| CPC G06T 9/00 (2013.01) [H04N 19/46 (2014.11)] | 11 Claims |

|
1. A coding method, comprising:
obtaining T1 first to-be-coded data units and T2 second to-be-coded data units, T1 and T2 being integers greater than or equal to 1;
coding the T1 first to-be-coded data units and the T2 second to-be-coded data units based on at least one of an extended reality data type corresponding to the T1 first to-be-coded data units or an extended reality data type corresponding to the T2 second to-be-coded data units, to obtain K coded data units, K being an integer greater than or equal to T1+T2; and
outputting the K coded data units, wherein:
coding the T1 first to-be-coded data units and the T2 second to-be-coded data units comprises:
obtaining a coding matrix based on the least one of the extended reality data type corresponding to the T1 first to-be-coded data units or the extended reality data type corresponding to the T2 second to-be-coded data units; and
coding the T1 first to-be-coded data units and the T2 second to-be-coded data units based on the coding matrix;
the coding matrix comprises a coding submatrix comprising T1 first coding vectors and T2 second coding vectors, the T1 first coding vectors respectively corresponding to the T1 first to-be-coded data units and the T2 second coding vectors respectively corresponding to the T2 second to-be-coded data units; and
the quantity of non-zero coefficients included in the any one of the T2 second coding vectors is less than a coding depth M, the coding depth being a maximum value of a quantity of to-be-coded data units that participate in coding when one coded data unit is obtained, M being an integer greater than 0 and less than or equal to T1+T2.
|