JPH0378259B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a typesetting method for a phototypesetting machine in which character outlines are stored as character generation data.

Conventionally, this type of phototypesetting machine samples a memorized character pattern specified by outline data at a pitch corresponding to the desired magnification of the output character to determine the outline position, and performs scanning reproduction based on this outline position information. The brightness of the means is controlled on and off, and the image displayed on the reproduction means is projected onto the sensitive material to print a character image at a desired magnification.

As the scanning reproduction means, a cathode ray tube (CRT) having a constant scanning line pitch, an optical fiber tube (OFT), a laser beam printer, etc. are used.

The principles and problems of such conventional character generation and printing methods using phototypesetting machines will be outlined below with reference to FIGS. 1 to 3.

FIGS. 1A and 1B are diagrams schematically showing the relationship between a contour 1 specified by stored data of a conventional phototypesetting machine developed on X and Y coordinates and a reproduced character image 2.

In FIG. 1, lines 3 and 3' are reference lines indicating an example of sampling positions, and lines 4 and 4' are scanning lines on the reproduction means corresponding to the reference lines 3 and 3'.

Further, the interval P between the reference lines 3 and 3' is a sampling pitch that changes depending on the desired magnification α of the reproduced character image, as described later, and the interval P between the scanning lines 4 and 4'.
P' is the scanning line pitch specific to the reproduction means.

In this case, for example, the Y coordinate values y ₁ and y ₂ of the intersection of the reference line 3 and the contour are determined by a contour data decoding unit (not shown) (this decoding operation is referred to as "sampling" in this application), and then this A desired bright line 5 can be obtained on the scanning line 4 by setting the deflection positions on the scanning line 4 corresponding to the sampling coordinate values as brightness on/off points S ₁ and S ₂ .

Similarly, a desired character image 2 is printed by sequentially generating bright lines on different scanning lines while moving the sampling position and projecting the bright lines onto a sensitive material (not shown).

Now, as shown in Figure 1, if the contour data is stored as an area (usually called full-width) where one character can exist, with the length of one side being EM, in units of N x N. For example, data corresponding to 2N scanning lines, N scanning lines, and N/2 scanning lines are obtained in the contour data decoding section, corresponding to sampling pitch P=1/2 unit, 1 unit, and 2 units, respectively. It turns out.

Therefore, the scanning line pitch inherent to the reproduction means
P' is 25μ (constant), and the full-width EM is N=400
Taking the case of decomposition into units as an example, data corresponding to 800, 400, and 200 scanning lines is obtained according to each pitch P, and in the end, the size of the full-width EM is 20 mm, 10 mm, Character images 2 with different magnifications, such as 5 mm, are reproduced.

In addition, in the following explanation, sampling pitch P=
The character image reproduced in the case of 1 unit is called a basic character image, and this state will be explained assuming that the magnification α=1.

In this way, changing the magnification α of the reproduced character image in the conventional example (strictly speaking, changing the magnification in the X direction) is carried out by changing the sampling pitch P, but more generally, P=1/ By sampling the stored data at a pitch of α,
A character image corresponding to α times the basic character image is reproduced and output.

The sampling operation described above is started from the end of the full-width area, and FIG. 1 shows a case where the left end of the full-width area is set as the sampling start position.

Note that to change the magnification in the Y direction (scanning line direction of the reproduction means), the sampled Y coordinate values y ₁ , y _{2 ,} etc. are further multiplied by α, and the deflection position corresponding to this multiplied value is set on the reproduction means. Actual brightness on/off point S ₁ ,
S ₂ is processed.

By the way, phototypesetting machines are usually required to be able to print characters of various different sizes. for example,
In the above embodiment, each integer percentage value between 20% and 200% of the basic character image (20, 21, 199, 200
%), the magnification α must take on 181 different values, and accordingly, the sampling pitch P=1/α must have 181 different values.

FIGS. 2 and 3 are explanatory diagrams showing problems in the principle of character generation in which sampling is performed at different pitches for each magnification α.

FIG. 2A shows the stored data of two character contours 21 and 22 with the same line width "W" and slightly shifted in the X direction, and the stored data in pitches of P=1 unit and 4 units. FIG. 3 is a diagram schematically showing a typical relationship with the reference line 3 when sampling is performed on the X and Y coordinates.

