JPH0481382B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a device that converts the pixel density of an image expressed in halftone or pseudo-halftone to adaptively binarize it.

(Prior Art and its Problems) Halftone processing, systematic dithering, and the like are often used as means for pseudo-expressing images containing continuous gradations. In addition to this, error diffusion method (Error
Diffusion) is M.R. Schroeder (M.Diffusion).
Proposed by R. Schroeder). (“Images
from Computers”IEEE Spectrum, vol.6,
1969) This saves the error between input and output when pixels surrounding the pixel of interest are binarized, and reflects that error when binarizing the pixel of interest to match the average brightness level between the input and output images. This is to create a pseudo-halftone effect.

Halftone printing is widely used as a pseudo-halftone expression in general printing, but when a halftone photograph is input from a facsimile or the like, moiré may occur. This becomes noticeable when the halftone dot period of the input image and the sample period of the image input device are close to each other or are integer multiples. Moire appears as a striped pattern that does not exist in the input image, which confuses the user and can be a major nuisance in some cases. This moiré is due to a mismatch in average brightness levels in the input and output images, and was reported by JMWhite in the Journal of Applied Photographic Engineering.
Recent Advances in Thresholding Techniques published in the April 1980 issue of
The mechanism of its occurrence is described in detail in Facsimile.

The error diffusion method described above is also effective in suppressing this moiré. In other words, this method reduces moire by matching the average brightness level between the input and output images.

However, the pixel densities of the image input device and the output device are generally different, and it is necessary to convert the pixel densities at an intermediate stage. However, since the error diffusion method focuses on the average brightness level between the input and output images within the reference area, for example, when subsampling is performed for pixel density conversion, the error diffusion method In some cases, it could not be removed even by using this method, resulting in a disturbance equivalent to moiré.

(Objective of the Invention) The object of the present invention is to eliminate the above-mentioned disadvantages, and to avoid disturbances such as moire even when pixel density conversion is performed using an image expressed in halftone or pseudo-halftone as input.
An object of the present invention is to provide a pseudo-halftone image processing device that outputs a value image.

(Structure of the Invention) According to the present invention, in an apparatus for adaptively binarizing an image expressed in halftone or pseudo-halftone, there is provided a means for suppressing a preset spatial frequency component of the input image; A means for converting the pixel density of an input image in which frequency components have been suppressed, and for binarizing a pixel of interest in the density-converted image,
A value obtained by multiplying the error between the input image signal level and output image signal level of the surrounding pixels that have already been binarized by a coefficient set by the user is added to the signal level of the pixel of interest, and compared with a preset threshold value. A pseudo-halftone image processing device is obtained, characterized in that it is equipped with a binarization means for performing the above processing and outputting a binary image signal.

Further, according to the present invention, in an apparatus for adaptively binarizing an image expressed in halftone or pseudo-halftone, there is provided a means for suppressing a preset spatial frequency component of the input image; means for converting the pixel density of an input image, and a level limit for input image signal level between a maximum and minimum level set by a user when binarizing a pixel of interest in the density-converted image; and a binarization means for outputting a binarized image signal using the error between the input image signal level and the output image signal level of the surrounding pixels that have already been binarized. A halftone image processing device is obtained.

(Principle of the Invention) An object of the present invention is to perform pixel density conversion on an image expressed in halftone or pseudo-halftone, and then to binarize it. First, a digital filter is applied to the input image to suppress frequency components that cause aliasing distortion using subsamples. The pixel density of the output image of this digital filter is converted. Next, using the above-described error diffusion method, when the pixel of interest is binarized, the pixel of interest is binarized by reflecting the error when the surrounding pixels that have already been binarized are binarized.

(Example) Examples of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an example of a pseudo halftone image processing apparatus of the present invention.

Input image data is input from a terminal 11, and frequency components that cause aliasing distortion are suppressed in a digital film 13. Subsample circuit 1
4 is main scanning. Both sub-scanning and sub-scanning are thinned out to 1/2. The error diffusion binarization circuit 15 stores the level difference between the input pixel and the output pixel as an error when surrounding pixels are binarized, multiplies them by a coefficient determined by the position, and adds them all to the pixel of interest. , and then compared with an externally applied threshold level to binarize and output to terminal 12.

FIG. 2 shows a block diagram of the digital filter 13. Let P(i,j) be the signal level of a pixel at coordinate i in the main scanning direction and coordinate j in the sub-scanning direction.
23A, 23B are delay elements for one pixel, 24A,
24B and 24C are multipliers whose respective coefficients are a, b, and c. 25A is an adder, and its output
P′(i,j) is P′(i,j)=a ^* P(i+1,j)+b ^* P(i,
j)+c ^* P(i-1,j). This becomes a filter in the main scanning direction.

26A and 26B are delay elements for one line, 24
D, 24E, 24F have respective coefficients d,
Multipliers e, f and 25B are adders, so the terminal 22 has P″(i, j)=d ^* P′(i, j+1)+e ^* P′(i,
j)+f ^* P'(i, j-1) will be output. This becomes a filter in the sub-scanning direction, and together with the filter in the main scanning direction mentioned earlier, constitutes a two-dimensional filter. Coefficients a, b, c, d, e, and f are selected so as to suppress frequency components that cause aliasing distortion due to subsampling. Specifically, the sampling frequency is fs
As an example, a=c=d=f=0.25 and b=e=0.5 are selected so as to suppress components of fs/4 or more.

Here, for simplicity, a configuration of 3×3 taps is explained, but it is possible to use a larger number of taps with a similar configuration.

In addition, instead of passing through a one-dimensional transversal filter in the main scanning and sub-scanning directions as described above, an equivalent filter can be constructed by simultaneously referring to neighboring pixels two-dimensionally.

