|
Mean curvature motion (MCM) equation is a nonlinear PDE model with special geometry in image processing, which can be implemented by many methods. For the currently used methods, explicit method and AOS method, their stability are bad and the calculation of the methods is severely restricted by time step, and they have low accuracy. In this paper, the compact alternating direction implicit (CADI) method, a high accuracy and unconditionally stable difference method based on alternating direction implicit (ADI) method, is constructed for MCM equation. The numerical experiments show that both the CADI method and ADI method of MCM equation is efficient for image denoising. In addition, it can be visually seen that the image denoised by our method is a little better than by ADI method, and it could be better to approximate the original image. |
|
Keywords: image denoising; mean curvature motion (MCM) equation; ADI method; compact alternating direction implicit (CADI) method; numerical experiment |
|