Abstract: A method of sample adaptive offset (SAO) compensation of video data is disclosed, where pixels in the video data are classified into SAO categories, each SAO category representing a possible edge artefact and defining a corresponding offset value to be applied to pixels in the respective SAO category to compensate for the edge artefact. In the method, a plurality of SAO categories (200) is provided (110).

Abstract: An image processing method involves determining a global motion between a reference frame and a current frame in a frame sequence. A reference block is identified in the reference frame for a current block based on the global motion. A classification parameter is calculated based on the pixel values of the current block and the reference block. The parameter can be used for classifying the block as belonging to the background or foreground of the current frame. The parameter is preferably also utilized in frame rate-up conversion when extrapolating or interpolating new frames.

Abstract: A method of encoding a plurality of adaptive filter coefficients (104, 107, 112) into a bitstream (110). The method comprises the steps of entropy encoding (109) the adaptive filter coefficients (104, 107, 112) into the bitstream (110), whereby a prediction (202) for an instance (204) of the adaptive filter coefficients is determined (201) based on at least one of the remaining adaptive filter coefficients (206), a prediction error (205) is determined (203) based on the difference between the instance (204) of the adaptive filter coefficients and the prediction (202), and wherein the step of entropy encoding (109) the plurality of adaptive filter coefficients (104, 107, 112) comprises entropy encoding (109) the prediction error (205) for the instance (204) of the adaptive filter coefficients and entropy encoding (109) the remaining adaptive filter coefficients (206).

Abstract: In video encoding and decoding predictions may be generated by intra-frame prediction. Intra-frame prediction uses reconstructed pixels in a reconstructed frame. Intra-frame prediction is performed by extending the reconstructed pixels into a predicted block using intra-frame prediction modes, each intra-frame prediction mode indicating a direction of the extension. In order to reduce the number of possible intra-frame prediction modes, a subset is selected from a predetermined set of possible intra-frame prediction modes. A subset of intra-frame prediction modes can be created by forming preselection sets of intra-frame prediction modes on the basis of similarity of their associated predicted blocks and selecting a representative intra-frame prediction mode from each preselection set.

Abstract: Blocking artifacts at a block boundary between a block and a neighboring block in a video frame are reduced by calculating an offset based on pixel values of pixels in a line of pixels in the block and based on pixel values of pixels in a corresponding line of pixels in the neighboring block. The offset is added to the pixel value of the pixel closest to the block boundary in the line of pixels and is subtracted from the pixel value of the pixel closest to the block boundary in the corresponding line of pixels. The resulting deblocking filter has good low-pass characteristics and is efficient for reducing blocking artifact.

Abstract: A first filter decision value is calculated for a block (10) of pixels (11, 13, 15, 17) in a video frame based on pixel values of pixels (11, 13, 15) in a first line (12) of pixels (11, 13, 15, 17) in the block (10). A second filter decision value is also calculated for the block (10) based on pixel values of pixels (21, 23, 25, 27) in a corresponding first line (22) of pixels (21, 23, 25, 27) in a neighboring block (20) in the video frame. The first filter decision value is used to determine how many pixels in a line (12) of pixels (11, 13, 15, 17) in the block (10) to filter relative to a block boundary (1) between the block (10) and the neighboring block (20). The second filter decision value is used to determine how many pixels in a corresponding line (22) of pixels (21, 23, 25, 27) in the neighboring block to filter relative to the block boundary (1).

Abstract: Current deblocking filters are using the same filters with the same filtering strength irrespective of the block size and the size of the transform used. However, in the new video coding standards such as emerging HEVC the PU sizes can vary from 4 to 64 and the TU sizes can vary from 4 to 32. Therefore, filtering the same amount of pixels (e.g. two or three) from the block boundary for the block of size 4 can be excessive, while for the block size 32 it may not be enough, with the result that the boundary between two blocks is still visible. Hence, there is a need for an efficient deblocking filter control that can be used to reduce blocking artifacts at block boundaries and that does not have the above mentioned drawbacks. It is a general objective to provide an efficient deblocking filter control. Thus, the objective is solved by applying different filters for different block sizes such as CU, PU or/and TU sizes.

Abstract: A method and device for determining an adaptive filter having multiple filter parameters, wherein a first filter parameter has a first level of adaptivity and a second filter parameter has a second, different level of adaptivity. Parameter values for the first filter parameter are determined among a first set of allowable filter parameter values. Parameter values for the second filter parameter are correspondingly determined among a second, different set of allowable filter parameter values. The different levels of adaptivity in the filter parameters are achieved because the second set includes more allowable filter parameter values than the first set. The adaptive filter is advantageously used in filtering in intra- or inter-predication during video encoding and decoding.

Abstract: A method of reducing blocking artifacts associated with consecutive pixels of a block boundary of an image, such as e.g. a video frame is provided. Pixels values of pixels selected from a first block and at least a neighboring block, being located on opposite sides of a block boundary are evaluated. A first offset for the two pixels of each block located next to the block boundary is calculated, after which the first offset is compared to a first threshold value.

Abstract: Blocking artifacts at a block boundary (1) between a block (10) and a neighboring block (20) in a video frame are reduced by calculating an offset based on pixel values of pixels (11, 13) in a line (12) of pixels (11, 13, 15, 17) in the block (10) and based on pixel values of pixels (21, 23) in a corresponding line (22) of pixels (21, 23, 25, 27) in the neighboring block (20). The offset is added to the pixel value of the pixel (11) closest to the block boundary (1) in the line (12) of pixels (11, 13, 15, 17) and is subtracted from the pixel value of the pixel (21) closest to the block boundary (1) in the corresponding line (22) of pixels (21, 23, 25, 27). The resulting deblocking filter has good low-pass characteristics and is efficient for reducing blocking artifact.

