JPH0257753B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is directed to various image processing devices, and relates to a reduction conversion method that reduces image quality deterioration in images expressed in black and white binary. .

[Conventional technology] For conventional reduction conversion of binary images, OPC
(Ordinal Processing Conversion) method, SPC
(Selective Processing Conversion) method, (both methods are in the proceedings of the 7th National Conference of the Institute of Image Electronics Engineers, Proceedings)
No. 10, 1975) and area discrimination method (IEICE Techniques,
PRL81-92, 1982) have been published.
The configuration of these devices includes an original image readout unit that reads pixels of the original image (original pixels) from the buffer memory, a converted pixel position calculation unit that calculates the positions of images (converted pixels) that make up the reduced and converted image, and an original pixel reference unit for obtaining white and black information; an area determination unit for determining in which area the converted pixel is located; a converted pixel value determination unit for determining the value of the black and white information of the converted pixel; It consists of a converted image writing section for storing reduced converted images. The OPC method is a method in which pixels of the original image (original pixels) are sequentially extracted and replaced with pixels in the converted image (converted pixels), and the SPC method is a method in which the converted pixels are replaced with the closest original pixel on the original image. Further, in the area discrimination method, the average density projected onto the converted pixel position by the four original pixels closest to the converted pixel is calculated, and the average density is subjected to threshold processing to be used as the value of the converted pixel. In these conversion methods, a maximum of four original pixels are selected in order from the one closest to the converted pixel position, and the value of the converted pixel is determined by a logical operation to obtain their values or densities. Furthermore, conventional methods have only been proposed for storing black thin lines, but there have been no proposals for white thin lines.

[Problems to be Solved by the Invention] If there are black or white lines on the original image and those lines are not the closest original pixels to the converted pixel positions, or if the average density of those lines to be projected onto the converted pixel positions is If the threshold value processing result does not match the line value, the corresponding line may disappear in the converted image. For this reason, these conversion methods have the drawback of deteriorating image quality, such as sharpness due to the disappearance of black lines and blurring due to the disappearance of white lines. In addition, in order to prevent image quality deterioration such as sharpness and blurring caused by the disappearance of the black line, a method can be considered that sets the value of the converted pixel to black even when the black line is not at the closest position to the converted pixel. . However, with this method, for example, when the line width is 1 pixel, two converted pixels located across the line may both become black, and as a result, the line width on the converted image becomes smaller. The line width may become thicker than the line width on the original image, or blurring may occur, resulting in deterioration in image quality.

[Means for Solving the Problems] In order to solve these drawbacks, the present invention identifies thin lines that are at risk of disappearing among the thin lines on the original image, and creates a method to prevent the thin lines from disappearing and to preserve them. This prevents thin lines from becoming thicker and saves them on the converted image to prevent image quality from deteriorating.

One aspect of the present invention is that by preserving thin lines,
This invention aims to obtain converted images of good quality. By preserving both the white thin line and the black thin line, sharpness, fading, and blurring can be prevented and a converted image of good quality can be obtained. The second aspect of the present invention is to prevent thin lines to be saved from becoming thicker in order to further improve the quality of the converted image that has been prevented from becoming sharp, blurred, or blurred in the first aspect. That is, this prevents the thin lines on the converted image from becoming thicker than the thin lines on the original image despite the reduction conversion, and prevents blur from occurring on the converted image.

[Operation] Identify thin lines that are at risk of disappearing among the thin lines on the original image, and save the thin lines on the converted image while preventing the thin lines from disappearing and preventing the saved thin lines from becoming thicker. Quality deterioration can be prevented.

[Example] FIG. 5 shows an example of the basic configuration of an apparatus using the present invention. In the figure, 1 is an original image reading unit, 2 is a converted pixel position calculation unit, 3 is an original pixel reference unit, 4 is an area determination unit, 5 is a thin line storage processing unit, 6 is a converted pixel value determination unit, and 7 is a converted image This is the writing section.

