This article only describes how to use the above pixels (not existing) (such as the vertical mode) for the upper edge block, and how to use the left pixel (not existing) (such as the horizontal mode) for the left edge block.
Processing of H.264 I frame edge Blocks
The first edge block on the top left of the I frame uses intra-frame prediction. There are no pixels on the left and top of the frame. Therefore, the pixels on the left and the top are both 128 (255/2 ), when the upper edge block is used for vertical prediction, there is no upper edge pixel.
The value 128 is used, while the value 128 is used in the horizontal prediction mode because there is no left edge pixel at the left edge. remember that the block prediction of the I frame is based on the blocks that have been rebuilt on the left and top of the surrounding area.
To make another prediction.