Method for determining an optimum gain response in a spatial frequency response correction for a projection system
View Patent ↗A method for determining an optimum gain response in a spatial frequency correction for a projection system comprises the steps of setting a code value ( 100 ) to a first code value; projecting a flat field image at the code value; capturing the flat field image ( 540 ) with a camera; creating a defect map ( 70 ) of defects in the flat field image; applying the defect map to the flat field image to form a corrected image; setting a gain table ( 80 ) to a first gain for the corrected image; applying the gain table to the corrected image; registering the corrected image; filtering the registered image; measuring the filtered image for a standard deviation ( 570 ); determining whether all gains for producing an under corrected image ( 60 ) and an over corrected image have been exhausted for the code value; if all gains have not been exhausted for the code value, set the gain table to the first gain plus n and go to the step of applying the gain table to the corrected image, if all gains have been exhausted go to determining an optimum gain for the code value.
1. A method for determining an optimum gain response in a spatial frequency response correction for a projection system comprising the steps of:
a) setting a code value to first code value;
b) projecting a flat field image at said code value;
c) capturing said flat field image with a camera;
d) creating a defect map of defects in said flat field image;
e) applying said defect map to said flat field image to form a corrected image;
f) setting a gain table to a first gain for said corrected image;
g) applying said gain table to said corrected image;
h) registering said corrected image;
i) filtering said registered image;
j) measuring said filtered image for a standard deviation;
k) determining whether all gains for producing an under corrected image and an over corrected image have been exhausted for said code value;
l) if all gains have not been exhausted for said code value, set said gain table to said first gain plus n and go to step g), if all gains have been exhausted determine the optimum gain for said code value.
2. A method as in claim 1 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a second-order curve fit for each of said standard deviations at each of said gains;
n) selecting a optimum standard deviation from said second-order curve fit; and
o) determining the optimum gain for said code value from said optimum standard deviation.
3. A method as in claim 1 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a cubic spline curve fit for each of said standard deviations at each of said gains;
n) selecting a optimum standard deviation from said cubic spline curve fit; and
o) determining the optimum gain for said code value from said optimum standard deviation.
4. A method as in claim 2 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
5. A method as in claim 3 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
6. A method as in claim 4 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
7. A method as in claim 5 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
8. A method as in claim 4 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
9. A method as in claim 5 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
10. A method as in claim 1 , finding the optimum gain using a successive approximation method, comprising the additional steps of:
m) selecting a optimum standard deviation from said gain values
n) determining an optimum gain from said optimum standard deviation;
o) determining a standard deviation tolerance from gains adjacent to said optimum gain;
p) if standard deviation tolerance is within a minimum tolerance for said optimum gain then go to r), if standard deviation is not within minimum tolerance for said optimum gain then go to q); and
q) set new gain increment n for step 1 ), and reset first gain as in step f) to said optimum gain minus sum of all new n gain increments divided by 2. Go to g) for continued successive approximation approach to determine optimum gain.
11. A method as in claim 10 additionally selecting gains at multiple said code values, comprising the additional steps of:
r) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to s).
12. A method as in claim 11 to generate a gain table from said gain values for all possible said code values comprising the step of:
s) filling a table of gain versus code value using all values of said gains from all possible said code values.
13. A method as in claim 11 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
s) curve fitting said gain values versus a subset of said code values; and
t) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
14. A method for determining an optimum gain response used in a spatial frequency response correction with high spatial frequency defects for a projection system comprising the steps of:
a) setting a code value to first code value;
b) projecting a flat field image at said code value;
c) capturing said flat field image with a camera;
d) creating a defect map of defects in said flat field image;
e) applying said defect map to said flat field image to form a corrected image;
f) setting a gain table to a first gain for said corrected image;
g) applying said gain table to said corrected image;
h) registering said corrected image;
i) filtering said registered image with a high pass filter;
j) measuring said filtered image for a standard deviation;
k) determining whether all gains for producing an under corrected image and an over corrected image have been exhausted for said code value;
l) if all gains have not been exhausted for said code value, set said gain table to said first gain plus n and go to step g), if all gains have been exhausted determine the optimum gain for said code value.
15. A method as in claim 14 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a second-order curve fit for each of said standard deviations at each of said gains;
n) selecting a minimum standard deviation from said second-order curve fit; and
o) determining the optimum gain for said code value from said minimum standard deviation.
16. A method as in claim 14 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a cubic spline curve fit for each of said standard deviations at each of said gains;
n) selecting a minimum standard deviation from said cubic spline curve fit; and
o) determining the optimum gain for said code value from said minimum standard deviation.
