JPH0934437A - Generating method of character pattern and character output device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a generating method of character pattern which can realize remarkable compression of a font data quantity without causing deterioration of quality of a character pattern, and provide an character output device which can reduce the hardware cost required for a memory means. SOLUTION: A font ROM 2 stores standard data expressing the shapes of standardized sized ones as for subcharacter patterns such as left-hand radicals, crown parts, and right-hand radicals constituting chinese characters, and character pattern data including part scaling data for changing multiplying factor of standard subcharacter patterns and difference data expressing difference between the basic strokes of a subcharacter pattern to be generated and the basic strokes of a standard subcharacter pattern expressed from the standard data. A CPU 1 changes multiplying factor of the standard subcharacter pattern by the part scaling data included in the character pattern data, correct the standard data based on the difference data in the character pattern data, and hence the shape of the character pattern is painted.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for generating a character pattern and a character output device, which are formed by combining subcharacter patterns such as partial, crown, and whistle that constitute a Chinese character.

[0002]

2. Description of the Related Art For example, in a Japanese word processor, since it is necessary to output a character pattern consisting of many kinds of kanji having a complicated structure, the amount of font data stored in the font ROM is large. Therefore, conventionally, it has been considered to compress such font data as much as possible to reduce the required capacity of the font ROM. Specifically, for example, paying attention to the fact that the Kanji character pattern is a combination of radicals such as bias, crown, and whiskers, prepare unit font data that expresses radicals,
It is considered to output (reproduce) a kanji character pattern by combining these unit font data.

[0003]

However, if the unit font data representing the radical as described above is simply combined, the quality of the reproduced character pattern deteriorates, for example, the balance of the reproduced character pattern is lost. There was a point. In addition, in order to solve such a problem, multiple variation shapes are prepared for each radical, multiple unit font data corresponding to those variation shapes are created, and the character pattern to be output is created. It is considered to combine the unit font data with the optimum balance for each, but even with such a structure, if the variation shape of each radical is small, the quality improvement of the reproduced character pattern is insufficient. However, if the variation shapes are increased in order to cope with such a situation, the problem that the compression of the font data amount becomes insufficient is caused.

The present invention has been made in view of the above circumstances, and an object thereof is to significantly reduce the amount of font data.
The object of the present invention is to provide a method for generating a character pattern that can be realized without deteriorating the quality of the character pattern. Moreover, the hardware required for the storage means can realize such a large compression of the font data amount with a reasonable configuration. An object of the present invention is to provide a character output device capable of reducing the cost.

[0005]

A method for generating a character pattern according to the present invention sets a standard size for a sub-character pattern such as a bias, a crown, and a whistle that composes a Chinese character.
A unique sub-character code is given to each of these standard sub-character patterns, and the shape of the standard sub-character pattern is described as standard data that is expressed by a combination of basic stroke shapes that make up the standard sub-character pattern. Character pattern data for expressing the target character pattern, the subcharacter code of the standard subcharacter pattern required for the character pattern, the size of the standard subcharacter pattern and the subcharacter pattern required for the generation target character pattern Scaling data indicating the size ratio, and difference data expressing the difference between the basic stroke of the sub-character pattern forming the character pattern to be generated and the basic stroke of the standard sub-character pattern represented by the standard data. Described in the state, and generate a character pattern from the character pattern data. In hitting, the standard subcharacter pattern corresponding to the subcharacter code included in the character pattern data is scaled by the component scaling data included in the character pattern data, and the standard data is the difference data in the character pattern data. The shape of the character pattern is drawn by performing correction based on (Claim 1).

With respect to the standard data, regarding the basic strokes forming the standard sub-character pattern, the contour line shapes on both sides of the main body except at least both end portions are used as core line data indicating the shape of a core line passing through the middle of each contour line. Described in combination with the sample value indicating the width of the main body portion corresponding to the core line, the difference data in the character pattern data, the difference with the width of the standard main body portion indicated by the sample value in the standard data The configuration may include a difference sample value indicating

In this case, of the standard data, core data corresponding to a curved basic stroke is approximated by a curve in which the shape of the core is represented by a control point method, and control points on the reference core. The main body data may be described by a combination with a sample value indicating the width of the main body at a position corresponding to (Claim 3).

Further, regarding the standard data, regarding the basic strokes forming the standard sub-character pattern, the contour line shapes on both sides of the main body portion excluding at least both end portions are defined as the contour line data indicating the shape of one contour line. Described in combination with a sample value indicating the width of the main body portion corresponding to the contour line, the difference data in the character pattern data is a difference from the standard main body portion width indicated by the sample value in the standard data. It is also possible to adopt a configuration including a difference sample value indicating

In this case, of the standard data, the one corresponding to the curved basic stroke is the contour line data obtained by approximating the shape of one contour line by a curve expressed by the control point method and the reference. It may be described by main part data which is a combination with a sample value indicating the width of the main part at a position corresponding to a control point on the contour line (claim 5).

In addition to the sub-character pattern, a standard size is set for a character pattern represented by an existing code system, and the shapes of those standard character patterns are set to the basic stroke shape that constitutes the character pattern. Described as standard data expressed by a combination, character pattern data for expressing the character pattern of the generation target, the sub-character code of the standard sub-character pattern required for the character pattern and the existing code of the standard character pattern, Component scaling data indicating the size ratio between these standard subcharacter patterns and standard character patterns and the subcharacter patterns and character patterns required for the generation target character patterns, and the subcharacter patterns and character patterns forming the generation target character patterns Of each basic stroke and the standard data Described in a state that includes the difference data expressing the difference between the quasi-subcharacter pattern and the basic stroke of the standard character pattern, in generating the character pattern from the character pattern data, the subcharacter included in the character pattern data The standard sub-character pattern corresponding to the code and the standard character pattern corresponding to the existing code are scaled by the component scaling data included in the character pattern data, and the standard data is corrected based on the difference data in the character pattern data. By doing so, the shape of the character pattern may be drawn (claim 6).

