Intra-Prediction Using a Cross-Component Linear Model in Video Coding

An encoder comprising:

• a memory with stored instructions;
• and
• a processor coupled to the memory and configured to:
• down-sample a reconstructed luma block to obtain a down-sampled luma block, the reconstructed luma block corresponding to a chroma block;
• determine a max luma value and a min luma value based on first down-sampled samples of reconstructed top neighboring luma samples that are above the reconstructed luma block or second down-sampled samples of reconstructed left neighboring luma samples within N columns which are left of the reconstructed luma block, wherein 0<N<=3 and N is a positive integer, and wherein the reconstructed top neighboring luma samples are within a single row being adjacent to the reconstructed luma block;
• determine a first chroma value and a second chroma value based on reconstructed neighboring chroma samples of the chroma block, wherein the first chroma value and the second chroma value are respectively associated with the max luma value and the min luma value;
• derive parameters of a linear model (LM) based on the max luma value, the min luma value, the first chroma value, and the second chroma value;
• and
• generate predicted chroma values of the chroma block based on the parameters of the LM and the down-sampled luma block.

A decoder comprising:

• a memory with stored instructions;
• and
• a processor coupled to the memory and configured to:
• down-sample a reconstructed luma block to obtain a down-sampled luma block, the reconstructed luma block corresponding to a chroma block;
• determine a max luma value and a min luma value based on first down-sampled samples of reconstructed top neighboring luma samples that are above the reconstructed luma block or second down-sampled samples of reconstructed left neighboring luma samples within N columns which are left of the reconstructed luma block, wherein 0<N<=3 and N is a positive integer, and wherein the reconstructed top neighboring luma samples are within a single row being adjacent to the reconstructed luma block;
• determine a first chroma value and a second chroma value based on reconstructed neighboring chroma samples of the chroma block, wherein the first chroma value and the second chroma value are respectively associated with the max luma value and the min luma value;
• derive parameters of a linear model (LM) based on the max luma value, the min luma value, the first chroma value, and the second chroma value;
• and
• generate predicted chroma values of the chroma block based on the parameters of the LM and the down-sampled luma block.