Study On The Removal Of High Density Salt And Pepper Noise Based On Decision-based Algorithm

Image is an important way for human to get information,a simple image can contain tremendous information.Tedious text information can be described with several simple pictures.Therefore,clear image quality is an important factor to ensure information acquisition.With the development of science and technology and the quickening of the pace of life,analog image processing fade out of people's field,digital image processing has gradually taken the lead.It can be polluted by all kinds of noise during the process of generation and transmission and so on.Salt and pepper noise cause the greatest damage to the image among them,therefore,the removal of this kind of noise has also become the focus of image denoising.The existing denoising methods are mainly divided into linear method and nonlinear method,the related research results show that the nonlinear method has better performance.This paper mainly studies on removing salt and pepper noise,two denoising methods are proposed based on the nonlinear median filter.One of them is salt and pepper noise segmentation and iteration removal method based on decision-based algorithm,and the other is the salt and pepper noise removal algorithm based on two-stage filtering and permeation.In the salt and pepper noise segmentation and iteration removal method based on decision-based algorithm,the paper adopt decision-based method for noise detection,then the noise points are divided into common noise points and full noise points according to their distribution.The full noise points is the center pixels in the filter window,which has all noise points in it.Unsymmetric trimmed median filter was used to process common noise points.For the full noise points,in order to make full use of the adjacent pixels information,we marked them without processing and handle them during the next scanning,then different processing ways are applied according to the specific location of the full noise pixels.Finally,the concept of noise dead angle was put forward,and the problem was solved by the iteration method.Simulation results illustrate that the proposed algorithm has good denoising ability in terms of both the qualitative and quantitative measures and outperforms the existing methods especially at high noise ratio.After a deeper study on the full noise distribution,the paper presents salt and pepper noise removal algorithm based on two-stage filtering and permeation.The algorithm analyzes the distribution of full noise points more deeply,finds out the noise-free pixels around the full noise pixels and uses their median value as result to prevents the introduction of noise and avoid to classify the full noise pixels while processing by using the two-stage filtering method,which improve the efficiency of the algorithm.Finally the method analyzes the available pixels in the filter window,and we put forward permeation method to effectively increase the available pixels.By the method,all the other full noise pixels in the filter window besides the center one are all processed,so the available pixels are in-creased.Through simulation,the algorithm performs better on both effectiveness and efficiency especially when noise density is high.