基于高斯—马尔可夫随机场模型的图像修补方法研究

Study of Digital Image Inpainting Methods Based on GMRF Model

实现基于高斯马尔可夫随机场(GMRF)模型的图像修补算法,详细讨论像素扫描顺序以及邻域大小对图像修补效果的影响。实验表明,基于GMRF的图像修补算法的性能依赖于像素扫描顺序和邻域大小的选取。

A digital image inpainting algorithm based on Gauss-Markov Random Field （GMRF） is completed. The pixel scanning order and neighborhood size is discussed in detail for the algorithm. The experiment results show that the perform- ante of inpaiuting algorithm based on GMRF model depends on the pixel scanning order and neighborhood size.