FIELD: data processing.
SUBSTANCE: invention relates to video coding, in particular, to internal video prediction. Video decoding method is proposed, comprising phases when: it is determined if predetermined number of adjacent pixels adjacent to current block are not available for internal prediction of current block; when first adjacent pixel from number of predetermined amount of adjacent pixels is unavailable, second adjacent pixel is searched from among predetermined number of adjacent pixels in predetermined direction from innermost adjacent pixel on left to uppermost adjacent pixel from left for adjacent pixels on left and from uppermost adjacent pixel to uppermost adjacent pixel on right for adjacent pixels from above; assigning pixel value of the second adjacent pixel to pixel value of lowermost adjacent pixel on left; inaccessible pixel located at left side of current block is replaced in sequence with pixel value of lower adjacent pixel located immediately below inaccessible pixel; inaccessible pixel located at current block top is replaced successively with pixel value of adjacent pixel on left, located directly leftwards from inaccessible pixel; and performing internal prediction with respect to current block by using predetermined number of adjacent pixels, first adjacent pixel is located at the very bottom leftwards from current block, in this case, when size of current block is nTxnT, where nT is integer, predetermined number is 4nT + 1.
EFFECT: technical result is reduction in complexity of process while determining reference pixel for internal prediction.
1 cl, 5 tbl, 32 dwg