An original image reading unit 1 reads an original image from an image buffer. Pixel P (m, n) with binary values of black and white
In the reduction conversion method of an image processing device, which reduces the original image expressed by horizontally at an arbitrary conversion ratio α smaller than 1 to create a converted image expressed by Q(k,l), The detection unit 2 has Q
When (k, l) is mapped onto P(m, n), assuming the distance between the pixels of the original image as 1, Q(k, l)
It has means (referred to as first means) for determining the position (x, y) in the horizontal and vertical directions. The original pixel reference section 3 has means for reading out pixel information necessary for the thin line storage processing section 5 from the original image. The area determining unit 4 determines that the position (x, y) of the converted pixel determined by the converted pixel position detecting unit 2 satisfies m≦x<m+1/
2 and n≦y<n+1/2, and m+
a second region where 1/2≦x<m+1 and n≦y<n+1/2, and m≦x<m+1/2 and n+1/2≦y<
a third region where n+1 and m+1/2≦x<m+
1, and means (referred to as second means) for determining which region of the fourth region where n+1/2≦y<n+1 is located. The thin line storage processing unit 5
It has third to sixth means for detecting thin lines and storing thin lines based on the output results of the converted pixel reference section 3 and the area determination section 4. In addition, the converted pixel value determination unit 6
In the case where thin line preservation processing is not performed, converted pixel values are determined by a method similar to the conventional method (referred to as the seventh means). The converted image writing unit 7 stores each pixel information of the reduced converted image in a buffer memory.

Details of the processing in the thin line storage processing section 5 are shown below.

FIG. 1 is a diagram showing the positional relationship between the original pixel and the converted pixel when the original image and the reduced converted image are made to correspond to each other. If the horizontal conversion ratio is α, then α=
There is a B/A relationship. Figure 1 shows an example where α=3/5. P(m, n), P(m+1, n), P(m
+2, n), P(m, n+1), P(m+1, n+
1), P(m+2, n+1), 〇, and ● mark are the original pixels, Q(k, l), and Q(k+1, l), ×
Each mark represents a converted pixel. In addition, the original pixels having the shortest distance in the original image are shown connected by solid lines, and the center line between the original pixels is shown by a dashed line.

Note that if the distance between pixels in the original image is normalized to 1 in Figure 1, the distance between pixels in the reduced converted image will be the reciprocal of the conversion ratio α, which is A/B (5/3 in the figure).
becomes. In FIG. 1, the shaded area is the original pixel P
When (m+1, n) and P(m+1, n+1) are black, this indicates an area where the value of the converted pixel is black when the position of the converted pixel is in this shaded area.

In order to prevent thin lines from disappearing, it is necessary to identify thin lines that are at risk of disappearing.

Figure 1 shows original pixels P(m+1,n) and P(m+
1, n+1) are black and have the same color, P(m, n), P
(m+2, n), P(m, n+1), P(m+2, n
+1) is white, which is the opposite color of P(m+1, n),
The horizontal and vertical positions (x, y) of the converted pixel Q (k, l) satisfy m≦x<m+1/2 and n≦y<
n+1/2 (first region), converted pixel Q (k
+1, l) horizontal position x′ is m+1/2≦x′<
The case where m+2 is shown.

In the conventional conversion method, P(m, n), P(m
+1, n), P(m, n+1), P(m+1, n+
1)) is used as a reference pixel, and the value of the original pixel closest to the converted pixel position is given the most weight to determine the value of the converted pixel. In other words, the white of P(m, n) is selected for Q(k, l) in the same figure, and the white of P(m+2, n) is selected for Q(k+1, l), so that the original image is P(m+1,n) and P that existed as thin lines
The thin black line connecting (m+1, n+1) disappears.

In determining the value of a converted pixel, the present invention starts from the pixel position of the original image closest to the converted pixel position and moves up to two pixels ahead of the original pixel in a horizontal direction approaching the converted pixel position. The value of the original pixel is also referred to to determine whether there is a thin line that may disappear. That is, in the case of FIG. 1, Q(k, l)
For Q(k+1, l), the value of P(m+2, n) is also referred to. A specific example of a method for determining the presence or absence of a thin line in consideration of the reference pixels described above will be shown below.

When determining the value of Q(k,l) in FIG. 1, the reference pixels are P(m+2,n) and P(m,
n), P(m, n+1), P(m+1, n), P(m+
1, n+1), P(m, n) = P(m+2, n) = P(m, n+1) ≠ P(m+1, n) = P(m+1, n+1)
......When (Equation 1) is satisfied, it is determined that there is a thin line that may disappear, and the value of P (m + 1, n) is substituted for Q (k, l) (referred to as the third means). ).

