Device and method for determining a key for clipping a subject moving against a colored background

• US 5,903,318 A
• Filed: 11/25/1996
• Issued: 05/11/1999
• Est. Priority Date: 11/23/1995
• Status: Expired due to Term

• Current Assignee: Thomson Broadband Systems
• Original Assignee: Thomson Broadband Systems
• Inventors: Le Lan, Michel, Demay, Alain
• Primary Examiner(s): Peng, John K.
• Assistant Examiner(s): Sajous, Wesner
• Application Number:
  US08/755,645
• Time in Patent Office: 897 Days
• Field of Search: 348/587, 348/590, 348/591, 348/592, 348/593, 348/586, 348/588, 348/584, 348/585, 358/22 CK, 358/537, 358/538, 358/540, 345/113, 345/114, 345/168, 345/118, 345/150, 382/284
• US Class Current: 348/592
• CPC Class Codes: H04N 9/75 Chroma key

Abstract

The invention relates to a method for determining a key KD for clipping a subject moving against a colored background, the said key making it possible to separate the color space into three regions, a first region defining a volume representing the colored background, a second region defining a volume representing the subject and a third region defining a transition region between the colored background and the subject, where the device and method defines the volume representing the colored background in the form of a cone of aperture angle α whose axis of symmetry (W) passes through the achromic and zero-luminance point of the color space and a point representing the color of the colored background. The invention also relates to any type of electronic apparatus implementing the abovementioned device and method, such as video mixers or autonomous devices for clipping and overlaying video images.

Claims

1. A method for determining a clipping key, KD, for clipping a subject from a first colored background and overlay the subject on a new colored background, comprising the steps of:

identifying a color space in which said subject and said first colored background are represented, said color space having an achromic and zero-luminance point;

defining a first volume as a first region in said color space corresponding to the first colored background, comprising the step of forming a cone having an aperture angle α

with an axis of symmetry /W/ that passes through the achromic and zero-luminance point of the color space and a point representing a color of the first colored background;

defining a second volume in said color space that is separate from said first volume, said second volume including a second region representing said subject; and

determining said clipping key based on at least one of said first volume and said second volume.

2. The method according to claim 1, wherein said step of defining a first volume comprises the step of:

defining a cross section of the cone in a plane (Uo, Vo) with polar coordinates r(i) and θ

(i) for plural pixels (Mi), comprising the step of identifying a right-handed trihedral (Uo, Vo, W), that complies with an equation,

space="preserve" listing-type="equation">k.sup.2 r(i).sup.2 cos.sup.2 (θ

(i)+R)+r(i).sup.2 sin.sup.2 (θ

(i)+R)=constant,

where θ

(i) is a polar angle of respective of the plural pixels M(i), k is an elliptical modulation factor for the coordinate r(i), and R is an angle of rotation of the cross-section of the cone.

3. The method according to claim 2, wherein said step of defining a first volume comprises the steps of:

linking the aperture angle α

with an adjustment parameter, CLIP, so that r_e (i)-α

w(i)=CLIP, where

space="preserve" listing-type="equation">r.sub.e (i)=r(i) (A).sup.1/2, ##EQU10## and w(i) is a coordinate on the axis of symmetry /W/ of the cone; and

defining a face of the cone opposite a vertex of the cone aperture angle α

with a coordinate wh on the axis of symmetry /w/ of the cone aperture angle α

.

4. The method according to claim 2, wherein said step of defining a cross section of the cone defines the cone with the elliptical modulation factor k set in a range 1≦

k≦

2, and the angle of rotation R set in a range from 0 to π

.

5. The method according to claim 1, wherein the step of defining the second volume comprises the steps of:

defining as a third region a transition region between a surface of the cone of aperture angle α and

another cone of aperture angle α

1 greater than α

that has the same axis of symmetry and the same vertex as the cone of aperture angle α

, and having a face opposite the vertex and in the same plane as the face opposite the vertex of the cone of aperture angle α

; and

defining the subject region as being separate from the cone of aperture angle a and the another cone of aperture angle α

1.

6. The method of claim 5, wherein said step of defining a transition region comprises the step of adjusting a size of the transition region with an adjustment parameter.

7. The method according to claim 1, wherein said step of forming a cone having an aperture angle α

, comprises the step of truncating the cone having an aperture angle α

at a distance, d, so as to separate a low luminance threshold region between a coordinate ws that defines a position of the vertex and axis of the cone of aperture angle α and

a coordinate wb defining the low luminance threshold, where wb is positioned between ws and wh, and wh being an abscissa of a face opposite the vertex of the cone of aperture angle α

.

8. The method according to claim 5, wherein said step of determining a clipping key comprises combining a chrominance key KC and a luminance key KL according to a formula,

9. The method according to claim 1, wherein said step of identifying a color space comprises converting into arithmetic zeros an origin of a frame of reference for the color space which corresponds to respective achromic levels of chrominance components of the color space and a zero level for a luminance component of the color space.

10. The method according to claim 1, further comprising the step of:

transposing a red color difference chrominance component and a blue difference chrominance component of the pixel into a common video format as a luminance component of the pixel, comprising the step of interpolating the red color difference component and the blue color difference component.

