Inter-layer video decoding method for performing subblock-based prediction and apparatus therefor, and inter-layer video encoding method for performing subblock-based prediction and apparatus therefor

An inter-layer video decoding method comprising:

• obtaining, from a bitstream, subblock size information of a second layer image, which indicates a value about base-2 logarithm of a size of a subblock within a range that is equal to or greater than base-2 logarithm of a minimum size of a coding unit minus 3 and is less than or equal to base-2 logarithm of a maximum size of the coding unit minus 3, wherein the base-2 logarithm of the minimum size of the coding unit is determined based on information indicating the minimum size of the coding obtained from the bitstream and the base-2 logarithm of the maximum size of the coding unit is determined based on the information indicating the minimum size of the coding unit and information indicating a difference between the maximum size of the coding unit and the minimum size of the coding unit obtained from the bitstream;
• determining a size of a subblock from a range that is equal to or greater than the minimum size of the coding unit and is less than or equal to the maximum size of the coding unit, by using the subblock size information obtained from the bitstream;
• determining a candidate block that corresponds to a current block included in the second layer image and is included in a first layer image different from the second layer image, wherein at least one subblock of the first layer is determined from the candidate block based on the size of the subblock and at least one subblock of the second layer is determined from the current block based on the size of the subblock;
• obtaining motion information of the at least one subblock included in the candidate block of the first layer image corresponding to the at least one subblock included in the current block of the second layer image;
• and
• obtaining or predicting motion information of the current block by using the motion information of the at least one subblock of the first layer image, and decoding the current block by using the obtained or predicted motion information of the current block.