Selection of quantization step size in accordance with predicted quantization noise

• US 5,654,760 A
• Filed: 03/27/1995
• Issued: 08/05/1997
• Est. Priority Date: 03/30/1994
• Status: Expired due to Fees

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Ohtsuki, Tomoyuki
• Primary Examiner(s): Chin, Tommy P.
• Assistant Examiner(s): VINCENT, DAVID ROBERT
• Application Number:
  US08/410,721
• Time in Patent Office: 862 Days
• Field of Search: 348/404, 348/405, 348/419
• US Class Current: 375/240.04
• CPC Class Codes: H04N 19/126 Details of normalisation or weighting functions, e.g. normalisation matrices or variable uniform quantisers
• CPC Class Codes: H04N 19/146 Data rate or code amount at the encoder output
• CPC Class Codes: H04N 19/149 by estimating the code amount by means of a model, e.g. mathematical model or statistical model
• CPC Class Codes: H04N 19/152 by measuring the fullness of the transmission buffer
• CPC Class Codes: H04N 19/159 Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/45 performing compensation of the inverse transform mismatch, e.g. Inverse Discrete Cosine Transform [IDCT] mismatch
• CPC Class Codes: H04N 19/60 using transform coding
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A video image encoder estimates a quantization noise curve which will result from quantizing a coefficient data block and uses the estimated quantization noise curve and a specified minimum noise value to select the quantization step size actually used to quantize the coefficient data block. The quantization step size is also selected as a function of transmission buffer occupancy so as to limit the amount of encoded data to a predetermined rate.

Claims

1. A quantization control apparatus comprising:

quantization noise curve estimating means for estimating a quantization noise curve by determining quantization noise as a function of a quantization step size for a block of image data and limiting said quantization noise curve to a quantization step size which becomes saturated such that said quantization noise is a maximum for larger values of said quantization step size; and

selecting means for selecting an actual quantization step size in accordance with the estimated quantization noise curve such that said actual quantization step size is not saturated.

2. The apparatus of claim 1, wherein said quantization noise curve estimating means includes:

means for predicting a saturation quantization noise value at which point the quantization noise is a maximum;

means for predicting a saturation quantization step size corresponding to said saturation quantization noise value; and

means for predicting said estimated quantization noise curve in accordance with said saturation quantization noise value and said saturation quantization step size.

3. The apparatus of claim 2, wherein said means for predicting is operative to predict said estimated quantization noise curve as a straight line when said saturation quantization step size is a maximum quantization step size, and to predict said estimated quantization noise curve as a polygonal line when said saturation quantization step size is less than said maximum quantization step size.

4. The apparatus of claim 2, wherein said means for predicting a saturation quantization noise value is operative to approximately calculate the portions of said block of image data which will be lost by quantization, to sum the lost portions in said block of image data and to multiply the sum by a constant.

5. The apparatus of claim 2, wherein said means for predicting a saturation quantization step size is operative to find a minimum quantization step size which results in quantized image data for said block of image data having values of zero.

6. The apparatus of claim 1, wherein said selecting means is operative to select said actual quantization step size also in accordance with a buffer occupancy of a buffer which stores quantized image data.

7. The apparatus of claim 6, wherein said selecting means includes:

first means for generating a first quantization step size in accordance with said estimated quantization noise curve;

second means for generating a second quantization step size in accordance with said buffer occupancy; and

third means for selecting a larger of said first and second quantization step sizes as said actual quantization step size.

8. The apparatus of claim 7, wherein said first means is operative to generate said first quantization step size also in accordance with a specified noise.

9. The apparatus of claim 8, wherein said first means is operative to generate said first quantization step size as a maximum quantization step size when A≧

a, and to generate said first quantization step size as (A/a)×

(b) when A<

a, where

A is said specified noise,

a is a saturation noise value, and

b is a saturation quantization step size.

10. A quantization control method comprising the steps of:

estimating a quantization noise curve by determining quantization noise as a function of a quantization step size for a block of image data and limiting said quantization noise curve to a quantization step size which becomes saturated such that said quantization noise is a maximum for larger values of said quantization step size; and

selecting an actual quantization step size in accordance with the estimated quantization noise curve such that said actual quantization step size is not saturated.

11. The method of claim 10, wherein said step of estimating includes:

predicting a saturation quantization noise value at which point said quantization noise is a maximum;

predicting a saturation quantization step size corresponding to the saturation quantization noise value; and

predicting said estimated quantization noise curve in accordance with said saturation quantization noise value and said saturation quantization step size.

12. The method of claim 11, wherein said step of predicting said estimated quantization noise curve predicts said estimated quantization noise curve as a straight line when said saturation quantization step size is a maximum quantization step size, and predicts said estimated quantization noise curve as a polygonal line when said saturation quantization step size is less than said maximum quantization step size.

13. The method of claim 11, wherein said step of predicting said saturation quantization noise value includes approximately calculating the portions of said block of image data which will be lost by quantization, summing the lost portions in said block of image data and multiplying the sum by a constant.

14. The method of claim 11, wherein said step of predicting a saturation quantization step size includes finding a minimum quantization step size which results in quantized image data for said block of image data having values of zero.

15. The method of claim 10, wherein said step of selecting selects said actual quantization step size also in accordance with a buffer occupancy of a buffer which stores quantized image data.

16. The method of claim 15, wherein said step of selecting includes:

generating a first quantization step size in accordance with said estimated quantization noise curve;

generating a second quantization step size in accordance with said buffer occupancy; and

selecting a larger of said first and second quantization step sizes as said actual quantization step size.

17. The method of claim 16, wherein said step of generating a first quantization step size generates said first quantization step size also in accordance with a specified noise.

18. The method of claim 17, wherein said step of generating a first quantization step size includes generating said first quantization step size as a maximum quantization step size when A≧

a, and generating said first quantization step size as (A/a)×

(b) when A<

a, where

A is said specified noise,

a is a saturation noise value, and

b is a saturation quantization step size.

19. An encoding apparatus comprising:

quantization noise curve estimating means for estimating a quantization noise curve by determining quantization noise as a function of a quantization-step size for a block of image data and limiting said quantization noise curve to a quantization step size which becomes saturated such that said quantization noise is a maximum for larger values of said quantization step size;

selecting means for selecting an actual quantization step size in accordance with the estimated quantization noise curve such that said actual quantization step size is not saturated; and

quantizing means for quantizing said block of image data using said actual quantization step size.