/Patent Tracker
SamsungGranted patentAI & softwareInterest 2/10

Samsung patents video codec tweaks for faster decoding

Samsung has secured a patent for a specific method to decode video files by adjusting how prediction values are calculated for rectangular blocks. The system determines an average pixel value based on surrounding data, using different weighting rules depending on whether the block is square or rectangular.

Why it mattersThis is a routine improvement to video compression standards that helps devices process video streams more efficiently without changing the user experience.

In the filing’s own words

Provided are a video encoding or decoding method and apparatus, which, during video encoding and decoding processes, determine whether an intra prediction mode of the current block is a DC mode, when the intra prediction mode of the current block is the DC mode, in a case where lengths of a height and a width of the current block are the same, determine an average of upper reference samples of the current block and left reference samples of the current block, as a DC value, and in a case where the lengths of the height and the width of the current block are different from each other, determine a weighted average value, which is determined by determining an average of the upper reference samples of the current block as an upper DC value, determine an average of the left reference samples of the current block as a left DC value, and combining the upper value and the left DC value with a predetermined weight, as a DC value, and perform intra prediction on the current block, based on the DC value.

What exactly they claim (claim 1)

1. A video decoding method comprising: when a quad split of a first block is not allowed, obtaining split direction information indicating whether the first block is split horizontally or vertically, and split type information indicating whether the first block is split using ternary split or binary split; when the split direction information indicates that the first block is split vertically and the split type information indicates that the first block is split using binary split, obtaining a second block and a third block by vertically splitting the first block; splitting the second block horizontally into a fourth block and a fifth block and splitting the third block horizontally into a sixth block and a seventh block; when an intra prediction mode of the seventh block is a DC mode and lengths of a width of the seventh block is greater than lengths of a height of the seventh block, determining an average of upper reference samples of the seventh block as a DC value; and performing intra prediction on the seventh block, based on the DC value, wherein the fourth block, the fifth block, the sixth block, and the seventh block are decoded according to a decoding order of the fourth block, the fifth block, the sixth block, and the seventh block, and wherein the second block is adjacent to a left side of the third block, the fourth block is adjacent to an upper side of the fifth block, and the sixth block is adjacent to an upper side of the seventh block, wherein the number of the upper reference samples of the seventh block is equal to the width of the seventh block.

Facts

Company
Samsung (SAMSUNG ELECTRONICS CO., LTD.)
Official title
Video decoding method and apparatus, and video encoding method and apparatus
Patent number
US12756841B2
Application number
18902138
Filed
Sep 30, 2024
Granted
Oct 6, 2026
Inventors
9
Claims
3

Classified under

H04N 19/159 Electricity & electronicsH04N 19/105 Electricity & electronicsH04N 19/119 Electricity & electronicsH04N 19/132 Electricity & electronicsH04N 19/176 Electricity & electronicsH04N 19/593 Electricity & electronics

More from Samsung in AI & software

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.