Note that the reference lines numbered 1, 2, etc. along the X-axis in FIG. This does not mean that it is a result.

Also in the case of FIG. 2, the sampling operation itself is performed from the end of the full-width EM to which the character contours 21 and 22 belong. In this way, the sampling itself always starts from the end of the full-width EM, so if the sampling pitch P is the same, that is, if the magnification α is the same, the same output can be obtained. Therefore, although FIG. 2 shows a typical example for explanation, it can be reproduced repeatedly if the magnification α is the same.

For the contour data 21 and 22 of FIG. 2A,
As shown in FIG. 2B, the basic character images 23 and 24 obtained when sampling is performed in units of pitch P=1 are approximately faithful reproductions of the original character contours.

However, when attempting to obtain a character image 1/4 times the basic character image by performing sampling in units of pitch P = 4, the reproduced character images 25 and 26 will consist of 3 and 3 lines, as shown in Figure 2C. The two scanning lines are formed to have different line widths.

In this way, even though the line width images are originally the same, the sampling positions corresponding to each image are relatively different, resulting in, for example, three pairs of line widths as shown in Figure 2C. The phenomenon 2 is extremely noticeable despite the difference in only one scanning line, as it is a thin line of 75μ and 50μ (when the scanning line pitch P' = 25μ). , resulting in a decrease in print quality.

In other words, this not only causes a significant imbalance in line width, but also makes it difficult to reproduce fine lines in the next process of plate making and printing, and in the worst case, may even cause thin lines to be missing. This results in low quality output.

The problem with FIG. 2 described above is related to reproducibility in the X direction, but a similar phenomenon occurs in the Y direction (scanning direction of the reproducing means) for different reasons. This will be explained below based on FIG.

FIG. 3 illustrates the reproduction state of two sets of images having the same length "L" and slightly shifted in the Y direction.

That is, points S ₁ to _{S 4} in the figure are the values obtained by multiplying each Y coordinate value obtained by the contour data decoding unit by α as described above for the magnification changing process (hereinafter referred to as the variable magnification Y value).
This point corresponds to .

Further, the length "U" in the figure is a unit amount of positioning accuracy in the scanning direction, and positionable positions 3 are set discretely apart by the length of this "U".
The grid point Q, which is the intersection of the scanning line 1 and the scanning line 4, is the point where the brightness can actually be controlled.

Therefore, the scaling Y value is rounded off, rounded up,
By performing appropriate calculations such as rounding down, the point normalized to the positionable position 31 becomes the actual brightness on/off point.

FIG. 3 exemplifies the case of rounding, and in this case, the points S ₁ to _{S 4} in the calculation are as follows:
As shown by the arrows in the figure, the images are normalized to points S' ₁ to S' ₄ , and as a result of controlling the brightness on and off at each of these points S', images (bright lines) 32 and 33 of different lengths are displayed on the reproduction means. It will be projected.

Note that a similar phenomenon may occur in rounding up and rounding down processing, but since the principle is the same as the above explanation, the explanation will be omitted.

As described above, with respect to the scanning direction of the reproducing means, images of originally the same length can be reproduced as images of different lengths due to normalization to the positionable position 31. This phenomenon ultimately causes the same adverse effects as in the X direction described with reference to FIG. 2, and becomes a factor in deteriorating print quality.

The above explanation based on Figures 2 and 3 outlines the conventional problems using simply abstracted images as an example.Next, we will discuss actual character images affected by the above phenomenon. FIG. 4 illustrates a comparison between the character image and the desired character image, and how the occurrence of the phenomenon changes as the magnification changes.

Each character shown in Figure 4 shows an example of the actually printed character enlarged approximately 14.5 times.
A, C, E, and F are enlarged views of character images printed by a conventional phototypesetting machine, and B, D are originally desired character images in comparison with A and C.

The arrows in the figure indicate the locations where the above-mentioned phenomenon occurs, and significant variations in line width can be seen in both the vertical and horizontal lines. In particular, arrows a and b, c and d,
The points pointed out in e and f, etc. originally have the same line width, and have an extremely negative effect on print quality.

