JPH0721164A - Character string image generating device - Google Patents

Abstract

PURPOSE:To provide the character string image generating device which can print characters like an autography in equipment such as a word processor. CONSTITUTION:Handwritten character images of a user which are previously registered by a standard handwritten character image registration part 2 are deformed by an image deformation part 5 on the basis of information in a deformation information storage part 6 to obtain invariably different shapes and sizes even for characters of the same category. Further, an output position determination part 7 makes the output position of character images not fixed, but variable in output position on the basis of information in an output position information storage part 8, and a deformed image output part 9 outputs the deformed character images at the output positions, thereby outputting a character string image which is close to a handwriting.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character string image generating device used for printing on a word processor or the like.

[0002]

2. Description of the Related Art A conventional character string image generating device used in a general word processor or the like will be described below.

FIG. 18 is a block diagram of this conventional character string image generating apparatus, in which 1 is a character code input section,
Reference numeral 200 is a font reading unit, 201 is a font storage unit, and 202 is an image output unit.

In the conventional character string image generating apparatus configured as described above, the character code input unit 1 sequentially outputs the codes representing the characters in the document to be printed to the font reading unit 200. An image of a character corresponding to each character code is stored in the font storage unit 201, and the font reading unit 200 reads a character image corresponding to an input code from the font storage unit 201, and the character image Is output to the image output unit 202. Next, the image output unit 202 outputs the character image input at a predetermined fixed position.

[0005]

In recent years, devices such as word processors have become widespread in ordinary households, and have come to be used for private letters and the like, which are more humane,
There is an increasing demand for printing in a handwritten typeface. However, in the above-described configuration, since a predetermined character image preset for each category of characters is aligned and printed as it is, there is a problem that it cannot be printed as if written by hand.

In view of the problems of the conventional character string image generating apparatus, it is an object of the present invention to provide a character string image generating apparatus capable of printing a character as if it were a handwriting.

[0007]

According to the present invention, a character code input section for inputting a code of a character desired to be output by a user,
An image that reads a standard handwritten character image and stores a character code corresponding to the image and a standard handwritten character image registration unit that adds writing information such as character strokes, and an image to which the information is added A storage unit, an image reading unit that reads a character image to which the character code input from the character code input unit is added from the image storage unit, and, for example, what kind of deformation the image transformation unit has made in the past. A transformation information storage unit that stores and retains transformation information, such as an image transformation unit that transforms a read image based on the transformation information, and a transformation that outputs a character image transformed by the image transformation unit. Determine where to output the image output section and the character image transformed by the image transformation section And an output position determination unit that, the output position information storage unit for storing and holding the information of the output position determined by the output position determination unit.

[0008]

According to the present invention, according to the above-mentioned configuration, the handwriting character image of the user registered in advance by the standard handwriting character image registering unit is deformed by the image deforming unit based on the information in the deformation information storage unit so that the same category of Characters always have different shapes and sizes.Furthermore, based on the information in the output position information storage unit, the output position determination unit does not make the output position of the character image constant but gives the output position fluctuation, and the modified image output unit By outputting the transformed character image to the output position, it is possible to output a character string image closer to handwriting.

[0009]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a character string image generating apparatus according to an embodiment of the present invention. The operation of the character string image generation device of this embodiment will be described below together with its configuration.

The character code input unit 1 is a document file that already stores and holds a document as a character code string, or
Character codes representing characters in the document printed by the user are sequentially output to the image reading unit 4.

The standard handwritten character image registration unit 2 is a means for the user to preregister a handwritten or printed character image as a standard character image. The operation will be described with reference to FIG. FIG. 2 is a configuration diagram of the standard handwritten character image registration unit 2, 20 is an image reading unit, 21 is an automatic writing information setting unit, and 22 is a manual writing information setting unit. First, the image reading unit 20 reads the handwritten character string written by the user, and inputs the image to the automatic writing information setting unit 21. The automatic writing information setting unit 21 cuts out an area for each character from the handwritten character string written by the user, performs character recognition processing on each cut out character image, and adds a corresponding character code. In addition, strokes are extracted for each character image, and based on the stroke extraction result and the character recognition result, the starting point and the end point of the character image, and the reference points for character transformation on each stroke of the character (hereinafter, this Points are called deformation points) and are added as writing information.
With this, the writing information can be automatically extracted from the characters written by the user and registered with the handwritten character image,
The burden on the user can be reduced.

