JP2019028094A - Character generation device, program and character output device - Google Patents

Abstract

To provide a character generation device capable of reproducing a handwritten character while taking individual's variation into consideration.SOLUTION: A character generation device 1 comprises: a storage part which stores a model font consisting of a character data group corresponding to a plurality of character codes; a character data acquisition part which acquires character data; a character specifying part which specifies a character that the acquired character data corresponds to; a deformation quantity extraction part which extracts a deformation quantity from difference values of feature quantities by comparing the acquired character data with character data of the model font corresponding to the character that the character specifying part specifies; and a character data generation part which uses a value obtained by increasing or decreasing the deformation quantity for each character data of the model font to generate new character data made to correspond to a corresponding character code.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a character generation device, a program, and a character output device.

  Techniques for creating font characters that reproduce handwritten characters written by humans are known. For example, Patent Document 1 discloses a terminal device that compares handwritten character images with a base font image to extract distortion data, expands the distortion data into a plurality of characters in the base font, and generates handwritten font data. It is disclosed.

Japanese Patent No. 5834822

  However, the invention according to Patent Document 1 can only generate font characters with the same font from the same handwritten character, and does not take into account individual variations that the font changes slightly each time it is written.

  In one aspect, an object of the present invention is to provide a character generation device or the like that can reproduce handwritten characters in consideration of intra-individual variation.

  In one proposal, the character generation device includes a storage unit that stores a model font including character data groups corresponding to a plurality of character codes, a character data acquisition unit that acquires character data, and a correspondence between the acquired character data. A character specifying unit for specifying a character to be compared, the acquired character data, and character data of the sample font corresponding to the character specified by the character specifying unit, A deformation amount extraction unit to extract, and a character data generation unit that generates new character data in association with a corresponding character code using a value obtained by adding or subtracting the deformation amount to each character data of the model font It is characterized by providing.

  In one plan, the character generation device is characterized in that the adjusted value is a random numerical value within a certain range.

  In one plan, the character generation device is characterized in that the character data generation unit generates new character data using the deformation amount extracted by the deformation amount extraction unit.

  In one plan, in the character generation device, the storage unit stores a plurality of the model fonts, and the deformation amount extraction unit uses the plurality of model fonts to resemble the acquired character data. This font is specified, the acquired character data is compared with the character data of the specified sample font, and the deformation amount is extracted from the difference value of the feature amount.

  In one proposal, the character generation device includes an input unit that receives an input of a character code, and the character data generation unit generates the new character data when the input unit receives the input of the character code. It is characterized by that.

  In one plan, the character generation device includes an output unit that outputs the new character data corresponding to the input character code.

  In one plan, in the character generation device, the character data generation unit generates a plurality of character data for one character using different deformation amounts, and each of the generated character data corresponds to each character unit. The character code is stored in association with the character code.

  In one plan, the character generation device includes an input unit that receives an input of a character code, and an output unit that randomly selects and outputs the new character data corresponding to the input character code. To do.

  In one plan, the character generation device is configured such that the acquired character data is drawing data of handwritten characters drawn by a user, and the output unit generates the new character in accordance with the writing order when the handwritten characters are drawn. Characters of character data are output.

  In one plan, the program acquires character data in a computer having a storage unit that stores a model font composed of character data groups corresponding to a plurality of character codes, and stores the character corresponding to the acquired character data. The obtained character data is compared with the character data of the sample font corresponding to the specified character, and the deformation amount is extracted from the difference value of the feature amount, and each character data of the sample font is extracted. On the other hand, a process of generating new character data in association with a corresponding character code is executed using a value obtained by adjusting the amount of deformation.

  In one proposal, the character output device includes a storage unit that stores each character of a plurality of fonts in association with one character code, an input unit that receives input of a character code, and a character code received by the input unit. And an output unit that selects and outputs a corresponding character from any of the plurality of fonts.

