像的边缘信息方面具有明显的效果,通过对受到较大噪声影响的输入图像进行滤波,可使输出图像的信噪比提高10dB以上。
Neural network is increasingly applied in image processing for its good performance on intelhgent information processing and nonlinear filtering, This paper proposes a new method which combines nonlinear filter with neural networks. The experiment results show its distinct effect in eliminating hybrid noises and protecting the edge of image, For the heavy noised image, The SNR of output image can be improved more than lOdB after using the new method.
出处
《信息技术》
2005年第11期55-58,149,共5页
Information Technology
关键词
神经网络
非线性滤波器
混合噪声消除
图像处理
neural network
nonlinear filter
hybrid noise eliminating
image processing
