Method of and apparatus for establishing highlight and shadow densities

• US 5,053,888 A
• Filed: 03/31/1989
• Issued: 10/01/1991
• Est. Priority Date: 03/31/1988
• Status: Expired due to Term

• Current Assignee: SCREEN Holdings Co., Ltd.
• Original Assignee: SCREEN Holdings Co., Ltd.
• Inventors: Nomura, Akihiro
• Primary Examiner(s): Orsino, Joseph A.
• Assistant Examiner(s): LEE, TOMMY D
• Application Number:
  US07/332,466
• Time in Patent Office: 914 Days
• Field of Search: 358/458, 358/455, 358/456, 358/457, 358/461, 358/464, 382/1, 382/18, 382/51, 382/54, 364/525, 364/526
• US Class Current: 358/3.22
• CPC Class Codes: H04N 1/4074 using histograms

Abstract

Highlight density (X_H) (or shadow density) is established on the basis of a cumlative histogram (H(X)) of density of an original. Computed first are reference densities (X_HR, X_HL and X_Hl -X_Hn) associated with prescribed reference statistic values (Y_HR, Y_HL and Y_Hl -Y_Hn), respectively, and threshold values (W_Hl -W_Hn). Then, the reference density (X_HR) is compared with a prescribed threshold density (X_HS), and a difference (X_Hi) concerning the reference density (X_Hi) is compared with the threshold value (W_Hi) in a procedure, to thereby select one of procedures for establishing the highlight density.

Claims

1. A method of establishing a gradation control density X_c employed as a highlight density or a shadow density on a gradation correction curve F(X) that is used in gradation correction of image data expressing density of each pixel in an image, the method comprising the steps of:

(a) on the basis of said image data, generating a statistical curve Y=S(X) expressing a cumulative density distribution of said image, where said statistical curve Y=S(X) is defined on an X-Y coordinate plane having an X-axis representing density levels and a Y-axis representing a cumulative number of pixels in an arbitrary scale;

(b) providing a threshold density X_TH on said X-axis;

(c) finding a terminal value Y_m on said Y-axis at which said statistical curve Y=S(X) is terminated;

(d) providing a first reference value Y_1R on said Y-axis;

(e) finding a terminal density X_m on said X-axis at which said statistical curve Y=S(X) reaches said terminal value Y_m ;

(f) converting said first reference value Y_1R through said statistical curve Y=S(X), to thereby obtain a first reference density X_1R ;

(g) comparing said first reference density X_1R with said threshold density X_TH, to thereby designate said statistical curve Y=S(X) as a first type corresponding to an extraordinarily-bright image or an extraordinarily-dark image when said first reference density X_1R is farther from said terminal density X_m than is said threshold density X_TH, or as a second type corresponding to an ordinary image other than said extraordinarily-bright image or said extraordinarily-dark image when said first reference density X_1R is closer to said terminal density X_m than is said threshold density X_TH ;

(h) selecting a first procedure for setting said gradation control density X_c at a terminal reference density X_TR predetermined between said terminal density X_m and said first reference density X_1R when said statistical curve Y=S(X) is said first type, or selecting a second procedure other than said first procedure for setting said gradation control density X_c when said statistical curve Y=S(X) is said second type; and

(i) establishing said gradation control density X_c through said first procedure or said second procedure selected at said step (h).

2. A method in accordance with claim 1, wherein:

said terminal reference density X_TR is computed by converting a terminal reference value Y_TR through said statistical curve Y=S(X), where said terminal reference value Y_TR is preset between said terminal value Y_m and said first reference value Y_1R.

3. A method in accordance with claim 2, wherein said second procedure comprises the steps:

of

(a) presetting a first difference D between said gradation control density X_c and a terminal density X_a of said gradation correction curve F(X);

(b) finding a line approximate to an end portion of said statistical curve Y=S(X);

(c) computing a terminal density X_im of said line at which said line reaches said terminal value Y_m ; and

(d) establishing said gradation control density X_c so that said first difference D agrees within a prescribed error with a second difference D_i between said gradation control density X_c and said terminal density X_im of said line.

4. A method in accordance with claim 3, wherein:

said line is a straight line drawn through a first point P₀ and a second point P_i, where said first point P₀ is specified by a combination of a prescribed second reference value Y_2R on said Y-axis and a second reference density X_2R computed by converting said second reference value Y_2R through said statistical curve Y=S(X); and

said second point P_i is specified by a combination of a prescribed third reference value Y_i on said Y-axis and a third reference density X_i computed by converting said third reference value Y_i through said statistical curve Y=S(X), while said third reference value Y_i is preset farther from said terminal value Y_m than said second reference value Y_2R is.

5. A method in accordance with claim 4, wherein:

said third reference value Y_i and said third reference density X_i include a plurality of candidate values Y₁ -Y_n and a plurality of candidate densities X₁ -X_n, respectively;

said plurality of candidate values Y₁ -Y_n are sequentially selected one by one in descending order of differences between said candidate values Y₁ -Y_n and said terminal value Y_m to specify said second point P_i ; and

a candidate density that is obtained by converting a selected candidate value through said statistical curve Y=S(X) is designated as said gradation control density X_c under the condition that said first differences D agrees within said prescribed error with said second difference D_i.

