Claims
- 1. A method of producing a reproduced image from an original image the method including generating a gradation correction curve Q=G(x) and using the same in a gradation correction circuit for gradation correction of the original image on an x-Q plane, where x is a first variable representing optical densities and Q is a second variable representing corrected optical densities or halftone dot percentages corresponding to the corrected optical densities, said method comprising the steps of:
- (a) preparing a plurality of sample images adapted for the same type of application as said original image, and generating data signals representative of such sample images, and in response to said data signals:
- (b) obtaining a plurality of Y-Q curves on a Y-Q plane through which optical densities corresponding to Y-values are converted into Q-values in experimental gradation correction rules applied to said plurality of sample images, respectively, where Y is a third variable representing accumulated occurrence of optical densities,
- (c) finding coordinate values (Y.sub.r, Q.sub.r) representing a region on said Y-Q plane at which said plurality of Y-Q curves concentrate,
- (d) defining a reference Y-value and a reference Q-value by components Y.sub.r and Q.sub.r in said coordinate values (Y.sub.r, Q.sub.r), respectively,
- (e) generating data signals representative of said original image, and with a histogram counting circuit, obtaining an accumulated histogram Y=h(x) of optical densities with respect to said original image,
- (f) converting said reference Y-value Y.sub.r into a reference x-value x.sub.r through a curve expressing said accumulated histogram Y=h(x),
- (g) generating said gradation correction curve Q=G(x) on said x-Q plane so that said gradation correction curve Q=G(x) passes through a point having coordinate values (x.sub.r, Q.sub.r) thereon, and storing data representative of said gradation correction curve in a memory in said gradation correction circuit, and
- (g') producing said reproduced image at least in part on the basis of said gradation correction curve.
- 2. A method of claim 1, wherein
- the step (b) includes the steps of:
- (b-1) obtaining a plurality of experiential x-Q curves for said plurality of sample images, respectively,
- (b-2) obtaining a plurality of accumulated histograms representing relations between the variables x and Y in said plurality of sample images, respectively, and
- (b-3) combining said plurality of experiential x-Q curves with curves expressing said plurality of accumulated histograms to obtain said plurality of Y-Q curves, respectively.
- 3. A method of claim 2, wherein
- first and second areas are defined on said Y-Q plane, where said first area has Y-values smaller than Y-values in said second are,
- the step (c) includes the steps of:
- (c-1) finding first coordinate values (Y.sub.r1, Q.sub.r1) representing a first region which exists in said first area and at which said plurality of Y-Q curves concentrate, and
- (c-2) finding second coordinate values Y.sub.r2, Q.sub.r2) representing a second region which exists in said second area and at which said plurality of Y-Q curves concentrate,
- the step (d and (f) are conducted for each of said first and second coordinate values (Y.sub.r1, Q.sub.r1) and (Y.sub.r2, Q.sub.r2), whereby first and second reference x-values x.sub.r1 and x.sub.r2 are obtained, and
- said gradation curve Q=G(x) is generated in the step (g) so as to pass through both of tow points which have coordinate values (x.sub.r1, Q.sub.r1) and (x.sub.r2, Q.sub.r2) on said x-Q plane, respectively.
- 4. A method of claim 3, wherein
- the step (c) includes the steps of:
- (c-3) dividing each of said first and second areas into a plurality of cells, and
- (c-4) drawing said plurality of Y-Q curves on said Y-Q plane,
- the steps (c-1) includes the steps of:
- (c-11) counting number of Y-Q curves passing through each cell belonging to said first area,
- (c-12) finding a first cell through which a maximum number of Y-Q curves are passing among said plurality of cells belonging to said first area and
- (c-13) finding coordinate values representing a position of said first cell on said Y-Q plane to obtain said first coordinate values (Y.sub.r1, Q.sub.r1), and
- the step (c-2) includes the steps of:
- (c-21) counting number of Y-Q curves passing through each cell belonging to said second area,
- (c-22) finding a second cell through which a maximum number of Y-Q curves are passing among said plurality of cells belonging to said second area and
- (c-23) finding coordinate values representing a position of said second cell on said Y-Q plane to obtain said second coordinate values (Y.sub.r2, Q.sub.r2).
- 5. A method of claim 4, further comprising the steps of:
- (h) determining lower and upper limit values of said third variable Y,
- (i) comparing said values Y.sub.r1 and Y.sub.r2 with said lower and upper limit values, respectively,
- (j) adding a first predetermined value .delta..sub.1 to said value Y.sub.r1 to obtain a first corrected value Y.sub.r1c when said value Y.sub.r1 is equal to said lower limit value, and
- (k) subtracting a second predetermined value .delta..sub.2 from said value Y.sub.r2 to obtain a second corrected value Y.sub.r2c when said value Y.sub.r2 is equal to said upper limit value,
- wherein the step (f) and (g) are conducted using said first and second corrected values Y.sub.r1c and Y.sub.r2c in place of said values Y.sub.r1 and Y.sub.r2, respectively.