The character output device according to the present invention comprises a storage means for storing font data and a control means for drawing a character pattern based on the font data read from the storage means. To
The standard size is set for the sub-character patterns that make up the Kanji characters, such as bias, crown, and straw, and a unique sub-character code is assigned to each of these standard sub-character patterns. A first storage area for storing standard data in which the shape of a character pattern is expressed by a combination of basic stroke shapes forming the standard subcharacter pattern;
Sub-character code of the standard sub-character pattern required for the character pattern of the generation target, parts scaling data indicating the size ratio of these standard sub-character pattern and the sub-character pattern necessary for the generation target character pattern, and the generation target And a second area storing difference data representing the difference between the basic stroke of the sub-character pattern forming the character pattern and the basic stroke of the standard sub-character pattern represented by the standard data as character pattern data. In a prepared state, the control means, when generating a character pattern from the character pattern data, a component included in the character pattern data is a standard sub-character pattern corresponding to a sub-character code included in the character pattern data. The scaling is performed by scaling data, and the standard data It is obtained by a structure for drawing the shape of the character pattern by correcting on the basis of the difference data in Ndeta (claim 7).

With respect to the standard data stored in the first storage area of the storage means, with respect to the basic strokes forming the standard sub-character pattern, the contour line shapes on both sides of the main body portion except at least both end portions are defined as the respective contour lines. Of the character pattern data stored in the second storage area of the storage means, which is described by a combination of core line data indicating the shape of the core line passing through the middle of the line and a sample value indicating the width of the main body portion corresponding to the core line. The difference data may include a difference sample value indicating a difference from the standard width of the main body portion indicated by the sample value in the standard data (claim 8).

In this case, of the standard data stored in the first storage area of the storage means, data corresponding to a curved basic stroke is approximated by a curve whose core line shape is represented by a control point method. The main body data may be described by a combination of core line data and a sample value indicating the width of the main body at a position corresponding to a control point on the reference core line (claim 9).

In addition, the standard data stored in the first storage area of the storage means is defined as follows: Described by a combination of contour line data indicating the shape of the contour line and a sample value indicating the width of the main body portion corresponding to the contour line, in the character pattern data stored in the second storage area of the storage means. The difference data may include a difference sample value indicating a difference from the standard width of the main body portion indicated by the sample value in the standard data (claim 1
0).

In this case, of the standard data stored in the first storage area of the storage means, the one corresponding to a curved basic stroke is represented by a control point method in which the shape of one contour line is expressed by a control point method. The main body data may be described as a combination of the contour line data approximated by (1) and a sample value indicating the width of the main body portion at the position corresponding to the control point on the reference contour line (claim) 1
1).

In the first storage area of the storage means, in addition to the sub-character pattern, a standard size is set for a character pattern expressed by an existing code system,
Standard data expressing the shape of these standard character patterns by a combination of basic stroke shapes forming the character pattern is stored, and the standard sub-character necessary for the character pattern to be generated is stored in the second storage area of the storage means. A sub-character code of a pattern and an existing code of the standard character pattern, and parts scaling data indicating a size ratio between the standard sub-character pattern and the standard character pattern and the sub-character pattern and the character pattern necessary for the generation target character pattern, and A sub-character pattern constituting the character pattern to be generated and each basic stroke of the character pattern and difference data expressing a difference between the basic stroke of the standard sub-character pattern and the standard character pattern represented by the standard data. Character pattern data is stored, and the control means controls the character When generating a character pattern from turn data, the standard subcharacter pattern corresponding to the subcharacter code included in the character pattern data and the standard character pattern corresponding to the existing code are scaled by the component scaling data included in the character pattern data. In addition, the shape of the character pattern can be drawn by correcting the standard data based on the difference data in the character pattern data (claim 12).

In the method of generating a character pattern according to the first aspect, sub-character patterns such as partial, crown, and whistle that constitute a kanji are combined to generate a kanji character pattern. In this case, the standard character pattern is used. A sub-character pattern of a typical size is set, the entire standard sub-character pattern is uniformly scaled, and the shape of the basic strokes that make up the standard sub-character pattern is corrected, so that the character pattern is It is supposed to be reproduced.

Specifically, a unique sub-character code is given to each of the standard sub-character patterns as described above, and the shape of the standard sub-character pattern is combined with the basic stroke shape forming the standard sub-character pattern. It is described as standard data expressed by. Further, the sub-character code of the standard sub-character pattern required for the character pattern of the generation target, and component scaling data indicating the size ratio of these standard sub-character pattern and the sub-character pattern required for the generation target character pattern, Described as character pattern data in a state including basic strokes of sub-character patterns constituting the character pattern to be generated and difference data expressing differences between the basic strokes of standard sub-character patterns represented by the standard data. deep.

In generating a Kanji character pattern by combining subcharacter patterns, a standard subcharacter pattern required by the subcharacter code in the character pattern data corresponding to the character pattern to be generated is selected and Uniform scaling of these standard sub-character patterns is performed based on the component scaling data in the character pattern data, and the shape of the basic strokes that make up the standard sub-character pattern is corrected to determine the difference data in the character pattern data. Based on the above, the shape of the character pattern is reproduced and drawn.

