Coding video data using adaptive affine block prediction

A device for decoding video data, the device comprising:

• a memory configured to store video data;
• and
• a processing system comprising one or more processors implemented in circuitry, the processing system being configured to:
• determine whether motion information of a block of the video data is for sub-blocks of the block larger than individual pixels of the block or for the individual pixels, wherein to determine whether the motion information is for the sub-blocks or for the individual pixels, the processing system is configured to determine whether overlapped block motion compensation (OBMC) is to be applied to the block, the block being associated with data indicating that the block is to be predicted using affine motion compensation;
• in response to determining that the motion information of the block is for the sub-blocks, perform sub-block-based affine motion compensation to form a prediction block for the block;
• in response to determining that the motion information is for the individual pixels, perform pixel-based affine motion compensation to form the prediction block for the block;
• and
• decode the block using the prediction block.

A device for decoding video data, the device comprising:

• means for determining whether motion information of a block of video data is for sub-blocks of the block larger than individual pixels of the block or for the individual pixels, wherein the means for determining whether the motion information is for the sub-blocks or for the individual pixels comprises means for determining whether overlapped block motion compensation (OBMC) is to be applied to the block, the block being associated with data indicating that the block is to be predicted using affine motion compensation;
• means for performing sub-block-based affine motion compensation to form a prediction block for the block in response to determining that the motion information of the block is for the sub-blocks;
• means for performing pixel-based affine motion compensation to form the prediction block for the block in response to determining that the motion information is for the individual pixels;
• and
• means for decoding the block using the prediction block.