新的彩色图像分层修复方法

New color image multi-level-inpainting method

基于全变分的彩色图像修复(CTV)模型,可以在保持边缘的同时达到去噪目的,但对纹理较复杂的区域修复效果不理想。利用分层修复的思想在对整幅彩色图像进行CTV修复后,再对有明显修复痕迹的局部待修复区域使用基于块的纹理合成算法。实验结果表明,这种分层修复的方法在彩色图像的结构修复和纹理修复两方面实现较好的统一,既能很好地降噪,又能保持彩色图像的清晰边界,而且在较大区域图像修复上表现出良好的效果。

The Color Total Variation（CTV） model can denoise while maintaining the edge of the damaged images,but it should be improved in domains of complicated texture.In this paper,a new image multi-level-inpainting method for color images was proposed.The main research topic is to improve the visual result of inpainted images of complicated texture.Damaged images were calculated on the basis of CTV model,and then the patch-based texture synthesis method was used to improve the inpainting results of complicated texture domain.The experimental results show that the inpainting performance is better both in the structure and the texture of color images.Moreover,the new image multi-level-inpainting method can be used in the image inpainting with bigger damaged domains.