Fourth-order partial differential equations for effective image denoising

This article concerns mathematical image denoising methods incorporating fourth-order partial differential equations (PDEs). We introduce and analyze piecewise planarity conditions (PPCs) with which unconstrained fourth-order variational models in continuum converge to a piecewise planar image. It has been observed that fourth-order variational models holding PPCs can restore better images than models without PPCs and second-order models. Numerical schemes are presented in detail and various examples in image denoising are provided to verify the claim.