IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0081538
(2011-04-07)
|
등록번호 |
US-8781004
(2014-07-15)
|
발명자
/ 주소 |
- Bankoski, James
- Wilkins, Paul
- Xu, Yaowu
|
출원인 / 주소 |
|
대리인 / 주소 |
Young Basile Hanlon & MacFarlane P.C.
|
인용정보 |
피인용 횟수 :
6 인용 특허 :
142 |
초록
▼
A system and method for encoding and decoding a video image having a plurality of frames is disclosed. The method includes a loop filter, which is used to reduce blocking artifacts in reconstructed frames. A quality value is determined for each block in the reconstructed frame and each associated pr
A system and method for encoding and decoding a video image having a plurality of frames is disclosed. The method includes a loop filter, which is used to reduce blocking artifacts in reconstructed frames. A quality value is determined for each block in the reconstructed frame and each associated predictor block. A variable incremental loop filter strength is calculated for each inter-predicted block using at least one of the quality values. The variable incremental loop filter strength is used during encoding by the loop filter to filter the reconstructed frame.
대표청구항
▼
1. A method for encoding a video signal, the method comprising: identifying, by a processor, a current quality value of a current block of a residual frame;identifying a predictor quality value associated with a predictor block that is associated with the current block;determining a baseline filter
1. A method for encoding a video signal, the method comprising: identifying, by a processor, a current quality value of a current block of a residual frame;identifying a predictor quality value associated with a predictor block that is associated with the current block;determining a baseline filter strength associated with the current block;generating a reconstructed block based on the current block and the predictor block;determining a loop filter strength for the reconstructed block, wherein the current block is an inter-predicted block, and wherein determining the loop filter strength includes modifying the baseline loop filter strength using a variable incremental loop filter strength based at least on the current quality value and the predictor quality value;filtering the reconstructed block using the loop filter strength; andusing the filtered reconstructed block as a predictor block. 2. The method of claim 1, wherein determining the loop filter strength for the reconstructed block based at least on the current quality value and the predictor quality value includes: determining a quality value difference between the current quality value and the predictor quality value. 3. The method of claim 2, wherein determining the loop filter strength for the reconstructed block based at least on the current quality value and the predictor quality value includes: determining a normalized loop filter strength based on the quality value difference. 4. The method of claim 1, wherein the current quality value is at least one of a quantization level, a base loop filter strength, a bit rate, or a difficulty metric of the current block. 5. The method of claim 1, wherein the predictor quality value is at least one of a quantization level, a loop filter strength, a bit rate, or a difficulty metric associated with the predictor block. 6. The method of claim 1, further comprising: identifying a second current quality value of a second current block of the residual frame;identifying a second predictor quality value associated with a second predictor block that is associated with the second current block;generating a second reconstructed block based on the second current block and the second predictor block;determining a second loop filter strength for the second reconstructed block based at least on the second current quality value and the second predictor quality value; andfiltering the second reconstructed block using the second loop filter strength. 7. The method of claim 1, wherein a current quality value of another block from the residual frame differs from the current quality value of the current block. 8. The method of claim 1, wherein the predictor block is one of a plurality blocks from a reference frame other than the residual frame. 9. A method for decoding compressed video information, the method comprising: generating a reconstructed block based on a block of a residual frame and a predictor block associated with the block of the residual frame, the reconstructed block associated with a base loop filter strength, the block of the residual frame associated with a first quality value, and the predictor block associated with a second quality value;identifying a variable incremental loop filter strength for the reconstructed block based on the first quality value and the second quality value;identifying a loop filter strength for the reconstructed block modifying the base loop filter strength using the variable incremental loop filter strength; andfiltering the reconstructed block using the loop filter strength; andusing the filtered reconstructed block as a predictor block. 10. The method of claim 9, wherein the first quality value is at least one of a quantization level, a base loop filter strength, a bit rate, or a difficulty metric of the block of the residual frame. 11. The method of claim 9, wherein the second quality value is at least one of a quantization level, a loop filter strength, a bit rate, or a difficulty metric associated with the predictor block. 12. The method of claim 9, wherein identifying the loop filter strength includes: determining a quality value difference between the first quality value and second quality value. 13. The method of claim 12, wherein determining the loop filter strength includes: determining a normalized loop filter strength based on the quality value difference. 14. An apparatus for filtering blocking artifacts between adjacent blocks reconstructed from a frame of compressed video information, the apparatus comprising: a memory; anda processor configured to execute instructions stored in the memory to: identify a current quality value of a current block of a residual frame,identify a predictor quality value associated with a predictor block that is associated with the current block,determine a baseline filter strength associated with the current block, generating a reconstructed block based on the current block and the predictor block,determine a loop filter strength for the reconstructed block, wherein the current block is an inter-predicted block, and wherein determining the loop filter strength includes modifying the baseline loop filter strength using a variable incremental loop filter strength based at least on the current quality value and the predictor quality value, andfilter the reconstructed block using the loop filter strength, anduse the filtered reconstructed block as a predictor block. 15. The apparatus of claim 14, wherein the processor is configured to determine the loop filter strength for the reconstructed block based on a quality value difference between the current quality value and the predictor quality value. 16. The apparatus of claim 15, wherein current quality value is at least one of a quantization level, a base loop filter strength, a bit rate, or a difficulty metric of the current block and the predictor quality value is at least one of a quantization level, a loop filter strength, a bit rate, or a difficulty metric associated with the predictor block. 17. The apparatus of claim 14, wherein the processor is configured to use the filtered reconstructed block as a predictor block.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.