Method and device for reconstructing a self-similar textured region of an image
View Patent ↗The invention proposes a method for reconstructing a self-similar textured region of an image. Said method comprises determining pixels of a part of the self-similar textured region by copying sample pixels from a sample part of the self-similar textured region, the sample pixels being selected using a neighborhood matching, wherein a size of neighborhoods used for matching is selected based on an analysis of descriptors computed from coefficients of OCT transform of differently sized blocks of the sample part. The analysis of descriptors computed from coefficients of DCT transform of differently sized blocks of the sample part allows for determining the neighborhood size close to a feature size of the texture.
1. A method for reconstructing a self-similar textured region of an image, said method comprising:
determining pixels of a part of the self-similar textured region by copying sample pixels from a sample part of the self-similar textured region, the sample pixels being selected using a neighbourhood matching,
wherein a size of neighborhoods used for matching is adaptively selected by:
determining the coefficients by applying a DCT transformation on each of the differently sized blocks,
determining, for each of the differently sized blocks, an associated descriptor vector wherein each descriptor is determined using sums of coefficients summed along parallel paths,
selecting the descriptor vector associated with the smallest block size among those descriptor vectors which are not monotone; and
determining the size of the neighborhoods using a size of the differently sized block associated with the selected descriptor vector.
2. The method of claim 1 , said method further comprising:
receiving a bit stream in which the sample part is encoded, and
decoding the sample part.
3. The method of claim 2 , wherein said method further comprises receiving side information including a selected neighbourhood size.
4. The method of claim 2 , wherein said method further comprises determining at least one descriptor vector from the received bit stream and use said at least one descriptor vector to determine a neighbourhood size.