Next, the manual handwriting information setting unit 22 shows the character image and the handwriting information added thereto to the user from the automatic handwriting information setting unit 21, and the user corrects the character image and the handwriting information. The user himself / herself adds the writing information by inputting correct information to the information that is confirmed and judged to be incorrect or the information that is judged to be insufficient. This ensures registration of error-free information.

Specific operations of the automatic writing information setting unit 21 and the manual writing information setting unit 22 will be described below.

FIG. 3 is a detailed block diagram of the automatic writing information setting unit 21 and the manual writing information setting unit 22.
Is a character cutout unit, 31 is a stroke extraction unit, 32 is a character recognition unit, 33 is an automatic start point end point determination unit, 34 is an automatic deformation point determination unit, 35 is a writing information image display unit, 36 is a character image determination unit, and 37 is A character code determination unit, 38 is a start point / end point determination unit, and 39 is a deformation point determination unit.

The character clipping unit 30 first obtains a histogram distribution when black pixels in the image input from the image reading unit 20 are projected on the vertical axis and the horizontal axis. Next, in the histogram distribution, a region having a large histogram value and forming one lump is determined, and the image of the region is determined as a character image for one character.

The character slicing unit 10 sequentially outputs the character images for each character cut out from the handwritten character string in this manner to the stroke extraction unit 31, the character recognition unit 32, and the handwritten information image display unit 35. The image for one character will be described below.

The character recognizing section 32 obtains a similarity with a handwritten character image learned in advance, and displays a writing information image as a character code for adding the character code corresponding to the standard character image having the highest similarity to the character image. Part 3
Output to 5. In addition, the characteristics of the starting point and the ending point of the corresponding writing of the corresponding character are determined as the automatic start point end point determination unit 33.
Output to. The characteristics of the start point and the end point are the characteristics of the stroke including those points, and are represented by the direction of the stroke and the relative length and position with respect to the circumscribed rectangle of the character image.

The stroke extracting unit 31 is a character cutting unit 3
The stroke position of the character image input from 0 is determined.

FIG. 4 shows the configuration and operation of the stroke extraction unit 31.
Will be explained. FIG. 4 is a configuration diagram of the stroke extraction unit 31, 40 is an image thinning unit, 41 is an end point extraction unit, 42 is an intersection point extraction unit, and 43 is a stroke position determination unit.

The image thinning unit 40 is a character cutting unit 3
A character image input from 0 is subjected to a thinning process to convert the character image into a line image having a line width of 1. Next, the end point extraction unit 41 detects the position of the point on the line image as a end point that satisfies the following conditions. <End Point Condition> Of the eight neighboring pixels, one is a black pixel and all the remaining pixels are white pixels. Similarly, the intersection extraction unit 42
Detects the position of a point on the line image with a point satisfying the following condition as an intersection. <Cross point condition> Among the eight neighboring pixels, there are three or more black pixels that are not adjacent to each other. FIG. 5 is a diagram for explaining the operation of the stroke position determination unit 43. First, the end point detected by the end point extraction unit 41 is used as a starting point to reach the next end point on the line image or the intersection point detected by the intersection point extraction unit 42. To explore. Next, when the length of the route is longer than a certain length, the position information of the searched route is output to the automatic start point end point determination unit 33 and the automatic deformation point determination unit 34 as a stroke component forming a character. If the length of the route is shorter than a certain length, it is regarded as noise generated during the thinning process and is not output.

The automatic start point / end point determination unit 33 detects the stroke component having the characteristics of the start point input from the character recognition unit 32 among the stroke components input from the stroke extraction unit 31, and positions the stroke component at the upper right position of the stroke component. The point to be written is determined as the starting point for writing, and the position information is output.
Similarly, the stroke component having the characteristic of the end point input from the character recognition unit 32 is detected, the point located at the lower left of the stroke component is determined as the start point of writing, and the position information thereof is written information image display unit 35. Output to.

The automatic deformation point determination unit 34 outputs the position information to the handwriting information image display unit 35, with the point dividing the stroke component input from the stroke extraction unit 31 into a fixed length as the deformation point.

FIGS. 6, 7, and 8 are explanatory diagrams of the operation of the manual writing information setting unit 22, and the operation will be described with reference to the drawings.

The writing information image display unit 35 displays the image input from the image reading unit 20 and presents it to the user. Further, the position of the image portion cut out according to the position information input from the character cutout unit 30 is presented to the user. The character code input from the character recognition unit 32, the start and end position information of the character image input from the automatic start / end point determination unit 33, and the position information of the deformation point of the character image input from the automatic deformation point determination unit 34. Is presented to the user in correspondence with the displayed image.