Above, the horizontal and
The vertical position (x, y) is m≦x<m+1/2
And when n≦y<n+1/2 (first region)
Similarly, (a) the position of Q (x, y) satisfies m+1/2≦x<m+
1 and n≦y<n+1/2 (second region), P(m+1, n)=P(m-1, n)=
P(m+1, n+1)≠P(m,n)=P(m,n
+1), and if this conditional expression is satisfied, a thin line exists, and Q(k, l) = P(m, n)
(b) When the position (x, y) of Q satisfies m≦x<m+1/2 and n+1/2≦y<n+1 (third region), P (m, n+1)=P(m+2, n+1)
=P(m,n)≠P(m+1,n+1)=P(m+
1, n), and if this conditional expression is satisfied, a thin line exists, and Q(k, l)=P(m+1,
n+1) (referred to as the fifth means), and (c) the position (x, y) of Q satisfies m+1/2≦x<m+
1 and n+1/2≦y<n+1 (fourth region), P(m+1, n+1)=P(m-1,
n+1)=P(m+1,n)≠P(m,n+1)=
Using P(m, n), assuming that a thin line exists if this conditional expression is satisfied, Q(k, l)=P(m,
n+1) (referred to as the sixth means), and if all of the above conditional expressions are not satisfied, the converted pixel value determination unit 6 calculates P(m, n), P(m+
1, n), P(m, n+1), P(m+1, n+1)
Therefore, the value determined by the conventional conversion method (seventh means) (for example, the value of the shortest distance original pixel in the SPC method) is substituted for Q(k, l).

A converted image is created by repeating the process of determining Q(k, l) using the third, fourth, fifth, sixth, and seventh means to prevent thin lines from disappearing. FIG. 6 shows the processing flow of the thin line storage processing section 5.

A conversion method for preserving vertical lines during horizontal reduction conversion has been described above, but by using a similar method for vertical reduction conversion, it is possible to preserve horizontal lines.

Next, a method for preserving thin lines while preventing them from becoming thicker will be described.

In the present invention, as described above, starting from the original pixel closest to the converted pixel, the original pixel two pixels ahead in the horizontal direction approaching the converted pixel position is added to the reference pixel to determine whether there is a thin line that may disappear. This is to determine the Therefore, when the positional relationship between the converted pixel and the original pixel is as shown in Figure 2 a, the reference pixel of Q (k, l) is to the horizontal right of the nearest original pixel (P (m, n)); It will be within the dotted line frame shown inside. Also, Q
The reference pixel of (k+1, l) is also the nearest neighbor pixel (P(m+
2, n)) in the horizontal right direction, and is within another dotted line frame in the same figure. Therefore, the target for determining the presence or absence of a thin line is the line connecting P(m+1, n) and P(m+1, n+1) with respect to Q(k, l), and Q(k+
1, l), P(m+3, n) and P(m+
3, n+1). Therefore, P(m+
The thin line connecting P(m+1, n+1) is
It is stored only in Q(k,l).

On the other hand, in the case of FIG. 2b, the reference pixels of Q(k, l) are the same as in FIG. 2a, but Q(k+1,
The reference pixel of l) is the nearest neighbor pixel P(m+2,
n), it falls within the dotted line frame in Figure 2b, overlaps with the reference pixel of Q(k,l), and P(m
The line connecting +1, n) and P(m+1, n+1) is
The presence or absence of a thin line in the two converted pixels Q(k,l) and Q(k+1,l) is to be determined. For this reason, P
For the thin line connecting (m+1, n) and P(m+1, n+1), the same color is preserved for both, black for the value of Q(k, l) and black for the value of Q(k+1, l), There is a problem where thin lines are saved as thick.

In order to prevent this phenomenon, in the present invention, the distance between the position of the converted pixel and the thin line connecting the original pixels is determined, and the thin line preservation process described above is applied only to the converted pixel closest to the thin line.

An example of a method for determining whether a converted pixel is the pixel closest to a thin line will be described below.

In FIG. 3, the conversion pixel of interest is Q(k,
l), and the converted pixel adjacent to the right side is Q(k
+1, l). Let d be the distance between Q(k, l) and the dashed-dotted line representing the center line between original pixels. Q(k,
l) is P(m+1, n) from Q(k+1, l)
The condition close to the thin line connecting and P(m+1, n+1) is
It will look like this: The distance between converted pixels is A/B,
That is, if the conversion ratio α is 1/α, then d+1/2<(1/α)·(1/2) (second formula). Solving for d yields d<{(1/α)-1}/2 (3rd equation). If the converted pixel position of Q(k, l) satisfies the condition of the third equation, perform the thin line preservation process described above for Q(k, l), and if the converted pixel position does not satisfy the third equation, add the shortest distance to the thin line. Since it is not a converted pixel, no thin line preservation processing is performed to prevent the thin line from becoming thicker. In this way, the thin line is saved only in the converted pixel closest to the thin line, and the thin line is prevented from becoming thicker. The position of Q is m≦x<m+1/2
When , the range of x that satisfies the third formula above is x>m+
A means (referred to as an eighth means) for determining whether 1-1/(2α) is satisfied is provided, and if the condition is satisfied, thin line preservation processing is performed. Also, the position of Q is m
When +1/2≦x<m+1, a means (referred to as a ninth means) is provided for determining whether the range of x that satisfies the third formula above satisfies x<m+1/2α, and if it is satisfied, the same applies. Perform thin line preservation processing. Note that the eighth and ninth means are included in the region determining section 4. FIG. 7 shows the flow of this process.