Also, although the character images shown as E and F are the same character, in E the upper and lower two strokes are thicker as shown by the arrows, and conversely in F the center stroke is thicker. This suggests that the ideas behind the character design were not utilized at all.

Moreover, in actual use, the number of commonly used characters in Japanese is several thousand characters per typeface.

The stroke structure of each character, especially kanji, is much more complex than that of European characters, and in the Ming style typeface, thin lines of about 50 to 70 μm are commonly used in ordinary books.

As described above, the value of the sampling pitch P also varies depending on the various magnifications α.

For these reasons, it is difficult to specify in which part of which character and at which magnification the line width difference phenomenon described above occurs due to a change in magnification.
Moreover, even if it were possible to identify it, conventional phototypesetting machines could not correct it at all.

The present invention has been made to eliminate the above-mentioned drawbacks of conventional phototypesetting machines that store character outlines as character generation data. The principle of the present invention will be explained below.

The first invention of the present application classifies the desired magnification α into a plurality of ranges (for example, about 4 ranges).
group into. Also, the contour memory data is sampled at a sampling pitch _PR specific to each range. This results in decoding data, ie contour position information, corresponding to a unique number of scanning lines for each range. Then, the same contour position information is used for various magnification changes belonging to one arbitrary range, and the scanning line pitch of the reproduction means is
By changing P' according to the ratio between the specific sampling pitch P _R and the theoretical sampling pitch P = 1/α corresponding to the desired magnification α, various magnification changes within the same range can be achieved. This makes it easier to identify the character type and location where the line width difference phenomenon occurs.

Furthermore, a second invention related to the present application is such that the structure of the data stored in the storage unit is partially selective,
If the line width difference phenomenon occurs when a specific range (particularly the range to which a small magnification belongs) is adopted in the first invention, apart from the contour data that is involved in the occurrence of the line width difference phenomenon, only when the range is used The selected modified contour data is added, thereby completely preventing the occurrence of the line width difference phenomenon at low magnification.

Hereinafter, the invention according to the present application will be explained based on examples.

FIG. 5 is a diagram showing an example of the configuration of a phototypesetting machine that realizes the first invention, in which 51 is a magnification command unit that commands a desired magnification α of output characters, and 52 is a diagram shown in FIG. 53 is a range selection unit that classifies the command magnification α in order of magnitude, groups it into a relatively small number of ranges (four ranges in the case shown), and outputs the corresponding range data R, and 53 corresponds to each range. Among the sampling pitches set appropriately in advance,
This is a sampling pitch command unit that outputs a pitch _PR specific to the range R selected by the range selection unit 52 to the decoding unit 54.

In addition, although FIG. 6 shows the case where sampling pitch 1/2, 1, 2, and 4 units are set as fixed pitches corresponding to each range R=1 to 4, the sampling pitch command The values preset in the section 53 are not necessarily limited to these, and may also include the type of range and the magnification α belonging to it.
However, the present invention is not limited to the illustrated example.

Furthermore, the decoding section 54 shown in FIG. 5 samples the contour data from the contour data storage section 55 at the command pitch P, and outputs the Y coordinate value of the contour at each sampling position. Therefore, when changing the magnification belonging to the same range, the same outline position information is used regardless of its value, and a character image is formed with a unique number of scanning lines for each range.

Furthermore, 56 is a multiplication unit that multiplies the Y coordinate value supplied from the decoding unit 54 by a magnification α and outputs the scaled Y value, and the obtained scaled Y value is set in a first-in first-out register (FIFO) 57. be done.

Reference numeral 58 denotes a scanning line pitch command section which outputs the scanning line pitch P' necessary to reproduce a character of a desired size according to the number of scanning lines specific to the assigned range as described above. The scanning line pitch command unit 58 is instructed to specify a desired magnification α, and thereby determines the ratio between the specific sampling pitch P _R and the theoretical sampling pitch P=1/α corresponding to the desired magnification α. A scanning line pitch P' corresponding to the scan line pitch P' is output. or,
59 is a pitch commanded by the command unit 58.
It is a variable pitch scanning type reproducing means that performs scanning at P', 60 is a brightness control section, and 61 is a deflection control section.