The user looks at the writing information and the character image displayed in the writing information image display section 35 as shown in FIG. 6, and confirms whether or not they are incorrect. If it is incorrect, correct the error as follows.

As shown in FIG. 7, the character image determining unit 36 allows the user to set a correct character image area for the character image displayed on the writing information image display unit 35 when the character image is cut out incorrectly. The character image is corrected by teaching. Further, the character image determination unit 36 outputs the corrected correct character image to the image storage unit 3.

The character code determination unit 37, as shown in FIG.
When the character code is incorrect, the user teaches the correct character code to correct the character code. Further, the character code determination unit 37 outputs the corrected correct character code to the image storage unit 3.

As shown in FIG. 8, the start-point / end-point determining unit 38 determines the correct start point if the position information of the start point and the end point is incorrect.
The position information of the start point and the end point is corrected by the user teaching the position information of the end point. Further, the start point / end point determination unit 38 outputs the corrected correct position information of the start point and end point to the image storage unit 3.

As shown in FIG. 8, when the position information of the deformation point is incorrect, the deformation point determination unit 39 corrects the position information of the deformation point by instructing the correct position information of the deformation point by the user. . Further, the deformation point determination unit 39 outputs the corrected position information of the correct deformation point to the image storage unit 3.

As shown in FIG. 1, in the image storage unit 3, the character image obtained as described above is used as a standard handwritten character image, the character code corresponding to the character image, and the start point of writing. The position information of the end point at the end of writing and the position information of the deformation point are stored and held as writing information.

The image reading unit 4 reads a character image having the same character code as the character code input from the character code input unit 1 and the writing information added thereto from the image storage unit 3 and sends it to the image transformation unit 5. Output.

FIG. 9 is a block diagram of the image transformation unit 5,
Reference numeral 71 is a linear conversion unit, 72 is a non-linear conversion unit, and 73 is a smoothing processing unit.

The linear conversion unit 71 uniformly processes the entire character image, and the non-linear conversion unit 72 processes each character stroke.

First, the structure and operation of the linear conversion unit 71 will be described.

FIG. 10 is a block diagram of the linear conversion unit 71.
Reference numeral 81 is a transformed coordinate system determination unit, 82 is a fluctuation value generation unit, and 83
Is a conversion parameter setting unit, and 84 is an image deformation point conversion unit. FIG. 11 is an explanatory diagram of the operation of the linear conversion unit 71, and the operation of each component will be described with reference to the drawing.

The conversion coordinate system determination unit 81 performs the conversion shown in (Equation 1) and (Equation 2) with respect to the standard coordinate system as shown in FIG.

[0038]

[Equation 1] x '= k ₁ · x · cos θ ₁ + k ₂ · y · sin θ ₂

[0039]

[Equation 2] y '= k ₁ · x · sin θ ₁ + k ₂ · y · cos θ ₂ where parameters k ₁ , k ₂ , θ ₁ and θ ₂ are conversion parameters, and k ₁ is expansion / contraction in the x-axis direction. Rate, k ₂ is the expansion / contraction rate in the y-axis direction,
θ ₁ represents the inclination from the x-axis, and θ ₂ represents the inclination from the y-axis.
For example, when k ₁ = 0.5, k ₂ = 1, θ ₁ = 2, θ ₂ = 1
The coordinate system is as shown in 1.

The value of the conversion parameter is set by the deformation information storage unit 6, the fluctuation value generation unit 82, and the conversion parameter setting unit 83 as follows.

The deformation information storage unit 6 stores and holds the value of a predetermined conversion parameter in the initial stage of the handwritten character image generation process, the handwritten character image generation process is started, and the generation of the image for one character is completed. Then, the content held in the value of the conversion parameter input from the conversion parameter setting unit 83 is updated.

The fluctuation value generator 82 generates values within a preset value range centered on 0 in time series. However, the probability of occurrence of that value is highest at 0, and becomes lower as the value is farther from 0. The fluctuation value generating unit 82 outputs the generated value to the conversion parameter setting unit 83.

The conversion parameter setting unit 83 reads out the values of the conversion parameters stored and held in the deformation information storage unit 6, adds the values input from the fluctuation value generation unit 82, and converts them as new conversion parameter values. The data is output to the system determination unit 81 and the deformation information storage unit 6.

