关键词： 图像去噪, TV^p模型, 阶梯效应

Abstract: The adaptive TV^p model can effectively avoid the staircase effect of TV model, but the nonlinear degree of the model is relatively high, so it is difficult to solve it. In order to solve this problem, a new discrete TV^p model is established by using the geometric mean of the forward and backward difference of the gradient of each pixel to define the gradient mode of the pixel. The model can effectively reduce the nonlinear degree of TV^pmodel, and the image recovered has better quality and higher signal-to-noise ratio than the TV^p model.

Key words: image denoising, TV^p model, staircase effect