JPH03218175A - Image processor - Google Patents

Abstract

PURPOSE:To reduce the processing quantity for binarization and to realize the image of a high quality at a low cost by calculating average density by using only binary data whose binarization processing is ended, setting its average density value as a threshold and binarizing input multivalued data. CONSTITUTION:An edge detecting means 23 detects the edge quantity of an object picture element of a binarization object, arithmetic means 13, 14 derive average density in the vicinity of the object picture element, based on binary data of the picture element binarized alreaby before the object picture element, a binarizing means 18 binarizes multivalued image data of the notice picture element, based on the average density value derived by the arithmetic means 13, 14, and distributing means 19-22 distribute a binarization error generated at the time of binarizing the object picture element by the binarizing means to the multivalued image data of the picture element before binarization in the periphery of the notice picture element, based on the edge quantity detected by an edge detecting means 23, and save the density. In such a way, the image whose gradation and resolution are both excellent is constituted of a hardware being simple and low in cost, and especially, the depiction of a highlight part can be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device, and for example, to an image processing device that quantizes multivalued image data into binary data.

[Prior Art] Conventionally, this type of device includes image processing devices such as facsimile machines and digital copying machines, and error diffusion methods and average density approximation methods have been proposed as pseudo halftone processing methods for each device. There is.

The former error diffusion method is described in the document R. FLOYD. L.
STEINBERG's ゜'AN ADAPTIVE A
LGORITHM FOR SPECIAL GRAY
SCALE”, SID75 DIGEST, PP3
6 to 37, the multi-valued image data of the pixel of interest is binarized (converted to the darkest level or the lightest level), and the binarized level and the multi-valued data before binarization are The error is given a predetermined weight and added to data of pixels near the pixel of interest.

Moreover, the latter average concentration approximation method is disclosed in JP-A-57-1043.
As described in No. 69, the pixel of interest is converted into black or white using the already binarized binary data near the pixel of interest.
When converted into values, a weighted average with each neighboring pixel is obtained, and the average of these two average values is used as a threshold to binarize the image data of the pixel of interest.

[Problems to be Solved by the Invention] The error diffusion method described above is a method for correcting errors between input image data and output image data, so it is possible to preserve the density of the input image and the output image, and it is possible to preserve the resolution and gradation. It is possible to provide images with excellent quality.

However, the error diffusion method requires calculation of many two-dimensional operations when correcting errors between input image data and output image data, and the large amount of processing requires a very complex hardware configuration. There were some drawbacks.

In addition, since the average density approximation method performs calculations using binary data after binarization, it is possible to simplify the hardware configuration and achieve faster processing due to the extremely small amount of processing. Is possible.

However, the average density approximation method approximates the pixel of interest to the average value of the area including the pixel of interest and performs binarization, which limits the number of gradations and is unique to images with gradual density changes. The problem was that low-frequency textures were generated, which degraded the image quality.

Furthermore, the method of binarizing the pixel of interest based on the conditions of neighboring pixels has the disadvantage that when actually printing on paper, dot connections occur in highlighted areas, resulting in a reduction in image quality.

The present invention has been made in view of the above-mentioned drawbacks of the prior art, and its purpose is to construct an image with excellent gradation and resolution using simple and low-cost hardware, and in particular, to construct an image with excellent gradation and resolution using simple and low-cost hardware. An object of the present invention is to provide an image processing device that provides excellent depiction of highlight parts and is free from chipping at edge parts by preventing dot connections in light parts and correcting error distribution in edge parts.

[Means for Solving the Problems] In order to solve the above-mentioned problems and achieve the objectives, an image processing apparatus according to the present invention has the following configuration. That is, in an image processing apparatus that quantizes multivalued image data into binary data, there is provided an edge detection means for detecting an edge amount of a pixel of interest to be binarized, and a pixel that has already been binarized before the pixel of interest. a calculation means for calculating an average density in the vicinity of the pixel of interest based on the binary data of the pixel of interest; and a binarization means for binarizing the multivalued image data of the pixel of interest based on the average density value obtained by the calculation means. Based on the edge amount detected by the edge detection means, the binarization error that occurs when the pixel of interest is binarized by the binarization means is calculated from the value of the area around the pixel of interest before binarization. and distribution means for distributing multivalued image data of pixels.

