Iterative Adaptive Median Filter for Impulse Noise Cancellation
Hongzhi Wu · Shandong kexue · 2010
This paper presents an iterative adaptive median filter(IAMF) approach based on the characteristics of impulse noise.The approach constructs the weight function of an impulse noise contaminated image according to its characteristics and then eliminates noise with an iterative algorithm.An IAMF processed image can not only reserve the micro information of its original image but also have better definition.This approach effectively avoids the influence of noise dots on the non-noise dots of an image and improves computation speed since those noise dots are not involved in the calculation.This approach has a better denoise effect for an impulse noise contaminated image,especially when the noise rate is above 0.5.This approach still has better satisfactory effect when the noise rate is above 0.9.