Restoration of images contaminated by mixed Gaussian and impulse noise using a recursive minimum--maximum method

摘要

A technique is proposed for removing impulse noise in images, called the recursive minimum--maximum method. Statistical analysis of this method indicates that it is good at preserving fine details and suppressing impulse noise at the same time. Experimental results show that the technique is robust and produces better restored images under various impulse noise conditions than other median filter-based methods.