The image deformation point conversion unit 84 receives the position information of the input deformation point and the pixel position of the character image according to the conversion coordinates represented by (Equation 1) and (Equation 2).
It is converted as shown in 1 and output to the non-linear conversion unit 72.

Next, the configuration and operation of the non-linear conversion section 72 will be described.

FIG. 12 is a block diagram of the non-linear conversion unit 72. 100 is an image blocking unit, 101 is a deformation point perturbation unit, 102 is a perturbation parameter setting unit, 103 is a perturbation fluctuation value generation unit, and 104 is perturbation information storage. Reference numeral 105 denotes a modified image generation unit.

The image blocking unit 100 receives the character image output from the image transformation point conversion unit 84 and the position information of the transformation point.

The image blocking unit 100 divides each stroke in the input character image into a plurality of pixel blocks each including one deformation point based on the input position information of the deformation point, and generates the result as a deformed image. Part 10
5 and outputs the position information of the deformation point in each pixel block to the deformation point perturbation unit 101.

The transformation point perturbation unit 101 performs transformation shown in (Equation 3) and (Equation 4) on the input position information, and outputs the transformed position coordinates of each transformation point to the transformation image generation unit 105. . FIG. 13 shows a conceptual diagram of this conversion.

[0050]

[Equation 3] x '= x + r · cos α

[0051]

[Mathematical formula-see original document] y '= y + r.sinα where (x, y) is position information before conversion, and (x',
y ') is the position information after conversion. Further, r and α are perturbation parameters, where r is the amount of movement and α is the direction of movement.

The value of the perturbation parameter is set by the perturbation information storage unit 104, the perturbation fluctuation value generation unit 103, and the perturbation parameter setting unit 102 as follows.

The perturbation information storage unit 103 is the deformation point perturbation unit 10.
The value of the perturbation parameter input from the perturbation parameter setting unit 102 is updated every time the deformation point is converted by 0, and the value of the predetermined perturbation parameter is updated every time the conversion of the deformation point of one stroke component is completed. Is initialized to.

The perturbation fluctuation value generator 103 generates values within a preset value range centered on 0 in a time series. However, the probability of occurrence of that value is highest at 0, and becomes lower as the value is farther from 0. The perturbation fluctuation value generation unit 103 outputs the generated value to the perturbation parameter setting unit 102.

The perturbation parameter setting unit 102 reads the values of the perturbation parameters stored and held in the perturbation information storage unit 104, adds the values input from the perturbation fluctuation value generation unit 103, and transforms them as new perturbation parameter values. It outputs to the point perturbation unit 101 and the perturbation information storage unit 104.

FIG. 14 is an explanatory diagram of the operations of the modified image generation unit 105 and the smoothing processing unit 73. Based on the transformation of each transformation point by the transformation point perturbation unit 101, the transformation image generation unit 105 determines the positions of all pixels in the pixel block input from the image blocking unit 100 corresponding to those transformation points as shown in FIG. Convert as shown in.

Next, the converted character image is output to the smoothing processing unit 73. The smoothing processing unit 73 is shown in FIG.
As shown in FIG. 4, the input image is subjected to the low-pass filter process represented by (Equation 5) to reduce the unevenness between the pixel blocks and smooth the edge portion of the character. The processed character image is output to the determination unit 7.

[0058]

[Formula 5] F (x, y) = G (Σ ⁿ _{a = 0} Σ ^m _{b = 0} {f (xa, yb) ・
h (a, b)}) where f (x, y) is the character image before processing when the pixel value of the black pixel is 1 and the pixel value of the white pixel is 0, and F (x, y) is the processing Later character image, G (x) is 1 when x is larger than the set value,
In the following cases, a function that outputs 0, h (a, b) is a template representing a low-pass filter having a value that monotonically decreases from the center position toward the periphery and that satisfies (Equation 6). = 3 and n = 3, a template as shown in FIG. 15 can be considered.

[0059]

## EQU6 ## Σ ⁿ _{a = 0} Σ ^m _{b = 0} h (a, b) = 1 FIG. 16 is a block diagram of the output position determination unit 7, where 140 is the preceding character information storage unit and 141 is the concatenated character dictionary. , 142 is a start point end point connection section, 143 is an output position fluctuation value generation section, 14
Reference numeral 4 is a black pixel position output unit.

The output position fluctuation value generator 143 generates values within a preset value range centered on 0 in a time series,
The probability of occurrence of that value is highest at 0, and becomes lower as the value is farther from 0. The output position fluctuation value generation unit 143 outputs the generated value to the black pixel position output unit 144.