The brightness control unit 60 receives current scanning position data supplied from the Y deflection control circuit 62 and the FIFO
When the magnification Y values sequentially read out from . Then, this reproduced image is projected onto the photosensitive material 65 via the optical system 64,
Execute printing.

As described above, the phototypesetting machine shown in FIG. 5, which implements the first invention of the present application, reads specified data from the outline data storage unit that stores characters as outline data, and performs calculation based on the read data. This is a typesetting method for a phototypesetting machine in which desired characters are printed on a photosensitive material by supplying character outline position information to a scanning reproduction means, and a desired magnification α for printing on the photosensitive material is given. Then, it is determined to which range the desired magnification α belongs among a plurality of ranges grouped in order of magnitude of magnification, and the data is read from the storage unit using a sampling pitch PR unique to each range. is sampled, and for various magnifications belonging to any one range, the same contour position information obtained by the sampling is used to control the brightness modulation position in the reproduction means, and further, the scanning line of the reproduction means is controlled. The pitch P′ is changed according to the ratio between the specific sampling pitch P _R and the theoretical sampling pitch P 〓 corresponding to the desired magnification α, so as to cope with various changes in magnification within the determined range to which it belongs. Because it is becoming
Compared to the conventional example, it is possible to limit the character type and the location where the line width difference phenomenon occurs, and therefore, it is also easier to specify it.

Next, a second invention related to the present invention will be explained based on an example.

This second invention is applicable to either a method of approximating the outline of a character with a set of straight line vectors (hereinafter simply referred to as vectors) or a method of approximating it with a set of curved lines. In the following description, an example of the former case will be described.

FIG. 7A shows contour data when the second invention of the present application is applied to a typical example of two character images 71 and 72 of the same shape that are slightly misaligned in both the X and Y directions. The relationship between the content and the reference line 3 for sampling the contour data, for example, in units of pitch P=4, is schematically developed on the X and Y coordinates, following the example of FIG. 2A.

In FIG. 7A, V ₁ to _{V 8} are vectors that approximate a desired character image, and in the conventional method, the X-direction component, Y-direction component, length, inclination, etc. of each of these vectors, etc. As a set of data specifying the outline character data, the outline character data is stored in the outline data storage unit 5 in the format shown in FIG. 8A, for example.
I remember it in 5. Note that FIG. 8A and FIG. 8B, which will be described later, are shown with control codes necessary for decoding these vector data omitted.

Therefore, in the conventional method, when a 1/4 times larger character image is reproduced based on decoding data sampled at a reference line position of pitch P = 4 units, the "○" mark in FIG. 7A or Bright lines are formed corresponding to the points indicated by "●" marks, and reproduced images 73 and 74 as shown in FIG. 7B are obtained.

As is clear from FIG. 7B, although the reconstructed image by the conventional method has the same shape as the original image, due to a slight difference in the sampling position in the X direction, and in the Y direction. Due to normalization processing such as rounding, the reproduced images 71 and 7 have different shapes with different line widths in both the X and Y directions.
It becomes 2. As described above, this phenomenon causes the inconvenience as illustrated in FIG. 4.

By the way, the vector indicated by the broken line in Figure 7A
In the second invention of the present application, V' ₂ and V' ₄ are determined by a specific range (in this case, the sampling pitch P=
When the range 4) shown in FIG. 6 corresponding to 4 is given a selection command, it is a vector that is selectively used in place of the conventional vector V ₂ and vector V ₄ , as illustrated in FIG. 8B, for example. It is added by a format such as

In this Figure 8, the selection code "00" indicates selection for all ranges, "01" indicates selection for ranges other than the specific range, and "10" indicates selection for the specific range. However, the setting of the selection data is not necessarily limited to the type shown in the drawings, and may be any type that can substantially perform the selection as described above.

Therefore, in the second invention of the present application, when a 1/4 times larger character image is reproduced based on decoding data sampled at a reference line position of pitch P = 4 units, "●" in FIG. A bright line is formed corresponding to the point marked with "," and a reproduced image 7 as shown in FIG. 7C is formed.
5,76 is obtained.

As is clear from FIG. 7C, the reproduced image according to the second invention reproduces the original image as faithfully as possible, and the reproduced images 75, 7 have the same shape.
6, which makes it possible to eliminate the inconvenience illustrated in FIG.

