Remove extremum median filtering and minimal absolute difference of four directional filtering on improved PCNN model
Yi-Bev Yu, Yulan Zhang · 2015
The pulse coupled neural network (PCNN) has been widely applied to remove the image impulse noise due to its characteristics of variable threshold and synchronous pulse bursts. However, the denoising effect will be significantly worse when the noise density is too big. In this paper, Remove Extremum Median Filtering and Four Directional Minimal Absolute Difference Filtering algorithms are proposed based on the contours and edges continuity of the images. Simulation experiments show that, compared with the median filter based on improved PCNN, the two proposed algorithms have better performance in denoising quality and computing speed. They can also be applied to other image processing problems such as image restoration and edge detection.