The output position information storage unit 8 stores and holds position information for the coordinate system of the circumscribing rectangle of the character image output one character before, and each time the output for one character is completed, the position information is stored. The contents are updated to the position information of the circumscribed rectangle of the character image.

The black pixel position output unit 144 first adds the position information of the black pixel of the character image input from the image transformation unit 5 and the value input from the output position fluctuation value generation unit 143 to obtain a new black pixel. Request location information. Next, the position information stored and held in the output position information storage unit 8 is read and added to the newly obtained position information of the black pixel to obtain the position of the black pixel in the coordinate system of the printing paper, and the start point end point connection unit 1
42 and the modified image output unit 9.

The previous character information storage section 140 stores and holds the character code of the character image converted one character before and the position coordinate in the print paper coordinate system of the end point of the writing of the character image, and the conversion of one character is completed. Each time, the contents are updated to the character code and the end point position coordinate at that time. Further, the concatenated character dictionary 141 previously stores and holds a combination of characters that may be concatenated as a combination of character codes. The combination of characters that may be connected is, for example, the case of 2 and 0 as shown in FIG.

The starting point / ending point concatenation unit 142 searches the concatenated character dictionary 141 for a combination of the character code stored and held in the previous character code storage unit 140 and the character code of the character image currently being converted, and if the combination is If present, the positional information of the end point of the previous character stored and held in the previous character information storage unit 140 is read out at a certain frequency, and the start point of the character input from the black pixel position output unit 144 from the end point of the previous character. A line segment connecting between the line segments is obtained, and position information of black pixels on the line segment is output to the modified image output unit 9.

The modified image output unit 9 of FIG. 1 prints dots of black pixels on the printing paper according to the position information of the black pixels input from the output position determination unit 7.

As described above, according to this embodiment, the standard handwritten character image and the handwriting information can be automatically registered from the character string written by the user by the automatic handwriting information setting unit 21,
The burden on the user can be reduced. Further, the manual handwriting information setting unit 22 can surely register the correct standard handwritten character image and the handwriting information. Further, the image transforming unit 5 transforms the standard handwritten character image based on the transforming information of the previous character stored and held in the transforming information storage unit 6 so that the characters of the same category always have different shapes and sizes, and further output. Position determination unit 7
Is output based on the output position of the previous character stored and held in the output position information storage unit 8 with a fluctuation in the output position, so that an image of a character string close to handwriting can be generated.

Each means of the present invention can be realized by software using a computer, or can be realized by using a dedicated hardware circuit having each of these functions.

[0068]

As is apparent from the above description,
According to the present invention, since the handwritten character image of the user registered in advance by the standard handwritten character image registration unit is used as the standard character image, a character image similar to the character shape written by the user can be generated.

Further, by modifying the standard character image by the image modifying unit based on the information in the modification information storage unit, characters in the same category always have different shapes,
It becomes the size.

Further, based on the information in the output position information storage unit, the output position determination unit does not make the output position of the character image constant but gives the output position fluctuation, and the modified image output unit is transformed to the output position. By outputting the character image, it is possible to output a character string image that is closer to handwriting.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a character string image generation device according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a standard handwritten character image registration unit according to an embodiment of the present invention.

FIG. 3 is a configuration diagram of an automatic writing information setting unit and a manual writing information setting unit according to an embodiment of the present invention.

FIG. 4 is a configuration diagram of a stroke extraction unit according to an embodiment of the present invention.

FIG. 5 is an explanatory diagram of an operation of a stroke position determination unit according to an embodiment of the present invention.

FIG. 6 is an explanatory diagram of a manual writing information setting unit according to an embodiment of the present invention.

FIG. 7 is an explanatory diagram of a manual writing information setting unit according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram of a manual writing information setting unit according to an embodiment of the present invention.

FIG. 9 is a configuration diagram of an image transformation unit according to an embodiment of the present invention.

FIG. 10 is a configuration diagram of a linear conversion unit according to an embodiment of the present invention.

FIG. 11 is an explanatory diagram of an operation of the linear conversion unit according to the embodiment of the present invention.

FIG. 12 is a configuration diagram of a non-linear conversion unit according to an embodiment of the present invention.

FIG. 13 is an explanatory diagram of an operation of a deformation point perturbation unit according to an embodiment of the present invention.

FIG. 14 is a configuration diagram of a modified image generation unit according to an embodiment of the present invention.