Furthermore, as is clear from the comparison of Fig. 8A and B, there is no need to significantly change the contents of the stored data compared to the conventional method, and moreover, it is necessary to consider only the parts where noticeable inconvenience occurs in a specific range. Since this is sufficient, implementation of the second invention has little effect on the character data compression rate.

As described above in detail, the second invention of the present application is
Specified data is read from a contour data storage unit that stores characters as contour data, and character contour position information determined based on the read data is supplied to a scanning type reproduction means to print a desired character on a photosensitive material. This is a typesetting method of a phototypesetting machine in which characters are printed one by one, and when a desired magnification α for printing on a photosensitive material is given, which range is selected among multiple ranges grouped separately in order of magnification. Determine whether the desired magnification α belongs, sample the data read from the storage unit at a sampling pitch P _R unique to each range, and calculate the sampling pitch for various magnifications belonging to any one range. The brightness modulation position in the reproduction means is controlled using the obtained same contour position information, and the scanning line pitch P' of the reproduction means is made to correspond to the unique sampling pitch P _R and the desired magnification α. The data is changed according to the ratio with the theoretical sampling pitch Pα to cope with various magnification changes within the determined range to which it belongs, and the structure of the data stored in the storage unit is partially selective, so that Modified contour data is added that is selected only when the image data is adopted, thereby completely preventing the occurrence of the line width difference phenomenon when the magnification is small.

[Brief explanation of drawings]

Figures A and B are memory contours 1 of a conventional phototypesetting machine.
Figures 2A-C and 3 are explanatory diagrams illustrating problems with the conventional character generation principle, and Figures 4A-F are diagrams schematically showing the relationship between the reproduced character image 2 and the reproduced character image 2. A diagram illustrating a comparison between an actual character image and a desired character image, and a change in character image due to a change in magnification,
FIG. 5 is a diagram showing an example of the configuration of a phototypesetting machine that realizes the first invention, FIG. 6 is a diagram illustrating the relationship between command magnification α, range R, and sampling pitch P, and FIGS. FIGS. 8A and 8B, which are diagrams for explaining the operation when the second invention is applied, are diagrams illustrating the data format of contour data. 3... Reference line, 4... Scanning line, P... Sampling pitch, P'... Scanning line pitch, U... Positioning unit amount, Q... Grid point, α... Magnification.

Claims (1)

[Scope of Claims] 1. Specified data is read from a contour data storage unit that stores characters as contour data, and character contour position information obtained based on the read data is supplied to a scanning type reproduction means. It is a typesetting method of a phototypesetting machine that prints desired characters on a photosensitive material, and when a desired magnification α for printing on the photosensitive material is given, multiple ranges are grouped separately in order of magnification. Find to which range the desired magnification α belongs, and with a sampling pitch P _R unique to each range,
sampling the data read from the storage unit, and controlling the brightness modulation position in the reproduction means using the same contour position information obtained by the sampling for various magnifications belonging to any one range; , the scanning line pitch P' of the reproducing means is changed according to the ratio between the specific sampling pitch P _R and the theoretical sampling pitch P 〓 corresponding to the desired magnification α, and A phototypesetting method characterized by adapting to various magnification changes. 2. Read specified data from the contour data storage section that stores characters as contour data, supply character contour position information obtained based on the read data to a scanning reproduction means, and print the desired character on the photosensitive material. This is a typesetting method of a phototypesetting machine that prints on a photosensitive material, and when a desired magnification α to be printed on a photosensitive material is given, which range among multiple ranges grouped separately in order of magnification is selected. Find out whether the desired magnification α belongs to, and with a sampling pitch P _R unique to each range,
sampling the data read from the storage unit, and controlling the brightness modulation position in the reproduction means using the same contour position information obtained by the sampling for various magnifications belonging to any one range; , the scanning line pitch P' of the reproducing means is changed according to the ratio between the specific sampling pitch P _R and the theoretical sampling pitch P 〓 corresponding to the desired magnification α, and A phototypesetting method that copes with various magnification changes, has a partially selective structure for the data stored in the storage section, and has added modified outline data that is selected only when a specific range is adopted.