JP2000330545A - Font forming device, font forming method, font service system and storage medium for forming font - Google Patents

Abstract

PROBLEM TO BE SOLVED: To form a beautiful personal font to reflect the personality even for people who can not write characters beautiful by extracting personal parameters of hand written characters, and by modifying parts reflecting personality of standard font according to the personal parameters. SOLUTION: A character recognition part outputs a character code of recognition result to a personality extraction means 2 by using character patterns outputted from a pattern fetching means 1 as an input. The personality extraction means 2 extracts personal parameters of hand written characters and outputs them to a signal line 100s. An extraction means 4 of personality reflecting parts reads out a standard font 3 for every single character from the standard font 3 via a signal line 201s, and extracts a part to be added with deformation to reflect personality from the standard character 3. A personality reflecting means 5 receives parameters of positions of a voiced sound symbol and a semivoiced sound symbol, a size of loop, and modification of roundness as the personal parameters, and converts the coordinate point of standard font 3 as a whole.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention can be used to create fonts suitable for creating and distributing documents and greeting cards.

[0002]

2. Description of the Related Art Conventionally, a handwritten font (hereinafter, referred to as Japanese Patent Application Laid-Open No. 10-27217 or Japanese Patent Application Laid-Open No. 5-165823) capable of reflecting individual personality has been disclosed in Japanese Patent Application Laid-Open Nos. , Personal fonts).

In Japanese Patent Application Laid-Open No. Hei 10-27217, the similarity between several individual dictionary characters and personal characters is measured, and how the average of the weights of several dictionary characters is calculated for each individual It examines whether it can be made to resemble a character, and creates a personal font from its weight parameter and dictionary characters.

In Japanese Patent Application Laid-Open No. 5-165823, a personal font is created by adding flesh to a character written by an individual.

[0005] In addition to the standard fonts supported by ordinary systems, there is an example of a personal font business in which various fonts designed by a user or a system provider are provided on the Internet.

[0006]

SUMMARY OF THE INVENTION However, Japanese Patent Application Laid-Open No. 10-27
In the method of Japanese Patent Publication No. 217, an outline font cannot be used as a dictionary character in order to measure the similarity between a standard dictionary character and a personal character and create a personal font using the same dictionary character. Also, since personal fonts are not outline fonts, they lack expression power.

The method disclosed in Japanese Patent Laid-Open No. 5-165823 is
If you are not a person who can write beautiful characters, personal fonts may not be beautiful.

Accordingly, the present invention provides a function for creating a personal font using an outline font.

Also, the present invention provides a function for creating a beautiful personal font that reflects personality even for a person who cannot write beautifully.

In the conventional personal font service business, a personal font can be used instead of a facsimile or electronic mail signature, or used for a greeting card. Also, by writing a custom font for yourself on a FAX form, you can create a custom font that resembles that character. As described above, systems that provide fonts other than the standard fonts supported by the system have appeared, but have the following problems.

1) When a personal handwritten character is sent, there is a font service for a character code corresponding to the sent character, but no font service for other characters. For this reason, services for kanji in Japan and China, which have thousands of characters, are not considered.

2) The capacity of a personal font set of kanji can be several MB, so if this information is transmitted via communication, at a normal communication speed (32 kbps-64 kbps),
It takes several hours. For this reason, exchange of mail documents using personal fonts has not been realized. Further, storing a plurality of personal font sets in a client PC requires a large-capacity disk.

3) Printing of a document using a personal font to a printer or facsimile connected to a network or F
AX transfer has not been considered in the past, and there is a problem in that it is not possible to output characters using unique characters.

Therefore, the present invention provides a function that can use personal fonts in a network environment such as output to a printer or a facsimile, transfer of a greeting card, and transfer of a mail document.

[0015]

For this purpose, a pattern capturing means for capturing a character pattern and personality parameters of handwritten characters (positions of voiced / semi-voiced voices, loop size, roundness, aspect ratio, inclination, rotation, Individuality extraction means for extracting a trapezoidal transformation parameter, a standard font consisting of a set of segments represented by coordinate points, and a portion to be transformed from the standard font to reflect the individuality (a segment of a cloud point / semi-voiced point, L Personality reflection location extraction means for extracting a segment having a character shape and a segment having a loop), and receiving the parameters of the position of the cloudy spot / semi-voiced point, the size of the loop, and the deformation of the roundness as individuality parameters. Therefore, the corresponding segment of the standard font is transformed, and the aspect ratio, tilt, rotation, and trapezoidal transformation parameters are changed to individuality parameters. Receiving a chromatography data, and a personality reflecting unit for converting the coordinate point of the entire standard fonts.

For this purpose, according to the present invention, a personality parameter extraction service means for extracting and servicing personality parameters of handwritten characters and a program for generating a personal font from personality parameters extracted by the standard font and the personality parameter extraction service means are provided. The personal font server comprises a personal font generation program service means for performing the above service on a network in response to a request from a client.

Further, in order to output a document using a personal font to a printer connected to a network,
Personal font generation means for receiving a personality parameter dictionary service from a personality parameter extraction service means of a personal font server, further receiving a personal font generation program service from a personal font generation program service means to generate a personal font, and a client The personal font print server includes personal font print service means for receiving and providing a printer request from a terminal and network management means for performing the service on a network.

Further, in order to provide a glitching card service using a personal font to another client connected to the network, the personal font server receives a personality parameter dictionary service from the personality parameter extraction service means of the personal font server. Personal font generating means for receiving a personal font generating program service from the font generating program service means and generating a personal font, further receiving a card issuance request from a client terminal;
It has a greeting card server composed of a greeting card service means for providing services and a network management means for providing the services on a network.

