FIG. 8 shows intra prediction modes used for 4×4 block in H.264/AVC. Intra prediction modes for 8×8 block are basically equal to those used for 4×4 block [not shown in the drawing]. Referring to FIG. 8, one most suitable prediction mode is selected from 9 kinds of prediction modes per 4×4 block. And, the selected prediction mode is then encoded by 4×4 block unit.

In intra prediction coding for 16×16 block in H.264/AVC, prediction coding is performed by selecting one of four methods including vertical prediction (mode 0), horizontal prediction (mode 1), average prediction (mode 3) and planar prediction (mode 3) from 16 pixels of a top side macroblock and 16 pixels of a left side macroblock [not shown in the drawing].

Yet, a high-resolution video of 4K (3840×2160) or 8B (7680×4320) much greater than a currently and widely used video of 2K level (1920×1080) has characteristics different from those of the 2K-level video due to high resolution. For instance, the current H.264/AVC uses a fixed 16×16 macroblock only but fails to support macroblocks in various sizes. And, a texture of high-resolution video may not be sufficiently reflected. If a macroblock having a size of 32×32 greater than a current size of 16×16 or a greater size is used, a compression ratio can be enhanced.