Samsung patents video compression method for faster image decoding
Samsung has filed a patent for a specific way to compress and decompress video images. The method focuses on how a decoder predicts missing pixels in a block of an image by checking available neighboring pixels in a specific order. It simplifies the process when certain reference pixels are missing by using default values or previous data.
In the filing’s own words
The present disclosure relates to apparatus and methods of image decoding and image encoding. In one or more embodiments, an image decoding method may comprise determining a first sample value of an upper left reference sample of the current block. The image decoding method may further comprise sequentially searching for a searched reference sample of at least one reference line from a plurality of reference lines, except for the upper left reference sample. The image decoding method may further comprise determining a second sample value of remaining reference samples of the current block, except for the upper left reference sample. The image decoding method may further comprise performing intra prediction on the current block to obtain a prediction block of the current block. The image decoding method may further comprise obtaining a residual block of the current block and a reconstruction block of the current block.
What exactly they claim (claim 1)
1. An image decoding method, comprising: obtaining at least one coding unit including a current block by hierarchically splitting a current image based on split type information, determining, when a current prediction mode of the current block is an intra mode, a first sample value of an upper left reference sample of the current block by identifying a first availability of the upper left reference sample of the current block; sequentially searching, for remaining reference samples of at least one reference line from a plurality of reference lines, except for the upper left reference sample, the plurality of reference lines comprising a left reference line of the current block, an upper reference line of the current block, and a right reference line of the current block; determining second sample values of the remaining reference samples of the current block by identifying a second availability of the searched remaining reference samples; performing intra prediction on the current block to obtain a prediction block of the current block, based on the first sample value of the upper left reference sample of the current block and the second sample values of the remaining reference samples of the current block; and obtaining a reconstruction block of the current block based on the prediction block of the current block, wherein the determining of the first sample value of the upper left reference sample of the current block comprises: determining, when the upper left reference sample is not available, a first value obtained based on a sample bit depth as the first sample value of the upper left reference sample without searching for available reference samples neighboring the upper left reference sample, wherein the determining of the second sample values of the remaining reference samples of the current block comprises: determining, when a current reference sample in a current search location is not available, a second value obtained based on a sample value of a reference sample in a previous search location as a current sample value of the current reference sample in the current search location, wherein a first search direction of the upper reference line is a right direction relative to the upper left reference sample, wherein a second search direction of the left reference line is a first lower direction relative to the upper left reference sample, wherein a third search direction of the right reference line is a second lower direction relative to an upper right reference sample of the current block, wherein when a coordinate value of an upper left corner sample of the current block is (0,0), a x-axis coordinate value of the upper right reference sample of the current block is a width of the current block, and a y-axis coordinate value of the upper right reference sample of the current block is −1, wherein the split type information indicates one of a plurality of split types including a quad split, a binary split and a tri split, and wherein when the intra prediction on the current block is performed by using only a reference sample reconstructed in the intra mode and a prediction mode of a reference block including the current reference sample corresponds to an inter mode, it is determined that the current reference sample is not available.
Facts
- Company
- Samsung (SAMSUNG ELECTRONICS CO., LTD.)
- Official title
- VIDEO ENCODING METHOD AND DEVICE, AND VIDEO DECODING METHOD AND DEVICE
- Publication number
- US20260303813A1
- Application number
- 19710909
- Filed
- Jun 17, 2026
- Published
- Oct 1, 2026
- Inventors
- 2
- Claims
- 4
Classified under
H04N 19/132 Electricity & electronicsH04N 19/105 Electricity & electronicsH04N 19/117 Electricity & electronicsH04N 19/159 Electricity & electronicsH04N 19/176 Electricity & electronicsH04N 19/82 Electricity & electronicsMore from Samsung in AI & software
- Samsung patents tool to erase objects from photos using AI Sep 24, 2026
- Samsung patents drawing objects into photos with AI Oct 15, 2026
- Samsung patents AI that creates personalized exercise videos Sep 10, 2026
- Samsung patents AI summaries for paused TV shows based on your viewing habits Sep 24, 2026
- Samsung patents turning chat conversations into image prompts Oct 1, 2026
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.
