Lens correction using processed YUV data
Lens correction using processed YUV data
 CN 1,617,597 A
 Filed: 10/22/2004
 Published: 05/18/2005
 Est. Priority Date: 11/06/2003
 Status: Active Grant
First Claim
Patent Images
1. , a kind of the view data that is associated with lens is implemented the method for a lens correction, described method comprises:
 If described view data not in a YUV color space, is converted to described YUV color space to form the YUV view data with described view data;
Image processing program is put on described YUV view data to form image processed YUV data;
AndDescribed image processed YUV data is implemented described lens correction.
Chinese PRB Reexamination
Abstract
A lens correction is applied to image data by first converting the image data to a YUV color space, if the image data is not already in the YUV color space. Image processing procedures are applied to the image data in the YUV color space to form image processed YUV data. A corresponding correction value is then applied to each component, i.e., Y, U and V component, of the image processed YUV data.

1 Citation
Image brightness chrominance adjustment method, device and system  
Patent #
CN 104,811,587 A
Filed 04/21/2015

Current Assignee

No References
42 Claims

1. , a kind of the view data that is associated with lens is implemented the method for a lens correction, described method comprises:

If described view data not in a YUV color space, is converted to described YUV color space to form the YUV view data with described view data; Image processing program is put on described YUV view data to form image processed YUV data;
AndDescribed image processed YUV data is implemented described lens correction.


2. method according to claim 1, wherein implement described lens correction and further comprise:

Y component to described image processed YUV data applies a Y corrected value; U component to described image processed YUV data applies a U corrected value;
AndV component to described image processed YUV data applies a V corrected value.


3. method according to claim 1, wherein implement described lens correction and further comprise:

One Y component of described image processed YUV data be multiply by a Y corrected value; A U component and a U corrected value addition with described image processed YUV data;
AndA V component and a V corrected value addition with described image processed YUV data.


4. method according to claim 2, wherein said U corrected value are based on one first distance value, and wherein said first distance value is relevant apart from the position of a reference point of described reference picture with the object pixel in the reference picture.

5.
Whether 5, method according to claim 2, wherein said U corrected value be based on a luminance parameter, wherein be in the brightness range of selecting in advance according to the described Y component of described image processed YUV data and determine described luminance parameter.

6. method according to claim 2, wherein said U corrected value is based on a maximum correction limit value and a minimum limit value of proofreading and correct.

7. method according to claim 6, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct can be selected by the user.

8. method according to claim 6, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct are all based on the character of described lens.

9. method according to claim 4 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rings that the described reference point from described reference picture exhales and calculates described first distance value.

10. method according to claim 4, wherein by making first distance value=Root (Dx*Dx+Dy*Dy) * NormalizeValue calculate described first distance value, wherein:

Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.


11. method according to claim 4 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rectangles that the described reference point from described reference picture exhales and calculates described first distance value.

12. method according to claim 4, wherein (Dx, Dy) * NormalizeValue calculates described first distance value, wherein by making first distance value=max
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.
 abs(HalfXx+XSHIFT)；

13. method according to claim 4 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rhombus that the described reference point from described reference picture exhales and calculates described first distance value.

14. method according to claim 4, wherein first distance value=(Dx+Dy) * NormalizeValue calculates described first distance value, wherein by making
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.
 abs(HalfXx+XSHIFT)；

15. method according to claim 4, wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric polygon that the described reference point from described reference picture exhales and calculates described first distance value, wherein said a plurality of concentric polygons are essentially annular.

16. method according to claim 4, wherein calculate described first distance value by following algorithm:

(if Dx＞
(Dy＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dx(Dy＜
＜
2))＞
＞
3))*NormalizeValueAnd (if Dy＞
(Dx＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dy(Dx＜
＜
2))＞
＞
3))*NormalizeValueAnd if (max (Dx, Dy)＞
(abs (DxDy)＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(max(Dx，
Dy)(abs(DxDy)＜
＜
2)＞
＞
3))*NormalizeValueOtherwise, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy))*NormalizeValueWherein; Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.


17. method according to claim 2, wherein said V corrected value be based on one first distance value, and the position of a reference point of the described reference picture of object pixel distance in wherein said first distance value and the reference picture is relevant.

18.
Whether 18, method according to claim 2, wherein said V corrected value be based on a luminance parameter, wherein be in the brightness range of selecting in advance according to the described Y component of described image processed YUV data and determine described luminance parameter.