- 6. A method of claim 5, wherein
- the step (g) includes the steps of:
- (g-1) preparing a gradation correction curve Q=F(x) whose shape is arbitrarily determined, and
- (g-2) correcting said gradation correction curve, Q=F(x) so as to pass through said two points having said coordinate values (x.sub.r1, Q.sub.r1) and (x.sub.r2, Q.sub.r2), respectively, to thereby obtain said gradation curve Q=G(x).
- 7. A method of producing a reproduced image from an original image the method including generating a gradation correction curve Q=G(x) and using the same in a gradation correction circuit for gradation correction of the original image on an x-Q plane, where x is a first variable representing optical densities and Q is a second variable representing corrected optical densities or halftone dot percentages corresponding to the corrected optical densities,
- (a) generating data signals representative of said original image, an with a histogram counting circuit, obtaining an accumulated histogram Y=h(x) of optical densities on said original image, where Y is a third variable representing accumulated occurrence of optical densities,
- (b) finding a critical x-value x.sub.m at which said accumulated histogram Y=h(x) reaches a limit value of said third variable Y,
- (c) converting a given reference Y-value Y.sub.0 to a reference x-value x.sub.0 through a curve expressing said accumulated histogram Y=h(x),
- (d) finding a degree of inclination of said curve expressing said accumulated histogram Y=h(x) in a section between said critical x-value x.sub.m and said reference x-value x.sub.0, to obtain a quantity expressing said degree of inclination,
- (e) converting said quantity to a Q-value according to a predetermined conversion rule which is obtained experimentally, to thereby obtain a reference Q-value Q.sub.0,
- (f) generating said gradation correction curve Q=G(x) so that said gradation correction curve Q=G(x) passes through a point having coordinate values (x.sub.0, Q.sub.0) on an x-Q plane, and storing data representative of said gradation correction curve in a memory in said gradation correction circuit, and
- (g') producing said reproduced image at least in part on the basis of said gradation correction curve.
- 8. A method of claim 7, wherein
- the step (d) includes the step of:
- (d-1) calculating a difference .DELTA.x.sub.0 between said reference x-value x.sub.0 and said critical x-value x.sub.m, said difference .DELTA.x.sub.0 being inversely proportional to said degree of inclination,
- said method further comprises the step of:
- (g) determining said conversion rule in the form of a conversion function Q=f(.DELTA.x), where .DELTA.x is a parameter representing a difference between said first variable x and said critical x-value x.sub.m, and
- the step (e) includes the step of:
- (e-1) substituting .DELTA.x=.DELTA.x.sub.0 into said conversion function Q=f(.DELTA.x) to obtain said reference Q-value Q.sub.0.
- 9. A method of claim 8, wherein
- the step (g) includes the steps of:
- (g-1) preparing a plurality of sample images,
- (g-2) finding a plurality of accumulated histograms of optical densities with respect to said plurality of sample images, respectively,
- (g-3) converting a given Y-value Y.sub.A through A plurality of curves representing said plurality of accumulated histograms, to thereby obtain a plurality of x-values {x.sub.A }
- (g-4) calculating respective differences {.DELTA.x.sub.A } between said plurality of x-values {.DELTA.x.sub.A } and said critical x-value x.sub.m,
- (g-5) determining respective Q-values {Q.sub.A } to which said x-values {x.sub.A } are to be converted according to an experiential gradation corrected rule for said plurality of sample images, respectively.
- (g-6) plotting a plurality of points having coordinate values {(.DELTA.x.sub.A, Q.sub.A)} on a x-Q plane,
- (g-7) finding an approximate function approximately representing a line or curve on said .DELTA.x-Q plane along which said plurality of points distribute, and
- (g-8) defining said conversion function Q=f(.DELTA.x) and said reference Y-value Y.sub.0 by said approximate function and said given Y-value Y.sub.A, respectively.
- 10. A method of claim 9, wherein
- the step (g) further includes the steps of:
- (g-9) calculating a correlation coefficient in distribution of said plurality of points from said line or curve expressing said approximate function on said .DELTA.x-Q plane,
- (g-10) repeating the steps (g-3) through (g-7) and (g-9) for a plurality of Y-values {Y.sub.A } different from each other, to thereby obtain a plurality of approximate functions and a plurality of correlation coefficients, and
- (g-11) comparing said plurality of correlation coefficients with each other to find a maximum correlation coefficient, and
- the step (g-8) includes the steps of:
- (g-81) defining said conversion function Q=f(.DELTA.x) by an approximate function which is selected from said plurality of approximate functions and has said maximum correlation coefficient, and
- (g-82) defining said reference Y-value Y.sub.o by a Y-value which is selected from said plurality of Y-values {Y.sub.A } and for which said maximum correlation coefficient is obtained.