[Operation] According to this configuration, the edge detection means detects the edge amount of the pixel of interest to be binarized, and the calculation means detects the edge amount of the pixel of interest based on the binary data of pixels that have already been binarized before the pixel of interest. The average density of the neighborhood is determined, the binarization means binarizes the multivalued image data of the pixel of interest based on the average density value determined by the calculation means, and the distribution means is based on the amount of edges detected by the edge detection means. Then, the binarization error generated when the pixel of interest is binarized by the binarization means is distributed to multi-valued image data of pixels surrounding the pixel of interest before binarization, and the density is preserved. .

[Example] Embodiments of the present invention will be described in detail below with reference to the drawings.

First, the principle of this method will be explained.

FIGS. 3A to 3D are diagrams for explaining the binarization method of this embodiment.

FIG. 3A shows the multilevel density of each pixel of the input image. In FIG. 3A, f (i, j) is the multilevel density data of the input pixel at the pixel position of interest to be binarized,
The values that the data can take are normalized values between 0 and 1. Furthermore, the data of each pixel located above the broken line in the figure is data that has already been binarized. After the pixel of interest is binarized, the subsequent f (i, j+1), f (i
, j+2), . . . are subjected to similar binarization processing as the pixel of interest.

FIG. 3B shows binarized image data. In the figure, B (i, j) is the density of the pixel of interest after binarization (0
or a value of 1). The area surrounded by the broken line in the figure is pixel data that has already been binarized when the pixel of interest is binarized. These binarized pixel data are used in the binarization process of the pixel of interest.

Figures 3C and 3D represent weighted masks.

In FIGS. 3C and 3D, R. R2 is an example of a weighting mask for calculating the average density, and is 3×5 or 3×2
Represented by a matrix of sizes. Here, the weight for unbinarized pixel data is R, (0, O) = R, (0
．． 1)=0,R2 (0,O)=R2 (0.1)
=O is used.

In this method, the weighted average density m, (z, j) + m2 (11 J) of the binary image near the pixel of interest is determined using the following equations (1) and (2). In addition, m. (i, J), mz
(1, j) will be discussed later.

That is, m+(i+.+)"Σ Σ R+(x,y)・B(
f-x. j−y) − (1)mzh+j)”
Σ Σ R2(x,y)・B(i-x.j-y)
- (2).

Also, according to equation (3) shown below, the pixel of interest f(i, j
) and the already assigned binarization correction value E(i, j), one of m+(1+j)+m2(1+j) is selected. That is, Jf(i,j)+E(i.
When j)≦β, m(i. j)=m+ (L j)lf
(i, j) + E (i, j) When > β, m (i, j)
:m2 (i, j)...(3). Note that β is a predetermined value determined in advance.

Here, m+ (11 J) and mx (1 1
J) and β will be explained.

In this embodiment, binarization is performed using m(ilJ), the average density around the pixel of interest, as a threshold, and good binarization is performed by preserving the average density by error allocation. The wider the area m(iIJ) for calculating the average density, the better the gradation will be, and the narrower it will be, the better the resolution will be. For this reason, a mask with a good balance between gradation and resolution is used, but dots tend to approach each other in highlighted areas. Therefore, in the highlighted area, the average density calculation area is widened and the dots are scattered. Therefore, in the highlighted part rl'l+
(i+-+), and m. The value of β is set to select (i, j).

Therefore, β, which distinguishes whether or not a pixel to be binarized belongs to the highlight area, is 7 to It is set to a range of 20 levels.

The pixel of interest f (i, j) is the selected pixel m (i, j)
Using the above-mentioned binarization correction value E (i, j), the following equation (4) is obtained.
It is binarized according to the following.

