determining a reference prediction unit (colPu) for a current block which is a prediction block covering a specific location (x, y) in a reference picture;

deriving a temporal motion vector predictor for the current block based on motion information of the reference prediction unit;

deriving a prediction sample of the current block based on the temporal motion vector predictor;

deriving a residual sample of the current block based on received residual information;

generating a reconstructed sample based on the prediction sample and the residual sample,

wherein the determining the reference prediction unit comprises:

determining whether a lower right neighboring sample of the current block belongs to a Largest Coding Unit (LCU) to which the current block belongs, wherein the lower right neighboring sample neighbors the current block and is a sample located at a lower right side of the current block;

when the lower right neighboring sample does not belong to the LCU, specifying the specific location (x, y) as a position ((xPCtr>>4)<<4, (yPCtr>>4)<<4), wherein a position (xPCtr, yPCtr) is a position of a sample located at the bottom right side among four central samples of a co-located block, wherein the co-located block is a block in the reference picture with a same position and a same size as the current block; and

determining the reference prediction unit (colPu) which is the prediction block covering the specific location (x, y) in the reference picture.