- 11. A method of claim 10, wherein
- The step (g-7) includes the step of:
- applying a least square method to said distribution of said plurality of points, to thereby obtain a linear function as said approximate function.
- 12. A method of claim 11, further comprising the step of:
- (h) repeating the steps (b) through (e) and (g) for a first critical x-value x.sub.m1 at which said accumulated histogram Y=h(x) rises from lower limit value of said variable Y and a second critical x.sub.m2 at which said accumulated histogram Y=h(x) reaches an upper limit value of said variable y while using different values for said plurality of Y-values {Y.sub.A } for each repetition, whereby two sets of coordinate values (x.sub.01, Q.sub.01) and (x.sub.02, Q.sub.02) are obtained,
- the steps (f) including the steps of:
- (f-1) generating said gradation correction curve Q=G(x) so that said gradation correction curve Q=G(x) passes through two points having coordinate values (x.sub.01, Q.sub.01) and (x.sub.02, Q.sub.02), respectively.
- 13. A method of claim 12, wherein
- the step (f-1) includes the steps of:
- (f-11) preparing a gradation correction curve Q=F(x) whose shape is arbitrarily determined, and
- (f-12) correcting said gradation correction curve, Q=F(x) so as to pass through said two points having said coordinate values (x.sub.01, Q.sub.01) and (x.sub.02, Q.sub.02), respectively, to thereby obtain said gradation curve Q=G(x).
- 14. A method of producing a reproduced image from an original image the method including generating a gradation correction curve Q=G(x) and using the same in a gradation correction circuit for gradation correction of the original image, where x is a first variable representing optical densities and Q is a second variable representing corrected optical densities or halftone dot percentages corresponding to the corrected optical densities, said method comprising the steps of:
- (a) detecting a density distribution on said original image to obtain a value .DELTA.x representing a density range of said original image,
- (b) generating a model curve Q=K(x) on an x-Q plane by determining a curvature of said model curve Q=K(x) according to said value .DELTA.x, wherein a sign of said curvature of said model curve is variable depending on said density range of said original image,
- (c) correcting said model curve Q=K(x) so as to pass through at least one point which is given on said x-Q plane, to generate said gradation correction curve Q=G(x), and storing data representative of said gradation correction curve in a memory in said gradation correction circuit, and
- (c') producing said reproduced image at least in part on the basis of said gradation correction curve.
- 15. A method of claim 14, further comprising the steps of:
- (d) preparing a plurality of reference curves {Q=F(x)} whose curvatures are different from each other,
- (e) determining a conversion rule for converting a first parameter representing given density ranges into a second parameter representing composing rates reference curves,
- the steps (b) including the steps of:
- (b-1) converting said value .DELTA.x according to said conversion rule to obtain a composing rate f(.DELTA.x) corresponding to said value .DELTA.x, and
- (b-2) composing said plurality of reference curves {Q=F(x)} with each other at said composing rate f(.DELTA.x) to thereby obtain said model curve Q=K(x).
- 16. A method of claim 15, wherein
- said plurality of reference curves {Q=F(x)} has first and second reference curves Q=F.sub.a (x), Q=F.sub.b (x) which are selected from a group consisting of an upward convex curve, a downward convex curve, and a straight line defined on an oriented x-Q plane which is oriented so that an x-axis is a horizontal axis and a Q-axis is a vertical axis directed upwardly.
- 17. A method of claim 16, wherein
- said conversion function is:
- a first constant for values in a first section of said first parameter,
- a second constant different form said first constant for values in a second section of said first parameter which is apart from said first section, and a non-constant function for values in a third section of said first parameter defined between said first and second sections.
- 18. A method of claim 17, wherein
- the steps (a) includes the steps of:
- (a-1) obtaining an accumulated histogram of optical densities on said original image to generate a curve Y=h(x) representing said accumulated histogram, where Y is a third variable representing accumulated occurence of optical images,
- (a-2) converting given two Y-values by said curve Y=h(x) to obtain two x-values corresponding to said given two Y-values, and
- (a-3) calculating a difference between said two x-values to obtain said value .DELTA.x.
- 19. A method of claim 18, wherein
- said gradation correction curve Q=G(x) is generated for a positive reproduction of said original image, and
- said plurality of reference curves {Q=F(x)} are composed with each other so that said gradation correction curve Q=G(x) is:
- a downward convex curve on said oriented x-Q plane when said value .DELTA.x is smaller than a predetermined a critical value .DELTA.x.sub.C, and
- an upward convex curve on said oriented x-Q plane when said value .DELTA.x is larger than said critical value .DELTA.x.sub.C.
