基于TV-Stokes方程的快速图像修复算法

Fast Image Inpainting Based on TV-Stokes Equation

提出一种快速的分裂Bregman TV-Stokes修复算法,对图像的缺损部分进行快速自动修复。该算法运用一个能量极小化模型和零散度条件得到一个非线性的TV-Stokes方程,获得图像的等照度线方向;再利用分裂Bregman方法使修复按此等照度线方向进行。实验结果表明:与基于TV-Stokes方程的图像修复算法相比,本文算法在修复速度、峰值信噪比和相对误差方面都更具优势。

In the paper,a fast split Bregman TV-Stokes inpainting algorithm is proposed.It can restore damaged regions in an image automatically and quickly.The algorithm uses an energy minimization model combined with the zero divergence condition to get a nonlinear TV-Stokes equation,from which the isophote directions are found.The image is then restored to fit those directions by using the split Bregman method.Experimental results show that the proposed algorithm is better than the algorithm based on TV-Stokes equation in terms of computation time,peak signal to noise ratio（PSNR） and relative error.