Back to results
US 7,068,850 B2Granted
Decoding of predicted DC coefficient without division
AssigneeEquator Technologies, Inc.
InventorJeongnam Youn
Life of this patent
Estimating the term…
Abstract
A technique for eliminating the division in decoding a predicted DC coefficient includes calculating and storing the values of 1//DC_scalar in a table. This allows one to convert a division into a multiplication.
Claims22
A processing circuit comprising:
- a storage circuit operable to store a lookup table;
- and
- a processor operable to:
- calculate all possible values for an inverted variable in an equation for decoding predicted zero-frequency transform coefficients;
- and
- store all said values in said lookup table.
3 more elements
A processing circuit comprising:
- a storage circuit operable to store a lookup table of all thirty-nine values of an inverted scalar value of a quantization parameter for a macro block, and
- a processor operable to calculate the zero-frequency coefficient for a macro block using said table to avoid using rounding division.
A processing circuit comprising:
- a storage circuit operable to store a lookup table of all thirty-nine values an inverted scalar value of a quantization parameter for a macro block calculated once off-line;
- and
- a processor having a divide unit and operable to calculate the zero-frequency coefficients for any number of macro blocks being decoded without accessing a divide unit by repeatedly accessing the stored look up table.
A processing circuit comprising:
- a storage circuit operable to store a lookup table;
- and
- a processor operable to:
- calculate the zero frequency coefficient of a macro block used for prediction purposes times an inverted variable in the equation for decoding predicted zero frequency transform coefficients for all values of the zero frequency coefficient of a macro block used for prediction purposes and all values of the scalar value of a quantization parameter for a macro block comprises:
- rounding to 32-bit precision in all cases where an integer variable ranging from (−(44×2048) to +(44×2047))×1/DC_scalar×2−32×232 is identically equal to an integer variable ranging from (−(44×2048) to +(44×2047) dividing DC-scalar for a macro block calculated using a rounding division, and otherwise modifying 1/DC_scalar to be 1/DC_scalar×232±1 which quantity is then multiplied by 231 32 and recalculated iteratively until equality is obtained;
- and
- store all said values in said lookup table.
4 more elements
A video processing circuit comprising:
- a storage circuit operable to store a lookup table;
- and
- a processor to receive encoded images, each in a standard format, the processor operable to decode said received encoded bit stream in software with said lookup table wherein a divide unit is not used when decoding predicted zero-frequency transform coefficients.
A video processing circuit comprising:
- a storage circuit operable to store a lookup table of all thirty-nine values of the inverted scalar value of a quantization parameter for a macro block;
- and
- a processor to receive encoded images, each in a standard formal, the processor operable to decode said received encoded bit stream in software with said lookup table wherein a divide unit is not used when decoding the predicted zero-frequency transform coefficient.
A video processing circuit comprising:
- a storage circuit operable to store a lookup table of all thirty-nine values of an inverted scalar value of a quantization parameter for a macro block;
- and
- a processor to receive encoded images, each in a standard format, the processor operable to decode said received encoded bit stream in software with said lookup table wherein a divide unit is not used when decoding any number of predicted zero frequency transform coefficients.
A video processing circuit comprising:
- a storage circuit operable to store a lookup table;
- and
- a processor to receive encoded images, each in a standard format, the processor operable to decode said received encoded bit stream in software with said lookup table avoiding a type of division rounding results to the nearest integer value in the equation for decoding predicted zero frequency transform coefficients.
A method comprising:
- calculating all possible values for an inverted variable in the equation for decoding predicted zero frequency transform coefficients;
- and
- storing all said values in a lookup table.
A method comprising:
- calculating the zero-frequency coefficient for a macro block using a table of all thirty nine values of an inverted scalar value of a quantization parameter for a macro block and to avoid using rounding division.
A method comprising:
- calculate the zero frequency coefficient of a macro block used for prediction purposes times an inverted variable in the equation for decoding predicted zero frequency transform coefficients for all values of the zero frequency coefficient of a macro block used for prediction purposes and all values of the scalar value of a quantization parameter for a macro block;
- rounding to 32-bit precision in all eases where an integer variable ranging from (−(44×2048) to +(44×2047))×1/DC_scalar×2−32×232 is identically equal to an integer variable ranging from (−(44×2048) to +(44×2047) dividing DC-scalar for a macro block calculated using a rounding division;
- otherwise modifying 1/DC_scalar to be 1/DC_scalar×232±1 which quantity is then multiplied by 2−32 and recalculated iteratively until equality is obtained;
- and
- store all said values in a lookup table.
2 more elements
Dependencies are inferred from the claim text.