Method of determining quantization parameters

Patent metadata

Field	Value
Publication number	US-10924741-B2
Application number	US-201916384907-A
Country	US
Kind code	B2
Filing date	Apr 15, 2019
Priority date	Apr 15, 2019
Publication date	Feb 16, 2021
Grant date	Feb 16, 2021

Abstract

A method of determining quantization parameters includes the steps of: receiving a block of image data; calculating a general variance of the block of image data; calculating a plane-based variance of the block of image data by subtracting an image moment of the block of image data from the general variance; and determining a quantization parameter for the block of image data according to the plane-based variance.

First claim

What is claimed is:

1. A method of determining quantization parameters,
comprising:
receiving a block of image data;
calculating a first variance of the block of image data;
calculating a second variance of the block of image data by subtracting an image moment of the block of image data from the first variance; and
determining a quantization parameter for the block of image data according to the second variance;
wherein the image moment reflects an image gradient in the block of image data.

2. The method of claim 1 , wherein the second variance is an average of difference values of the block of image data correlated to a mathematical function other than a mean of the block of image data.

3. The method of claim 2 , wherein the mathematical function is a minimum mean square error estimator of the block of image data.

4. The method of claim 1 , wherein the step of calculating the second variance of the block of image data by subtracting the image moment of the block of image data from the first variance comprises:
subtracting a first raw image moment along a first direction multiplied by a first moment parameter and a second raw image moment along a second direction multiplied by a second moment parameter from the first variance, to obtain the second variance.

5. The method of claim 4 , further comprising:
setting values of the first moment parameter and the second moment parameter to control a correlation of an image gradient in the block of image data with the quantization parameter.

6. The method of claim 5 , wherein the block of image data is an N×N block, and each of the values of the first moment parameter and the second moment parameter is within a range from 0 to m, wherein m is equal to:

m
=

3
⁢

N
2

N
2

-
1

.

7. The method of claim 5 , wherein each of the values of the first moment parameter and the second moment parameter is within a range from 0 to 3.

8. The method of claim 1 , further comprising:
determining that the quantization parameter for the block of image data has a first value when the second variance has a first value; and
determining that the quantization parameter for the block of image data has a second value smaller than the first value of the quantization parameter when the second variance has a second value smaller than the first value of the second variance.

9. A method of determining quantization parameters,
comprising:
receiving a block of image data;
calculating a second variance of the block of image data by using an image moment of the block of image data; and
determining a quantization parameter for the block of image data according to the second variance;
wherein the image moment reflects an image gradient in the block of image data.

10. The method of claim 9 , wherein the second variance is an average of difference values of the block of image data correlated to a mathematical function other than a mean of the block of image data.

11. The method of claim 10 , wherein the mathematical function is a minimum mean square error estimator of the block of image data.

12. The method of claim 9 , wherein the step of calculating the second variance of the block of image data by using the image moment of the block of image data comprises:
calculating a first variance of the block of image data; and
subtracting a first raw image moment along a first direction multiplied by a first moment parameter and a second raw image moment along a second direction multiplied by a second moment parameter from the first variance, to obtain the second variance.

13. The method of claim 12 , further comprising:
setting values of the first moment parameter and the second moment parameter to control a correlation of an image gradient in the block of image data with the quantization parameter.

14. The method of claim 13 , wherein the block of image data is an N×N block, and each of the values of the first moment parameter and the second moment parameter is within a range from 0 to m, wherein m is equal to:

m
=

3
⁢

N
2

N
2

-
1

.

15. The method of claim 13 , wherein each of the values of the first moment parameter and the second moment parameter is within a range from 0 to 3.

16. The method of claim 9 , further comprising:
determining that the quantization parameter for the block of image data has a first value when the second variance has a first value; and
determining that the quantization parameter for the block of image data has a second value smaller than the first value of the quantization parameter when the second variance has a second value smaller than the first value of the second variance.

Assignees

• Novatek Microelectronics Corp

Inventors

• Jou Fandi

Classifications

• H04N19/42

  characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation (H04N19/635 takes precedence) · CPC title

• H04N19/189

  characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding · CPC title

• H04N19/14

  Coding unit complexity, e.g. amount of activity or edge presence estimation (H04N19/146 takes precedence) · CPC title

• H04N19/139

  Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title