The above is a case of reduction conversion in the horizontal direction, but the same applies to reduction conversion in the vertical direction.

Assuming that the horizontal conversion ratio is B/A and the vertical conversion ratio is D/C, dividing the area according to the conversion pixel position to determine whether or not it is necessary to take line preservation into account when determining the conversion pixel value is as follows. The result will be as shown in Figure 4. When the converted pixel is included in the area, there is no need to consider the disappearance of thin lines. In the case of the area and, it is necessary to prevent thin line disappearance only in the horizontal direction and the vertical direction, respectively. In the area, it is necessary to prevent thin lines from disappearing in both the horizontal and vertical directions.

Above, we have described a method for preserving thin lines while preventing them from becoming thicker. However, when thin lines cannot be identified or when there is no need to prevent thin lines from disappearing, such as in areas, the converted pixel value determination unit In 6, by the seventh means (e.g. SPC method),
Determine the value of the converted pixel.

In FIG. 5, the part indicated by the dotted line is the part according to the present invention. The other parts except the dotted line part are the same as the conventional device. That is, the area determination unit 4
In this step, the second, eighth, and ninth means determine where in the area shown in FIG. 4 the conversion pixel of interest is included. Next, the thin line storage processing section 5
In this step, thin line detection and storage processing are performed by the third to sixth means. The converted pixel determination unit 6 determines the value of each pixel for all converted pixels based on the processing results of the respective units described above and the seventh means. The converted pixel values obtained in this manner are stored in the converted image writing section 7 to obtain a reduced converted image.

[Effects of the Invention] As described above, the present invention includes a means for determining a thin line that is likely to disappear on a binary image, and a means for storing the thin line without making it thicker. Therefore, there is an advantage that it is possible to prevent cuts, blurring, blurring, etc. from occurring on the converted image during image reduction conversion of a facsimile image, etc., and to create a converted image with good image quality.

[Brief explanation of the drawing]

Fig. 1 is a layout diagram of original pixels and converted pixels to explain the principle of the present invention and reference pixels, and Fig. 2 a and b show thin lines on the converted image that are thicker than lines on the original image. FIG. 3 is a diagram for explaining an example of a method for preventing thin lines on a converted image from becoming thicker. FIG. 4 is a diagram for explaining the preservation of thin lines when the present invention is applied. 5 is a diagram showing an example of the basic configuration of an apparatus using the present invention, FIG. 6 is a diagram showing the flow of the thin line preservation processing section, and FIG. FIG. 3 is a diagram showing a flow of an area determination unit. 1 is an original image reading section, 2 is a converted pixel position calculation section, 3 is an original pixel reference section, 4 is an area determining section, 5 is a thin line storage processing section, 6 is a converted pixel determining section, and 7 is a converted image writing section.

Claims (1)