  In one aspect, handwritten characters can be reproduced in consideration of individual variations.

It is a block diagram which shows the structural example of a character production | generation apparatus. It is explanatory drawing regarding a font production | generation process. It is explanatory drawing regarding a character output process. It is a flowchart which shows an example of the process sequence of a font production | generation process. It is a flowchart which shows an example of the process sequence of a character output process. It is explanatory drawing which shows the process outline | summary of the character generation apparatus which concerns on Embodiment 2. FIG. 12 is a flowchart illustrating an example of a processing procedure of font generation processing according to the second embodiment. 10 is a flowchart illustrating an example of a processing procedure of character output processing according to the second embodiment. FIG. 10 is an explanatory diagram showing an outline of processing executed by a character generation device according to Embodiment 3. 14 is a flowchart illustrating an example of a processing procedure of font generation processing according to the third embodiment. 12 is a flowchart illustrating an example of a processing procedure of character output processing according to Embodiment 3. It is a functional block diagram which shows operation | movement of the character production | generation apparatus of the form mentioned above.

Hereinafter, the present invention will be described in detail with reference to the drawings illustrating embodiments thereof.
(Embodiment 1)
FIG. 1 is a block diagram illustrating a configuration example of the character generation device 1. In the present embodiment, a character generation device 1 that generates handwritten character data (character image) having a different shape each time output is described. In the following description, for convenience, a character font that reproduces a user's handwritten character is referred to as a “handwritten font”.

  The character generation device 1 is an information processing device that performs various types of information processing, such as a personal computer, a server device, a smartphone, a tablet terminal, or the like. The character generation device 1 learns the characteristics of the user's handwritten characters and generates a handwritten font.

The character generation device 1 includes a control unit 11, a storage unit 12, a RAM 13, a display unit 14, an input unit 15, and a communication unit 16.
The control unit 11 has an arithmetic processing unit such as one or a plurality of CPUs (Central Processing Units) and MPUs (Micro-Processing Units), and reads and executes a program P stored in the storage unit 12 to execute characters. Various information processing, control processing, and the like related to the generation device 1 are performed. The storage unit 12 is a large-capacity memory, a hard disk, or the like, and stores a program P and other data necessary for the control unit 11 to execute processing. The storage unit 12 stores model font data 121, learning data 122, and reference font data 123. The model font data 121 is data of an existing character font that is generally set as a standard in a computer, and is font data composed of character data groups respectively corresponding to a plurality of character codes.

  The learning data 122 is data constructed by learning the characteristics of the user's handwritten characters by deep learning described later. The reference font data 123 is handwritten font data generated using the learning data 122, and is font data serving as a reference for character data to be output. As will be described later, the character generation device 1 generates and outputs character data obtained by randomly deforming characters of the handwritten font with reference to the handwritten font generated from the learning data 122.

  The RAM 13 is an SRAM (Static Random Access Memory), a DRAM (Dynamic Random Access Memory), a flash memory, or the like, and temporarily stores data necessary for the control unit 11 to execute arithmetic processing. The display unit 14 is, for example, a liquid crystal display, and displays an image given from the control unit 11. The input unit 15 is a touch panel, a keyboard, a mouse, or the like, for example, and accepts drawing of handwritten characters and input of character codes. The communication unit 16 includes a processing circuit for performing processing related to communication, and transmits / receives information to / from other devices via the network N.

FIG. 2 is an explanatory diagram regarding font generation processing. Below, the outline | summary of the process which the character production | generation apparatus 1 performs is demonstrated.
The character generation device 1 performs a process of generating a handwritten font that reflects the characteristics of the handwritten characters by learning the characteristics of the user's handwritten characters from the character data that is image data of the handwritten characters written by the user. FIG. 2 shows an outline of the processing. FIG. 2 illustrates a state in which a reference font that reflects the characteristics of the user's handwritten characters is generated.

