Patent US5699119A successfully refreshed from Google Patents.
US5699119A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Quantization method and circuit therefor
Assignee: Samsung Electronics Co Ltd
•
Filing: 1995-03-30
•
Granted: 1997-12-16
•
Expires: 2015-03-30
Abstract
Abstract In a method for quantizing a moving picture signal and a circuit therefor, the method includes the steps of calculating horizontal and vertical gradients of first pixel values in a macroblock, determining each maximum value of the horizontal and vertical gradients, classifying the moving picture signal into complexity class if the maximum value is greater than a first threshold value and into edge class if the maximum value is greater than a second threshold value and is less than a third threshold value, calculating an activity value with respect to the first pixel values in the macroblock by calculating the sum of absolute values of DCT coefficients for four blocks in the macroblock, calculating an activity normalization coefficient, using the activity value of the macroblock and an average activity value of all pixels of the same picture type; and calculating a reference quantization step size produced by a backward rate control through a forward estimation considering coding efficiency and a visual characteristic which depends on the class of the moving picture signal and the energy of DCT coefficients. Therefore, a high efficiency and high picture quality moving picture compressor can be implemented.
Inventors
- 👤 Tae-yun Chung
- 👤 Gyu-hwan Jung
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-04 18:17
CPC Classifications
H; H04; H04N; H04N19/00; H04N19/10; H04N19/169; H04N19/18; H04N19/102; H04N19/124; H04N19/103; H04N19/105; H04N19/112; H04N19/115; H04N19/126; H04N19/134; H04N19/136; H04N19/137; H04N19/14; H04N19/17; H04N19/172; H04N19/176; H04N19/60; H04N19/61; H04N19/13; H04N19/146; H04N19/152; H04N19/90; H04N19/91
Claims Analysis
Total Claims: 4
Claim 1
Independent Claim
182 words
A method for quantizing a moving picture signal comprising the steps of: calculating horizontal and vertical gradients of first pixel values in a macroblock; determining a final gradient by comparison of said horizontal and vertical gradients; classifying said moving picture signal into a complexity class if said final gradient is greater than a first threshold value and into an edge class if said final gradient is greater than a second threshold value and is less than a third threshold value; calculating an activity value with respect to the first pixel values in said macroblock by calculating the sum of absolute values of DCT coefficients for four blocks in said macroblock; calculating an activity normalization coefficient, using said activity value of said macroblock and an average activity value of all pixels of a same picture type; and calculating a reference quantization step size produced by a backward rate control through a forward estimation considering coding efficiency and a visual characteristic which depends on the class of the moving picture signal and the energy of DCT coefficients as represented by said activity normalization coefficient.
Claim 2
Independent Claim
182 words
A method for quantizing a moving picture signal comprising the steps of: calculating horizontal and vertical gradients of first pixel values in a macroblock; determining a final gradient by comparison of said horizontal and vertical gradients; classifying said moving picture signal into a complexity class if said final gradient is greater than a first threshold value and into an edge class if said final gradient is greater than a second threshold value and is less than a third threshold value; calculating an activity value with respect to the first pixel values in said macroblock by calculating the sum of absolute values of DCT coefficients for four blocks in said macroblock; calculating an activity normalization coefficient, using said activity value of said macroblock and an average activity value of all pixels of a same picture type; and calculating a reference quantization step size produced by a backward rate control through a forward estimation considering coding efficiency and a visual characteristic which depends on the class of the moving picture signal and the energy of DCT coefficients as represented by said activity normalization coefficient.
Claim 3
Dependent Claim
219 words
A quantizer, comprising: means for calculating horizontal gradients of first pixel values of a moving picture signal in a macroblock and calculating a mean value of the respective gradients; means for calculating vertical gradients of said first pixel values in said macroblock and calculating a mean value of the respective vertical gradients; means for determining a final gradient by comparison of said horizontal and vertical gradients; means for classifying said moving picture signal into a complexity class if said final gradient is greater than a first threshold value and into an edge class if said final gradient is greater than a second threshold value and less than a third threshold value; means for calculating an activity value with respect to said first pixel values in said macroblock by calculating the sum of absolute values of DCT coefficients for four blocks in said macroblock; means for calculating an activity normalization coefficient, using said activity value of said macroblock and an average activity value of all pixels of a same picture type; and means for calculating a reference quantization step size produced by a backward rate control through a forward estimation considering coding efficiency and a visual characteristic which depends on the class of the moving picture signal and the energy of DCT coefficients as represented by said activity normalization coefficient.
Claim 4
Dependent Claim
219 words
A quantizer, comprising: means for calculating horizontal gradients of first pixel values of a moving picture signal in a macroblock and calculating a mean value of the respective gradients; means for calculating vertical gradients of said first pixel values in said macroblock and calculating a mean value of the respective vertical gradients; means for determining a final gradient by comparison of said horizontal and vertical gradients; means for classifying said moving picture signal into a complexity class if said final gradient is greater than a first threshold value and into an edge class if said final gradient is greater than a second threshold value and less than a third threshold value; means for calculating an activity value with respect to said first pixel values in said macroblock by calculating the sum of absolute values of DCT coefficients for four blocks in said macroblock; means for calculating an activity normalization coefficient, using said activity value of said macroblock and an average activity value of all pixels of a same picture type; and means for calculating a reference quantization step size produced by a backward rate control through a forward estimation considering coding efficiency and a visual characteristic which depends on the class of the moving picture signal and the energy of DCT coefficients as represented by said activity normalization coefficient.
References Cited (Backward Citations)
US5307163A
Cited
US5543848A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2000-12-15 | 2001-06-16 | 2001-12-16 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2004-12-15 | 2005-06-16 | 2005-12-16 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2008-12-15 | 2009-06-16 | 2009-12-16 | $7700.00 | Window Closed / Lapsed |