JPS6293764A - Image reducing method - Google Patents

Abstract

PURPOSE:To obtain an image reducing method which scarcely generates the omission of an image, in s short processing time, by setting an are which is formed by the picture element of N points in the periphery of a projection point on the original image of a picture element on a reduced image, dividing this area into M pieces of small areas, deriving the density of the area from the density of each small area, and determining white/black of the picture element. CONSTITUTION:An original image 10 is reduced to 1/2, and converted to a reduced image 11. Picture elements contained in an area 12 are 16 points of I1-I6, but these picture elements are further subdivided into four areas of small areas 20-23, and by the value of these small areas 20-23, the representative picture element of the area 12 is determined. This representative picture element becomes, namely, the picture element P of the reduced image 11, and stores in a memory in a work area. In the same way, when areas 13-15 are considered, it may suffice that the value of each small area is derived once.

Description

[Detailed description of the invention] [Field of interest of the invention] The present invention relates to an image reduction method.

[Background of the invention]

Conventionally, image reduction methods for reducing binary images include:
The following two methods are known. The second method is to select a specific pixel in the original image according to a predetermined algorithm to determine the density of the missing pixel in the reduced image, and then determine the pixel in the reduced image based on the density of that pixel. be. An example of this first method is
In the shortest distance pixel determination method, when determining the density of a pixel after reduction, it is determined by the density of the W4 element on the original image that is closest to the point where the pixel point after reduction is projected onto the original image. It is something.

However, this shortest distance pixel determination method means that even if most of the surrounding pixels are black (or white), if the closest pixel is white (or black), the pixel after reduction will be white (or black).
The problem is that it becomes. That is, since it is affected by only one pixel information, deterioration of image quality after reduction becomes a problem.

The fg2 method is JWii! Some pixels are selected from J7m and a reduced image is determined as the result of the calculation. As an example of the second method, a typical method is to determine the density of a pixel after reduction by taking a number of pixels around the point where the pixel point after reduction is projected onto the original image, and moving from the projection point to the original image. In this method, a value weighted according to the distance to a pixel on the image is calculated, and the density of both speeds after reduction is determined by comparing that value with a certain threshold value.

However, unlike the first method, this method is not affected by only one pixel information, but it requires searching for the density of surrounding pixels and calculating the weight of that pixel information. First, there was a problem that the processing time was long. Therefore, regarding this second method,
As disclosed in Publication No. 71, a method has been proposed that improves the method of referring to the four pixels around the point where the reduced pixel point is projected onto the original image to obtain a reduced image with less image loss. has been done.

However, the first and second methods described above generally have first-order problems. Figure 6(a).

(b) is an explanatory diagram of the original image reduced to - by the conventional method, and the original image shown in @6 (a) (pixel dotting, ~
I I&) is reduced to - and the reduced image (
It is a diagram made to correspond to pixel points P, Q, R, S). now,
Suppose a black image is drawn on a white background (black is significant information)
. As shown in FIG. 6(b), the pixel P of the reduced image is projected onto the original image.

It corresponds to N element p. Therefore, both prime points 1 included in area 1
1. L, I,, 1. If the pixel P of the -Tsugu reduced image is determined based on the pixel information of , black is significant information, so 1. ,L,L,1. If any one of them is black, the pixel P is determined to be black. Similar algorithms are employed in many image reduction methods to prevent information loss in reduction techniques. Incidentally, the pixel Q is the pixel point I, ~■ in the area 2. (pixel q is the projection point of pixel Q) 1 pixel R is the pixel point in area 3, determined by ~■□2 (pixel r is the projection point of pixel R), pixel S is the pixel point in area 4 Pixel points I13 to I1. l+ (pixel S is the projection point of pixel S).

When the image is reduced by the method described above, the image shown in FIG.
), (b), the original image 5 is converted into a reduced image 6. In the example shown in FIGS. 7(a) and 7(b), there is a diagonal line of 45'' from the upper left of image 5.

This is clearly reproduced on the reduced image 6.

However, in the case shown in FIGS. 8(a) and (b), the original image 7
When converted to the reduced image 8, there is a problem in that the line diagonally 45' from the upper left appears twice on the reduced image 8 (the line appears thicker on the actual image). . The above phenomenon of double diagonal lines occurs with a probability of approximately 50%.

In order to solve this problem, it can be easily inferred that an algorithm for determining reduced pixels by expanding the reference points on the original image over a wider area may be used.

