Method and apparatus for video codec quantization
View Patent ↗An encoder, a decoder, and corresponding methods are provided for encoding and decoding video signal data for an image block. The encoder includes a quantizer for receiving transform coefficients for the image block, and for adaptively performing dead-zone quantization based on coefficient positions and coefficient distributions of the transform coefficients.
1. An encoder for encoding video signal data for an image block, comprising:
a quantizer for receiving transform coefficients for the image block, and for adaptively performing dead-zone quantization based on coefficient positions and coefficient distributions of the transform coefficients to form quantization levels, wherein said quantizer adaptively performs the dead-zone quantization by determining a preliminary quantization level using a fixed dead-zone/rounding control, and repeating the dead-zone quantization based on the preliminary quantization level and using a coefficient quantization level dependent dead-zone/rounding control; and
an inverse quantizer for de-quantizing the quantization levels to generate reconstructed levels, wherein a rounding parameter used for de-quantizing a quantization level of a transform coefficient is determined by the quantization level.
2. The encoder as defined in claim 1 , wherein said quantizer computes the level dependent dead-zone/rounding control based upon a piecewise distribution within a corresponding level.
3. The encoder as defined in claim 2 , wherein said quantizer computes the level dependent dead-zone/rounding control using distribution concentrations with a reconstructed area of the image block.
4. The encoder as defined in claim 1 , wherein said quantizer computes the level dependent dead-zone/rounding control using an arithmetic integral process.
5. The encoder as defined in claim 4 , wherein the arithmetic integral process comprises at least one of a trapezoid approximation process and a simplex approximation process.
6. The encoder as defined in claim 1 , wherein said quantizer computes the level dependent dead-zone/rounding control based on distortion present in the transform coefficients.
7. The encoder as defined in claim 1 , wherein said quantizer forces at least some of the transform coefficients to a higher quantization level to improve subjective quality, when the at least some of the transform coefficients were originally set to a level zero.
8. The encoder as defined in claim 7 , wherein said quantizer determines whether to force the at least some of the transform coefficients to the higher level based on rate distortion optimized criteria.
9. A method for encoding video signal data for an image block, comprising the steps of:
receiving transform coefficients for the image block;
adaptively performing dead-zone quantization based on coefficient positions and coefficient distributions of the transform coefficients to form quantization levels, wherein said step of adaptively performing the dead-zone quantization comprises the steps of:
initially performing the dead-zone quantization by determining a preliminary quantization level using a fixed dead-zone/rounding control; and
repeating the dead-zone quantization based on the preliminary quantization level and using a coefficient quantization level dependent dead-zone/rounding control; and
de-quantizing the quantization levels to reconstructed levels, wherein a rounding parameter used for de-quantizing a quantization level of a transform coefficient is determined by the quantization level.
10. The method as defined in claim 9 , further comprising the step of computing the level dependent dead-zone/rounding control based upon a piecewise distribution within a corresponding level.
11. The method as defined in claim 10 , further comprising the step of computing the level dependent dead-zone/rounding control using distribution concentrations with a reconstructed area of the image block.
12. The method as defined in claim 9 , further comprising the step of computing the level dependent dead-zone/rounding control using an arithmetic integral process.
13. The method as defined in claim 12 , wherein the arithmetic integral process comprises at least one of a trapezoid approximation process and a simplex approximation process.
14. The method as defined in claim 9 , further comprising the step of computing the level dependent dead-zone/rounding control based on distortion present in the transform coefficients.
15. The method as defined in claim 9 , further comprising the step of forcing at least some of the transform coefficients to a higher quantization level to improve subjective quality, when the at least some of the transform coefficients were originally set to a level zero.
16. The method as defined in claim 15 , wherein said forcing step comprises the step of determining whether to force the at least some of the transform coefficients to the higher level based on rate distortion optimized criteria.