偏微分方程去噪模型中方向扩散与应用探讨

Research on directional diffusion and its application denoising model of partial differential equation

数字图像在成像、传输和处理过程中不可避免地会受到噪声影响,如何去除噪声是图像处理领域的关键问题之一.因此,通过对基于偏微分方程的图像去噪模型,建立经典的偏微分方程去噪模型,将其表示成沿图像边缘和跨图像边缘方向扩散方程的形式,总结出设计方向扩散模型的关键影响因素.研究发现偏微分方程的图像去噪模型同传统模型相比有更加灵活可控的结构.在方向扩散模型中,沿边缘的扩散系数应大于跨边缘的扩散系数.

Digital images inevitably exist noises in the process of imaging,transmission and processing.One of the key issues is how to remove noise in the field of image processing.Therefore,a classical de -noising model of partial differential equation is established by means of the image de -noising model based on partial differential equation,which is expressed as the diffusion equation along the edge of the image and across the edge of the image.Moveovev,the key factors of the design of the directional diffusion model are summarized.It is found that the image denoising model of partial differential equation has more flexible and controllable structure than the traditional image denoising model.The diffusion coefficient along the edge should be larger than that across the edge in the directional diffusion model.