  For example, the character generation device 1 acquires drawing data in which a user draws characters on a touch panel with an electronic pen, a finger, or the like. Further, for example, the character generation device 1 may acquire scan data of the paper on which handwritten characters are written from an external scanner device (not shown). The character generation device 1 acquires character data of handwritten characters for a plurality of types of characters in order to perform deep learning described later. The character generation device 1 performs character recognition for character data, and specifies the type of character (character code) for each character data.

  The character generation device 1 extracts features of handwritten characters from the acquired character data of handwritten characters. Specifically, the character generation device 1 compares character data of a preset model font with character data of handwritten characters, and extracts a deformation amount from a difference value of feature amounts.

  As described above, the character generation device 1 stores a plurality of model fonts. The character generation device 1 compares the handwritten character of the user with the characters of each example font corresponding to the handwritten character, and selects the example font most similar to the handwritten character. In the case of the example illustrated in FIG. 2, the character generation device 1 compares the handwritten “A” character with the “A” character of each font. The character generation device 1 selects the model font determined to be the most similar as a material for generating a handwritten font.

  The character generation device 1 compares the character of the model font determined to be similar to the handwritten character, and extracts the deformation amount from the difference value of the feature amount of each character. The character feature amount is, for example, a coordinate value related to a feature point such as an end point, a refraction point, an intersection, or a branch point of a character stroke. The character generation device 1 specifies the coordinate position of each feature point for handwritten characters and characters of the same model font as the handwritten characters. For example, in the example illustrated in FIG. 2, the character generation device 1 specifies the position of each feature point from the handwritten character “A” and also specifies the position of each feature point from the font character “A”. The character generation device 1 takes the difference of the coordinate values of the feature points between the handwritten character and the character of the model font, and extracts the deformation amount.

  Although illustration is omitted, the character generation device 1 compares the handwritten character with the model font for characters other than “A”. The character generation device 1 sequentially stores deformation amounts extracted from various characters in the storage unit 12, and constructs learning data 122 related to deep learning. For example, the character generation device 1 constructs a neural network related to GAN (Generative Adversarial Networks). GAN is a kind of data structure constructed by deep learning, and is a network that obtains an output having properties similar to a sample group to be learned. Since the GAN technique is known, detailed description thereof is omitted. The character generation device 1 learns the amount of deformation between the handwritten character and the model font, thereby constructing learning data 122 for outputting a handwritten font that reproduces the handwritten character.

  The character generation device 1 applies the deformation amount extracted above to each character data of the model font, and generates new character data reflecting the characteristics of the user's handwritten characters. For example, as conceptually shown in FIG. 2, the character generation device 1 applies the deformation amount related to the character “A” to the characters “I”, “U”, “E”. Handwritten fonts “I”, “U”, “E”,... Reflecting the characteristics of the user's handwritten characters are generated. In this way, the character generation device 1 generates handwritten-style font characters for the characters “I”, “U”, “E”,... The character generation device 1 stores a character data group generated based on the learning data 122, that is, handwritten font data in the storage unit 12 as reference font data 123.

  FIG. 3 is an explanatory diagram regarding character output processing. The character generation device 1 outputs a font character in which a character code arbitrarily input by a user is reproduced as a handwritten character based on the reference font generated above. Note that the character generation device 1 may accept input of a character code at the input unit 15 or may accept input of a character code from a client terminal. Here, the character generation device 1 does not simply output the reference font as it is, but generates and outputs new character data obtained by randomly deforming the character data of the reference font each time it is output.

  In general, characters handwritten by a person are not always the same, and each time they are written, the shapes are different. However, when the learned user's handwritten character is output as it is, the shape of the character is constant and impersonal. Therefore, in the present embodiment, the character generation device 1 transforms the character data of the reference font every time it is output.

