Reduction of impulse noise in MRI images using block-based adaptive median filter

A block-based adaptive median filter, which aims to purify impulse noise from an MRI image, is discussed in the paper. The proposed filter blocks the image into several sub-images, and calculates the standard deviation of each sub-image. During filtering process, each point of the sub-image needs to be judged whether it is noise point or not, according to the average value of the filtering window and the standard deviation of the sub-image. Thereafter, the filtering process is the same with traditional adaptive median filter. Experimental results indicate that the proposed filter can maintain the original image details nicely while reducing impulse noise effectively.