| 摘要 |
Provided is a method for transforming, by a graphic processing unit (GPU), coefficient information generated through video encoding operation. In detail, the GPU divides the coefficient information into at least one block constituting a video frame. The GPU generates a valid block map indicating whether each block is a valid block including a valid coefficient. The GPU generates a block scanning method map indicating a scanning method applied to each valid block. The GPU generates an end of block (EOB) position map indicating a position of a final valid coefficient of each valid block according to a scanning method applied to each valid block. The GPU generates frame coefficient information by aligning coefficients of each valid block according to the scanning method applied to each valid block. |
