System and method of reduced-temporal-resolution update for video coding and quality control
View Patent ↗A novel system and method of Reduced-Temporal-Resolution Update (RTRU) and quantizing video transform coefficients is provided for use using hybrid coding standards when receiving a frame from within a group of video frames, where the amplitude of the quantization parameter is defined for the quantization of the frame according to a reduced temporal resolution update process; where the amplitude of the quantization parameter for macroblocks in the frame is defined according to a reduced temporal resolution update process; and where a bitstream is composed for transmission and subsequent decoding by a standard compliant decoder.
1. A method of Reduced-Temporal-Resolution Update and quantizing video transform coefficients in hybrid coding standards, comprising:
receiving with a video encoder a frame from within a group of video frames; defining an amplitude of a quantization parameter and a plurality of quantization parameter adjustment values, wherein the plurality of quantization parameter adjustment values is associated with a range of values selected from a plurality of ranges, and each of the plurality of ranges is associated with a distinct frame coding type;
adjusting the amplitude of the quantization parameter based upon a coding type of the frame from a quantization parameter of a prior frame within the group of video frames using at least one of the plurality of quantization parameter adjustment values;
defining a macroblock quantization parameter adjustment value and further adjusting the amplitude of the quantization parameter used by said video encoder for macroblocks in the frame based upon said macroblock quantization parameter adjustment value;
composing a bitstream for transmission and subsequent decoding by a standard compliant decoder; and
the amplitude of the quantization parameter for a P frame other than a first P frame in said group of video frames is determined by adjusting the amplitude of the quantization parameter for a prior P frame by a value of said plurality of quantization parameter adjustment values.
2. A method according to claim 1 , wherein smaller quantization parameters are used to code I frames and larger quantization parameters are used to code P/B frames.
3. A method according to claim 1 , wherein the reduced temporal resolution update is performed in a compliant manner to hybrid coding standards including H.263, MPEG-2, MPEG-4, and H.264/AV.
4. A method according to claim 1 , wherein default quantization parameters for frame types, including predicted frames and bi-directionally predicted frames, in the same GOP follow a temporal pattern.
5. A method according to claim 4 , wherein the default quantization parameter decreases with respect to a distance from an I frame in the GOP for the same types of frames.
6. A method according to claim 1 , wherein quantization parameters in a frame are adjusted on a macroblock by macroblock basis.
7. A method according to claim 6 , wherein the quantization parameter adjustment for a macroblock is based on the macroblock data corresponding to visual importance.
8. A method according to claim 6 , wherein the quantization parameter adjustment for a macroblock is based on the macroblock data corresponding to visual importance, including brightness, SAD value and motion vector size.
9. A method according to claim 6 , wherein the quantization parameter adjustment for a macroblock is based on the macroblock data corresponding to brightness.
10. A method according to claim 6 , wherein the quantization parameter adjustment for a macroblock is based on the macroblock data corresponding to SAD value.
11. A method according to claim 6 , wherein the quantization parameter adjustment for a macroblock is based on the macroblock data corresponding to motion vector size.
12. The method according to claim 1 , wherein the amplitude of the quantization parameter for an I frame is determined by adjusting a default quantization parameter value by a first value of said plurality of quantization parameter adjustment values.
13. The method according to claim 12 , wherein said first value ranges from 2 to 4, inclusive.
14. The method according to claim 12 , wherein:
the amplitude of the quantization parameter for the first P frame in said group of video frames is determined by adjusting the amplitude of the quantization parameter for the I frame by a second value of said plurality of quantization parameter adjustment values; and
the amplitude of the quantization parameter for a P frame other than said first P frame in said group of video frames is determined by adjusting the amplitude of the quantization parameter for a prior P frame by a third value of said plurality of quantization parameter adjustment values.
15. The method according to claim 14 , wherein the second value is in the range of 4 to 8, inclusive.
16. The method according to claim 14 , wherein, for third value is in the range of 0 to 2, inclusive.
17. The method according to claim 14 , wherein the amplitude of the quantization parameter for a B frame is determined by adjusting the amplitude of the quantization parameter for a prior P frame by a fourth value of said plurality of quantization parameter adjustment values.
18. The method according to claim 17 , wherein, for fourth value is in the range of 1 to 2, inclusive.
19. The method according to claim 1 , wherein the macroblock quantization parameter adjustment value is in the range of −6 to 2, inclusive.
20. The method according to claim 1 , wherein adjustment of the amplitude of the quantization parameter comprises either incrementing or decrementing the amplitude of the defined quantization parameter using one of the quantization parameter adjustment values and incrementing the resulting amplitude of the quantization parameter using the macroblock quantization parameter adjustment value.