CPC H04N 19/105 (2014.11) [G06T 9/004 (2013.01); H04N 19/159 (2014.11); H04N 19/176 (2014.11); H04N 19/186 (2014.11); H04N 19/189 (2014.11); H04N 19/51 (2014.11)] | 20 Claims |
1. A method of decoding video data, the method comprising:
reshaping a pixel domain reference template block using a forward mapping function into a mapped domain reference template block, wherein the forward mapping function comprises:
FwdMap(Y_pred)=((b2−b1)/(a2−a1))*(Y_pred−a1)+b1,
where Y_pred is a luma prediction signal, i is a piece index, a1 is an input pivot point of i, a2 is an input pivot point of i+1, b1 is a mapped pivot point of i, and b2 is a mapped pivot point of i+1;
deriving local illumination compensation (LIC) model parameters from the mapped domain reference template block and a mapped domain neighboring reconstruction template block;
applying the LIC model parameters to motion-compensated prediction signals; and
decoding the video data based on the application of the LIC model parameters.
|