That is, ... (4). E (i, j) is the multilevel density f (i.j-
Error E2 (iIJ) that occurs when 1) is binarized into binary density B (i, j - 1) and the pixel (i - 1, j) that is one line before the pixel of interest (i, j). ) multilevel density f(i-
1,j) into a binary density B(i-1.j), and is the sum of the error E+(t+J), which is expressed by the following equation (5).

That is, E(i. .+)=E+ (i, j)+Ez (i.
j) ...(5).

In this way, by adding the binarization error to the pixel of interest and binarizing the corrected value, it is possible to save the image density after binarization as the average density over the entire input image.

In this embodiment, by performing processing that takes binarization errors into account, the halftone reproduction ability is significantly improved compared to the above-mentioned average density approximation method. Although the amount of processing is smaller than that of the error diffusion method, image reproduction performance equivalent to that of the error diffusion method can be obtained.

Furthermore, in areas where the density is low, by widening the area for calculating the weighted average density, it is possible to prevent dots from connecting in highlighted areas, resulting in a high-quality image.

Now, in this embodiment, before performing the binarization process, the edge amount edge (i, j) is detected in advance, and the error during binarization is calculated by the following equation (6) using the detected edge amount 6dge (i+j).
) is determined.

That is, shall be.

In this method, since the average density calculation area is widened in the highlight portion, image loss is likely to occur in the highlight portion near the edge. Therefore, in the place judged to be the edge part,
Negative binarization errors are not distributed. This improves the image quality at the edges.

Next, a configuration for performing the above processing will be described.

FIG. 1 is a block diagram of an image processing apparatus showing an embodiment of the present invention. In this embodiment, a copying machine is taken as an example, but it goes without saying that the binarization process that is the gist of the present invention is applicable to other facsimile machines and the like.

In FIG. 1, reference numeral 100 indicates an input sensor section, which is composed of a photoelectric conversion element such as a CCD and a driving device for scanning the photoelectric conversion element, and reads and scans an original. Reference numeral 101 indicates an A/D converter that converts the output from the input sensor unit 100, that is, an analog signal into a digital signal, and 102 indicates an A/D converter.
A correction circuit that corrects the digital signal output from the converter 101 is shown. Reference numeral 102 indicates a binarization circuit that binarizes the correction signal output from the correction circuit 102 according to the binarization method of this embodiment. 104 is a binarization circuit 10
3 shows a printer that visibly forms an output image on recording paper (not shown) based on binary data output from the printer 3.

Next, the overall operation of the above-mentioned copying machine will be explained.

The image data of the original read by the input sensor section 100 is sequentially sent to the A/D converter 101. Here, the data of each pixel is quantized into digital data. Next, in the correction circuit 102, shading correction and the like for correcting sensitivity unevenness of the CCD sensor and illuminance unevenness due to the illumination light source are performed by digital calculation processing. This corrected data is then sent to the binarization circuit 103. The binarization circuit 103 performs a process of quantizing input multivalued image data into binary data using the method described above. The printer 104 is a printer configured using a laser beam or an inkjet method, and reproduces a visible image on recording paper while controlling dots on/off based on binary data sent from the binarization circuit 103.

Next, the binarization circuit 103 will be explained.

FIG. 2 is a block diagram showing the detailed configuration of the binarization circuit 103 of this embodiment. In the same figure, 1 and 2 are delay RAs that hold the binarized binary data with a delay of one line.
M, 3 to 12.19 are D-type flip-flops (DF/F) that hold binary data with a one-pixel delay. 13. 14 is an average density calculation circuit that calculates the average density of a predetermined area from binary data around the pixel of interest, and 15 is a comparator that compares the multi-value data with a predetermined value (here, "β"). , 16 is a selector that selects one of the plurality of weighted average values as a threshold based on the selection signal from the comparator 15;
17 is a subtracter that calculates the difference between the threshold output from the selector 16 and the multi-value data of the pixel of interest; 18 is the selector 16;
A comparator 20 that compares the threshold output from the multi-value data of the pixel of interest with the multi-value data of the pixel of interest, adds the multi-value data sent from the correction circuit 102 and the error data output from the error ROM 2 1 and the error memory 22. Adder, 21 is subtracter 1
Error ROM 22 calculates error data based on the difference between the multivalued data of the pixel of interest and the threshold value sent from 7, and the edge amount also sent to the edge amount detection device 23; 22 stores error data for one line; An error memory 23 stores an edge amount detection circuit for detecting an edge amount at a pixel of interest.

