NeuralNetworkAdaptive Switching Median Filter forImage Denoising
暂无分享,去创建一个
A new neuralnetworkadaptive switching median(NASM)filter isproposed toremovesalt-and-pepper impulse noise fromhighly corrupted image. Thealgorithm is developed by combiningadvantages of the known progressive medianfilter withimpulse detection scheme. Neuralnetwork wasincluded intoimpulse detection stepto improve itscharacteristics. Comparison ofthegiven method withtraditional median-type filters isprovided.
[1] Moncef Gabbouj,et al. Weighted median filters: a tutorial , 1996 .