6. A method of establishing a gradation control density X_c employed as a highlight density or a shadow density on a gradation correction curve F(X) that is used in gradation correction of image data expressing density of each pixel in an image, the method comprising the steps of:

(a) on the basis of said image data, generating a statistical curve Y=S(X) expressing a cumulative density distribution of said image, where said statistical curve Y=S(X) is defined on an X-Y coordinate plane having an X-axis representing density levels and a Y-axis representing a cumulative number of pixels in an arbitrary scale;

(b) finding a line approximate to an end portion of said statistical curve Y=S(X);

(c) presetting a first difference D between said gradation control density X_c and a terminal density X_a of said gradation correction curve F(X);

(d) finding a terminal value Y_m on said Y-axis at which said statistical curve Y=S(X) is terminated;

(e) computing a terminal density X_im of said line at which said line reaches said terminal value Y_m ; and

(f) establishing said gradation control density X_c so that said first difference D agrees within a prescribed error with a second difference D_i between said gradation control density X_c and said terminal density X_im of said line.

7. A method in accordance with claim 6, wherein

said line is a straight line drawn through a first point P₀ and a second point P_i, where said first point P₀ is specified by a combination of a prescribed first reference value Y_2R on said Y-axis and a first reference density X_2R computed by converting said first reference value Y_2R through said statistical curve Y=S(X), and said second point P_i is specified by a combination of a prescribed second reference value Y_i on said Y-axis and a second reference density X_i computed by converting said second reference value Y_i through said statistical curve Y=S(X), while said second reference value Y_i is preset farther from said terminal value Y_m than said first reference value Y_2R is.

8. A method in accordance with claim 7, wherein:

said second reference value Y_i and said second reference density X_i include a plurality of candidate values Y₁ -Y_n and a plurality of candidate densities X₁ -X_n, respectively;

said plurality of candidate values Y₁ -Y_n are sequentially selected one by one in descending order of differences between said candidate values Y₁ -Y_n and said terminal value Y_m to specify said second point P_i ; and

a candidate density which is obtained by converting a selected candidate value through said statistical curve Y=S(X) is designated as said gradation control density X_c under the condition that said first difference D agrees within said prescribed error with said second difference D_i.

9. An apparatus for establishing a gradation control density X_c employed as a highlight density or a shadow density on a gradation correction curve F(X) that is used in gradation correction of image data expressing density of each pixel in an image, the apparatus comprising:

(a) a first processor means for generating a statistical curve Y=S(X) expressing a cumulative density distribution of said image on the basis of said image data, where said statistical curve Y=S(X) is defined on an X-Y coordinate plane having an X-axis representing density levels and a Y-axis representing a cumulative number of pixels in an arbitrary scale;

(b) a second processor means for finding a terminal density X_m on said X-axis at which said statistical curve Y=S(X) is terminated;

(c) a third processor means for computing a first reference density X_1R by converting a prescribed first reference value Y_1R on said Y-axis through said statistical curve Y=S(X);

(d) a fourth processor means for comparing said first reference density X_1R with a prescribed threshold density X_TH, to thereby designate said statistical curve Y=S(X) as a first type corresponding to an extraordinarily-bright image or an extraordinarily-dark image when said first reference density X_IR is farther from said terminal density X_m than is said threshold density X_TH, or as a second type corresponding to an ordinary image other than said extraordinarily-bright image or said extraordinarily-dark image when said first reference density X_IR is closer to said terminal density X_m than is said threshold density X_TH ; and

(e) a fifth processor means for establishing said gradation control density X_c through a first procedure for setting said gradation control density X_c at a terminal reference density X_TR predetermined between said terminal density X_m and said first reference density X_1R when said statistical curve Y=S(X) is said first type, or through a second procedure other than said first procedure for setting said gradation control density X_c when said statistical curve Y=S(X) is said second type.

10. An apparatus in accordance with claim 9, wherein:

said fifth processor means further comprises sub-processor means for executing said second procedure, said sub-processor means including;

(a) a first sub-processor means for substantially finding a line approximate to an end portion of said statistical curve Y=S(X);

(b) a second sub-processor means for computing a terminal density X_im at which said line is terminated, and

(c) a third sub-processor means for establishing said gradation control density X_c so that a prescribed first difference D between said gradation control density X_c and a terminal density X_a of said gradation correction curve F(X) agrees within a prescribed error with a second difference D_i between said gradation control density X_c and said terminal density X_im of said line.

11. An apparatus for establishing a gradation control density X_c employed as a highlight density or a shadow density on a gradation correction curve F(X) that is used in gradation correction of image data expressing density of each pixel in an image, the apparatus comprising:

(a) a first processor means for generating a statistical curve Y=S(X) expressing a cumulative density distribution of said image on the basis of said image data, where said statistical curve Y=S(X) is defined on an X-Y coordinate plane having an X-axis representing density levels and a Y-axis representing a cumulative number of pixels in an arbitrary scale;

(b) a second processor means for finding a line approximate to an end portion of said statistical curve Y=S(X);

(c) a third processor means for computing a terminal density X_im at which said line is terminated; and

(d) a fourth processor means for establishing said gradation control density X_c so that a prescribed first difference D between said gradation control density X_c and a terminal density X_a of said gradation correction curve F(X) agrees within a prescribed error with a second difference D_i between said gradation control density X_c and said terminal density X_im of said line.