11. The method according to claim 10, further comprising the step of:

coding the red color difference chrominance component with N bits, the blue color difference chrominance component with N bits, and the luminance component of the pixel with bits N, where N is an integer greater than a number of bits of video bits for respective of said red color difference chrominance component, blue color difference chrominance component, and the luminance component.

12. The method according to claim 11, wherein said step of coding comprises coding with N equal to 16 bits among, said 16 bits comprising 10 video bits, 4 splitting bits intended to define the accuracy of the video, 1 sign bit and 1 overflow bit.

13. The method according to claim 5, further comprising the steps of:

detecting into which region of the first region, the second region, and the third region the pixel is located; and

producing a signal DZ indicative of the respective region in which the pixel is located.

14. The method according to claim 13, wherein said detecting into which region step comprises detecting if the pixel is located in a shadow region.

15. The method according to claim 14, wherein said detecting into which region step comprises:

comparing a luminance value y corresponding to the pixel with a threshold value ya, if the clipping key, KD, is between 0 and 1;

assigning a first value to a signal DZ signifying that the pixel is located in the transition region, if the luminance value y is greater than the threshold value ya;

assigning a second value to the signal DZ signifying that the pixel is located in the shadow region if the luminance value y is less than threshold value ya and if a polar coordinate angle θ

(i) corresponding to the pixel is between two values θ

min and θ

max whose difference θ

max-θ

min defines an angular aperture within which the angle θ

(i) must lie in order for a hue of the pixel associated with θ

(i) to be likened to a hue of the colored background; and

assigning a third value to the signal DZ signifying the pixel is located in the transition region if the luminance value y is less than the threshold value ya and if the angle θ

(i) is not between the said values θ

min and θ

max.

16. The method according to claim 1, wherein said step of determining said clipping key comprises masking and filtering the clipping key so as to provide a final clipping key /KDF/.

17. A device that forms a clipping key, KD, used to clip a subject moving against a first colored background and overlay the subject on a new colored background, comprising:

means for identifying a color space in which said subject and said first colored background are represented, said color space having an achromic and zero-luminance point;

means for defining a first volume comprising a first region corresponding to the first colored background, comprising means for forming a cone having an aperture angle α

with an axis of symmetry /W/ that passes through the achromic and zero-luminance point of the color space and a point representing a color of the first colored background;

means for defining a second volume in said color space that is separate from said first volume, said second volume including a second region representing said subject; and

means for determining said clipping key based on at least one of said first volume and said second volume.

18. The device of claim 17, further comprising means for elliptically modulating the cross-section of the cone of aperture angle α

.

19. The device of claim 17, wherein said means for defining a second volume comprises means for defining as a third region a transition region positioned between the first region and the second region in a space between a surface of the cone of aperture angle α

and a cone of aperture angle α

1 greater than α

, with the same axis of symmetry and the same vertex as the cone of aperture angle α

, said transition region having a face opposite the vertex and coplanar with a face opposite the vertex of the cone of aperture angle α

.

20. The device of claim 19, wherein said means for defining a first volume comprises means for truncating the cone of aperture angle α

at a distance, d, so as to separate a low luminance threshold region between a coordinate ws that defines a position of the vertex and axis of the cone of aperture angle α and

a coordinate wb defining the low luminance threshold, where wb is positioned between ws and wh, and wh being an abscissa of the face opposite the vertex.

21. The device of claim 17, further comprising:

means for detecting into which region of the first region, the second region, and the third region the pixel is located; and

means producing a signal DZ indicative of the respective region in which the pixel is located.

22. The device according to claim 21, wherein:

said means for detecting detects if the pixel is located in a shadow region of the subject; and

said means for producing produces the signal DZ indicative of the pixel being located in the shadow region when said means for detecting detects that the pixel is located in the shadow region.

23. The device of claim 17, further comprising means for overlaying the clipped subject onto the new colored background.

24. A video mixer configured to form a clipping key, KD, and use the clipping key, KD, to clip a subject from a first colored background and overlay the subject on a new colored background, comprising:

input means for inputting the source video image;

means for identifying a color space in which said subject and said first colored background are represented, said color space having an achromic and zero-luminance point;

means for defining a first volume in said color space comprising a first region corresponding to the first colored background, comprising means for forming a cone having an aperture angle α

with an axis of symmetry /W/ that passes through the achromic and zero-luminance point of the color space and a point representing a color of the first colored background;

means for defining a second volume in said color space that is separate from said first volume, said second volume including a second region representing said subject;

means for determining said clipping key based on at least one of said first volume and said second volume; and

mixing means for mixing the clipping key with the video source image.

25. An autonomous video image clipping and overlaying device, configured to form a clipping key, KD, and use the clipping key, KD, to clip a subject from a first colored background and overlay the subject on a new colored background, comprising:

input means for inputting the source video image;

means for identifying a color space in which said subject and said first colored background are represented, said color space having an achromic and zero-luminance point;

means for defining a first volume in said color space representing a first region corresponding to the first colored background, comprising means for forming a cone having an aperture angle α

with an axis of symmetry /W/ that passes through the achromic and zero-luminance point of the color space and a point representing a color of the first colored background;

means for defining a second volume in said color space that is separate from said first volume, said second volume including a second region representing said subject;

means for determining said clipping key based on at least one of said first volume and said second volume;

means for extracting the subject from the source video image based on the clipping key, KD; and

means for overlaying the subject on the new colored background.