Image Denoising Using Cross-Validation Method with RBF Network Representation
Hong Li · 2004
This paper presents an image denoising method based on RBF neural network (RBFNN), shows the method's advantages and disadvantages and gives a comparison between this method and other classical image denoising methods. RBFNN is good dat function approximation, especially approximating to smooth functions. Experiments show that RBFNN can well approach the image while repressing the noise.