US 11,985,322 B2
Encoder, a decoder and corresponding methods of filter modification on general intra prediction process
Biao Wang, Munich (DE); Semih Esenlik, Munich (DE); Anand Meher Kotra, Munich (DE); Han Gao, Munich (DE); and Elena Alexandrovna Alshina, Munich (DE)
Assigned to Huawei Technologies Co., Ltd., Shenzhen (CN)
Filed by HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed on Mar. 1, 2022, as Appl. No. 17/684,224.
Application 17/684,224 is a continuation of application No. PCT/CN2020/112865, filed on Sep. 1, 2020.
Claims priority of application No. PCT/EP2019/073370 (WO), filed on Sep. 2, 2019.
Prior Publication US 2022/0286684 A1, Sep. 8, 2022
Int. Cl. H04N 19/105 (2014.01); H04N 19/132 (2014.01); H04N 19/159 (2014.01); H04N 19/176 (2014.01); H04N 19/46 (2014.01)
CPC H04N 19/132 (2014.11) [H04N 19/105 (2014.11); H04N 19/159 (2014.11); H04N 19/176 (2014.11); H04N 19/46 (2014.11)] 9 Claims
OG exemplary drawing
 
1. A method for intra prediction, comprising:
during an intra predicted sample generation stage, obtaining a filtering indication indicating whether a filtered value or an unfiltered value of reference samples of a current coding block is to be used;
in response to determining that a value of the filtering indication is true, generating a prediction sample by using the filtered value of the reference samples based on a planar mode; or
in response to determining that the value of the filtering indication is false, generating a prediction sample by using the unfiltered value of the reference samples based on the planar mode; and
during the intra predicted sample generation stage, obtaining the filtered value of the reference samples by applying a filter to the reference samples, wherein:
when the filter is [1 2 1],
predV[x][y]=((nH−1−y)*((p[x−1][−1]+2*p[x][−1]+p[x+1][−1]+2)>>2)+(y+1) * p[−1][nTbH])<<Log2 (nW),
predH[x][y]=((nW−1−x)*((p[−1][y−1]+2*p[−1][y]+p[−1][y+1]+2)>>2)+(x+1)*p[nTbW][−1])<<Log2 (nH),
wherein predV[x][y] is a vertical component of the prediction sample, predH[x][y] is a horizontal component of the prediction sample, and p[x][y] (x=−1, y=−1 . . . nTbH and x=0 . . . nTbW, y=−1) is the reference sample; and
when the filter is [1 6 1],
predV[x][y]=((nH−1−y)*((p[x−1][−1]+(p[x][−1]*6)+p[x+1][−1]+4)>>3)+(y+1)*p[−1][nTbH])<<Log2 (nW),
predH[x][y]=((nW−1−x)*((p[−1][y−1]+(p[−1][y]*6)+p[−1][y+1]+4)>>3)+(x+1)*p[nTbW][−1])<<Log2 (nH),
wherein predV[x][y] is a vertical component of the prediction sample, predH[x][y] is a horizontal component of the prediction sample, and p[x][y] (x=−1, y=−1 . . . nTbH and x=0 . . . nTbW, y=−1) is the reference sample.