FIGS. 9(a) and 9(b) are explanatory diagrams showing both image reduction using the above algorithm, and show the original image 10 being reduced to one size and converted into a reduced image 11. The point where the pixel P on the reduced image 11 is projected onto the original @ image 10 is 1
This is pixel P(It+). In order to determine the darkness 1!1 (black or white) of pixel P, B surrounding pixels 4X4 (pixel dotting in area 12, ~I
, 9), and is determined using a predetermined algorithm. Similarly, the density of the pixel Q at the reduced image level 11 is the density of the area 13 of the original image 10 (pixel I 1+ IJI Iy+ l1ll
Ding t1+ 11. i+LQ+I is+I
The density of pixel R in the reduced image 11 is determined by 1.
1. , -112), and the density of the pixel S of the reduced image is determined by the original image area 15 (pixel I Ill I!□t
Itst Itst Ll~I 241 1 JT~
I3. ) to be determined.

However, this method searches a 4x4 matrix to determine one pixel in a reduced image, so there are 216 black and white patterns, and the determination algorithm (for example, table search) requires a lot of effort. The problem was that it required time.

[Purpose of the invention]

The present invention was developed in view of the above-mentioned problems of the prior art, and prevents the 45' line of the original image from becoming thicker, reduces the processing time, and produces an image with fewer missing r@ images. The purpose is to provide a reduction method.

[Summary of the invention]

In the image reduction method of the present invention, a pixel (one point) on the reduced image is formed by N points (including the projection point, N is a natural number that is not a prime number) around a projection point (pixel) on the original image. Set an area, divide this area into M small areas (M is a natural number), which is a divisor of N, and divide each small area according to the white/black of each pixel included in each of the M small areas. Find the concentration of the area,
The present invention is characterized in that the density of the area is determined from the density of each small area, and based on this, the white/black of the W4 element on the reduced image is determined.

[Embodiments of the invention]

Next, the present invention will be described with reference to one drawing.

Figures 1(a) and (b) are similar to Figure 8(a) and (b).
FIG. 3 is an explanatory diagram illustrating that the original image 10 is reduced to − and converted into a reduced image 11, similarly to FIG. In determining the pixel P of the reduced image 11, referring to the area 12 (pixel area, ~■0.) of the original image 1o is as shown in FIG. 8(a), (
This is the same as explained in b), but there is a difference in the way the search is performed.

The pixels included in area 12 are 16 points from 11 to ■ts, but these pixels are further divided into small area 20 to area 2.
Subdivide into 4 areas of 3, and add 2x2 to each of these small areas.
contains pixels. According to this figure, the small area 20 has pixels 1. ,I,,I,,1. .

The small area 21 has 13 pixels. Engineering. I LI L, small area 22 is pixel I, 1. . , I, ffl 114%
Small area 23 is pixel Itt+It□t l1st L8
including.

Since each of the small areas 20 to 23 is composed of four 2×2 pixels, there are 24 black and white patterns. This 2
' Recognize the value according to the algorithm determined by (=8) different references and store it in the memory in the work area.

The representative plane element of area 12 is determined based on the values of these small areas 20 to 23. This representative plane element is the reduced image 1
1 pixel P and is stored in the memory in the work area.

Next, when considering area 13, the small areas are area 21,
There will be four areas, areas 23-25. In the same manner as described above, the substitute pixels of the area are determined, and the pixel Q of the reduced image 11 is determined.
is calculated and stored in memory in the work area. that time,
Since the values of the small area 21 and the small area 23 have already been determined when determining the representative plane elements of the area 12, it is sufficient to refer to the memory in the work area.

Considering area 14, there are four small areas: small area 22, small area 23, small area 26, and small area 27. In this case, since the small area 22 and the small area 23 have been determined when determining the representative surface element of the area 12, it is sufficient to refer to the memory in the work area.

Considering area 15, there are four small areas: small area 23, small area 25, small area 27, and small area 28. In this case, the small area 23 is used when determining the representative surface elements of area 12, the small area 25 is used when determining the representative surface elements of area 13, and the small area 27 is used when determining the representative surface elements of area 14.
Since each is required, all you have to do is refer to the memory in the work area.

As is clear from the above example, the value of each small area is.

You only need to ask once. That is, the small area 23.

The algorithm is applied as a small area at the lower right of area 12, a small area at the lower left of area 13, a small area at the upper right of area 14, and a small area at the upper left of area 15, respectively.

In the above explanation, we have described a general method of referring to four 2x2 small areas on the original image when determining one pixel of a reduced image.Next, we will explain an example of that algorithm. . For simplicity, this algorithm uses only two 2×2 small areas in the left and right directions to determine one pixel.

