Band-based Image and Video Inpainting

다국어 초록 (Multilingual Abstract)

We propose a novel image inpainting method composed of two parts: band matching and seamless cloning. In band matching, a band enclosing the boundary of a missing region is compared to those from all other parts of the image. The inner area of the minimum difference band is then copied to the missing region. Even though this band matching results in successful inpainting, brightness discontinuity (a seam) may appear between the filled missing region and its neighborhood. We apply seamless-cloning to remove such discontinuity between the two regions. Examples show that this two step approach can provide a very fast and effective image inpainting. However, since this basic method using one patch may not deal with cases where there are abrupt changes of color or brightness along the boundary, we furthermore devise one more step: target sub-division. The band matching and seamless cloning are applied to small sub-areas. This sub-division is done also when the missing region is too large or the user wants to see more candidates to choose a better one. We propose a video inpainting based on homography: We combine sequence images using homography, select the objects to be removed, apply the proposed inpainting and separate the combined images. The inpainting operation is done on the stitched image once. Our algorithm is demonstrated with various experiments using real images.