Method and system for color correction using three-dimensional information

Patent metadata

Field	Value
Publication number	US-9508190-B2
Application number	US-30813708-A
Country	US
Kind code	B2
Filing date	Jun 9, 2006
Priority date	Jun 9, 2006
Publication date	Nov 29, 2016
Grant date	Nov 29, 2016

Abstract

A system and method for the color correction of an image using three-dimensional, geometrical information of the capture environment of the image includes determining geometrical properties of at least a portion of the image and modifying a look of at least the portion of the image by altering a value of at least one of the determined geometrical properties and using image formation theory. In one embodiment of the present invention, the geometrical properties of the image include at least one of light properties, surface color, reflectance properties, and scene geometry of the at least one portion of the image. In accordance with the present invention, the geometrical properties of the image are alternatively determined by using sensing devices, by inferring the geometrical properties from the image itself, or by user input.

First claim

The invention claimed is:

1. A method for color correction of an image, comprising:
selecting a portion of the image to be modified, wherein the selected portion includes a normal map that is continuous and smooth within the regions such that

F
⁡

(

N
⁡

(
x
)

)

=

∫



L
r

⁡

(

x
,

ω
r

)

-

ρ
d

⁡

(
x
)

⁢

∑
i

⁢

(

N
⁡

(
x
)

·

ω
i

⁡

(
x
)

)

⁢

L
i



2

⁢

ⅆ
x

+

λ
⁢

∫



∇
x

⁢

N
⁡

(
x
)



2

⁢

ⅆ
x

and determining at least one geometrical property of the selected portion of the image to be modified by:
(a) estimating a value for at least one of said at least one geometrical property; wherein said estimated value is computed by changing the light source or geometry of said selected portion;
(b) rendering said portion of said image using said estimated value; and
(c) comparing said rendered portion of said image to a corresponding portion of said original image to determine if a match exists therebetween; and
(d) modifying an appearance of said portion of said image by altering a value of the at least one determined geometrical property when a match exists; and
(e) when a match does not exist, re-estimating the value of the at least one of said at least one geometrical property before returning to rendering (b) and repeating subsequent (c)-(e) using the re-estimated value.

2. The method of claim 1 , wherein said at least one geometrical property comprises at least one of light properties, surface color, reflectance properties, and scene geometry of said portion of said image.

3. The method of claim 1 , wherein said at least one geometrical property is determined from a user input.

4. The method of claim 1 , wherein said determining comprises inferring at least one of a light source position and object geometry based on a surface in said portion of said image.

5. The method of claim 4 , wherein said inferring comprises employing an image formation theory.

6. The method of claim 1 , wherein said at least one geometrical property is determined using at least one sensing device.

7. The method of claim 6 , wherein said at least one sensing device is located in a capture environment of said image during im

Assignees

• Zhang Dong-Qing
• Benitez Ana Belen
• Fancher James Arthur
• Thomson Licensing

Inventors

• Zhang Dong-Qing
• Benitez Ana Belen
• Fancher James Arthur

Classifications

• G06T15/50 Primary

  Lighting effects · CPC title

• G06T15/20

  Perspective computation · CPC title

• G06T15/506

  Illumination models · CPC title