In the above configuration, the comparator 18 has 2
The converted data B (i, j) is output. This output binary data is delayed by R for each line.
AM2, input to delay RAMI, 2f direct data B (i-1, j+2
) is delayed by 2 lines by RAMI, resulting in binary data B (
i-2, j+2) is output. Furthermore, DF/F3 is B
(i-2, j+1), DF/F 4 is B (i-2,
j). DF/F 5 is B (i-2, j-1).
DF/F 6 is B(i-2, j 2). DF
/F 7 is B(i-1,j+1). DF/F8 is B (
i-1, j), DF/F9 is B (i-l, j-
1) DF/F 10 is B(i-1, j-2),
DF/F 1 1 is B (IIJ-1) DF/F12
outputs B(i, J-2).

The above binary data is input to the average concentration calculation circuit 13, of which B (i-1, j+1), B (i-1,
j) B (i-1, j-1) , B (i, j
-1) are also input manually to the average concentration calculation circuit 14.

The average density calculation circuits 13 and 14 input binary data of pixels around the pixel of interest, and each calculates a weighted average density value m.
+(1+J)+m2(i,j) is output.

On the other hand, the comparator 15 receives f (i, j) from the DF/F 19.
+E (i, j) is input and compared with a predetermined value (β), and a select signal is output to the selector 16 based on the comparison result. The selector 16 uses the select signal sent from the comparator 15 to determine the weighted average density value m.
+(1+J)+rn2(i,j) is output as the threshold rn(x+j).

The above threshold m(i, j) and f output from DF/F19
(i, j) + E (i, j) is both subtractor 1
7 and comparator 18. Subtractor 17 calculates the difference between two inputs. Also, the comparator 18 receives the above two inputs f
(i, j) 10E (i, j) and m(t, j) are compared and binarized data B (i, j) is output.

The difference between the multivalued data of the pixel of interest output from the subtracter 17 and the threshold value is input to the error ROM 21. At the same time, the edge amount obtained from the edge amount detection circuit 23 is also input to the error ROM 21. The error ROM21 is calculated using the above formula (6).
Accordingly, the error E2(i, j+1) is output. This E
2 (i, j+1) is input to the adder 20 and also to the error memory 22. The error memory 22 stores and holds the error for one line, and instead of E2 (i, j+1), E1 (i+1, j) already stored in the previous stage processing is output.

The above E2 (i, j+1) and E + ('i, j +1) outputted from the error memory 22 are added to the input image data f (i, j+1) by the adder 20. The DF/F 19 delays the added value output from the adder 20 by one data clock period, that is, by one pixel.

Thereafter, the binarization process is sequentially performed by repeating the above process.

In this way, the average density is calculated using only the binary data that has been binarized, and the input multivalued data is binarized using the average density value as a threshold, so the processing for binarization is The amount can be reduced compared to the error diffusion method. Further, by widening the area in which the average density is calculated in the highlight part, it is possible to prevent dot connections that appear in the highlight part. Furthermore, by not distributing a negative error to the edge portion, it is possible to prevent image loss, etc. at the edge portion.

As explained above, according to this embodiment, high-quality images can be obtained at low cost.

[Other Examples] Next, other embodiments will be described.

In the embodiment described above, a plurality of weighted average values are calculated,
A method is used in which one of a plurality of weighted average values is selected based on the level of the pixel of interest, and a negative error at an edge portion is not distributed when calculating the weighted average value.

On the other hand, in other embodiments, a method of allocating negative errors at edges is used in calculating the weighted average value.

