A Median Filtering Method for Pulse Noise Detection
Yamei Zhang · 2017
This paper presents a method for detecting image impulse noise and using median filter to filter out impulse noise according to the test results.The method will contain the subspace of the subimage of the impulsive noise, and the kernel function is mapped into a hypersphere in the high dimensional space, and the radius R of the sphere and the corresponding spherical vector a are calculated.The median filter method is used to filter the detected noise points.The proposed algorithm is more accurate in judging noise, and the way to filter noise is more reasonable.The applicable image range is more extensive and has better filtering performance.