  Specifically, the character generating device 1 applies the value obtained by adding or subtracting the above deformation amount to the character data of the reference font, and deforms new character data for output. For example, the character generation device 1 applies the random numerical value obtained by randomly adjusting the deformation amount within a certain numerical range to the reference font character, with the learned deformation amount as an average, and character data obtained by randomly deforming the reference font character. Is generated. As a result, as shown in the center of FIG. 3, the character generation device 1 randomly changes the coordinate positions of the feature points of the characters of the reference font, and generates character data in which the character shape is changed little by little. The character generation device 1 generates character data in a random form for each character indicated by the input character code. Thereby, as shown in the lower part of FIG. 3, the character generating device 1 outputs handwritten characters having a different pattern for each output. For example, the character generation device 1 displays randomly deformed character data on the display unit 14. For example, the character generation device 1 may output character data to the client. As described above, the character generation device 1 can generate handwritten font characters reflecting the tendency of handwritten characters by generating different handwritten fonts for each output.

FIG. 4 is a flowchart illustrating an example of a processing procedure of font generation processing. The processing contents of the handwritten font generation processing will be described with reference to FIG.
The control unit 11 of the character generation device 1 acquires character data of handwritten characters written by the user (step S11). For example, the control unit 11 acquires drawing data of handwritten characters drawn on the touch panel. Further, for example, the control unit 11 may acquire scan data obtained by reading a paper surface on which a user has written handwritten characters with a scanner device. The control part 11 acquires the character data of a user's handwritten character about several types of characters.

  The control unit 11 performs character recognition on the character data (image data) and specifies a character corresponding to the acquired character data (step S12). And the control part 11 specifies the model font similar to the character data acquired by step S11 from a some sample font (step S13). Specifically, the control unit 11 compares the character data of each example font corresponding to the character specified in step S12 with the character data of the user's handwritten character, and the example relating to the character data having the closest feature amount. Identify the font.

  The control unit 11 compares the character data of the handwritten character acquired in step S11 with the character data of the model font specified in step S13, and extracts the deformation amount from the difference value of the feature amount (step S14). Specifically, the control unit 11 compares the character of the model font corresponding to the character specified in step S12, that is, the character of the same model font as the handwritten character, and the coordinate value of each feature point of the handwritten character. And the coordinate value of each feature point of the character of the model font, and the deformation amount is extracted from the difference between the coordinate values. The control unit 11 compares all the handwritten characters acquired in step S11 with the model font and extracts the deformation amount. The control unit 11 sequentially stores the deformation amount extracted for each character in the storage unit 12 and constructs learning data 122.

  The control unit 11 applies the deformation amount extracted in step S14 to each character data of the model font, and generates new character data related to the reference font (step S15). That is, the control part 11 applies the learning result by deep learning to the character data of a model font about the character which has not acquired the sample image of the handwritten character, and produces | generates the character data reflecting the characteristic of a user's handwritten character. For example, the control unit 11 generates characters using a GAN algorithm. The control unit 11 stores the reference font data in which the generated new character data is associated with the corresponding character code in the storage unit 12 (step S16), and ends the series of processes.

FIG. 5 is a flowchart illustrating an example of a processing procedure for character output processing. The character data output process will be described with reference to FIG.
The control unit 11 of the character generation device 1 receives an input of a character code via the input unit 15 (step S31). The control unit 11 generates new character data using the value obtained by adding or subtracting the amount of deformation to the character data of the reference font corresponding to the input character code (step S32). Specifically, the control unit 11 applies a random numerical value obtained by randomly adjusting the deformation amount between the user's handwritten character and the character of the model font within a certain numerical range, and randomly determines the character data of the reference font. Generate character data transformed into. The control unit 11 generates corresponding random-shaped character data for each input character code. The control part 11 outputs the character data produced | generated by step S32 (step S33), and complete | finishes a series of processes.

  In the above description, the model font to be compared with handwritten characters is an existing font that is standardly set in the computer. However, any font that can be compared with handwritten characters may be used. The content is not particularly limited.

