Anisotropic diffusion filtering is a widely used partial differential equation (PDE) based technique, which works effectively for removal of noise and preserving edges. This technique is highly dependent on diffusion coefficient and threshold parameter. In this work, we propose a new diffusion coefficient and image dependent threshold parameter which has a faster rate of convergence than those found in literature. The new technique outperforms the traditional models of anisotropic diffusion. Experimental results show the superiority of the proposed work over traditional anisotropic diffusion models.

中文翻译：

---

具有边缘保留的图像去噪的高效各向异性扩散模型

各向异性扩散滤波是一种广泛使用的基于偏微分方程（PDE）的技术，该技术有效地去除了噪声并保留了边缘。该技术高度依赖于扩散系数和阈值参数。在这项工作中，我们提出了一种新的扩散系数和与图像有关的阈值参数，其收敛速度比文献中的收敛速度更快。这项新技术优于传统的各向异性扩散模型。实验结果表明，所提出的工作优于传统的各向异性扩散模型。