Selective adaptive weighted median filter
Caiquan Xiong · Optical Engineering · 2008
A new impulsive noise removal algorithm, the selective adaptive weighted median filter (SAWMF), is introduced. The proposed solution is a class of adaptive weighted median filters with incorporation of a switching mechanism. Using a median-based comparison technique to classify each image pixel as an impulse or a noise-free one, the new algorithm employs a weighted median filter where the weights are adaptively selected from two fixed values to restore the detected noisy pixels and keep the noise-free ones unchanged. The experimental results indicate that the SAWMF provides a significant performance improvement over many of the existing filtering techniques in suppressing impulsive noise with different contamination ratios.