一种保持边缘线的图像修补技术

Image inpainting technology based on edge-directed holding

图像修补是对图像损毁区域进行适当修补以满足人眼视觉要求的一种技术.对于边缘清晰的图像,大部分基于偏微分方程修补模型的修复结果可能会造成边缘的模糊扩散或产生锯齿.为此,提出一种基于拟合和插值的边缘保持方法,通过对边缘线预测保持后再用TV修复模型进行修补.该算法实现速度快,实验证明对边缘的修复效果较好.

Image inpainting is a technique which fills proper information into the breakages of an image so that it can satisfy the demands of human vision. For the images with clear edges, most PDE inpainting models would make edges blurry or sawtooth. In this paper a new edge - directed holding method is proposed based on approximation and interpolation. The method first estimates the edge information in the missing area and then inpainting the other area with TV model. The experiments show that the new method can improve the inpainting effect with rapid convergence.