为了提高受损数字图像修复质量,将整体变分模型用于图像修复。采用Matlab编程工具对修复过程进行了仿真实现,克服了直接求解偏微分方程的困难。发现随着迭代次数的增加,被修复区域接近原图质量,当受损区域较小且受损区域灰度梯度不大时,看不出明显的修复痕迹。仿真实验结果表明,该算法复原图像的视觉质量好,算法收敛快,特别适合于线状裂痕的图像修补。
The total variation model is applied to image inpainting to improve restoration quality of the damaged images.In order to overcome the difficulty of directly resolving the partial differential equation,Matlab software is used to simulate the process of image inpainting.It is found that the repaired area in an image gradually approximates to the original one with the increase of iteration.When the damaged area is small and its gray gradient is not great,the repaired trace of the image is not obvious.Experimental results show that the inpainting algorithm performs well in visual quality and has a rapid convergence.This algorithm is especially suitable for the linear information lost area.