  Moreover, although the character production | generation apparatus 1 decided to change the deformation amount between the handwritten character and the character of a model font at random above, this Embodiment is not limited to this. The character generation device 1 may perform character deformation under certain rules such as changing the character deformation pattern according to the output time so that the output font characters appear to be randomly deformed. Good. That is, the character generation device 1 only needs to be able to generate a handwritten font that takes into account individual variations by changing the difference value. The character generation device 1 may perform random character deformation or pattern character deformation. May be performed.

  As described above, according to the first embodiment, handwritten characters can be reproduced in consideration of individual variation.

  Further, according to the first embodiment, handwritten characters can be reproduced in a form closer to the intra-individual variation by using a value obtained by randomly adjusting the amount of deformation within a certain numerical range.

  In addition, according to the first embodiment, by generating an intermediate reference font using the deformation amount itself instead of the deformation amount extracted by comparing the handwritten character with the model font, A more preferable implementation is possible.

  Further, according to the first embodiment, reproducibility of handwritten characters can be improved by selecting a model font similar to handwritten characters as the material of the handwritten font.

  Further, according to the first embodiment, it is possible to reproduce characters close to handwriting by generating character data in real time after receiving an input of a character code.

  Further, according to the first embodiment, it is possible to output a handwritten style font character obtained by randomly deforming a character indicated by an arbitrarily input character code.

(Embodiment 2)
In the present embodiment, a mode will be described in which a plurality of patterns of handwritten fonts are generated in advance, and fonts are randomly selected and output at the time of output. In addition, about the content which overlaps with Embodiment 1, the same code | symbol is attached | subjected and description is abbreviate | omitted.
FIG. 6 is an explanatory diagram showing a processing outline of the character generation device 1 according to the second embodiment. In the first embodiment, the character generation device 1 deforms the reference font at the time of output to generate character data having a random shape. In this embodiment, the character generation device 1 generates character data in advance, not sequentially. Thereby, the processing speed is improved.

  Specifically, as illustrated in FIG. 6, the character generation device 1 generates a plurality of patterns of handwritten fonts in advance from a reference font generated by deep learning. The character generation device 1 applies a plurality of different deformation amounts to one character of the reference font, and generates a plurality of character data per character. For example, in the example of FIG. 6, the character generation device 1 applies different amounts of deformation to the character “A” of the reference font, and generates a plurality of patterns of character data “A”. For example, as in the first embodiment, the character generation device 1 generates character data of each handwritten font using a value obtained by adding or subtracting the deformation amount within a certain numerical range. Similarly, the character generation device 1 generates a plurality of patterns of character data for each of the characters “I”, “U”, “E”.

  The character generation device 1 stores a plurality of character data generated from one character data of the reference font in the storage unit 12 in association with the corresponding character code. In the case of the above example, the character generation device 1 stores a plurality of “A” character data in association with the “A” character code. The same applies to other characters. Thereby, the storage unit 12 stores character data of a plurality of handwritten fonts in association with one character code. In this manner, the character generation device 1 stores the generated character data of the handwritten font of each pattern in the storage unit 12 in association with the corresponding character code in character units.

  When the character generation device 1 accepts input of a character code and outputs a character, the character generation device 1 randomly selects one of the handwritten fonts of each pattern generated in advance and outputs the character. For example, as shown in FIG. 6, when three patterns of fonts are generated in advance, the character generation device 1 randomly selects one of the fonts and outputs character data. As a result, the character generation device 1 can output different character data each time it is output, as in the first embodiment.

FIG. 7 is a flowchart illustrating an example of a processing procedure of font generation processing according to the second embodiment. After executing the process of generating the reference font (step S15), the control unit 11 of the character generation apparatus 1 executes the following process.
The control unit 11 applies a plurality of different deformation amounts to each character of the reference font, and generates a plurality of character data for one character of the reference font (step S201). For example, as in the first embodiment, the control unit 11 applies a random numerical value obtained by randomly adjusting a parameter related to the deformation amount to generate character data related to a handwritten font. In the present embodiment, the control unit 11 applies a plurality of random numerical values having different values to one character of the reference font, and generates character data of a plurality of handwritten fonts from one character data.

