JPH0329466A - Facsimile equipment - Google Patents

Abstract

PURPOSE:To prevent generation of a pattern specific to binarization and to avoid deterioration in the resolution for a line drawing and a character by adaptively changing a comparator threshold level in the case of quantizing an output picture data in which a quantized error around an area of a noted picture data is fed back to a multi-gradation digital picture data noted. CONSTITUTION:An intermediate tone original of multigradation is read by a read section 1, an analog picture data outputted from the read section 1 is read and corrected by a picture signal correction section 2 based on a value referenced from a gamma correction table 3 in response to the characteristic of a reader and recorder and converted into a 6-bit multi-gradation digital picture data by an A/D converter section 4. The multi-gradation digital picture is binarized by a pseudo intermediate binarizing section 5 and a binarized data is sent to a coding section 7. Then the threshold level decided adaptively when the signal is binarized by the pseudo intermediate binarizing section 5 is inputted from a threshold level decision section 6 to the pseudo intermediate binarizing section 5. Thus, in the case of binarizing coding, the production of the specific pattern is avoided and sufficient resolution is obtained when a line drawing or a character is binarized.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is applied to a facsimile machine. In particular, it relates to adaptive means for pseudo-halftone binarization using the minimum average error method.

〔overview〕

The present invention provides adaptive pseudo-halftone binarization means using the minimum average error method, by adaptively changing the threshold value for comparison with multi-tone digital image data on which quantization errors are superimposed. This method is designed to prevent the occurrence of patterns and thin lines that are characteristic of the minimum average error method.

[Conventional technology]

In the conventional example, as shown in FIG. 5, a multi-tone halftone original is read by the reading section 18, and the analog image data outputted from the reading section 18 is read and recorded at T according to the characteristics of the recording device.
The analog image data is corrected in the image signal correction section 19 according to the value referenced from the correction table 20, and converted into multi-gradation digital image data in the AD conversion section 21. This multi-gradation digital image data is binarized by a pseudo-halftone binarization unit 22,
Binarized data is sent to the encoding unit 23.

FIG. 6 shows that error data is superimposed based on the following formula before being input to the device 27 where pseudo-halftone binarization is performed in the conventional example, and
, , are input to the comparator 27 as .

l XLj"' X++ , + a+-1+j-1' e
+-1. j-1+ ai-1+j ” Ni-1+J
” aj-1+j+1' fi-1.j+l+ al.
j-1' e+,,-+However, al-In J-1
+ a l-1, J+al-1n j+1+ a i
, j−+ ”” 1 (1) xij
is compared with the fixed threshold value Th output from the fixed threshold value generating section 26, and '/+. Output j.

However, XL. t M A X and Xi, 4 MI
N is multi-gradation digital image data x 1 . , and Th means the maximum and minimum values of conditions Xi. jM I N < T h < X,, J
Satisfy M.A.X.

The 31'l. J is conditional expression (3)
It is binarized based on and outputs zi1.

Quantization error data ei,j to be superimposed on xi,J manually input to the comparator 27 is defined by the following equation.

e1. j=Xi. j'/I. j- (4) Output e
1. , are stored in the error data storage memory 29 and the delay device 24, and e,1 necessary for calculating the error superimposed on xi,J is extracted under the control of the memory control unit 30.

In this way, multi-gradation digital image data Xi,j
Minimize the quantization error for the entire screen of a manuscript when binarizing it.

[Problem that the invention seeks to solve]

In such a conventional example, appropriate reproducibility of the gradation level of the thick conversation can be obtained, but a peculiar pattern occurs when binarized, and it is also The disadvantage is that sufficient resolution cannot be obtained.

SUMMARY OF THE INVENTION An object of the present invention is to eliminate such drawbacks, and to provide a facsimile machine that prevents the occurrence of patterns peculiar to binarization and that does not impair the resolution of line drawings, characters, etc.

[Means for solving problems]

The present invention provides a comparator that compares pixel values included in second image data, in which error data is superimposed on first image data corresponding to a multi-tone halftone original, with a given threshold; a binarization unit that performs a binarization process on the comparison result value; and an error data generation means that generates error data based on one of the pixels of the first image data and the comparison result value of the comparator corresponding to this one pixel. In a facsimile machine equipped with
An edge detection unit temporarily stores pixels in the surrounding area of one pixel of interest included in the first pixel data, and extracts pixels in the area surrounding this pixel of interest, and an edge detection unit that extracts pixels in the area surrounding this pixel of interest, and a value of the pixel of interest extracted by this edge detection unit. an edge parameter determination unit that generates a correction value corrected by a weighting coefficient for quantization error; a random number generator that generates a random number; a random number generated by this random number generator and a correction value generated by the edge parameter determination unit; The present invention is characterized by comprising a calculation means for determining the value of the difference between and providing the calculated value to the comparator as a threshold value.

[Effect]

Input a multi-gradation halftone original into the reading unit, read the analog image data, and
After correction, AD conversion is performed to convert to multi-tone digital image data, and when pseudo halftone binarization is performed using the minimum average error method, fine patterns peculiar to the minimum average error method are removed. In order to do this, a random number generator is used, and black pixels are concentrated on edge areas where density value level changes are significant, and where humans are strongly impressed.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the structure of this embodiment.