In this case, since the character pattern is reproduced through the scaling of the entire standard sub-character pattern and the shape correction of the basic strokes constituting the standard sub-character pattern, the balance of the reproduced character pattern may be lost. It is possible to reliably prevent the situation where the quality is deteriorated, and the standard data, which has a relatively large amount of data, only needs to be prepared for each sub-character pattern one by one. The required character pattern data has a configuration that requires a relatively small amount of data, including sub-character codes, component scaling data, and difference data. Consequently, many variation shapes are prepared for sub-character patterns. Compared with the conventional configuration, it is possible to realize a large compression of font data.

In the character pattern generating method according to the second aspect of the present invention, the standard data as described above is such that the basic stroke forming the standard sub-character pattern is a main body portion excluding at least both end portions, that is, a core line shape and its line. Concerning a portion suitable for describing in a form in which width and width are combined, contour line shapes on both sides thereof are shown, core line data indicating a shape of a core line passing through the middle of each contour line, and width of a main body portion corresponding to the core line A difference sample having a structure described by a combination with a sample value and showing difference between the difference data in the character pattern data and the standard width of the main body indicated by the sample value in the standard data. Since they are represented by values, the standard data and character pattern data are simplified, and in the above-mentioned difference sample value, Becomes smaller the dynamic range to be able to express a small amount of data, it becomes possible to contribute to improving the compression ratio of the font data from this plane.

In the character pattern generating method according to the third aspect of the present invention, among the standard data, the one corresponding to the curved basic stroke is the core line data obtained by approximating the shape of the core line by the curve expressed by the control point method. And the main part data consisting of a combination of the sample value indicating the width of the main part at the position corresponding to the control point on the reference core line, the curve-shaped basic stroke is a relatively simple standard. The data can be represented, and the amount of data can be further reduced.

In the character pattern generating method according to the fourth aspect, the standard data is contour line data indicating a contour line shape on both sides of the main body portion excluding at least both end portions of a basic stroke forming the standard sub-character pattern. The configuration is described by a combination with a sample value indicating the width of the main body portion corresponding to the contour line, and the difference data in the character pattern data is a standard value indicated by the sample value in the standard data. Since it is expressed by the difference sample value indicating the difference from the width of the main body portion, it is possible to contribute to the improvement of the compression ratio of the font data in this case as well.

According to a fifth aspect of the present invention, in the method for generating a character pattern, one of the standard data corresponding to a curved basic stroke is approximated by a curve represented by a control point method to the shape of one contour line. In this case, the standard data is also used because it is described by the main body data that is a combination of the contour line data and the sample value indicating the width of the main body portion at the position corresponding to the control point on the reference contour line. Becomes relatively simple, and the amount of data can be further reduced.

According to a sixth aspect of the present invention, in the method for generating a character pattern, the standard data includes, in addition to the basic stroke shape of the standard sub-character pattern, the standard shape of a character pattern represented by an existing code system. It is described by a combination of basic stroke shapes that compose. In generating a Kanji character pattern by combining these standard sub-character patterns and standard character patterns, the standard sub-character patterns and standard character patterns required by the sub-character codes and existing codes in the character pattern data are selected, and A uniform scaling of
Performed based on the component scaling data in the character pattern data, by correcting the shape of the basic stroke forming the standard sub-character pattern and the standard character pattern, by performing based on the difference data in the character pattern data,
The shape of the character pattern is reproduced and drawn. Therefore, in this case, the data relating to the existing character pattern can be effectively used.

According to the character output device of the present invention, the control means draws the character pattern based on the font data read from the storage means. In this case, the storage means stores the first pattern. Since the font data including the standard data and the character pattern data that requires a small amount of data may be stored in the storage area and the second storage area, the storage capacity required for the storage means is small. And, as a result, the hardware cost can be suppressed.

According to the character output device of the eighth aspect, the standard data stored in the first storage area and the character pattern data stored in the second storage area are the character pattern of the second aspect. The compression ratio of the font data including the respective data is improved, and the storage capacity required for the storage means can be further reduced.

According to the character output device of the ninth aspect, among the standard data stored in the first storage area, the one corresponding to the curved basic stroke is represented by the control point method in the shape of the core line. It is described by the main body data that is a combination of the core line data approximated by the curve and the sample value indicating the width of the main body portion at the position corresponding to the control point in the reference core line. The basic stroke can be expressed by relatively simple standard data, and the storage capacity required for the storage means can be further reduced.

According to the character output device of claim 10,
Since the standard data stored in the first storage area and the character pattern data stored in the second storage area are configured like the character pattern generating method described in claim 4, each of them is stored. Since the compression rate of font data including data is improved, the storage capacity required for the storage means can be further reduced.

According to the character output device of claim 11,
Among the standard data stored in the first storage area, the one corresponding to the curved basic stroke is the contour line data obtained by approximating the shape of one contour line by the curve expressed by the control point method and the reference. Since it is described by the main body data consisting of a combination with the sample value indicating the width of the main body at the position corresponding to the control point in the contour line,
The curved basic stroke can be expressed by relatively simple standard data, and the storage capacity required for the storage means can be further reduced.

According to the character output device of claim 12,
The standard data stored in the first storage area of the storage means and the character pattern data stored in the second storage area are configured as in the method for generating a character pattern according to claim 6. Therefore, the data related to the existing character pattern can be effectively used.

[0032]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment in which the present invention is applied to a Japanese word processor will be described below with reference to the drawings. In FIG. 1 showing an outline of the hardware configuration, a CPU 1 corresponding to the control means in the present invention includes a font ROM 2, a main memory 3, a hard disk device 4, a keyboard 5, a printer 6 and a display corresponding to the storage means in the present invention. Data is exchanged between the video RAM 7 for 7a and other input / output devices (not shown) through the bus line 8 and an interface (not shown).

The font ROM 2 stores font data corresponding to a large number of character patterns in a description format as described later, and outputs the font data requested by the CPU 1 through the bus line 8.

