Method and apparatus for low complexity quarter pel generation in motion search

A method of encoding video data, comprising:

• storing full-pixel position information related to a reference frame in a memory;
• when a searched position is not located on a full-pixel row within the reference frame:
• applying a vertical interpolation filter to every other row of the full-pixel position information for video information related to the reference frame to determine a first sub-pel position information, and
• storing the first sub-pel position information in the memory;
• when a searched position is not located on a full-pixel column within the reference frame:
• applying a horizontal interpolation filter to the full-pixel position information and the first sub-pel position information to determine a second sub-pel position information for the every other line of video information;
• generating a syntax element indicating pixel motion of a current frame based on the every other row of the full-pixel position information, the first sub-pel position information, and the second sub-pel position information;
• and
• encoding a block based on the generated syntax element.

A device for encoding video, comprising:

• a memory configured to store full-pixel position information related to a series of frames in a memory for motion search;
• a processor operably coupled to the memory and configured to:
• when a searched position is not located on a full-pixel row of a reference frame, apply a vertical interpolation filter to the full-pixel position information for video information related to the series of frames to determine a first sub-pel position information for the video information;
• and
• store the first sub-pel position information in the memory with the same precision as full-pixel position information;
• and
• when a searched position is not located on a full-pixel column of the reference frame, apply a horizontal interpolation filter to every other row of the full-pixel position information and the first sub-pel position information to determine a second sub-pel position information for the every other line of video information;
• generate a syntax element indicating pixel motion of the series of frames based on the full-pixel position information, the first sub-pel position information and second sub-pel position information;
• and
• encode a block based on the generated syntax element.

An apparatus for encoding video data, comprising:

• a storage means configured to store full-pixel position information related to a series of frames in a memory for motion search;
• a processing means operably coupled to the storage means and configured to:
• apply a vertical interpolation filter to the full-pixel position information when a searched position is not located on a full-pixel row of a reference frame, to determine a first sub-pixel position information related to the series of frames;
• and
• store the first sub-pel position information in the memory with the same precision as full-pixel position information;
• and
• apply a horizontal interpolation filter to every other row of the full-pixel position information and the first sub-pel position information when a searched position is not located on a full-pixel column of the reference frame, to determine a second sub-pel position information for the every other line of video information;
• a generation means for generating a syntax element indicating pixel motion of the series of frames based on the full-pixel position information, first and second half-pixel position information, or first and second quarter-pixel position information;
• and
• an encoding means configured to encode a block based on the generated syntax element.