Research On Fractional Image Filtering Algorithm

With the rapid development of information technology,the way of communication between people is no longer only satisfied with voice and text.Images can vividly express the information to be transmitted,and it has gradually become one of the important means of people's daily communication.However,image signals are inevitably subject to noise pollution during the acquisition and transmission process,and contaminated images will cause errors and impacts on subsequent studies,such as image recognition,feature extraction,image coding,and target detection.In order to obtain higher quality images,many scholars have proposed many theories and algorithms for the problem of image denoising.Among them,the ROF model can effectively denoise images,which is a classic algorithm in the field of image denoising.In this paper,using fractional calculus theory and total variation model,taking image signals as the research object,we have made some explorations in the field of image denoising.The main work is as follows:Firstly,the research progress and development status of the full variational model,Curvelet transform are briefly introduced.The theory of Curvelet transform and fractional calculus are introduced.The fast algorithm of Curvelet transform and several algorithms for variation denoising are summarized.Secondly,a method of denoising medical images by combining fractional variation model,curvelet transform and twin support vector machine is proposed.This method first performs a Curvelet transform on the noisy image,and the transformed coefficients are divided into two categories using a twin support vector machine,one is not processed,and the other is processed by a fractional variation model machine.The processed coefficients are subjected to Curvelet inverse transform together with the unprocessed coefficients to obtain a denoised image.The experimental data and simulation results show that this method is effective in denoising.Finally,the first-order variation model is extended to the fractional order,and an adaptive function is used to replace the regular term parameters,and an adaptive fractional-order variation model is proposed.The adaptive parameters in the new model are composed of image gradients and local entropy.Because the values of image gradient and local entropy are smaller in the flat area of the image,and larger in the edge detail area of the image,during processing,the processing effect is different for different features of the image,and adaptive processing is realized.