The main memory 3 is provided to store a program for word processing executed by the CPU 1 and to provide a program work area, and a part of the area is for font data read by the CPU 1. It is also used as a buffer memory for developing a contour line into a bit image and developing a bit image for one page. Although the word processor program is configured to be read from the hard disk device 4, for example, a dedicated program memory may be provided.

First to third storage areas are set in the font ROM 2, the sub-character pattern dictionary and the element dictionary describing the standard data Ds are stored in the first storage area, and the second storage area is stored in the second storage area. The storage area stores the character dictionary describing the character pattern data Dc, and the third storage area stores the decoration part dictionary describing the detailed shape data Dm of the decoration part of the basic stroke forming the character. Has become. Of these, the standard data Ds and the detailed shape data Dm have the following description form.

That is, the character patterns that make up the kanji are:
It can be decomposed into multiple types of basic strokes such as "vertical bar", "horizontal bar", "left side", etc. Also, many character patterns have a combination of the above basic strokes, such as bias, crown, and straw. The standard data Ds is composed of a combination of sub-character patterns (including those not specified as radicals in the Kanji characters (for example, the streak part of the character ""), and the standard data Ds is as described above. It represents the shape of the pattern.

In this case, the standard data Ds is composed of subcharacter data Ds0 described in the subcharacter pattern dictionary and main body data Ds1 and DS2 described in the element dictionary.

The sub-character data Ds0 is described in the sub-character pattern dictionary, and for a group of character patterns including Chinese characters, the bias, crown, and
Set the standard size for subcharacter patterns such as Straw and the character patterns expressed by the existing code system, and then set the standard size for each of these standard subcharacter patterns and standard character patterns. On the other hand, it is necessary to add a unique sub-character code (for convenience, the code given to the standard character pattern is also called a sub-character code), and to configure the standard sub-character pattern and standard character pattern as described above. A combination of basic strokes, that is, a plurality of basic component codes indicating the types of the basic strokes are added.

For a standard character pattern that constitutes a Kanji character by itself, an existing code that expresses the standard character pattern is used as the above subcharacter code. Standard character patterns will also be treated as part of subcharacter patterns.

The main body data Ds1 and DS2 described in the element dictionary are basically of a plurality of types (about 21 types in the case of Mincho typeface) necessary for reproducing a group of character patterns forming a Chinese character. For the strokes, set the standard size, give each of these standard basic strokes a unique basic part code, and for each of the above standard basic strokes, remove the end decoration The structure is described as data indicating the contour line shapes on both sides of the part.

Specifically, for example, the Chinese character "sa" shown in FIG. 4 is based on the Mincho typeface, and among the basic strokes that form the character pattern, the uneven portion is a curved line of "left side". It is composed of a basic stroke 11 and a linear basic stroke 12 of a "vertical bar", and each basic stroke 11 and 12 is a decorative portion 11a, 11b and 12a, 12b at both ends (solid line in FIG. 4). , And between the decorative portions 11a and 11b and 12a,
Main body parts 11c and 12c forming each part between 12b
(Indicated by a broken line in FIG. 4).

In describing such a character pattern as font data, the main body portions 11c and 12c in the basic strokes 11 and 12 forming the uneven portion,
The decorative portions 11a, 11b and 12a, 12b are expressed in different formats, respectively, and the straw portion of the character pattern is also expressed in the same format.

Specifically, for example, with respect to the main body portion 11c of the basic stroke 11, the shapes of the contour lines 11d and 11e on both sides of the main stroke portion 11c are defined by a core line 11f (shown by a one-dot chain line) passing through the middle of the contour lines 11d and 11e. It is described as main body data Ds1 that is a combination of core line data indicating the shape and a sample value indicating the width of the main body 11c corresponding to the core line 11f.

In this case, in the main body data Ds1, the core wire 11f is a cubic Bezier curve C (0
Control points p0, p1, p2 for expressing ≤t≤1),
The coordinate point information indicating p3 is used. Further, as the sample value, each control of the parameter t = 0, 1/3, 2/3, 1 in the core line c (corresponding to the core line 11f shown in FIG. 4) represented by a cubic Bezier curve as shown in FIG. Points p0 and p1
, P2, p3, sample values w0, w1, w2, w3 corresponding to the width of the main body portion 11c are used.
In FIG. 6, curves corresponding to the upper and lower contour lines 11d and 11e of the main body portion 11c are indicated by c'and c ", respectively.

On the other hand, regarding the main body portion 12c of the linear basic stroke 12 shown in FIG.
The shapes of 2d and 12e are defined by their contour lines 12d and 12
Core line data indicating the shape of a straight core line (not shown) passing through the middle of e (coordinate point information indicating the positions of two points at both ends of the core line)
And a sample value indicating the width of the main body portion 12c corresponding to the core line, are described as main body portion data Ds2.

On the other hand, the detailed shape data Dm described in the decoration part dictionary includes contour line shapes of the decoration parts 11a, 11b and 12a, 12b corresponding to the standard basic stroke, and a large number of coordinate point information. The structure is represented by the outline method using the provided cubic Bezier curve.

However, the subcharacter data Ds0 actually includes, in addition to the subcharacter code and the basic part code described above, data representing the size of the bounding box corresponding to the corresponding subcharacter pattern. It is composed of.

Then, the subcharacter data Ds0 as described above.
A combination of the standard data Ds consisting of the main body data Ds1 and DS2 and the detailed shape data Dm for one sub-character pattern is prepared for all sub-character patterns required to express a group of Kanji character patterns. The standard data Ds and the detailed shape data Dm are stored in the font ROM 2 as font data for the sub-character pattern dictionary, the element dictionary and the decoration part dictionary which are divided for each standard sub-character pattern.