The sub-sample circuit 14 is realized with a simple configuration. That is, one register may be provided to store the output of the digital filter 13, and reading may be performed by dividing each of the write clock, line, and synchronization clocks into 1/2.

Figure 3 shows a binarization circuit 15 using the error diffusion method.
FIG. The image signal inputted from the terminal 32 is added to the sum of errors in surrounding pixels in the adder 37A, and compared with the threshold supplied from the terminal 33 in the comparator 38. The result is output to terminal 12 and also supplied to multiplexer 40. If the comparison result is white, the multiplexer 40 outputs the white representative value supplied from the terminal 35A, and if the comparison result is black, the multiplexer 40 outputs the representative value of white supplied from the terminal 35A.
Outputs the representative black color supplied from 5B. The adder 37B calculates the difference between the image signal to which the error has been added and the output image signal. That is, the error due to the pixel of interest is calculated. This error is 1 pixel delay element 3
The signals 9A to 39J are stored in an error memory consisting of a delay element 38A, which has two pixels less than one line, and a one-line delay element 38B. The stored errors are stored in multipliers 36A to 36L corresponding to the respective positions.
After the coefficients s1 to s12 are multiplied by the coefficients s1 to s12, the sum is calculated in the adder 37C. The coefficients used here include, for example, s1=0.03, s2=0.06, s3=0.10, s4=0.06, s5=
0.03 s6=0.06, s7=0.10, s8=0.15, s9=0.10, s10=
Use 0.06s11=0.10, s12=0.15, etc.

Here, we are looking at 12 pixels as surrounding pixels,
Of course, it is not necessary to limit the number to this number. In addition to the above configuration, for example, by focusing on the symmetry of the coefficients, parts corresponding to coefficients that take the same value can be added first, and then multiplied by the coefficients, thereby simplifying the circuit. You can also do it.

In any case, an image signal binarized by the error diffusion method is output to the terminal 12.

Fig. 4 shows a system in which a multiplier 42 is added to the binarization circuit using the error diffusion method shown in Fig. 3, so that it is possible to control the amount of error in surrounding pixels reflected during binarization. . The coefficient at this time is terminal 4
The user inputs from 1. By doing so, it becomes possible to control the image quality according to the input image.

Fig. 5 shows a circuit in which a limiter 53 is added to the binarization circuit using the error diffusion method shown in Fig. 4, so that the input image signal level can be limited between the maximum and minimum levels set by the user. It is. This limiter outputs levels between the maximum and minimum levels as they are, values exceeding the maximum level to the maximum level, and values smaller than the minimum level to the minimum level. The maximum and minimum levels at this time are terminals 35A and 35.
These are representative levels of white and black supplied from B. In the error diffusion method, even when the color is completely white or black, if even a slight error occurs, an isolated black symbol is created on a white background to eliminate the error, or vice versa. White dots may be output. In order to avoid such inconveniences, the input image signal level is limited by the representative level of error calculation to suppress the occurrence of errors. By doing this, background noise is suppressed and an easily viewable image can be obtained.

(Effects of the Invention) As described above, the pseudo halftone image processing device according to the present invention performs binarization using the error diffusion method after suppressing signal components that cause aliasing distortion by pixel density conversion. , it is possible to obtain a binary image signal free from disturbances such as moiré.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention;
Figure 2 is a block diagram of the digital filter, Figure 3
The figure shows a block diagram of a binarization circuit that uses the error diffusion method. Figure 4 is a block diagram of a binarization circuit that uses the error diffusion method and adds a multiplier to control the amount of error feedback. , FIG. 5 is a block diagram of a binarization circuit that uses an error diffusion method and adds a limiter for limiting the level of an input image signal. In the figure, 13...digital filter, 1
4...Subsample circuit, 15...Binarization circuit using error diffusion method, 23A, 23B...1 pixel delay element, 24A, 24B, 24C, 24D, 2
4E, 24F... Multiplier, 25A, 25B... Adder, 26A, 26B... 1 line delay element, 3
6A, 36B, 36C, 36D, 36E, 36
F, 36G, 36H, 36I, 36J, 36K,
36L... Multiplier, 37A, 37B, 37C...
Adder, 38A, 38B...Delay element with 2 pixels less than 1 line, 38B...1 line delay element,
39A, 39B, 39C, 39D, 39E, 39
F, 39G, 39H, 39I, 39J...1 pixel delay element, 40...Multiplexer, 42...
Multiplier, 46A, 46B, 46C, 46D, 46
E, 46F, 46G, 46H, 46I, 46J,
46K, 46L... Multiplier, 47A, 47B, 4
7C...Adder, 48A...Delay element with 2 pixels less than 1 line, 48B...1 line delay element,
49A, 49B, 49C, 49D, 49E, 49
F, 49G, 49H, 49I, 49J... 1 pixel delay element, 50... Multiplexer, 52...
Multiplier, 53...Limiter, 56A, 56B,
56C, 56D, 56E, 56F, 56G, 56
H, 56I, 56J, 56K, 56L... Multiplier, 57A, 57B, 57C... Adder, 58A
...Delay element with 2 pixels less than 1 line, 58B
...1 line delay element, 59A, 59B, 59
C, 59D, 59E, 59F, 59G, 59H,
59I, 59J...1 pixel delay element, 60... multiplexer.

Claims (1)

[Claims] 1 In a device that adaptively binarizes an image expressed in halftones or pseudo-halftones, means for suppressing a preset spatial frequency component of the input image, and means for converting the input image with the suppressed spatial frequency component into pixels. When the density conversion means and the pixel of interest in the density-converted image are binarized, the input image signal level is limited between the maximum and minimum levels set by the user. A pseudo-halftone image processing device comprising: binarization means for outputting a binarized image signal using an error between an input image signal level and an output image signal level of digitized surrounding pixels.