论文部分内容阅读
针对不同的自然图像去噪，现有方法的处理结果往往都含有吉布斯效应，目前很难找到非常理想的方法来进行处理。文中提出了一种基于图像特征的偏微分方程图像去噪方法。文中在研究TV模型和PM模型的基础上提出了基于每幅图像具体特征的去噪模型。该模型能够自适应的根据图像每个区域内的细节特征来调节扩散系数的大小，使其能在消除高梯度噪声的同时较好的保留边缘信息。我们证明了该模型的理论性。实验表明改进后的方法在消除噪声的同时也消除了吉布斯现象。“,”For the special different nature images, we could hardly find particularly desirable approach, and there always exist Gibbs-type artifacts in the results of most methods. A novel Partial Differential Equation (PDE) model is proposed based on image feature for images denoising. The PDE model is adaptive within each region according to the details of the image feature to adjust the size of the diffusion coefficient. So it can be disposed the high gradient noise at the same time better to retain the edge information. We also analyze the performance of the PDE model method. Numerical results show that our algorithm competes favorably with state of the-art TV projection methods to eliminate noise and reduce Gibbs-type artifacts.