As shown in FIG. 1, this embodiment includes a reading section l, an image signal correction section 2, a T correction table 3, an AD conversion section 4,
Pseudo-halftone binarization section 5, threshold value determination section 6, and encoding section 7
In other words, this embodiment compares the values of pixels included in the second image data, in which error data is superimposed on the first image data corresponding to a multi-tone halftone original, with a given threshold value. A comparator 12, a binarization unit 13 that performs binarization processing on the comparison result value of this comparator 12, and a comparison result value of one of the pixels of the first image data and the comparison result of the comparator l2 corresponding to this one pixel. It is equipped with a delay device 8 and a multiplier 9, which are error data generation means for generating error data based on the image data. The image data memory 14
an edge detection unit l5 that temporarily stores pixels in the surrounding area of the pixel of interest, and an edge detection unit that generates a correction value by correcting the value of the pixel of interest extracted by this edge detection unit with a weighting coefficient for quantization error. A parameter determination unit 16;
A random number generator 17 that generates random numbers, and a calculation means that calculates the difference between the random number generated by the random number generator and the correction value generated by the edge parameter determining section and supplies it to the comparator 12 as a threshold. Equipped with.

Next, the operation of this embodiment will be explained. A multi-gradation halftone original is read by the reading unit 1, the analog image data output from the reading unit 1 is read, and the image signal correction unit 2 is based on the value referenced from the γ correction table 3 according to the characteristics of the recording device. The image data is corrected by the AD converter 4 and converted into 6-bit multi-gradation digital image data. This multi-gradation digital image data is binarized by a pseudo-halftone binarization unit 5 and the binarized data is sent to an encoding unit 7. When the pseudo-halftone binarization unit 5 performs binarization, the adaptively determined threshold is input from the dark value determination unit 6 to the pseudo-halftone binarization unit 5 .

FIG. 2 is a block diagram of a circuit in which adaptive pseudo-halftone binarization using the minimum average error method is performed. The 6-bit multi-gradation digital image data X, , J is superimposed with error data before being input to the comparator l2 to become X, , .
The weighting coefficient ai,J for the quantization error calculated by equation (1) is determined to be the value shown in FIG. Regarding the weighting coefficient 31, J, if the value of the four adjacent pixels of the pixel of interest xi,j is increased, a diagonal pattern will occur, while if it is decreased, a vertical and horizontal pattern will occur, so a value that is approximately intermediate therebetween is used.

Xl. J is Di,j outputted from the edge parameter determination unit 16 and R.j outputted from the random number generator l7. 'Ji. Output t.

(5) However, Q<Th=Rt. j D+. ．． + <63 Rl. calculated by the random number generator 17. j is 27≦R. ，，
≦37(6) However, a pseudorandom number is generated that satisfies (average value of R.1)=32.Furthermore, for D.,, which is output from the edge parameter determination unit l6, the manually generated image data .,j are stored in the image data memory 14, and the edge detection section 15 executes the following judgment based on the small areas A, B, C, and D shown in FIG. 4 and the pixel of interest XI,j.

(7) If two or more of the conditions in equation (7) are satisfied, then D, ,, If one is satisfied, DI, If none is satisfied, D, ,,.

'lr. j is output through the comparator 12, and the binarization unit l3 outputs Z under the following conditions.
i'' is output.

In this example, the parameter D. , j to find 1
Line delays occur.

〔Effect of the invention〕

As explained above, the present invention adaptively adjusts the comparator threshold when quantizing the output image data fed back to the multi-gradation digital image data of interest by the quantization error in the peripheral area of the image data of interest. By changing to Further, there is an effect that blurring of thin lines in the scanning line direction can be removed. In addition, it can be realized with a simple structure, and it can reduce the size of facsimile machines.
This also has the effect of contributing to lower prices.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the structure of an embodiment of the present invention. FIG. 2 is a block diagram showing the structure of an adaptive pseudo-halftone binarization unit using the minimum average error method of the present invention. FIG. 3 is a diagram showing a weighting coefficient matrix for quantization errors. FIG. 4 is a diagram showing reference image data for edge detection. FIG. 5 is a block diagram showing the structure of a conventional example. FIG. 6 is a block diagram showing the configuration of an adaptive pseudo-halftone binarization unit using the conventional minimum average error method. 1, 18...reading unit, 2, l9...image signal correction unit,
3, 20... r correction table, 4, 21... AD conversion section, 5, 22... pseudo halftone binarization section, 6... darkness value determination section, 7, 23... code conversion unit, 8, 24...delay unit, 9, 25...multiplier, 10, 29...error data storage memory, 11, 30...memory control unit, 12
, 27... Comparator, l3, 28... Binarization unit, 14
. . . Image data memory, l5 . . . Edge detection section, 16
. . . edge parameter determination section, 17 . . random number generator, 26 . . . fixed threshold generation section. Figure 3: Weighting coefficient matrix of 1 × 1, overall configuration of the embodiment

Claims (1)

[Claims] 1. A comparator that compares the value of a pixel included in second image data on which error data is superimposed on first image data corresponding to a multi-gradation halftone original with a given threshold value; , a binarization unit that performs binarization processing on the comparison result value of this comparator, and generates error data based on one of the pixels of the first image data and the comparison result value of the comparator corresponding to this one pixel. an edge detection unit that temporarily stores pixels in a peripheral area of one pixel of interest included in the first image data and extracts pixels in a peripheral area of the pixel of interest; A facsimile apparatus comprising: an edge parameter determination section that generates a correction value by correcting the value of the pixel of interest extracted by the edge detection section using a weighting coefficient for quantization error;