Image processing method and apparatus for performing sample adaptive offset processing

Patent metadata

Field	Value
Publication number	US-2018234702-A1
Application number	US-201715814668-A
Country	US
Kind code	A1
Filing date	Nov 16, 2017
Priority date	Feb 10, 2017
Publication date	Aug 16, 2018
Grant date	—

Abstract

An image processing method and apparatus for performing sample adaptive offset (SAO) processing are provided. The image processing method includes: performing, based on a largest coding unit (LCU) and a coding unit (CU), an encoding operation on a frame image; setting a maximum offset having a value adjusted according to a maximum transformation size in the encoding operation; and calculating, based on the set maximum offset, a sample adaptive offset (SAO) parameter for use in SAO compensation.

First claim

1 . An image processing method comprising:
performing, based on a largest coding unit (LCU) and a coding unit (CU), an encoding operation on a frame image; setting a maximum offset having a value that is adjusted according to a maximum transformation size in the encoding operation; and calculating, based on the maximum offset, a sample adaptive offset (SAO) parameter for use in sample adaptive offset (SAO) compensation.

2 . The image processing method of claim 1 , wherein the maximum transformation size is corresponding to a maximum processing unit of a transformation operation and a quantization operation in the encoding operation, and
as the maximum transformation size increases, the maximum offset decreases, and as the maximum transformation size decreases, the maximum offset increases.

3 . The image processing method of claim 2 , wherein the maximum transformation size is corresponding to a number of pixels in the largest coding unit (LCU).

4 . The image processing method of claim 1 , wherein the calculating, based on the maximum offset, the sample adaptive offset (SAO) parameter comprises:
classifying pixels in the largest coding unit (LCU) or the CU into one or more SAO blocks; and allocating an offset value less than or equal to the maximum offset to each of the one or more SAO blocks.

5 . The image processing method of claim 1 , wherein the setting of the maximum offset comprises calculating the maximum offset by further using information related to a bit depth of pixels of the frame image, and
with respect to the maximum transformation size, as the bit depth of pixels of the frame image increases, the maximum offset increases, and as the bit depth of pixels of the frame image decreases, the maximum offset decreases.

6 . The image processing method of claim 1 , wherein the setting the maximum offset comprises calculating the maximum offset by further using a quantization parameter in a quantization operation of the encoding operation, and
with respect to the maximum transformation size, as a quantization parameter value increases, the maximum offset increases, and as the quantization parameter value decreases, the maximum offset decreases.

7 . The image processing method of claim 1 , wherein the setting the maximum offset comprises calculating the maximum offset via a predetermined functional formula operation based on an input of information related to the maximum transformation size.

8 . The image processing method of claim 1 , wherein a table comprising information regarding the maximum offset corresponding to the maximum transformation size is stored in an image processing apparatus, and
the setting the maximum offset comprises calculating the maximum offset by reading the information from the table based on an input of information related to the maximum transformation size.

9 . The image processing method of claim 1 , further comprising:
reconstructing, with respect to the frame image having the encoding operation performed thereon, the frame image to generate a reference image for use in inter prediction; and performing SAO compensation by using the sample adaptive offset (SAO) parameter, with respect to a reconstructed image.

10 . The image processing method of claim 1 , further comprising transmitting a bitstream comprising the sample adaptive offset (SAO) parameter and image data generated according to the encoding operation,
wherein the sample adaptive offset (SAO) parameter comprises a parameter indicating an offset value that is to be used in SAO compensation, and when the maximum offset decreases, an amount of information of the parameter indicating the offset value decreases.

11 . An image processing method comprising:
performing, based on a first maximum transformation size, a first encoding operation on a first image; setting a first maximum offset corresponding to the first maximum transformation size; performing a sample adaptive offset (SAO) compensation on the first image by calculating a first offset value that is less than or equal to the first maximum offset; performing, based on a second maximum transformation size, a second encoding operation on a second image; and setting a second maximum offset corresponding to the second maximum transformation size, wherein a first value of the first maximum offset and a second value of the second maximum offset are different.

12 . The image processing method of claim 11 , wherein the first image is corresponding to a first frame image and the second image is corresponding to a second frame image.

13 . The image processing method of claim 12 , wherein, when the first maximum transformation size is greater than the second maximum transformation size, the first maximum offset is less than the second maximum offset.

14 . The image processing method of claim 11 , wherein the first image is in a first region of a first frame, and the second image is in a second region of the first frame.

15 . The image processing method of claim 11 , wherein the first maximum transformation size is corresponding to a first maximum processing unit for performing a first transformation operation and a first quantization operation on the first image, and the second maximum transformation size is corresponding to a second maximum processing unit for performing a second transformation and quantization operation on the second image.

16 . The image processing method of claim 11 , wherein the performing of the sample adaptive offset (SAO) compensation comprises:
with respect to one or more coding units in the first image, classifying pixels of each of the one or more coding units into one or more SAO blocks; allocating the first offset value less than or equal to the first maximum offset with respect to each of the one or more SAO blocks; and adding up offset values allocated with respect to pixels in each of the one or more SAO blocks.

17 . The image processing method of claim 11 , further comprising performing SAO compensation on the second image by calculating a second offset value that is less than or equal to the second maximum offset,
wherein, when the second maximum transformation size is less than the first maximum transformation size, a second maximum value of the second offset value applied to the second image is greater than a first maximum value of the first offset value applied to the first image.

18 - 19 . (canceled)

20 . An image processing apparatus comprising:
a transformation and quantization unit configured to perform, based on a processing unit less than or equal to a maximum transformation size, a transformation operation and a quantization operation on an input image; and a sample adaptive offset (SAO) encoder configured to receive a reconstructed image generated via an inverse quantization process and an inverse transformation process with respect to the input image on which the transformation operation and the quantization operation has been performed, and to perform sample adaptive offset (SAO) compensation on the reconstructed image by using a maximum offset having a value adjusted according to a change of the maximum transformation size.

21 . The image processing apparatus of claim 20 , further comprising an entropy encoder configured to generate a bitstream by performing entropy encoding on encoded image data from the transformation and quantization unit and a sample adaptive offset (SAO) parameter from the sample adaptive offset (SAO) encoder.

22 . The image processing appar

Assignees

• Samsung Electronics Co Ltd

Inventors

• Jung Young-Beom
• BYUN Ju-won
• Jang Hyuk-Jae
• Yang Jung-Yeop

Classifications

• H04N19/61

  in combination with predictive coding · CPC title

• H04N19/91

  Entropy coding, e.g. variable length coding [VLC] or arithmetic coding · CPC title

• H04N19/172

  the region being a picture, frame or field · CPC title

• H04N19/13

  Adaptive entropy coding, e.g. adaptive variable length coding [AVLC] or context adaptive binary arithmetic coding [CABAC] · CPC title

• H04N19/132

  Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking · CPC title