  The control unit 11 stores a plurality of character data generated for each character of the reference font in the storage unit 12 in association with the corresponding character code in units of characters (step S202). Thereby, each character of a plurality of handwritten fonts is stored in the storage unit 12 in association with the corresponding character code. The control unit 11 ends the series of processes.

FIG. 8 is a flowchart illustrating an example of a processing procedure of character output processing according to the second embodiment. After receiving the input of the character code (step S31), the control unit 11 of the character generation device 1 executes the following process.
The control unit 11 randomly selects one of a plurality of character data stored in the storage unit 12 in association with the character code input in step S31 (step S221). That is, the control unit 11 randomly selects one of a plurality of handwritten fonts. The control unit 11 outputs the selected character data (step S222), and ends a series of processes.

  In the above description, the character generation device 1 performs all the processes from generation to output of a plurality of handwritten fonts, but the present embodiment is not limited to this. For example, a plurality of handwritten font data generated by the character generation device 1 may be installed in a predetermined terminal device, and the terminal device may randomly select a handwritten font and output characters when outputting (displaying) characters. Good. In this case, the terminal device functions as a character output device that randomly outputs characters of the handwritten font. Thus, it is not necessary for the character generation device 1 to perform all processing from generation to output of the handwritten font, and character generation processing and output processing may be performed in a distributed manner.

  As described above, according to the second embodiment, the same effects as those of the first embodiment can be obtained by generating data of a plurality of handwritten fonts in advance.

  Further, according to the second embodiment, it is possible to improve the processing speed by selecting one of a plurality of handwritten fonts generated in advance at the time of character output.

(Embodiment 3)
In the present embodiment, a mode is described in which font characters are output by reproducing the writing order of handwritten characters.
FIG. 9 is an explanatory diagram illustrating an outline of processing executed by the character generation device 1 according to the third embodiment. In the present embodiment, the character generation device 1 reproduces the writing order when the user writes handwritten characters and outputs the characters.

  Specifically, the character generation device 1 acquires drawing data of handwritten characters drawn on a predetermined touch panel as character data of handwritten characters. And the character production | generation apparatus 1 reads the writing order of each stroke which comprises a character from the drawing data of a handwritten character.

  The character generation device 1 outputs characters according to the writing order by the user when outputting character data. For example, as shown in FIG. 9, the character generation device 1 sequentially outputs each stroke of font characters according to the writing order. Thereby, the character production | generation apparatus 1 can improve the reproducibility of a user's handwritten character.

FIG. 10 is a flowchart illustrating an example of a processing procedure of font generation processing according to the third embodiment. Based on FIG. 10, the processing content of the font generation processing according to Embodiment 3 will be described.
The control unit 11 of the character generation device 1 acquires drawing data of handwritten characters drawn by the user (step S301). The drawing data is line drawing data drawn on a predetermined touch panel with an electronic pen, for example, and is data that can identify the writing order of handwritten characters. After executing the processes of steps S12 to S14, the control unit 11 stores information on the writing order in the storage unit 12 in association with the corresponding character data together with the reference font generated based on the drawing data (step S302). The process ends.

FIG. 11 is a flowchart illustrating an example of a processing procedure of character output processing according to the third embodiment. After executing the process of generating character data in which the characters of the reference font are randomly deformed (step S32), the control unit 11 of the character generation device 1 executes the following process.
The control unit 11 refers to the storage unit 12 and outputs character data according to the writing order when handwritten characters are drawn (step S321). Specifically, the control unit 11 sequentially outputs each stroke constituting the character according to the writing order of the handwritten character. The control unit 11 ends the series of processes.

