JPS616980A - Binarization processing method - Google Patents

Abstract

PURPOSE:To eliminate false contour while making the change in a recording pattern not visually eminent by adding a signal whose polarity of the level is fluctuated in bipolarity to a picture element level of each picture element, comparing the result with each figure of a dither matrix and applying binary conversion. CONSTITUTION:A picture signal (m) of a picture is branched by a branching element 5, one is inputted to a function generator 1 using the picture data (m) as a variable. An output signal of the function generator 1 and a picture signal (r) whose level is changed in bipolarity are multiplied by a multiplier 2 and the output is added to the picture signal (m) at an adder 3 to obtain binary-coded outputs Yi, (j) by an arithmetic unit 4 of the organizing dither method. Thus, the change in the recording pattern in the unit of a dither matrix when the number of recording picture elements is comparatively less is made not eminent visually.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a binarization processing method for improving the halftone reproducibility of grayscale images in a binary recording apparatus using the Infusiera recording method or the thermal transfer method.

Conventional Structure and Problems The systematic dither method is widely known as a halftone processing method for binary recording devices. This dither method determines the presence or absence of a recording pixel by comparing each numerical value of a dither matrix, which has the number of rows and columns according to the number of gradations to be reproduced, with the pixel level of each pixel of image information, and then distributes the discrete distribution of recording pixels. It expresses intermediate tones depending on the state.

In other words, as shown in FIG. 1, an image signal x containing grayscale information
01. and each threshold value C0 of the dither matrix C.

and v comparison t, , x□2. >Cil, throaty VJ,
When Y＞, 4 Ka' K rx, , <c, Yi, 9 becomes O.'' Divide into two.

1+] ”+1 The presence or absence of a recording pixel is determined according to Y of and.

”+1 Therefore, if each recorded pixel is viewed from a distance that is not visually noticeable, a recorded image that approximates the shading of the original image can be obtained.0 This dithering method is a simple and relatively accurate halftone reproduction method. Good gradation can be obtained, but since the gradation is determined independently of the image signal of the image, depending on the image, the turns of the dither matrix may be visually noticeable.Also, When there are relatively few recorded pixels in the area of the dither matrix unit, changes in the recorded pattern are easily noticeable and this occurs as false contours.Objective of the InventionThe present invention improves the problems of the systematic dither method as described above. The purpose is to improve visual gradation characteristics.

Structure of the Invention The present invention adds, to the pixel level m of each pixel of an image signal containing grayscale information, the product of a signal γ whose level fluctuates between positive and negative, and a function with m as a variable (-1). DESCRIPTION OF EMBODIMENTS Below, embodiments of the present invention will be described in detail.

γ is added to m as shown in equation (1), and systematic dithering is applied to this signal dm as shown in FIG.

dm = m 10F (m) X r
−・−(1) F(m) is a function with m as a variable,
For example, it can be expressed as equation (2).

F (m) = aX (MmaX-m)b-=
(2) Here, 0≦m≦Mmax a and b are arbitrary coefficients, where Mmax=1.0. When a=0.5 and γ is a uniform random number, the envelope of dm becomes as shown in FIG.

Solid line for b-1, dashed line for b-2, b-3
The case is shown by a broken line. When the pixel level, that is, the value of m, is large, the influence of the additional signal on the pixel signal m of the image is small when binarized, and conversely, when the pixel level is small, that is, the number of recorded pixels is small when viewed in dither matrix units. This indicates that the influence after binarization processing is sometimes large. Therefore, it is possible to make dithering (1) changes in the recording pattern in units of blocks visually less noticeable when the number of recorded pixels is relatively small.

The signal γ can be a uniform random number between -1 and 1, a signal whose polarity is inverted every fixed number of pixels, a signal that gives one level value of a uniform random number in one cycle of positive or negative polarity, or a fixed number of uniform random numbers. Various signals can be used, such as a sine wave with a period for each pixel. FIG. 3 shows a case where the polarity is reversed for each pixel and one level value of a uniform random number is given in one cycle of positive and negative polarity. In the figure, a, a',...
・b, b',...C, C',... are uniform random numbers, and O≦a, a',...

b, b',...C, C',...≦1.

The dm obtained in this way is compared with each darkness value of the dissa matrix C to obtain a binarized output signal. FIG. 4 is a block diagram for carrying out the above calculation. A signal m is branched by a branching element 5, and one of the signals is inputted to a function generator 1 which uses pixel data m as a variable. The function generator 1 generates the function F(m) of m in equation (2). The function generator 1 uses a dedicated arithmetic unit, or
Alternatively, it can be implemented as a ROM and configured as a table reference system. The output signal γ of the function generator 1 is multiplied by the multiplier 2, and this output is added pixel by pixel by the adder 3, and the binarized output Y is obtained by the systematic dither method operator 4. , , is obtained. As mentioned above, if the signal γ is a uniform random number between -1 and 1 inch, or if the polarity is reversed by 1 for every fixed number of pixels, the signal γ will change to 1 of the uniform random number in one cycle of positive or negative polarity. There are various cases, such as a case in which one level value is given, and a case in which a sine wave is given with a period for each fixed number of pixels. In either case, the number of recorded pixels is relatively small. In the case of /1, the change in the recording pattern in dither matrix units can be made visually inconspicuous.

The recording quality depends on how the function F(m) of m in equation (2) is set, that is, the magnitude of the setting values of a, b2Mmax, etc.
, and how the signal γ is selected, these depend on the type of image,
By setting the optimum value depending on the content, it is possible to obtain image quality with visually good gradation.

Effects of the Invention As described above, the present invention adds a signal whose level polarity fluctuates between positive and negative to the pixel level of each pixel in an image signal, and performs binary conversion by comparing the signal with each numerical value of the dither matrix. With the binarization processing method, changes in the recorded pattern are visually inconspicuous, and it is possible to obtain a high-quality binarized halftone image without false contours.

[Brief explanation of drawings]

Fig. 1 is a pattern perspective view explaining the principle of the conventional systematic dithering method, Fig. 2 is a signal characteristic diagram in an embodiment of the binarization processing method according to the present invention, and Fig. 3 is a random number used in the present invention. FIG. 4 is a pattern perspective view showing an embodiment of the signal, and FIG. 4 is a block diagram showing an embodiment of the binarization processing method according to the present invention. 1...Function generator, 2-...Multiplier, 3...
- Adder, 4... Arithmetic unit, 5... Branch. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure υ 1 Figure 2 Figure 3 Figure 4

Claims (6)

[Claims] (1) Pixel level m of each pixel of the image signal including grayscale information
After adding the product of the signal γ whose level fluctuates between -n and n (n is a positive integer) and the function F(m) with m as a variable, the binary conversion is performed by comparing it with a dither matrix. A binarization processing method characterized by performing. (2) The function F(m) is F(m)=a×(M_m_a_x-m)^bwhere, 0
≦m≦M_m_a_x; The binarization processing method according to claim 1, wherein a and b are expressed by arbitrary coefficients. (3) The binarization processing method according to claim 1, wherein the signal γ is a uniform random number varying between -1 and 1. (4) The binarization processing method according to claim 1, wherein the signal γ is a uniform random number whose polarity is reversed for every fixed number of pixels. (5) The binarization processing method according to claim 1, wherein the absolute value of the level of the signal γ is fixed at one level of a uniform random number during one cycle of positive and negative polarities. (6) The binarization processing method according to claim 1, wherein the signal γ is a sine wave having a period of a fixed number of pixels.