In the basic stroke such as the "left side" illustrated, the contour lines on both sides can be represented by one main body data Ds1, but the basic strokes forming the character pattern include the contours on both sides. Since some lines require more than one main body data to represent a line, a plurality of main body data are prepared for such a basic stroke.

Of the basic strokes that make up the typeface, basic strokes whose total length is relatively short with respect to the line width, such as "points" and "reverse points", follow the cubic Bezier curve similar to the decorative portion. It can also be described by an outline method.

By the way, since the main body data Ds1 and DS2 in the standard data Ds as described above represent basic strokes having a standard size, for example, the kanji character "Sa" shown in FIG. In the case of generating the basic stroke 13 of "left flare" in which the stroke length existing in the part is longer than the standard, the main body data Ds1 in the standard data Ds is generated.
It is necessary to perform scaling and correction (corresponding to a curved basic stroke), and the scaling and correction method will be described below in advance.

That is, FIG. 7A shows a standard basic stroke relating to "left side", that is, a curved basic stroke including the main body described as the main body data Ds1, on both sides of the main body. As described above, the shapes of the contour lines L1 and L2 are the control points p0, p1, p2, p3 for expressing the core line L3 by a cubic Bezier curve, and the sample value w0 indicating the width of the main body, It is represented by a combination of w1, w2, and w3.

The contour line L of such a standard basic stroke
In order to correct the shapes of 1 and L2 for the relatively long basic stroke shown in FIG. 7B (the contour lines on both sides of the main body are indicated by L1 'and L2', and the core line is indicated by L3 '), , The starting point A and the ending point B (corresponding to the control points p0 and p3, respectively) of the core L3 of the standard basic stroke are changed to the starting points A'and B '(control points p0', p3 'of the core L3') of the core L3 '. The sample value w0, w1, w2, w3 indicating the width of the standard main body portion is scaled while mapping is performed while multiplying the sample value w0 ′ indicating the width of the main body portion of the basic stroke shown in FIG. 7B. , W1 ′, w2 ′, w3 ′ need to be converted and corrected, and in order to perform the correction more accurately, the control points p1 and p2 are also control points according to the actual shape of the core line L3 ′. It is necessary to move to p1 'and p2'.

The starting point A and the ending point B of the core wire L3 are the core wire L3 '.
To map to the starting points A'and B'of
For example, an affine transformation without rotation (parallel movement, combination of enlargement / reduction for each of the x-axis and y-axis) can be used, and in order to perform such transformation, the starting points A ′ and B ′ of the core line L3 ′ can be used. The coordinate point information indicating each position is determined as stroke scaling data for each basic stroke forming each character pattern. Also, the control point p
Control points p1 * and p2 * obtained by applying the above affine transformation to coordinate points p1 and p2 in order to move 1 and p2 to control points p1 ′ and p2 ′ (see FIG. 7B) ) And the control points p1 'and p2' of the movement destination, the movement amount data ΔP1 and ΔP2 (which are vector values as shown in FIG. 7C) are used. Data ΔP
1 and ΔP2 are determined for each basic stroke forming each character pattern.

Further, in order to convert the sample values w0 to w3 indicating the standard width of the main body into w0 'to w3', it is usual to perform the affine transformation similar to the above. In this embodiment, both sample values w0 to w3 and w
Difference sample values Δw0 to Δw3 indicating differences between 0 ′ to w3 ′ are used, and these difference sample values Δw0 to Δw3 are used.
Is determined for each basic stroke forming each character pattern.

The stroke scaling data, the movement amount data and the difference sample value as described above indicate the difference between the basic stroke of the sub-character pattern forming the character pattern to be generated and the basic stroke of the standard sub-character pattern. It is used as the expressed difference data. Incidentally, with respect to the main body data Ds2 corresponding to the linear basic stroke, simple scaling is sufficient, and the data for the scaling is prepared as difference data.

On the other hand, the character pattern data Dc stored in the character dictionary has the following description form.
The character pattern data Dc is for expressing a character pattern to be generated, and is a sub-character code (the sub-character data for specifying the standard data Ds corresponding to the standard sub-character pattern required for the character pattern). Parts scaling data showing the size ratio (corresponding to the sub-character code in Ds0) and the size of the above-mentioned standard sub-character pattern and the sub-character pattern required for the generation target character pattern (corresponding to the magnification in both the x-axis and y-axis directions). And a state in which the difference data expressing the difference between the shape of the basic stroke of the sub-character pattern constituting the character pattern to be generated and the shape of the basic stroke of the standard sub-character pattern indicated by the standard data Ds is included. Has become.

In this case, as shown in FIG. 8, the case of the "tree" bias which is a sub-character pattern will be described as an example. Even if the "tree" bias is the same, the width dimension differs depending on the type of straw. Is. That is, in the example of FIG. 8, the width dimension x1 of the "tree" uneven portion in the "lin" character pattern and the width dimension x2 of the "tree" uneven portion in the "persimmon" character pattern are
There is a relationship of x1> x2. Therefore, the component scaling data as described above is described in the character pattern data Dc, and when the sub-character pattern is developed, the size of the sub-character pattern is corrected based on the component scaling data.

Specifically, in the character dictionary, the character pattern data Dc is described in the format as shown in FIG. That is, the character pattern data Dc describes the required number of the sub-character codes indicating the standard sub-character patterns forming the character pattern to be generated, and changes the standard sub-character patterns in correspondence with the sub-character codes. The component scaling data required twice and a predetermined number of stroke data representing the shape of the basic standard stroke forming the standard sub-character pattern are arranged. In this case, in addition to the basic part code and difference data described above for each basic standard stroke, the stroke data specifies the detailed shape data Dm in the ornament dictionary corresponding to the basic standard stroke. It is configured to include the decoration code of.

