Telefonaktiebolaget Lm Ericsson (publ)

Velos Media, LLC

Andrey Norkin

Kenneth Andersson

Rickard Sjöberg

Thomas Rusert

Clinton Priddle

Zhuangfei Wu

Martin Pettersson

Per Fröjdh

Apostolos Georgakis

A first filter decision value is calculated for a block of pixels in a video frame based on pixel values of pixels in a first line of pixels in the block. A second filter decision value is also calculated for the block based on pixel values of pixels in a corresponding first line of pixels in a neighboring block in the video frame. The first filter decision value is used to determine how many pixels in a line of pixels in the block to filter relative to a block boundary between the block and the neighboring block. The second filter decision value is used to determine how many pixels in a corresponding line of pixels in the neighboring block to filter relative to the block boundary.

Deblocking filtering control

A first filter decision value is calculated for a block of pixels in a video frame based on pixel values of pixels in a first line () of pixels in the block. A second filter decision value is also calculated for the block based on pixel values of pixels in a corresponding first line of pixels in a neighboring block in the video frame. The first filter decision value is used to determine how many pixels in a line of pixels in the block to filter relative to a block boundary between the block and the neighboring block. The second filter decision value is used to determine how many pixels in a corresponding line of pixels in the neighboring block to filter relative to the block boundary.

A method of reducing blocking artifacts associated with pixels of a block boundary of an image. Pixel values of pixels from a first block and 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. If abs[first offset]&#x3c;first threshold, the pixel values of consecutive pixels from the first block and the pixel values of consecutive pixels from the second block are modified by applying normal filtering on the respective pixels, while if instead abs[first offset]&#x3e;=first threshold, the pixel values of consecutive pixels from the first block and the pixel values of consecutive pixels from the second block are modified by applying weak filtering or no filtering at all on the respective pixels.

Deblocking filtering

Pixel values of pixels () in a line () of pixels () are filtered with a strong deblocking filter to obtain filtered pixel values. Each filtered pixel value is clipped off to a respective clipping parameter value defined based on a position of the pixel () relative to a block boundary () between two adjacent blocks () of pixels (). The clipping parameter values change at least linearly depending in the pixel position relative to the block boundary () so that pixels () in the line () of pixels () having different positions from the block boundary () will have different clipping parameter values.

Controlling deblocking filtering

A deblocking filtering control comprises checking whether pixel values of four pixels () in a line () of pixels () in a block () of pixels () and in a neighboring block () of pixels () form an approximate line. If the pixel values of the four pixels () in the line () of pixels () in both the block () of pixels () and in the neighboring block () of pixels () form an approximate line the deblocking filtering control selects to apply strong deblocking filtering to pixel values in the line () of pixels (). The subject and objective quality of the deblocking filtering is thereby improved by applying strong deblocking filter to areas where the signal in a picture () not only has the form of a flat line but also form a ramp.

Strong deblocking filtering decisions

Visible artifacts in a video stream of pictures with slices are reduced by having a separate maximum transform size for intra coding units in inter coded slices as compared to intra coding units in intra coded slices and/or inter coding units or by penalizing the usage of large transform size for such intra coding units in inter coded slices as compared to intra coding units in intra coded slices and/or inter coding units.

Encoding and decoding of slices in pictures of a video stream using different maximum transform sizes

A value of a deblocking parameter is determined for a picture () based on a depth of the picture () in a hierarchical coding structure of multiple pictures () in a video sequence (). The determined value is encoded to form an encoded value that is sent to a decoder () for use therein during decoding. The embodiments thereby reduces blocking artifacts, which otherwise can occur in video sequences with hierarchical coding structures, such as for QP toggling and multilayer/view video, by determining deblocking parameter values based on picture depth in the hierarchical coding structure.

Hierarchical deblocking parameter adaptation

Arrangements and methods therein for supporting view synthesis. One method is performed by an arrangement which is operable to receive an input video stream comprising an input view associated with a first camera position. The method includes receiving a measure M_D of a maximum difference in depth or disparity value between two pixels in a depth or disparity map representing the depth or disparity information of a video frame of the input view. The method further includes determining a maximal distance d_cam, from the first camera position to a second camera position, based on the received measure; and indicating the determined distance d_cam to a view synthesizer. This can enable the view synthesizer to synthesize a view associated with a second camera position so that disocclusions due to the distance between the first and second camera position may be handled.

Methods and arrangements for supporting view synthesis

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.

A method () of constructing at least one list of reference pictures for inter-layer prediction of a current picture is provided. The method comprises inserting () reference pictures into a first set of reference pictures or a second set of reference pictures, based on respective values of a scalability identifier associated with the reference pictures and a value of the scalability identifier associated with the current picture, and inserting () the first set of reference pictures and the second set of reference pictures into the at least one list of reference pictures. By taking indications for similarities between reference layers and the current layer into account, a more efficient multi-layer video compression is achieved. Further, a corresponding computer program, a corresponding computer program product, and a corresponding device are provided.

Growing community of inventors

Solna, Sweden

Andrey Norkin