A New Fractional Partial Differential Equation Model for Image Denoising
Jun Liu, Yi‐Fei Pu, Jiliu Zhou · 2016
To overcome the “staircase effect” brought by the traditional integer order ROF image denoising model in smooth regions and the poor visual effect brought by LLT image denoising model although which can suppress the staircase effect, we proposed a new image denoising model with the fractional partial differential equations in this paper, and discussed the influence of fractional order v and p value in the new model on the denoising results which was then well evaluated by comparison from the aspects of subjective visual effects and PSNR. Experiments show that compared with the two-order ROF model, the four-order LLT model as well as the fractional order ROF model, the new model not only can suppress the staircase effect, but also can improve the image of the peak signal to noise ratio, and can better preserve the image edge and texture details under the action of the optimal fractional order v and appropriate p values.