It should be noted that each sub-character code and stroke data may be arranged in the stroke order of the corresponding sub-character pattern. However, when the character dictionary is in the push-down stack format,
It is desirable to arrange them in the reverse order.

Further, as in the example of "Hayashi" shown in FIG. 8A, when the sub-character pattern of the partial portion and the sub-character pattern of the straw portion are similar (both are in the fourth stroke). The character strokes differ only in that one is a "point" and the other is a "right side". These subcharacter patterns have character pattern data Dc having the same subcharacter code.
(In this example, the data corresponding to the sub-character pattern of the partial part that appears to have a high appearance frequency)
The stroke data may be omitted by including a minor code expressing the difference between the character pattern data Dc corresponding to the other sub-character pattern (in this example, the portion that is a straw).

Now, FIG. 2 shows a processing routine for expanding a sub-character pattern based on the sub-character pattern dictionary, the element dictionary, the character dictionary and the decoration part dictionary in the font ROM 2 among the control contents by the CPU 1. Only the main parts are shown and will be described below.

That is, when the CPU 1 newly develops a character pattern, first, the character pattern data Dc (see FIG. 3) corresponding to the character pattern is stored in the font ROM.
It is read from the character dictionary in 2 (step A1).

After this, among the standard data Ds in the sub-character pattern dictionary and element dictionary in the font ROM 2, sub-character data Ds0 (sub-character pattern dictionary) corresponding to the sub-character code in the read character pattern data Dc. ), And the main body data Ds1 and DS2 (element dictionary) corresponding to the basic part code included in each stroke data in the read character pattern data Dc are sequentially read, and the read character pattern data Dc from the decoration part dictionary is read. A dictionary pointer is set to sequentially read out the detailed shape data Dm corresponding to the decoration code included in each stroke data (step A2). still,
The dictionary pointer is set to read out the data in the sub-character pattern dictionary, the element dictionary, and the decoration part dictionary in a predetermined order.

Next, the component coefficient (coefficient corresponding to the scaling ratio) corresponding to the component scaling data in the character pattern data Dc read from the character dictionary is set (step A3), and the data restoration processing routine A4 is executed.

In this data restoration processing routine A4, while reading the dictionary content pointed to by the dictionary pointer, the read data is scaled based on the component coefficient, and the scaled data is stored in each stroke in the character pattern data Dc. By repeating the operation of reproducing the basic stroke by correcting with the difference data included in the data, and performing such operation for each sub-character pattern corresponding to all sub-character codes in the read character pattern data Dc. Finally, the character pattern corresponding to the character pattern data Dc is drawn.

In short, according to the above configuration, the font data (standard data D stored in the font ROM 2
s, the character pattern data Dc, and the detailed shape data Dm), the standard data D
scaling of the entire standard subcharacter pattern represented by s,
Since the shape correction of the basic strokes constituting the standard sub-character pattern is performed, the balance of the reproduced character pattern does not become unbalanced as in the conventional configuration. As a result, it is possible to reliably prevent a situation where the quality of the reproduced character pattern is deteriorated.

Moreover, in this case, the standard data Ds having a relatively large amount of data need only be prepared for each sub-character pattern, and at the same time, the required number of character patterns are required. Data Dc
Is a configuration that requires a comparatively small amount of data that includes sub-character codes, component scaling data, and difference data, and as a result, when compared to the conventional configuration in which many variation shapes for sub-character patterns are prepared. , It will be possible to realize a large compression of the required font data. Therefore, the storage capacity required for the font ROM 2 can be reduced, and the hardware cost can be suppressed.

Further, the standard data Ds includes the main part of each basic stroke forming the standard sub-character pattern, that is, the part suitable for describing in a form in which the core wire shape and its line width are combined, on both sides thereof. The contour line shape is described by main body part data Ds1 and Ds2, which is a combination of core line data indicating the shape of a core line passing through the middle of each contour line and a sample value indicating the width of the main body part corresponding to the core line. And the difference data in the character pattern data Dc is represented by a difference sample value indicating a difference from the standard width of the main body indicated by the sample value in the standard data Ds. Standard data D
In addition to simplifying s and the character pattern data Dc, the differential sample value as described above can be expressed with a small amount of data because the dynamic range becomes small, and from this aspect, the compression rate of font data is improved. Therefore, it is possible to contribute to the reduction of the storage capacity required for the font ROM 2.

In this case, of the standard data Ds, the main body data Ds1 corresponding to the curved basic stroke is
It is described by a combination of core line data obtained by approximating the shape of the core line with a cubic Bezier curve represented by the control point method and a sample value indicating the width of the main body portion at the position corresponding to the control point in the reference core line. Therefore, the standard data Ds becomes relatively simple, and the amount of data and hence the storage capacity required for the font ROM 2 can be further reduced.

Further, in addition to the basic stroke shape of the standard sub-character pattern, the standard data Ds includes the standard shape of the character pattern expressed by the existing code system by combining the basic stroke shapes of the character pattern. As described above, the standard character pattern is combined with the standard sub-character pattern to reproduce the target character pattern. Therefore, the data related to the existing character pattern can be effectively used, which is advantageous in practical use. Become. In addition, since the standard sub-character patterns include those that are not specified as radicals in the Kanji dictionary (such as the streak part of the character ""), the versatility can be improved.

Now, in the Japanese word processor configured as described above, the font data stored in the font ROM 2 is created by sequentially decomposing the existing outline font data. A method of creating pattern data will be described.

The above font data conversion work is carried out, for example, using a workstation as a man-machine interface, and the outline of the conversion program is as follows. That is, as shown in an example in FIG. 9, the character pattern to be converted is displayed on the display 14 of the workstation, and when one of the sub-character patterns j1 and j2 to be decomposed is selected by the cursor k, the selection is made. Subcharacter pattern (j2 in the example in the figure)
Is displayed in a state of being surrounded by a frame n including a control point m for alignment.

Next, when the corresponding radical pattern or other Chinese character pattern is input in response to the input of the JIS code by the key operation, the input pattern is superimposed on the sub-character pattern j2 on the display 14. Display with. In this case, the relative size of the sub-character pattern j2 and the input pattern can be adjusted by moving the control point with the cursor k. After this adjustment,
For example, when the return key is operated, the difference value between the sub-character pattern j2 and the input pattern is automatically calculated,
The calculation result is sequentially recorded as font data.

In the above embodiment, the curved stroke 11
11d and 11e on both sides of the main body portion 11c of
The main part data D consisting of a combination of core data indicating the shape of the core 11f and a sample value indicating half the width of the main part 11c corresponding to the core 11f.
Although the configuration is described as s1, the contour lines 11d and 11e have the shape of one contour line represented by, for example, a cubic Bezier curve, and the width of the main body portion 11c corresponding to the one contour line. It is also possible to adopt a configuration in which it is described as main body part data consisting of a combination with a sample value indicating. In this case, it is desirable to use a contour line having a large curvature (the contour line 11c in the case of the curved stroke shown in FIG. 4) as the reference contour line because the error that occurs in the approximate curve becomes smaller.

In the above embodiment, the cubic Bezier curve is used to express the main body data Ds1 and the detailed shape data Dm for expressing the curve. However, instead of this, a spline curve or the like may be used. It is also possible to use a function curve of.

[0077]

As is apparent from the above description, according to the character pattern generating method of the present invention, standard sizes are set for the sub-character patterns such as bias, crown, and whistle that constitute a Chinese character. , A unique sub-character code is given to each of these standard sub-character patterns, and its shape is described as standard data expressed by a combination of basic strokes, and when a character pattern occurs, the standard required for that character pattern. A sub-character pattern is selected by the sub-character code, and the standard sub-character pattern is scaled by component scaling data indicating the size ratio with the sub-character pattern of the generation target, and the standard data is changed to the character of the generation target. The basic stroke of the sub-character pattern that constitutes the pattern and the standard sub-character pattern represented by the standard data. Since the character pattern is drawn by correcting it based on the difference data expressing the difference from this stroke, it is possible to realize a large compression of the font data amount without deteriorating the quality of the character pattern. .

Further, according to the character output device of the present invention, since the font data in a state of being largely compressed as described above may be stored, the storage capacity required for the storage means can be reduced, and the hardware cost can be reduced. Can be suppressed.

[Brief description of drawings]

FIG. 1 is a functional block diagram schematically showing a hardware configuration of an embodiment of the present invention.

FIG. 2 is a flowchart showing a main part of control contents by a CPU.

FIG. 3 is a diagram showing a format example of character pattern data.

FIG. 4 is a diagram showing an example of a character pattern for explaining a description format of main body data that forms standard data.

FIG. 5 is a diagram for explaining a description format of main body data part 1

FIG. 6 is a diagram for explaining a description format of main body data part 2

FIG. 7 is a diagram for explaining a method of scaling and correcting main body data.

FIG. 8 is a diagram showing an example of a character pattern for explaining a difference in size of sub-character patterns.

FIG. 9 is a diagram for explaining the content of font data conversion work.

[Explanation of symbols]

In the drawing, 1 is a CPU (control means), 2 is a font ROM
(Memory means) 3, main memory, 6 printer, 11 curved basic stroke, 12 linear basic stroke, 11
a, 11b, 12a, 12b are decorative parts, 11c, 12c
Is a main body portion, 11d, 11e, 12d and 12e are contour lines,
11f shows a core wire.

Claims (12)