FIG. 4 is a block diagram showing the configuration of a binarization circuit according to another embodiment of the present invention. Another embodiment, as shown in FIG. 4, has substantially the same construction as the previous embodiment, and circuits having the same function and configuration have the same reference numerals as shown in FIG. Use numbers and omit descriptions. Therefore,
In FIG. 4, numeral 30 indicates an average density calculation circuit, which is connected to the output of the comparator 15 and the output of the edge detection circuit 23 (
i, j) are also included to calculate the average density. From this configuration,
Only the output from the average density calculation circuit 30 is provided, and a selector for selecting a plurality of average density calculation circuits is not required.

Therefore, in another embodiment, the average concentration calculation circuit 30 includes:
In addition to the binarized peripheral pixel data, the level of the pixel of interest and the edge 11 e d g e (i, j) are input, and calculations are performed as shown in the following equation (7). That is,. Here, α is a level for determining whether or not it is an edge portion. For example, if the above process (widening the average density calculation area in the highlight part) is performed at a place where the image suddenly changes from a background part to a highlight part, the image will be missing at the edge part. In order to prevent this image from being missing, even if the pixel of interest is determined to be a highlight part, if it is an edge part, a normal average density calculation area is used.

As a result, it is possible to obtain an output image in which the width of the defect in the image is suppressed to be narrow and the defect is not visually noticeable.

In this way, by reducing the size of the weighting mask at the edge area independently of the density of the pixel of interest, the response of binarization is improved, and the highlight area can be improved without causing image loss at the edge area. Can prevent dots from connecting.

Now, in the above two embodiments, the explanation was given using a copying machine as an example, but it goes without saying that the binarization processing that is the gist of the present invention can be applied to other facsimile machines and the like.

[Effects of the Invention] As explained above, according to the present invention, an image with excellent gradation and resolution can be obtained with a simple and inexpensive hardware configuration, while suppressing image chipping at edges. , it is possible to obtain high-quality images with improved image quality in the highlight areas by preventing dot connections that occur in the highlight areas.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an image processing apparatus showing an embodiment of the present invention, FIG. 2 is a block diagram showing a detailed configuration of the binarization circuit 103 of this embodiment, and FIGS. FIG. 4 is a block diagram showing the configuration of a binarization circuit according to another embodiment of the present invention. In the figure, 1.2...delay RAM, 3-12, 19...
D F/F, 13, 14.30... Average concentration calculation circuit, 15.18... Comparator, 16... Selector, 17
...Subtractor, 20...Adder, 21...Error RO
M, 22...Error memory, 23...Edge detection circuit, 100...Input sensor section, 101...A/D converter, 102...Correction circuit, 103...Binarization circuit ,
104: Printer.

Claims (4)

[Claims] (1) An image processing device that quantizes multivalued image data into binary data, comprising an edge detection means for detecting an edge amount of a pixel of interest to be binarized, and a pixel of interest that has already been binarized before the pixel of interest. a calculation means for calculating an average density in the vicinity of the pixel of interest based on binary data of the pixel; and 2 for binarizing the multivalued image data of the pixel of interest based on the average density value found by the calculation means. and a binarization error generated when the pixel of interest is binarized by the binarization means, based on the edge amount detected by the edge detection means, of the pixel of interest. and distribution means for distributing to multivalued image data of surrounding pixels before binarization, and preserving density by distributing the binarization error by the distribution means. Processing equipment. (2) The calculation means has a plurality of weight masks of different sizes, and a masking means for masking binary data of pixels that have already been binarized before the pixel of interest based on the plurality of weight masks. a calculating means for calculating an average density based on the results for each weighted mask obtained by the masking means; and a selecting means for selecting one of the plurality of average densities calculated by the calculating means. The image processing apparatus according to claim 1, characterized in that the image processing apparatus includes: (3) The image processing apparatus according to claim 2, wherein the selection means uses the level of the pixel of interest at the time of selection. (4) The image processing apparatus according to claim 1, wherein the average density value is used as a threshold value.