New adaptive hybrid filter for image de-noising

The paper proposed an adaptive hybrid filter for digital image de-noising.Based on analyzing the eight basic gradient and its 3×3 neighborhood window pixels' structure and connected properties of a pixel in the image,firstly mentioned nine rules to classify each image pixel as one of the three different type pixels(impulse noise pixel,edge or detail pixel,flat area pixel).Then,this paper adopted different sub-filter to process the three types pixels respectively,and the filter window dimensions could adaptively choice according to the processing pixel's neighborhood structure and connected properties.Through extensive simulation experiments conducted using a wide range of test images,the filter demonstrates good ability to suppress several distinct types of noise commonly considered in digital image de-noising and preserves edge detail.