FIG. 15 is a diagram illustrating an example of a template for low-pass filter processing of a smoothing processing unit according to an embodiment of the present invention.

FIG. 16 is a configuration diagram of an output position determining unit according to an embodiment of the present invention.

FIG. 17 is a diagram of an example of a connected character.

FIG. 18 is a schematic configuration diagram of a conventional image generation device.

[Explanation of symbols]

1 Character code input unit 41 End point extraction unit 2 Standard handwritten character image registration unit 42 Intersection point extraction unit 3 Image storage unit 43 Stroke position determination unit 4 Image reading unit 71 Linear conversion unit 5 Image transformation unit 72 Non-linear transformation unit 6 Deformation information storage Part 73 Smoothing processing part 7 Output position determination part 81 Transformation coordinate system determination part 8 Output position information storage part 82 Fluctuation value generation part 9 Deformed image output part 83 Conversion parameter setting part 20 Image reading part 84 Image deformation point conversion part 21 Automatic writing Information setting unit 100 Image block diagram 22 Manual writing information setting unit 101 Deformation point perturbation unit 30 Character cutout unit 102 Perturbation parameter setting unit 31 Stroke extraction unit 103 Perturbation fluctuation value generation unit 32 Character recognition unit 104 Perturbation information storage unit 33 Automatic start point End point determination part 1 5 Deformed image generation unit 34 Automatic deformation point determination unit 140 Pre-character information storage unit 35 Writing information image display unit 141 Concatenated character dictionary 36 Character image determination unit 142 Start point / End point coupling unit 37 Character code determination unit 143 Output position fluctuation value generation unit 38 Start point end point determination unit 144 Black pixel position output unit 39 Transformation point determination unit 201 Font storage unit 40 Image thinning unit 202 Font reading unit 203 Image output unit

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G09G 5/24 8121-5G 9192-5L G06F 15/72 355 U

Claims (5)

[Claims] 1. A standard for inputting a code of a character desired to be output by a user, a standard handwritten character image is read, and a character code corresponding to the image and writing information such as a character stroke are added. A handwritten character image registration unit, an image storage unit that stores and holds the image to which the information is added, and a character image to which the character code input from the character code input unit is added is read from the image storage unit. An image reading unit, a deformation information storage unit that stores and holds deformation information, an image deformation unit that deforms the read image based on the deformation information, and a character image deformed by the image deformation unit is output. And a deformed image output unit for performing the character string image generation device. 2. The standard handwritten character image registration unit extracts an image reading unit for reading an image of a user's handwritten character, extracts a character region from the read image, and stores a character image stored in the image storage unit. A character slicing unit that determines the character image obtained by the character slicing unit and a character recognizing unit that determines the code of the corresponding character, and a stroke for each character from the character image obtained by the character slicing unit. A stroke extraction unit for extracting and outputting the position information, and an automatic setting of the position information of the start point and the end point of the writing of the character as the writing information based on the input from the character recognition unit and the input from the stroke extraction unit. The start point and end point determination section, and the points that divide each stroke are the transformation points when transforming the character,
The character string image generation device according to claim 1, further comprising: an automatic writing information setting unit including an automatic deformation point determination unit that sets position information of the deformation point as writing information. 3. A standard handwritten character image registration unit, an image reading unit for reading an image of a user's handwritten character, and an image display unit for displaying the read character image.
The character image determination unit that sets the character image that the user stores and holds in the image storage unit by looking at the displayed image, and the position of the start point and the end point of the writing when the user looks at the displayed image The start point and end point determination unit that sets the information as writing information, the deformation point determination unit that sets the position information of the deformation point when the user deforms the character looking at the displayed image as the writing information, and the displayed image The character string according to claim 1 or 2, further comprising: a manual writing information setting unit including a character code determination unit that allows a user to determine a character code corresponding to the image. Image generator. 4. The image transforming unit includes a linear transforming unit that transforms the entire character image by applying uniform coordinate transformation to the entire character image, and a non-linear transforming unit that transforms each stroke component of the character image. The character string image generation according to claim 2 or 3, further comprising: a smoothing processing unit that smoothes an edge portion of the character image converted by the non-linear conversion unit and removes fine unevenness. apparatus. 5. An output position deciding unit for deciding at which position a character image transformed by the image transforming unit is to be output, and output position information for storing and holding information on the output position decided by the output position deciding unit. The character string image generation device according to claim 2, 3, or 4, further comprising: a storage unit.