19. method according to claim 2, wherein said V corrected value is based on a maximum correction limit value and a minimum limit value of proofreading and correct.

20. method according to claim 19, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct can be selected by the user.

21. method according to claim 19, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct are all based on the character of described lens.

22. method according to claim 17 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rings that the described reference point from described reference picture exhales and calculates described first distance value.

23. method according to claim 17, wherein by making first distance value=Root (Dx*Dx+Dy*Dy) * NormalizeValue calculate described first distance value, wherein:

Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.


24. method according to claim 17 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rectangles that the reference point from described reference picture exhales and calculates described first distance value.

25. method according to claim 17, wherein (Dx, Dy) * NormalizeValue calculates described first distance value, wherein by making first distance value=max
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.
 abs(HalfXx+XSHIFT)；

26. method according to claim 17 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rhombus that the described reference point from described reference picture exhales and calculates described first distance value.

27. method according to claim 17, wherein first distance value=(Dx+Dy) * NormalizeValue calculates described first distance value, wherein by making
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction
 abs(HalfXx+XSHIFT)；

28. method according to claim 17, wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric polygon that the described reference point from described reference picture exhales and calculates described first distance value, wherein said a plurality of concentric polygons are essentially annular.

29. method according to claim 17, wherein calculate described first distance value by following algorithm:

(if Dx＞
(Dy＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dx(Dy＜
＜
2))＞
＞
3))*NormalizeValueAnd (if Dy＞
(Dx＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dy(Dx＜
＜
2))＞
＞
3))*NormalizeValueAnd if (max (Dx, Dy)＞
(abs (DxDy)＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(max(Dx，
Dy)(abs(DxDy)＜
＜
2)＞
＞
3))*NormalizeValueOtherwise, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy))*NormalizeValueWherein; Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.


30. method according to claim 2, wherein said Y corrected value are based on a second distance value, and wherein said second distance value is again based on one first distance value and one or more brightness based on a F value of described lens.

31. method according to claim 2, wherein said Y corrected value are based on a smoothing parameter, and wherein said smoothing parameter can be selected according to a required level and smooth amount by the user.

32. method according to claim 2, wherein said Y corrected value is based on a maximum correction limit value and a minimum limit value of proofreading and correct.

33. method according to claim 32, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct can be selected by the user.

34. method according to claim 32, wherein said maximum correction limit value and the described minimum limit value of proofreading and correct are all based on the character of described lens.

35. method according to claim 30 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rings that the described reference point from described reference picture exhales and calculates described first distance value.

36. method according to claim 30, wherein by making first distance value=Root (Dx*Dx+Dy*Dy) * NormalizeValue calculate described first distance value, wherein:

Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.


37. method according to claim 30 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rectangles that the described reference point from described reference picture exhales and calculates described first distance value.

38. method according to claim 30, wherein (Dx, Dy) * NormalizeValue calculates described first distance value, wherein by making first distance value=max
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.
 abs(HalfXx+XSHIFT)；

39. method according to claim 30 is wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric rhombus that the described reference point from described reference picture exhales and calculates described first distance value.

40. method according to claim 30, wherein first distance value=(Dx+Dy) * NormalizeValue calculates described first distance value, wherein by making
Dx＝  abs(HalfXx+XSHIFT)；
Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.
 abs(HalfXx+XSHIFT)；

41. method according to claim 30, wherein by supposing that object pixel in the described reference picture is arranged in a plurality of concentric polygon that the described reference point from described reference picture exhales and calculates described first distance value, wherein said a plurality of concentric polygons are essentially annular.

42. method according to claim 30, wherein calculate described first distance value by following algorithm:

(if Dx＞
(Dy＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dx(Dy＜
＜
2))＞
＞
3))*NormalizeValueAnd (if D3＞
(Dx＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(abs(Dy(Dx＜
＜
2))＞
＞
3))*NormalizeValueAnd if (max (Dx, Dy)＞
(abs (DxDy)＜
＜
2))Then, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy)+(max(Dx，
Dy)(abs(DxDy)＜
＜
2)＞
＞
3))*NormalizeValueOtherwise, Function_Distance(x，
y)＝
(Dx+Dy+max(Dx，
Dy))*NormalizeValueWherein; Dx＝
abs(HalfXx+XSHIFT)；Dy＝
abs(HalfYy+YSHIFT)；HalfX is half of the length of described reference picture on a directions X;
AndHalfY is half of the width of described reference picture on a Y direction.

Specification(s)