17. A method as in claim 15 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
18. A method as in claim 16 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
19. A method as in claim 17 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
20. A method as in claim 18 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
21. A method as in claim 17 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
22. A method as in claim 18 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
23. A method as in claim 14 , finding the optimum gain using a successive approximation method, comprising the additional steps of:
m) selecting a minimum standard deviation from said gain values;
n) determining an optimum gain from said minimum standard deviation;
o) determining a standard deviation tolerance from gains adjacent to said optimum gain;
p) if standard deviation tolerance is within a minimum tolerance for said optimum gain then go to r), if standard deviation is not within minimum tolerance for said optimum gain then go to q); and
q) set new gain increment n for step 1 ), and reset first gain as in step f) to said optimum gain minus sum of all new n gain increments divided by 2. Go to g) for continued successive approximation approach to determine optimum gain.
24. A method as in claim 23 additionally selecting gains at multiple said code values, comprising the additional steps of:
r) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to s).
25. A method as in claim 24 to generate a gain table from said gain values for all possible said code values comprising the step of:
s) filling a table of gain versus code value using all values of said gains from all possible said code values.
26. A method as in claim 24 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
s) curve fitting said gain values versus a subset of said code values; and
t) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
27. A method for determining an optimum gain response used in a spatial frequency response correction with low spatial frequency defects for a projection system comprising the steps of:
a) setting a code value to first code value;
b) projecting a flat field image at said code value;
c) capturing said flat field image with a camera;
d) creating a defect map of defects in said flat field image;
e) applying said defect map to said flat field image to form a corrected image;
f) setting a gain table to a first gain for said corrected image;
g) applying said gain table to said corrected image;
h) registering said corrected image;
i) filtering said registered image with a low pass filter;
j) measuring said filtered image for a standard deviation;
k) determining whether all gains for producing an under corrected image and an over corrected image have been exhausted for said code value;
l) if all gains have not been exhausted for said code value, set said gain table to said first gain plus n and go to step g), if all gains have been exhausted determine the optimum gain for said code value.
28. A method as in claim 27 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a second-order curve fit for each of said standard deviations at each of said gains;
n) selecting a zero crossing standard deviation from said second-order curve fit; and
o) determining the optimum gain for said code value from said zero crossing.
29. A method as in claim 27 , finding the optimum gain using a curve fit method, comprising the additional steps of:
m) determining a cubic spline curve fit for each of said standard deviations at each of said gains;
n) selecting a zero crossing standard deviation from said cubic spline curve fit; and
o) determining the optimum gain for said code value from said zero crossing.
30. A method as in claim 28 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
31. A method as in claim 29 additionally selecting gains at multiple said code values, comprising the additional steps of:
p) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to q).
32. A method as in claim 30 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
33. A method as in claim 31 to generate a gain table from said gain values for all possible said code values comprising the step of:
q) filling a table of gain versus code value using all values of said gains from all possible said code values.
34. A method as in claim 30 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
35. A method as in claim 31 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
q) curve fitting said gain values versus a subset of said code values; and
r) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
36. A method as in claim 27 , finding the optimum gain using a successive approximation method, comprising the additional steps of:
m) selecting a nearest zero standard deviation from said gain values;
n) determining an optimum gain from said nearest zero standard deviation;
o) determining a standard deviation tolerance from gains adjacent to said optimum gain;
p) if standard deviation tolerance is within a minimum tolerance for said optimum gain then go to r), if standard deviation is not within minimum tolerance for said optimum gain then go to q); and
q) set new gain increment n for step l), and reset first gain as in step f) to said optimum gain minus sum of all new n gain increments divided by 2. Go to g) for continued successive approximation approach to determine optimum gain.
37. A method as in claim 36 additionally selecting gains at multiple said code values, comprising the additional steps of:
r) if all code values have not been used to determine said gain for said code value, set said code value to said first code value plus n and go to step b), if all code values have been used to determine said gain for all said code values go to s).
38. A method as in claim 37 to generate a gain table from said gain values for all possible said code values comprising the step of:
s) filling a table of gain versus code value using all values of said gains from all possible said code values.
39. A method as in claim 37 to generate a gain table from said gain values for a subset of all possible code values comprising the steps of:
s) curve fitting said gain values versus a subset of said code values; and
t) filling a table of gain versus code value using all said curve fit gain values for all possible code values.
40. A method as in claim 1 where said projection system is used for printing on photo sensitive print media.
41. A method as in claim 40 where photo sensitive print media is Silver Halide based.
42. A method as in claim 14 where said projection system is used for printing on photo sensitive print media.
43. A method as in claim 42 where photo sensitive print media is Silver Halide based.
44. A method as in claim 27 where said projection system is used for printing on photo sensitive print media.
45. A method as in claim 44 where photo sensitive print media is Silver Halide based.
46. A method as in claim 1 where said projection system is used for digital cinema projection.
47. A method as in claim 14 where said projection system is used for digital cinema projection.
48. A method as in claim 27 where said projection system is used for digital cinema projection.
49. A method as in claim 1 where said projection system is used for an immersive imaging display.
50. A method as in claim 14 where said projection system is used for an immersive imaging display.
51. A method as in claim 27 where said projection system is used for an immersive imaging display.