[Claims] 1. An original image expressed by a pixel P (m, n) having two values of black and white is reduced and transformed in one axis direction of an orthogonal coordinate system at an arbitrary transformation ratio α smaller than 1. Q(k,
In the reduction conversion method in an image processing device that creates a transformed image expressed by a first means for determining the position (x, y) of (k, l); and a position (x, y) determined by the first means.
y) is a first region where m≦x<m+1/2 and n≦y<n+1/2, m+1/2≦x<m+1 and n
A second region where ≦y<n+1/2, m≦x<m+
a third region where 1/2 and n+1/2≦y<n+1;
a second means for determining in which region of the fourth region m+1/2≦x<m+1 and n+1/2≦y<n+1 the image is located; ,
P(m, n), P(m+2, n), P(m, n+1),
Select P(m+1,n), P(m+1,n+1) and apply the conditional expression P(m,n)=P(m+2,n)=P(m,
n+1)≠P(m+1,n)=P(m+1,n+1)
If it is satisfied, it is assumed that a thin line exists, and Q(k,l)=P(m+1,
a third means for outputting P(m+1,n), P(m-) from the original image when the position (x, y) is located in the second area;
1, n), P(m+1, n+1), P(m, n), P
(m, n+1) and conditional expression P(m+1,
n)=P(m-1,n)=P(m+1,n+1)≠P
A fourth method that determines whether (m, n) = P (m, n + 1) is satisfied, and if it is satisfied, a thin line exists and outputs Q (k, l) = P (m, n). means, P(m, n+1), P(m+1) from the original image when the position (x, y) is located in the third area;
2, n+1), P(m, n), P(m+1, n+1),
Select P(m+1, n) and apply the conditional expression P(m, n+
1)=P(m+2,n+1)=P(m,n)≠P(m
+1, n+1) = P(m+1, n), and if it is satisfied, a thin line exists and outputs Q(k, l) = P(m+1, n+1). and when the position (x, y) is located in the fourth area, P(m+1, n+1), P
(m-1, n+1), P(m+1, n), P(m, n
+1), P(m, n), and conditional expression P(m+
1, n+1)=P(m-1, n+1)=P(m+1,
Determine whether n)≠P(m, n+1)=P(m, n) is satisfied, and if it is, a thin line exists and output Q(k, l)=P(m, n+1). and if the position (x, y) does not satisfy all of the conditional expressions, P(m, n), P(m+1, n),
Seventh means for determining the value of the original pixel closest to the position (x, y) among the four original pixels P(m, n+1) and P(m+1, n+1) as the value of Q(k, l). An image reduction conversion method, comprising: determining Q(k, l) from the third, fourth, fifth, sixth, and seventh means to create a converted image. 2. The original image expressed by a pixel P (m, n) having two values of black and white is reduced and transformed in one axis direction of the orthogonal coordinate system at an arbitrary conversion ratio α smaller than 1, and then Q (k,
In the reduction conversion method in an image processing device that creates a transformed image expressed by a first means for determining the position (x, y) of (k, l); and a position (x, y) determined by the first means.
y) is a first region where m≦x<m+1/2 and n≦y<n+1/2, m+1/2≦x<m+1 and n
a second region where ≦y<n+1/2, m≦x<m+
a third region where 1/2 and n+1/2≦y<n+1;
a second means for determining in which region of the fourth region m+1/2≦x<m+1 and n+1/2≦y<n+1 the image is located; ,
P(m, n), P(m+2, n), P(m, n+1),
Select P(m+1,n), P(m+1,n+1) and apply the conditional expression P(m,n)=P(m+2,n)=P(m,
n+1)≠P(m+1,n)=P(m+1,n+1)
If it is satisfied, a thin line exists, and Q(k,l)=P(m+1n)
a third means for outputting P(m+1,n), P(m-) from the original image when the position (x, y) is located in the second area;
1, n), P(m+1, n+1), P(m, n), P
(m, n+1) and conditional expression P(m+1,
n)=P(m-1,n)=P(m+1,n+1)≠P
A fourth method that determines whether (m, n) = P (m, n + 1) is satisfied, and if it is satisfied, a thin line exists and outputs Q (k, l) = P (m, n). means, P(m, n+1), P(m+1) from the original image when the position (x, y) is located in the third area;
2, n+1), P(m, n), P(m+1, n+1),
Select P(m+1, n) and apply the conditional expression P(m, n+
1)=P(m+2,n+1)=P(m,n)≠P(m
+1, n+1) = P(m+1, n), and if it is satisfied, a thin line exists and outputs Q(k, l) = P(m+1, n+1). and when the position (x, y) is located in the fourth area, P(m+1, n+1), P
(m-1, n+1), P(m+1, n), P(m, n
+1), P(m, n), and conditional expression P(m+
1, n+1)=P(m-1, n+1)=P(m+1,
Determine whether n)≠P(m, n+1)=P(m, n), and if so, assume that a thin line exists and output Q(k, l)=P(mn+1). and if the position (x, y) does not satisfy all of the conditional expressions, P(m, n), P(m+1, n),
Seventh means for determining the value of the original pixel closest to the position (x, y) as the value of Q(k, l) among the four original pixels P(m, n+1) and P(m+1, n+1) and when m≦x<m+1/2, x>m+1−1/2α
an eighth means for determining whether m+
A ninth means for determining whether x<m+1/2α is satisfied when 1/2≦x<m+1, and when the condition of the eighth or ninth means is satisfied, the third, fourth, and fifth means , Q(k,l) from the sixth and seventh means
is determined, and if the conditions of the eighth or ninth means are not satisfied, P(m, n), P(m+1, n), P
Among the four original pixels (m, n+1) and P (m+1, n+1), the value of the original pixel closest to the position (x, y) is determined as the value Q (k, l) and a converted image is created. An image reduction conversion method characterized by: