JP2000242633A - Typeface data preparation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily prepare typeface data with a feature. SOLUTION: In this typeface data preparation method, a normal outline font is analyzed by an analysis software 4 and a knowledge type font data base 5 including basic knowledge about the writing order of a center line and the constituting elements of a left side and a right side, etc., is constructed. Handwritten samples 2 for plural characters set beforehand are read by a scanner 3, the features are extracted by the analysis software 4 and a feature data base 6 is constructed. The center line is automatically recognized from the image of a handwritten character and compared with the shape of the center line to be a reference stored in the knowledge type font data base 5 and the feature is extracted. An operation software 7 reads the knowledge type font data base 5 and the feature data base 6, deforms the shape of the character and generates the character corresponding to the attributes.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a typeface data creating method for displaying or printing characters as font data in a specific typeface.

[0002]

2. Description of the Related Art Conventionally, printed or handwritten characters are each characterized by a unique typeface. As for the characters to be printed, a large number of typefaces are set and used as typefaces, including basic typefaces such as Mincho and Gothic. In recent years, many types of fonts used for display on computer devices and the like have been developed using outline fonts.
Even in the case of handwritten characters, it is naturally written in a typeface that reflects the personality of the writer.

[0003] Today, the chances of creating letters and texts by handwriting are decreasing, and texts are created by reading software such as word processors into a dedicated computer device such as a word processor or a general-purpose personal computer. In many cases, printing or printing is performed. In a word processor, a personal computer, or the like, a sentence created on a screen can be displayed in the middle of creating a sentence, or a sentence can be printed during or after the creation of a sentence.

[0004]

[Problems to be Solved by the Invention] Currently 40% of ordinary households
Have purchased a personal computer, and it is said that the most common use form is as a word processor. Dedicated word processors have also become widespread. However, texts and letters created using a personal computer or word processor should have the same typeface no matter who uses them, and should eliminate the possibility of giving the impression of "bad taste" or "heartlessness". Can not. For this reason, many people create special letters such as New Year's cards by handwriting with a brush. However, address books such as New Year's cards issued to many people require a great deal of effort, and there is a strong demand for "I want to easily write New Year's cards with my own handwriting."

If font data provided in a personal computer or a word processor is designed to be able to represent personal handwriting, even if a sentence is expressed using a personal computer or a word processor, the personality of the individual is also reflected in the typeface. It is possible to avoid giving a uniform impression. However, in the kanji cultural sphere such as Japan, many characters are required so that they cannot be compared with the alphabet, and it is very difficult to create font data in a typeface that reflects individual characteristics. In order to use personal handwriting as it is as font data, there are 1945 characters for even common kanji,
Kanji, which would be 2229 characters when including personal name kanji, is handwritten one by one and read with a scanner etc.
Must be converted to font data. If data is converted into an outline font, the size of characters can be changed in various ways. However, fonts of characters having different typefaces, such as line, square, and cursive, need to be created even with the same characters.

It is an object of the present invention to provide a typeface data creating method capable of creating font data having specific characteristics such as personal handwriting without directly inputting many characters one by one. is there.

[0007]

According to the present invention, for a group consisting of a plurality of characters, for each character, a guideline representing a reference shape, a writing order of the guideline, and a unit element constituting the character are defined. At the database construction stage that collects basic knowledge and builds a knowledge database, and the guidelines and outline of typeface data to be created,
Refer to the typeface characteristic setting stage to set the basic characteristics and the knowledge database constructed in the database construction stage,
The guideline that is the standard for each character is divided into a guideline transformation stage that transforms the stroke order and unit elements according to the basic characteristics set in the font type feature setting stage, and a guideline for each character transformed in the guideline transformation stage. A typeface data creating step of forming typefaces in accordance with the basic features of the outline set in the typeface feature setting step to create typeface data representing the shape of the character. .

According to the present invention, a knowledge database is constructed in a database construction stage by collecting basic knowledge including a guideline representing a basic shape of a character, a writing order of the guideline, and a unit element constituting the character. You. In the typeface characteristic setting stage, basic characteristics of guidelines and outlines of typeface data to be created are set. Refer to the knowledge base constructed during the database construction stage, and set the basic guidelines for each character.
With respect to the writing order and unit elements, the guideline is transformed in the guideline transformation stage in accordance with the basic features set in the font feature setting stage. With respect to the guideline of each character deformed in the guideline deformation step, the outline is formed in accordance with the basic characteristics of the outline set in the font characteristic setting step, and typeface data representing the shape of the character is created in the typeface creation step.
With regard to the characters collected in the knowledge database, many types of typeface data can be created only by setting basic guidelines and outline features at the typeface characteristic setting stage.

In the present invention, in the feature setting step, a plurality of characters are selected so that a basic unit element set in advance is included in any of the characters, and handwriting is performed on the selected characters. Create a sample, read the created handwritten sample, refer to the knowledge database, extract guidelines from the sample, based on the comparison result between the extracted guidelines and the guidelines included as basic knowledge in the knowledge database Basic characteristics of the guideline of the typeface data to be created are set.

According to the present invention, a plurality of characters are selected so that a basic unit element set in advance is included in any one of the characters, and a handwritten sample is prepared using the plurality of selected characters. , Read the sample, refer to the knowledge database, extract the guideline from the sample, and based on the comparison result between the extracted guideline and the guideline included as basic knowledge in the knowledge database, the guideline of the typeface data to be created Since the basic characteristics are set, it is possible to create typeface data for all of a plurality of characters to be output to the base group while reflecting the handwritten characteristics obtained for the sample. By changing the basic characteristics of the guideline set in the font characteristic setting stage, it is possible to create font data having different fonts such as, for example, a square style and a line style. Further, by changing the basic features of the outline, it is possible to easily convert a typeface such as a Gothic style or a Mincho style.

In the feature setting step of the present invention, a basic feature of the contour is set as a rule for forming a contour of a character according to a guideline.

According to the present invention, in the typeface characteristic setting step,
Since the basic characteristics of outlines are set as rules for forming outlines of characters according to the guidelines, for example, create typeface data such as Gothic font with a constant thickness, or make the font thicker and thinner. With such rules, it is possible to create typeface data close to Mincho style.

Further, in the present invention, the rule is defined in relation to a writing instrument for writing characters.

According to the present invention, rules for forming a contour of a character according to a guideline are defined in relation to a writing implement for writing a character, so that typeface data as if handwritten using a writing implement is created. Can be.

Further, according to the present invention, for a group of a plurality of characters, basic knowledge including a guideline representing a reference shape, a writing order of the guideline, and a unit element forming a character is collected for each character. Then, as a database construction stage for constructing a knowledge database and a font type feature setting stage for setting basic characteristics for guidelines and outlines of typeface data to be created, one of the basic unit elements set in advance is A plurality of characters are selected so as to be included in the character, a handwritten sample is created for the selected characters, the created handwritten sample is read, the guidelines are referred to the knowledge database, and guidelines are created from the sample. The extracted guidelines are compared with the guidelines included as basic knowledge in the knowledge database. Based on, it sets the basic features of the guidelines typeface data to said forming,
In relation to the writing implement that forms the outline of the character according to the guidelines, a typeface characteristic setting step of setting basic characteristics of the outline and a knowledge database constructed in the database construction stage are referred to. The guideline transformation stage that transforms the reference guideline according to the basic characteristics set in the font design feature setting stage with respect to the script order and unit elements, and the font design feature setting for the guideline of each character transformed in the guideline transformation stage Forming a contour in accordance with the basic characteristics of the contour set in the step, and forming a font data representing the shape of the character. It is a readable medium.

According to the present invention, it is not necessary to create and prepare all character data of a number larger than the number of characters included as a sample based on a sample of a character created by handwriting. Labor can be greatly reduced.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a schematic system configuration from character generation to output as an embodiment of the present invention. The normal outline font 1 is an outline font for a typeface provided as standard in a personal computer, a word processor, or the like, and is formed in advance in the form of vector data indicating an outline shape corresponding to a character code for each character. ing. The handwritten sample is formed by handwriting a predetermined character by hand by an individual in advance. The handwritten sample is read by the scanner 3 as image data.

The analysis software 4 automatically recognizes the center line of the shape and the like as a guideline for each character of the normal outline font 1 and constructs a standard font dictionary. Based on the continuous state about the center line of the shape,
It is divided into radicals, such as flat and side, which are the unit elements that make up a character. In addition, for each character, the stroke order is input in an interactive format. Since the handwritten sample 2 read by the scanner 3 is determined in advance as to what character is to be written, the center line of the code corresponding to the normal outline font 1 and the center line of the handwritten sample 2 Then, the difference between the center lines is extracted and recognized as a feature of the handwritten character. It automatically recognizes the distance from the center line to the outline as the thickness, and also recognizes features such as stop and splash, such as the start and end points of the center line. In addition, the automatically recognized element can be modified interactively.

Based on the result of the analysis by the analysis software 4, a knowledge type font database 5 and a feature database 6 are constructed. The knowledge-based font database 5 contains the typeface of each character in the group of characters that can be created.
Components represented by the basic center line, radicals,
Basic knowledge such as a stroke order is stored corresponding to the character code. The feature database 6 has a feature on how the center line of the typeface data to be created is deformed from the center line of each character stored in the knowledge-based font database 5, and the type data of the typeface data based on the center line. The feature about the change in thickness when forming the contour is stored.

The operating software 7 reads the knowledge-based font database 5 and the feature database 6 and, for each character stored in the knowledge-based font database 5, stores the basic shape of the character based on the center line in the feature database 6. The font data is deformed based on the stored features, and font data is generated so as to represent the shape of the character according to the thickness information stored in the feature database 6. When a character to be output is input using a code such as text data, the character corresponding to the code is laid out in consideration of the stroke order and the like, and output to the printer 8 or the like. The printer 8 can output characters in a typeface in which characteristics such as handwriting expressed in a handwritten sample are reflected on the recording paper 9 or the like.

FIG. 2 shows a basic concept of extracting a center line from a character outline. Assuming a circle inscribed in the figure represented by the outline, the center of the circle is connected, and the guideline 10 is extracted.

FIG. 3 shows the basic idea that the analysis software 4 of FIG. 1 extracts a center line from a character typeface as a guideline 10 and recognizes it as a "knowledge typeface" different from a conventional typeface. As shown in FIG. 3A, a conventional typeface is handled in an outline font data format in which the shape of an outline is represented by vector data. As shown in FIG. 3B, in the knowledge typeface of the present embodiment, the center line of the outline font is extracted as the guide line 10, and is treated as vector data representing the extracted guide line 10.

FIG. 4 shows the basic concept of recognizing the features of the guideline extracted from the image obtained by reading the handwritten sample 2 of FIG. The example guideline 10 is automatically extracted from the normal outline font 1 based on the corresponding character code.
The handwritten guideline 11 is extracted from the character image 11a obtained by reading the handwritten sample 2 by the scanner 3 based on the contour 11b formed by the image processing as shown in FIG. When the guideline 10 of the example is compared with the guideline 11 of the handwriting, the feature of the handwritten character is recognized at least for the character of “God”.

FIG. 5 shows an example of the handwritten sample 2 of FIG. The handwritten sample 2 includes a plurality of kanji selected to include hiragana, katakana, and various radicals. The total number of characters is, for example, about 150 characters.
Typeface data for zero or more characters can be created.

FIG. 6 shows an example of the radicals of the kanji which are radicals. The character "God" is formed from a flat 12 called "Ne" called Shimesen and a side 13 called "Shin". Since the flat 12 and the side 13 are also used as components of other kanji, respectively,
If the character is used as a handwritten sample, it is possible to recognize features applicable not only to the character “Kami” but also to other Chinese characters that include the flat 12 and the side 13 as constituent elements.

FIG. 7 shows an example of another radical of the flat 12 and the side 13 shown in FIG. The other radicals include "Ukanuri" as "Kanri" shown in FIG. 7A, "Satsukoro" as "Tsukuashi" shown in FIG. 7B, and "Saret" shown in FIG. 7C. (E), "Shinnyo" as "Nyo" shown in (d), and "Mongae" as "Kamae" shown in (e). In the handwritten sample 2 shown in FIG. 5, characters are selected so that the radicals do not overlap much.

When the feature database 6 based on the handwritten sample 2 as described above is used, even characters not included in the handwritten sample 2 can be converted into a combination of radicals by using typeface data having similar characteristics. Can be output. In some cases, radicals can be formed by deformation between similar radicals without being directly input. For example, from "Shimesuhen" to "Koromohen" or "Madore"
To "Yamaidare". Further, when the concept of the present invention is applied, various attributes can be given to the guideline data extracted from the handwritten character, and the character can be displayed in a different expression form.

FIG. 8 shows another embodiment of the present invention.
The concept of changing the contour shape by changing the attribute for one guideline will be described. As shown in FIG. 8A, for example, a center line 21 is determined from a handwritten character 20, and
The guideline data 22 as shown in (b) is extracted. In FIG. 8, a relatively simple character “large” is shown as an example, but for a character having a more complicated configuration, information on radicals such as flat and side is added to the guideline data 22 as data. Further, the order of the strokes is added to the guideline data 22 as basic data. FIG.
As shown in (c), the guideline data 22
When a square Gothic attribute having a constant thickness is given, an outline font 23 similar to a square Gothic font can be obtained. As shown in FIG. 8 (d), when a brush character attribute whose thickness changes is given to the guideline data 22, an outline font 24 similar to a brush character such as a text font can be obtained. As shown in FIG. 8E, when the attribute of the marker pen is given to the guideline data 22, an outline font 25 having a different thickness depending on the direction written by the marker pen can be obtained. That is, the outline font 25 can be formed along the same locus as when the rectangular pen tip of the marker pen is moved obliquely to the paper and moved along the guide line data 22. As described above, by providing attribute information to the guideline data 22, it is possible to form infinitely many types of outline fonts.

FIG. 9B shows a state in which an outline font is formed by changing the type of the brush as shown in FIG. 9A for the same guideline data. See Handwriting sample 2 for how to apply force when using a brush.
Is collected as operation information when extracting the guideline data. As attributes of the brush, for example, hardness, length, and the like are used.

In the embodiment shown in FIG. 1, a normal outline font 1 is automatically analyzed by analysis software 4 and a knowledge type font database 5 is constructed, but the knowledge type font database 5 is commonly used. Therefore, it can be created in advance and used. Also, instead of the handwritten sample 2, a newly designed typeface sample can be input. In this way, without creating data in a new typeface for all characters, all characters for which basic knowledge is provided in the knowledge-based font database 5 only by creating a common sample are written in a new typeface. Data can be obtained.

When the present invention is applied as described above, one character is represented as a font based on a plurality of guideline groups, so that the stroke order can be held as data. When the stroke order is used, it is possible to perform character arrangement and character deformation utilizing the connection of characters unique to Japanese.

By giving the character thickness according to the guideline as the attribute information, the external shape of the character can be freely controlled, and a change in the attribute of the writing implement from a writing character to a hard brush can be generated based on the same guideline information. Infinite character shapes can be created with font data according to one guideline.

The guidelines are classified as a few image elements, and individual characteristics of each image element are analyzed as data, and the characteristic database 6 is constructed. . The guideline representing the basic font contained in the knowledge-based font database 5 is automatically edited according to the feature data contained in the feature database 6, and a font can be freely created according to personal handwriting and the like. In addition, by analyzing the guideline data in a hierarchical structure for each radical and analyzing individual characteristics as data in each radical and constructing a handwriting file in the characteristic database 6, individual characteristics can be made into a database. By automatically deforming the data of the standard font dictionary according to the characteristic data, a font can be freely created according to the personal handwriting.

When the guideline is used as a reference line and three-dimensionalization is performed for each image element, a three-dimensional character can be automatically generated. If the three-dimensional character is rendered, it is possible to perform three-dimensional printing and character expression on a virtual background. By using the guideline as a reference line, integrating each image element, and outputting the data to a high-speed molding apparatus or a CAM system, a three-dimensional character can be created.

[0035]

As described above, according to the present invention, it is possible to set a guideline representing a character shape and a contour based on a knowledge database constructed in advance for each character of a group of a plurality of characters. If the shape of each character is determined according to the contents of the above and the basic characteristics are changed, the shape of the character can be expressed in a different typeface.

Further, according to the present invention, in the typeface characteristic setting means, basic elements composed of a predetermined number are extracted from a sample created by handwriting, and typeface data reflecting handwriting of handwritten characters is created and used. can do.

According to the present invention, the rule for forming the outline of a character along the guideline is set as a basic feature of the outline by the typeface characteristic setting means, and the outline of the character is set along the guideline. Creates data based on the same knowledge database, even for character data of different typefaces, such as Gothic type that makes the thickness the same in the vertical direction and horizontal direction, and Mincho type that changes the vertical direction and the horizontal direction. can do.

Further, according to the present invention, since the rule for forming the outline of the character is defined in accordance with the guideline in relation to the writing implement, the typeface data is written in the same character shape as when writing the character by hand using the writing implement. Can be created.

Further, according to the present invention, it is possible to make a computer read a program, read handwritten characters, refer to a knowledge database, and easily create typeface data representing personal handwriting.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic system configuration according to an embodiment of the present invention.

FIG. 2 is a diagram showing a concept of extracting a guideline according to the present invention.

FIG. 3 is a diagram showing a concept of a knowledge typeface handled in the present invention in comparison with a conventional typeface.

FIG. 4 is a diagram illustrating a concept of recognizing a feature by comparing a guideline of a model, which is a reference in the present invention, with a handwritten guideline.

FIG. 5 is a diagram showing a part of a handwritten sample 2 of FIG. 1;

FIG. 6 is a diagram illustrating an example of a kanji character and a tangential character.

FIG. 7 is a diagram showing an example of radicals constituting a kanji.

FIG. 8 is a diagram showing a concept of forming an outline font by giving various attributes using a guideline as a basic font.

FIG. 9 is a diagram illustrating an example in which a brush type is given as an attribute using guideline data to form a character.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Normal outline font 2 Handwritten sample 3 Scanner 4 Analysis software 5 Knowledge type font database 6 Feature database 7 Operation software 8 Printer 9 Recording paper 10 Example guidelines 11 Handwriting guidelines 11a Character image 11b Outline 12 Flat 13 20 Handwritten characters 21 Center line 22 Guideline data 23, 24, 25 Outline font

Claims (5)

[Claims] For a group consisting of a plurality of characters, for each character, collect basic knowledge including a guideline representing a reference shape, a writing order for the guideline, and a unit element constituting the character, Refer to the knowledge construction database created in the database construction stage, the typeface characteristics setting stage to set the basic characteristics for the guidelines and outline of the typeface data to be created, and each character. Guidelines for
Regarding the script order and unit elements, the guideline transformation stage that transforms according to the basic features set in the font feature setting stage, and the guidelines for each character transformed in the guideline transformation stage were set in the font feature setting stage. Forming a contour in accordance with the basic characteristics of the contour to create font data representing the shape of the character. 2. In the feature setting step, a plurality of characters are selected so that a basic unit element set in advance is included in any one of the characters, and a handwritten sample is selected for the plurality of selected characters. Create, read the created handwritten sample, refer to the knowledge database, extract guidelines from the sample, and, based on the comparison result between the extracted guidelines and the guidelines included as basic knowledge in the knowledge database,
2. The typeface data creating method according to claim 1, wherein basic characteristics of the guideline of the typeface data to be created are set. 3. The typeface data creation according to claim 1, wherein in the feature setting step, basic features of the contour are set as rules for forming a contour of a character according to a guideline. Method. 4. The typeface data creating method according to claim 3, wherein the rule is determined in relation to a writing instrument for writing characters. 5. For a group consisting of a plurality of characters, for each character, collect basic knowledge including a guideline representing a reference shape, a writing order for the guideline, and a unit element forming the character, A basic unit element that is set in advance is included in any of the characters as a database construction stage to build a knowledge database, and a typeface characteristic setting stage to set basic characteristics for guidelines and outlines of typeface data to be created. A plurality of characters are selected so that a handwritten sample is created for the selected characters, the created handwritten sample is read, a guideline is extracted from the sample by referring to the knowledge database, Based on the comparison between the extracted guidelines and the guidelines included as basic knowledge in the knowledge database ,
Setting a basic characteristic of the guideline of the typeface data to be formed, and a typeface characteristic setting step of setting a basic characteristic of the outline in relation to a writing implement that forms an outline of a character according to the guideline; , Refer to the knowledge database constructed in the database construction stage,
A guideline transformation stage that transforms the script order and unit elements according to the basic features set in the font style feature setting stage, and a contour set in the font style feature setting stage for the guideline of each character transformed in the guideline transformation stage Forming a typeface data representing a shape of a character by forming an outline in accordance with the basic characteristics of the above.