[Claims] 1. A standard size is set for subcharacter patterns such as bias, crown, and whistle constituting a Chinese character, and a unique subcharacter code is given to each of these standard subcharacter patterns. At the same time, the shape of the standard sub-character pattern is described as standard data that is expressed by a combination of the basic stroke shapes that make up the standard sub-character pattern, and the character pattern data for expressing the character pattern to be generated is written as the character pattern. Sub-character code of the standard sub-character pattern required for, component scaling data indicating the size ratio of the standard sub-character pattern and the sub-character pattern required for the generation target character pattern, and the generation target character pattern A basic stroke of a sub-character pattern and a basic stroke of a standard sub-character pattern represented by the standard data Described in a state including the difference data expressing the difference, when generating a character pattern from the character pattern data, the standard sub-character pattern corresponding to the sub-character code included in the character pattern data is the character pattern data. A method of generating a character pattern in which the shape of the character pattern is drawn by performing scaling with the component scaling data included in, and correcting the standard data based on the difference data in the character pattern data. 2. The standard data is a core line indicating a shape of a core line passing through the middle of each of the contour lines on both sides of the main body part excluding at least both end portions of a basic stroke forming the standard sub-character pattern. Described by a combination of the data and the sample value indicating the width of the main body portion corresponding to the core line, the difference data in the character pattern data is the width of the standard main body portion indicated by the sample value in the standard data. The method for generating a character pattern according to claim 1, wherein the character pattern generation method is configured to include a difference sample value indicating a difference of the above. 3. The standard data corresponding to a curved basic stroke includes core line data obtained by approximating the shape of the core line with a curve expressed by a control point method and control points in the reference core line. 3. The character pattern generating method according to claim 2, wherein the character pattern is described by main body data that is a combination with a sample value indicating the width of the main body at a corresponding position. 4. The standard data includes contour line data on both sides of a main body portion except at least both end portions of a basic stroke forming the standard sub-character pattern, and contour line data indicating a shape of one contour line. Described in combination with the sample value indicating the width of the main body portion corresponding to the contour line, the difference data in the character pattern data is the difference from the standard main body width indicated by the sample value in the standard data. 2. The method for generating a character pattern according to claim 1, wherein the method includes a difference sample value indicating 5. The standard data corresponding to a curved basic stroke includes contour line data obtained by approximating the shape of one contour line with a curve represented by a control point method and a contour serving as a reference. The character pattern generating method according to claim 4, wherein the character pattern is described by main body data which is a combination with a sample value indicating a width of the main body at a position corresponding to a control point on a line. 6. In addition to the sub-character pattern, a standard size is set for a character pattern represented by an existing code system, and the shape of the standard character pattern is set to a basic stroke that constitutes the character pattern. Described as standard data expressed by a combination of shapes, character pattern data for expressing the character pattern to be generated is a subcharacter code of the standard subcharacter pattern and an existing code of the standard character pattern required for the character pattern. And component scaling data indicating the size ratio between these standard subcharacter patterns and standard character patterns and the subcharacter patterns and character patterns required for the generation target character patterns, and the subcharacter patterns that form the generation target character patterns and Expressed by each basic stroke of the character pattern and the standard data Described in a state of including the standard subcharacter pattern and the difference data expressing the difference between the standard stroke of the standard character pattern and the basic stroke of the standard character pattern, and in generating the character pattern from the character pattern data, the subpattern included in the character pattern data is included. The standard sub-character pattern corresponding to the character code and the standard character pattern corresponding to the existing code are scaled by the component scaling data included in the character pattern data, and the standard data is based on the difference data in the character pattern data. The method for generating a character pattern according to claim 1, wherein the shape of the character pattern is drawn by performing correction. 7. A character output device comprising a storage means for storing font data and a control means for drawing a character pattern based on the font data read from the storage means, wherein the storage means stores kanji characters. After setting the standard size for the sub-character patterns such as bias, crown, and streak to be configured, assign a unique sub-character code to each of the standard sub-character patterns, A first storage area that stores standard data that represents the shape of the above by a combination of basic stroke shapes that form the standard sub-character pattern, and a sub-character code of the standard sub-character pattern necessary for the character pattern to be generated, Parts scaling data indicating the size ratio of the standard subcharacter pattern and the subcharacter pattern required for the generation target character pattern, A second area in which difference data representing a difference between a basic stroke of a subcharacter pattern forming a character pattern to be generated and a basic stroke of a standard subcharacter pattern represented by the standard data is stored as character pattern data. And, in generating the character pattern from the character pattern data, the control means includes a standard subcharacter pattern corresponding to a subcharacter code included in the character pattern data, and component scaling data included in the character pattern data. A character output device which draws the shape of a character pattern by scaling the standard data and correcting the standard data based on the difference data in the character pattern data. 8. The standard data stored in the first storage area of the storage means includes a contour stroke shape on both sides of the main body excluding at least both end portions of a basic stroke forming the standard sub-character pattern. Character pattern data described by a combination of core line data indicating the shape of the core line passing through the middle of the contour line and a sample value indicating the width of the main body portion corresponding to the core line, and stored in the second storage area of the storage means. 8. The character output according to claim 7, wherein the difference data in is configured to include a difference sample value indicating a difference from the standard width of the main body indicated by the sample value in the standard data. apparatus. 9. Among the standard data stored in the first storage area of the storage means, the one corresponding to a curved basic stroke is obtained by approximating the shape of the core line with a curve represented by a control point method. 9. The character according to claim 8, which is described by main body data that is a combination of core line data and a sample value indicating a width of the main body at a position corresponding to a control point on the reference core line. Output device. 10. The standard data stored in the first storage area of the storage means includes, for a basic stroke forming the standard sub-character pattern, one of contour line shapes on both sides of the main body portion except at least both end portions. Of the character pattern data stored in the second storage area of the storage means, which is described by a combination of the contour line data indicating the shape of the contour line and the sample value indicating the width of the main body portion corresponding to the contour line. 8. The character output device according to claim 7, wherein the difference data is configured to include a difference sample value indicating a difference from the standard width of the main body portion indicated by the sample value in the standard data. 11. The standard data stored in the first storage area of the storage means, which corresponds to a curved basic stroke, is a curve in which the shape of one contour line is expressed by a control point method. It is described by main body data which is a combination of approximated contour line data and a sample value indicating the width of the main body portion at a position corresponding to a control point on the reference contour line. Item 10
A character output device as described. 12. The first storage area of the storage means comprises:
In addition to the sub-character patterns, the standard size is set for the character patterns expressed by the existing code system, and the shapes of these standard character patterns are expressed by combining the basic stroke shapes that make up the character patterns. Stored in the second storage area of the storage means, the sub-character code of the standard sub-character pattern and the existing code of the standard character pattern required for the character pattern to be generated, and these standard characters. Component scaling data indicating the size ratio between the character pattern / standard character pattern and the subcharacter pattern / character pattern required for the generation target character pattern, and each basic stroke of the subcharacter pattern / character pattern forming the generation target character pattern And the standard sub-character pattern and standard character pattern represented by the standard data Character pattern data including difference data expressing a difference from a basic stroke of a character is stored, and the control unit generates a character pattern from the character pattern data, and the control unit includes a sub pattern included in the character pattern data. The standard sub-character pattern corresponding to the character code and the standard character pattern corresponding to the existing code are scaled by the component scaling data included in the character pattern data, and the standard data is based on the difference data in the character pattern data. The character output device according to claim 7, wherein the shape of the character pattern is drawn by the correction.