基于四方向导数信息的图像非线性扩散去噪

Nonlinear Diffusion Model Based on Four Directional Derivatives for Image Denoising

分析图像去噪的非线性扩散模型的离散解,针对其中的梯度获取问题,考虑一种新的基于四方向导数信息的图像非线性扩散去噪方法。该方法将原有非线性扩散去噪中的二阶扩散矩阵扩展为四阶扩散矩阵,并由边缘的实际取向来决定像素点邻域内八个点对该点去噪的贡献大小,有利于更好的保护边缘和对边缘定向。数值计算结果表明,新方法较原有方法具有更高的峰值信噪比。

Have analyzed the discrete solution of the nonlinear diffusion model for image denoising, and put forward a new model based on four directional derivatives aimed to solve the problem of gradient-obtaining in the old model, the new model adapts the two-order diffusion tensor in the old model to a four-order diffusion tensor, and decides the power of the eight pixels in the neighborhood based on the direction of the edge, which is propitious to preserve the edges and to direct them. Numerical results show that the new model has higher peak signal to noise ratio than the known model.