Abstract: A solution according to one or more embodiments of the present invention is to reconstruct image blocks where only the frequency coefficient that is most sensitive to changes of an average value of the residual image block, typically the lowest frequency coefficient, is present with a flat residual image. The other image blocks are reconstructed by applying an inverse spatial 2D transform. Hence, the other image blocks comprise image blocks with multiple coefficients and image blocks with a single coefficient and wherein the single coefficient is not the coefficient determining that is most sensitive to changes of an average value of the residual image block.

Abstract: A border region is identified in an image by calculating an average of pixel values in a row or column of the image. Differences in property values are determined between each pixel in the row or column and a neighboring pixel present on a same column or row but in a neighboring row or column. An average difference is calculated based on these differences. The pixels in the row or column are classified as belonging to a border region or internal region of the image based on the average pixel value and the average pixel difference.

Abstract: The invention relates to encoding and decoding pixel blocks of a video frame through a hybrid mode involving usage of a first prediction of a pixel block and at least a second prediction of the pixel block. An initial first weighting factor is modified using a factor modifier to generate a first weighting factor comprising multiple different factor values that are assignable to the different pixel-based first prediction values of the first prediction. The first weighting factor is applied to the first prediction and a second weighting factor is applied to the second block prediction. The at least two weighted predictions are then combined to form a hybrid prediction of the current pixel block.

Abstract: An adaptive filter to use in connection with prediction-based pixel block encoding and decoding is determined independently at the encoder and decoder side through a template-based procedure. A pixel block (12) has an identified reference pixel block (22) in a reference frame (20). A template (16) of multiple pixels (18) adjacent the pixel block (12) and a reference template (26) of multiple pixels (28) adjacent the reference pixel block (22) are used for determining the filter parameters of the adaptive filter. The determined adaptive filter is then applied to the reference pixel block (22) and is used for generating an encoded representation (40) of the pixel block (12) during encoding and or generating a decoded representation of the pixel block (12) during decoding.

Abstract: A block-specific filter decision value is calculated for a pixel block (10) in a video frame. If the block-specific filter decision value is below a block-specific threshold, each row or column (12) in the block (10) is individually processed in order to select between a strong and a weak de-blocking filter. A respective line-specific filter decision value is thereby calculated for each row or column (12) in the block (10) and compared to a line-specific threshold. If the line-specific filter decision value calculated for a row or column (12) is below the line-specific threshold a strong de-blocking filter is selected for the row or column (12), otherwise a weak de-blocking filter is instead selected to combat any blocking artifacts.

Abstract: Blocking artifacts at a block boundary (1) between a block (10) and a neighboring block (20) in a video frame are reduced by calculating an offset based on pixel values of pixels (11, 13) in a line (12) of pixels (11, 13, 15, 17) in the block (10) and based on pixel values of pixels (21, 23) in a corresponding line (22) of pixels (21, 23, 25, 27) in the neighboring block (20). The offset is added to the pixel value of the pixel (11) closest to the block boundary (1) in the line (12) of pixels (11, 13, 15, 17) and is subtracted from the pixel value of the pixel (21) closest to the block boundary (1) in the corresponding line (22) of pixels (21, 23, 25, 27). The resulting deblocking filter has good low-pass characteristics and is efficient for reducing blocking artifact.

Abstract: Methods and arrangements in video encoding and decoding entities. The methods and arrangements relate to the joint encoding of reference information associated with encoded video. The method and arrangement in a decoding entity relate to obtaining (402) a single syntax element associated with an encoded block Be, and identifying (404) a reference mode and one or more reference pictures based on the obtained syntax element. The method and arrangement further relate to-decoding (406) of the block Be based on the identified reference mode and one or more reference pictures, thus providing a decoded block, B, of pixels.

Abstract: Methods and arrangements in video encoding and decoding entities. The methods and arrangements involve determining (804) the frequency of occurrence of a plurality of reference pictures associated with an obtained (802) set of blocks, which are neighbors of a block B. The methods and arrangements further involve selecting (806) a reference picture or combination of reference pictures having the highest determined frequency of occurrence to be a prediction, Cpred, of the reference picture or combination of reference pictures, C, to be used when encoding/decoding the block, B. The methods and arrangements further involve providing/obtaining (610,612, 808) an indication specifying whether the prediction, Cpred, corresponds to C, and when the prediction, Cpred, is indicated to correspond to C, the encoded block, Be, is decoded (812) based on the prediction Cpred.

Abstract: A solution according to one or more embodiments of the present invention is to reconstruct image blocks where only the frequency coefficient that is most sensitive to changes of an average value of the residual image block, typically the lowest frequency coefficient, is present with a flat residual image. The other image blocks are reconstructed by applying an inverse spatial 2D transform. Hence, the other image blocks comprise image blocks with multiple coefficients and image blocks with a single coefficient and wherein the single coefficient is not the coefficient determining that is most sensitive to changes of an average value of the residual image block.

Abstract: A method and arrangement for prediction of pixel values in an image decoder. In an image decoder, a reference vector which is provided by an image encoder is provided 500. An initiation region of pixels is determined 502, which corresponds to a reference region of pixels at the image encoder. The initiation region is spatially displaced in relation to the prediction region according to the reference vector, and a part of the initiation region overlaps a part of the prediction region. Pixel values are assigned 504 to pixels of the prediction region, whose corresponding pixel values in the initiation region are known. Pixel values of the overlapping region of the initiation region are assigned 506 to the corresponding pixels in the prediction region, the pixel values being assigned 504.