/Patent Tracker
SamsungGranted patentChips & hardwareInterest 4/10

Samsung patents video compression using flexible block shapes

Samsung has been granted a patent for a method to compress video files by breaking images into flexible square and non-square blocks. The system uses specific mathematical tools, called transform kernels, to reduce data size based on the shape and prediction method of each block.

Why it mattersThis improves how video is stored and streamed, potentially leading to smaller file sizes or better quality at the same bandwidth.

In the filing’s own words

Provided is a method of decoding motion information characterized in that information for determining motion-related information includes spatial information and time information, wherein the spatial information indicates a direction of spatial prediction candidates used for sub-units from among spatial prediction candidates located on a left side and an upper side of a current prediction unit, and the time information indicates a reference prediction unit of a previous picture used for prediction of the current prediction unit. Further, an encoding apparatus or a decoding apparatus capable of performing the above described encoding or decoding method may be provided.

What exactly they claim (claim 1)

1. A video decoding method comprising: splitting a current picture into a plurality of maximum coding units; splitting a maximum coding unit, from among the plurality of maximum coding units, into a plurality of coding units; obtaining transform shape information about whether a current coding unit, from among the plurality of coding units, is split in a vertical direction or a horizontal direction or is not split, from a bitstream; determining a horizontal length and a vertical length of a transform block with respect to the current coding unit based on the transform shape information; obtaining an encoded data from the bitstream; obtaining transform kernel information indicating whether a non-separable transform is used or which transform kernel is used from among transform kernel candidates; when the transform kernel information indicates that the non-separable transform is used, determining a transform kernel among the transform kernel candidates based on the horizontal length and the vertical length of the transform block, an intra prediction mode used in intra prediction, and the transform kernel information; performing inverse transformation on the encoded data by using the determined transform kernel in order to obtain a residual sample; and reconstructing the current coding unit based on the residual sample and a prediction sample obtained by performing intra prediction, wherein one from among the plurality of coding units has one of a square shape and a non-square shape.

Facts

Company
Samsung (SAMSUNG ELECTRONICS CO., LTD.)
Official title
Image encoding method and apparatus, and image decoding method and apparatus
Patent number
US12732610B2
Application number
18921608
Filed
Oct 21, 2024
Granted
Sep 8, 2026
Inventors
3
Claims
3

Classified under

H04N 19/122 Electricity & electronicsH04N 19/136 Electricity & electronicsH04N 19/157 Electricity & electronicsH04N 19/159 Electricity & electronicsH04N 19/176 Electricity & electronics
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Samsung in Chips & hardware

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.