  As described above, according to the third embodiment, it is possible to output characters closer to handwriting by reproducing the writing order of handwritten characters.

(Embodiment 4)
FIG. 12 is a functional block diagram illustrating the operation of the character generation device 1 having the above-described form. When the control unit 11 executes the program P, the character generation device 1 operates as follows. The storage unit 1201 stores a model font composed of a character data group corresponding to a plurality of character codes. The character data acquisition unit 1202 acquires character data. The character specifying unit 1203 specifies a character corresponding to the acquired character data. The deformation amount extraction unit 1204 compares the acquired character data with the character data of the sample font corresponding to the character specified by the character specifying unit 1203, and extracts the deformation amount from the difference value of the characteristic amount. . The character data generation unit 1205 generates a character code corresponding to new character data by using a value obtained by adding or subtracting the deformation amount to each character data of the model font.

  The fourth embodiment is as described above, and the other parts are the same as those of the first to third embodiments. Therefore, the corresponding parts are denoted by the same reference numerals, and detailed description thereof is omitted.

  The embodiment disclosed this time is to be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the meanings described above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

DESCRIPTION OF SYMBOLS 1 Character generator 11 Control part 12 Memory | storage part P Program 121 Example font data 122 Learning data 123 Reference | standard font data 13 RAM
14 Display unit 15 Input unit 16 Communication unit

Claims (11)

A storage unit for storing a model font composed of character data groups corresponding to a plurality of character codes;
A character data acquisition unit for acquiring character data;
A character identification part for identifying the corresponding character of the acquired character data;
A deformation amount extraction unit that compares the acquired character data with the character data of the example font corresponding to the character specified by the character specifying unit, and extracts a deformation amount from a difference value of the feature amount;
A character data generation unit that generates new character data in association with a corresponding character code using a value obtained by adding or subtracting the amount of deformation to each character data of the model font. Character generator. The character generation apparatus according to claim 1, wherein the adjusted value is a random numerical value within a certain range. The character generation device according to claim 1, wherein the character data generation unit generates new character data using the deformation amount extracted by the deformation amount extraction unit. The storage unit stores a plurality of the model fonts,
The deformation amount extraction unit
Identifying the example font similar to the acquired character data from the plurality of example fonts;
4. The deformation amount is extracted from a difference value of feature amounts by comparing the acquired character data with the character data of the specified example font. 5. The described character generator. It has an input unit that accepts input of character codes,
The character generation device according to any one of claims 1 to 4, wherein the character data generation unit generates the new character data when the input unit receives the input of the character code. . The character generation device according to claim 5, further comprising an output unit that outputs the new character data corresponding to the input character code. The character data generation unit
A plurality of character data is generated for one character using different deformation amounts,
5. The character generation device according to claim 1, wherein the generated character data is stored in association with a corresponding character code in character units. An input unit that accepts input of a character code;
The character generation device according to claim 7, further comprising: an output unit that randomly selects and outputs the new character data corresponding to the input character code. The acquired character data is drawing data of handwritten characters drawn by the user,
The character generation device according to claim 6 or 8, wherein the output unit outputs characters of the new character data according to a writing order when the handwritten characters are drawn. In a computer having a storage unit for storing a model font consisting of character data groups corresponding to a plurality of character codes,
Get character data,
Identify the corresponding character in the acquired character data,
Comparing the acquired character data with the character data of the model font corresponding to the specified character, and extracting the deformation amount from the difference value of the feature amount;
A program for executing a process for generating new character data in association with a corresponding character code using a value obtained by adding or subtracting the amount of deformation to each character data of the model font. A storage unit for storing each character of a plurality of fonts in association with one character code;
An input unit that accepts input of a character code;
A character output device comprising: an output unit that selects and outputs a character corresponding to a character code received by the input unit from any of the plurality of fonts.

Images