- 20. A method of claim 18, wherein
- said gradation correction curve Q=G(x) is generated for a negative reproduction of said original image, and
- said plurality of said reference curves {Q=F(x)} are composed with each other so that said gradation correction curve Q=G(x) is:
- an upward convex curve on said oriented x-Q plane when said value .DELTA.x is smaller than a predetermined critical value .DELTA.x.sub.C, and
- a downward convex curve on said oriented x-Q plane when said value .DELTA.x is larger than said critical value .DELTA.x.sub.C.
- 21. A method of producing a reproduced image from an original image the method including generating a gradation correction curve Q=G(x) and using the same in a gradation correction circuit for gradation correction of an original image, where x is a first variable representing optical densities and Q is a second variable representing corrected optical densities or halftone dot percentages corresponding to the corrected optical densities, said method comprising the steps of:
- (a) generating a first gradation curve Q=F(x) passing through designated highlight and shadow points on an x-Q plane which is oriented so that an x-axis is a horizontal axis and a Q-axis is a vertical axis being directed upward,
- (b) designating a saturation critical density range CR of said first variable x which includes a saturation critical density value x.sub.c at which said first gradation correction Q=F(x) reaches a saturation value of said second variable Q,
- (c) generating a correction curve Q=G.sub.c (x) satisfying all of the following conditions (i) through (iii):
- (i) said correction curve Q=G.sub.c (x) is defined in said saturation critical density range CR,
- (ii) said correction curve Q=G.sub.c (x) changes more gradually than said first gradation correction curve Q=F(x) in said saturation critical density range CR,
- (iii) said correction curve Q=G.sub.c (x) is smoothly connectable to said first gradation correction curve Q=F(x) at a terminal point of said saturation critical density range CR,
- (d) replacing a part of said first gradation correction curve Q=F(x) located in said saturation critical density range CR by said correction curve Q=G.sub.c (x) in order to correct said first gradation correction curve Q=F(x), whereby a second gradation correction curve serving as said gradation correction curve Q=G(x) is generated, and storing data representative of said second gradation correction curve in a memory in said gradation correction circuit, and
- (d') producing said reproduced image at least in part on the basis of said gradation correction curve.
- 22. A method of claim 21, wherein
- the step (b) includes the steps of:
- (b-1) finding a limit density value x.sub.m designating a limit of optical densities in said original image, and
- (b-2) designating said saturation critical density range CR so as to include both of said saturation critical density value x.sub.C and said limit density value x.sub.m.
- 23. A method of claim 22, wherein
- the step (b-2) includes the step of:
- (b-21) finding a value x.sub.T which is deviated from said limit density value x.sub.m by a predetermined deviation .DELTA.x, and
- (b-22) determining said saturation critical density range CR So that one of terminal points of said range CR has said value x.sub.T, and
- the condition (iii) includes the condition that said correction curve Q=G.sub.C (x) is smoothly connectable to said first gradation correction curve Q=F(x) at both terminal points of said range CR.
- 24. A method of claim 21, wherein
- the step (b) includes the step of:
- (b-1) designating said saturation critical density range CR so that a first terminal point of said range CR is located at a limit point of said first variable x and a second terminal point of said range CR is located at a point other than said limit point of said first variable x, and
- said terminal point in said condition (iii) is said second terminal point.
- 25. A method of claim 21, wherein said terminal point said condition (iii) is one of said highlight and shadow points.
- 26. A method of claim 25, wherein
- said correction function Q=G.sub.C (x) has no inflection point in said saturation critical density range CR.
- 27. A method of claim 26, wherein
- said correction function Q=G.sub.C (x) is a polynomial function of said first variable x.
- 28. A method of claim 26, wherein
- said correction function Q=G.sub.C (x) is an integral of an exponential function of said first variable x.
Priority Claims (4)
Number |
Date |
Country |
Kind |
63-165364 |
Jun 1988 |
JPX |
|
63-165365 |
Jun 1988 |
JPX |
|
63-165366 |
Jun 1988 |
JPX |
|
63-208820 |
Aug 1988 |
JPX |
|
Parent Case Info
This is a continuation of Application Ser. No. 07/373,090 filed Jun. 28, 1989, now abandoned.
US Referenced Citations (3)
Number |
Name |
Date |
Kind |
4472736 |
Ushio et al. |
Sep 1984 |
|
4791678 |
Iwase et al. |
Dec 1988 |
|
4792979 |
Nomura et al. |
Dec 1988 |
|
Continuations (1)
|
Number |
Date |
Country |
Parent |
373090 |
Jun 1989 |
|