Exemplar-based inpainting algorithm constrained by continuity
连续性约束下基于样例的图像修复算法*

To improve image inpainting effect, in this paper, an exemplar-based inpainting algorithm constrained by continuity is proposed. The filling order is optimized by establishing an energy function and introducing a novel partial differential equation ( PDE) constraint. The algorithm can overcome the disconnecting phenomenon when filling the strong structure image using traditional algorithm. Structure is well preserved when filling the texture. Experiments have verified that this algorithm can inpainting the large-scale damaged images better.