Methods and apparatuses for encoding an HDR images, and methods and apparatuses for use of such encoded images

Patent metadata

Field	Value
Publication number	US-10779013-B2
Application number	US-201816112901-A
Country	US
Kind code	B2
Filing date	Aug 27, 2018
Priority date	May 28, 2014
Publication date	Sep 15, 2020
Grant date	Sep 15, 2020

Abstract

Producing in an image signal a codification of the pixel colors of the lower luminance dynamic range image and outputting in the image signal values encoding the functional behavior of the above color conversions as metadata, or values for the inverse functions.

First claim

The invention claimed is:

1. A method of encoding a high dynamic range image, comprising:
converting the high dynamic range image to a lower luminance dynamic range image by applying:
normalization of the high dynamic range image to a scale of the luma axis being (0,1) yielding a normalized high dynamic range image with normalized colors having normalized luminances;
calculating a gamma function on the normalized luminances yielding gamma-converted luminances;
applying a first tone mapping which boosts the gamma-converted luminances of pixels of the normalized high dynamic range image which lie below 0.1 with a predetermined amount lying between 1.5 and 5.0, yielding lumas; and
applying an arbitrary monotonically increasing tone mapping function, wherein the arbitrary monotonically increasing tone mapping function maps the lumas to output lumas of the lower dynamic range image;

outputting in an image signal a codification of the pixel colors of the lower luminance dynamic range image; and
outputting in the image signal values encoding the function shapes of the calculating of the gamma function, the applying of the first tone mapping and the applying of the arbitrary monotonically increasing tone mapping as metadata, or values for their inverse functions, wherein the metadata allows a receiver to reconstruct a reconstructed high dynamic range image from the lower luminance dynamic range image.

2. The method as claimed in claim 1 ,
wherein the first tone mapping is defined as

v
=

log
⁡

(

1
+

(

RHO
-
1

)

*
xg

)

log
⁡

(
RHO
)

,

wherein RHO has a predetermined value,
wherein RHO or a value which is a function of RHO is outputted in the metadata.

3. The method as claimed in claim 1 , wherein the gamma function calculation uses a gamma value equal to 1/(2.4).

4. The method as claimed in claim 1 , further comprising:
determining a gain value, wherein the gain value is arranged to map the maximum luma of the lower dynamic range image to a specific value of the possible values in the reconstructed high dynamic range image; and
encoding the gain value in the image signal.

5. The method as claimed in claim 4 , further comprising:
after applying one or more of the arbitrary monotonically increasing tone mapping function and the mapping the maximum luma of the lower dynamic range image to a specific value of the possible values to determine the lower dynamic range image, applying a technical tone mapping to determine a second lower dynamic range image,
wherein the second lower dynamic range image can be used to drive LDR displays as an alternative driving image alternative to the lower luminance dynamic range image,
wherein technical tone mapping is determined by:
determining a first geometrical region of the lower luminance dynamic range image for which the visibility of banding in the corresponding reconstructed high dynamic range image is above an acceptable level;
determining a first range of lumas for the first geometrical region;
determining a second range of lumas adjacent on the luma axis to the first range of lumas,
wherein the second range of lumas is identified to fulfill the conditions that it has a number of lumas above a minimum number,
wherein the second range of lumas corresponds to a second geometrical region,
wherein the second geometrical region contains a texture which can be represented using less than the minimum number of codes in an LDR image upon which to apply the functions yielding a reconstructed high dynamic range image of sufficient visual quality for that second geometrical region; and

determining a redistribution mapping function, wherein the redistribution mapping function redistributes the lumas of the first range of lumas and the second range of lumas, so that additional codes are available for the first range of lumas; and
outputting in the image signal values encoding the function shape of the redistribution mapping.

6. The method as claimed in claim 5 , wherein the amount of additional codes for the first range of lumas is determined based on a banding visibility criterion for the first geometrical region.

7. The method as claimed in claim 5 ,
wherein an identification of the first geometrical region is performed by a human grader via a user interface circuit,
wherein the amount of banding of the first geometrical region in the reconstructed high dynamic range image, and the visual quality of reconstruction of the second geometrical region in the reconstructed high dynamic range image are judged by the human grader as acceptable or unacceptable,
wherein in case of the acceptable judgement the values encoding the function shape of the redistribution mapping function or its inverse are encoded in the image signal, or in case of the inacceptable judgement the steps as claimed in claim 5 are repeated with different parameters to come to an alternative redistribution mapping function.

8. The method as claimed in claim 7 ,
wherein the pixel colors of the lower luminance dynamic range image are encoded as a luma channel,
wherein and u′ and v′ color coordinates are calculated as

u
′

=

4
⁢
X

X
+

15
⁢
Y

+

3
⁢
Z

⁢

,

and
⁢

⁢

v
′

=

9
⁢
Y

X
+

15
⁢
Y

+

3
⁢
Z

,

Assignees

• Koninklijke Philips Nv

Inventors

• Van Der Vleuten Renatus Josephus
• Mertens Mark Jozef Willem

Classifications

• G06T5/00 Primary

  Image enhancement or restoration · CPC title

• G06T5/92

  based on global image properties · CPC title

• H04N7/01

  Conversion of standards {, e.g. involving analogue television standards or digital television standards processed at pixel level} · CPC title

• H04N19/60

  using transform coding · CPC title

• H04N19/46

  Embedding additional information in the video signal during the compression process (H04N19/517, H04N19/68, H04N19/70 take precedence) · CPC title