In order to determine the pixel S of the reduced image 11 shown in FIG. 1(b), refer to the small area 1' 27 and the small area 28 of the original image. ), the pixel S is determined by the following logical formula.

S=(了、,nr、、n(了、Jury1z)) +(L、
+I3g) In this example, when referring to each 2×2 small area, for the small area shown in Figure 2, (
! =attnatzn (azt+atz)β=a11
Calculate 0azt to find the values of α and β in advance. Then, calculate these α and β to each small area 2 shown in Figure 1.
By calculating values from 0 to 28 and storing them in the memory in the work area, you can save yourself the trouble of having to calculate them over and over again.

FIG. 3 shows a flowchart for determining and displaying a reduced image. It is assumed that the original image is divided into 2x2 small areas in advance. As shown in FIG. 3, in step 31, the 2X2
In order to input the pixel information of the small area of
Next, in step 3, set the 2×
2 small area pixels (aLlp a, □ in Figure 2)
r azt+a22), and in step 33.

(z,,V,: a1xqa□z口(〒21.+amriβ-・Y):a1alsoguchi821. Next, in step 34, the obtained value α(X, V) rβ.2,) is stored in the memory in the work area.

In step 35, it is determined whether or not the processing of steps 32 to 33 has been completed for all the small areas on the original image. If it is determined that the processing has been completed, the process proceeds to step 36. In step 36.

The leftmost column of pixels of the reduced image is determined, and in step 37, the image 1g to the right of the second column from the left of the reduced image is determined.

In step 38, it is determined whether or not pixel determination has been made for all pixels of the reduced image, and if it has been completed, display of the reduced pixels is executed in step 39, and
Display on the screen above. Above.

The reduction process ends.

Figure 4(a) shows the reduced size after applying this algorithm.
, (b) and FIGS. 5(a) and (b).

4(a), (b) and FIG. 5(a), (b) show the reduced size of a 45° diagonal line on the original image 10, which is a clear reduced image as the reduced image 11. is obtained. In particular, in the conventional technology, when the original image is as shown in FIG. 5(a), the 45° diagonal lines in the reduced image become thicker, as shown in FIGS. 8(a) and (b). However, it can be seen that such a phenomenon does not occur in this example. Therefore, according to the present embodiment, in a relatively short processing time, it is possible to solve the drawback of the prior art, that is, "the line becomes thick", which occurs when reducing the line at an angle of 45 degrees.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, it is possible to prevent the 45° line in the original image from becoming thicker on the reduced image. Moreover, in the present invention.

In an image search that refers to N points (N is a natural number that is not a prime number), N points are divided into M (M is a natural number that is a prime number of N),
By referencing, it becomes possible to perform high-speed processing with 2N/'XMC searches, whereas conventional searches required 29 times. Therefore, a high-quality reduced image can be obtained in a short processing time.

[Brief explanation of drawings]

FIGS. 1(a) and (b) are explanatory diagrams showing one embodiment of the image reduction method of the present invention, FIG. 2 is an explanatory diagram showing one small area in the original image, and FIG. 3 is an explanatory diagram showing an embodiment of the image reduction method of the present invention. Flowchart showing an example of the algorithm used in the implementation, FIG. 4(a)
, (b) and FIGS. 5(a) and (b) are diagrams showing the relationship between the reduced image obtained by the present invention and the original image, and FIG. 6(a),
(b) is an explanatory diagram showing an example of a conventional image reduction method;
Figures (a), (b) and Figures 8 (a), (b) are diagrams showing the relationship between the reduced image and the original image obtained by the conventional image reduction method, and Figures 9 (a) and (b). ) is an explanatory diagram showing an example of a conventional image reduction method. 5.7.10...Original image, 6,8.11...Reduced image. ], 2-15...area, 20-28...small area, P. Q, R, S... Pixel (reduced double-sided) -1'l ql
rr S...pixel (original image), I, ~I 3R...
pixel point.

Claims (1)

[Claims] 1. An image reduction method that refers to a binarized two-dimensional original image and obtains a binarized two-dimensional reduced image, in which pixels on the reduced image are projected onto the original image. Set an area formed by N pixels (including the projection point, N is a natural number that is not a prime number) surrounding a point (pixel), and divide the area into M pixels (M is a natural number) that are divisors of N. Divide into areas, calculate the value indicating the density of each small area according to the white/black of each pixel included in each of the M small areas, and calculate the density of the area from the value showing the density of each small area. An image reduction method characterized in that the whiteness/blackness of a pixel on the reduced image is determined by obtaining a value indicated by the reduced image. 2. The image reduction method according to claim 1, wherein the area is a rectangular area.