The pattern capturing means captures a character pattern. The individuality extraction means extracts individuality parameters of handwritten characters (parameters of voiced / semi-voiced points, loop size, roundness, aspect ratio, inclination, rotation, and trapezoidal deformation). The standard font holds a set of segments represented by coordinate points. The individuality reflection point extraction means extracts a portion to be deformed (a segment having a clouded point / semi-voiced portion, a segment having an L-shaped shape, a segment having a loop) from a standard font to reflect the individuality. The individuality reflection means receives the parameters of the deformation of the position of the voiced / semi-voiced point, the size of the loop, and the degree of roundness as personality parameters, and deforms the corresponding segment of the standard font according to these parameters. The parameters of rotation and trapezoid deformation are received as individual parameters, and the coordinate points of the entire standard font are converted.

The present invention is effective in a network system having a client terminal and a server. First, when personal handwritten characters are sent from the client PC to the personal font server in advance, the personality parameter extraction service means of the personal font server extracts personality parameters and registers them in the personality parameter dictionary. Next, when the client PC requests the print server to print a document using a personal font, the print server receives the services of the personality parameter dictionary and the personal font generation program from the personal font server, and is sent from the client terminal. The printer outputs the character code of the document in a personal font.

Similarly, the greeting card develops the character code of the sent greeting card document into image data with a personal font, and as a greeting card,
Attach as a mail file.

On the other hand, when a mail document is to be received in a personal font, a personal font server
The personalized parameter dictionary of the corresponding ID and the service of the personal font generation program are received, and the character code of the mail document is similarly output on the screen in the personal font.

As described above, the personal font server mainly performs services of extracting personality parameters, providing a personality parameter dictionary, and providing a personal font generation program.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment in which the present invention is applied to a font creating apparatus will be described in detail.

FIG. 1 is a system configuration diagram of the font creating apparatus.

The pattern capturing means 1 is composed of, for example, a tablet or a scanner for reading a sample point sequence per unit time, captures a character pattern, and normalizes the size and position. Then, the normalized character pattern is output to the signal line 101s together with the character code. If the character code is received by a keyboard or through a character recognizing unit in the middle of the signal line 101s, the character recognizing unit can determine the character code using only the character pattern and pass it to the personality extracting unit 2. . When the character recognition unit is used, the character recognition unit receives the character pattern output from the pattern capturing unit 1 as an input, and outputs the character code of the recognition result to the individuality extraction unit 2.

The personality extracting means 2 includes personality parameters of handwritten characters as shown in FIG. 2 (positions of voiced / semi-voiced dots, loop size, roundness, aspect ratio, inclination, rotation, trapezoidal deformation parameters, etc.). Is extracted and output to the signal line 100s.

The standard font 3 of FIG. 1 holds a set of segments represented by coordinate points.

The individual reflection point extraction means 4 is connected to the signal line 201.
A standard font for each character is read from the standard font 3 via the s, and a portion to be transformed to reflect individuality from the standard font (having a voiced / semi-voiced segment, an L-shaped segment, and a loop) Segment).

The individuality reflecting means 5 receives the signal from the signal line 100 s
The parameters of deformation of the position of the voiced / semi-voiced point, the size of the loop, and the degree of roundness are received as individuality parameters.
From 202s, it is possible to add a transformation to reflect individuality by receiving information on which part of the standard font, transforming the corresponding segment of the standard font according to these information, and changing the aspect ratio, tilt, rotation, trapezoidal The transformation point is received as an individuality parameter, and the coordinate points of the entire standard font are transformed. The converted font is output as a personal font.

FIG. 3 is a detailed block diagram of the individuality extracting means 2. A character code and a character pattern (hereinafter, input pattern) can be received from the signal line 101s.

The dictionary 2-5 holds several characters or all characters in pairs of character codes and standard character patterns. The standard character pattern has the same format as the input pattern read by a tablet or a scanner, or the same format as the standard font 3.

If the standard character pattern has the same format as an input pattern read by a tablet or a scanner, the entire system can be shared by using a dictionary used for character recognition if there is a character recognition unit and a matching processing unit. Occupied memory capacity can be reduced.

When the standard character pattern has the same format as the standard font 3, it can be shared with the standard font 3. In this case, since the standard character pattern and the input pattern have different coordinate systems, it is necessary to match them. For example, after converting the position and size of the standard character pattern so as to match the average position and average size of the input pattern, the following features are compared. In addition, images, loops, and patterns of standard character patterns and input patterns for comparing features
The positions of the corners are associated with each other based on the stroke order and the flag attached to the standard pattern.

If the stroke order of the input pattern is erroneously input, the n-th stroke of the input pattern is not a loop even if the n-th stroke of the standard pattern has a loop flag. You can check if the stroke order is incorrect. If a stroke order error is detected, feature extraction is not performed.

Alternatively, a standard pattern having the same format as that of the standard font 3 and an image, loop,
Also in the correspondence of the corner positions, if the mediation is performed using a mediation standard character pattern having the same format as the input pattern read by the tablet or the scanner, the feature can be extracted from the input pattern having a different stroke order. The correspondence between the input pattern and each element of the intermediate standard character pattern is performed using the function of the matching processing unit of the character recognition unit, and the correspondence of each element of the intermediate standard character pattern and the standard character pattern is assigned to the stroke order and the standard pattern. Perform based on the flag.

The entire individuality extracting unit 2-1 is connected to the signal line 101.
The pattern of the character code received from s is read from the dictionary 2-5, this dictionary pattern is compared with the input pattern received from the signal line 101s, and parameters of aspect ratio, inclination, rotation, and trapezoidal deformation are used as individual parameters. Extract.

For example, as shown in FIG. 4, each input / dictionary pattern is divided into four regions with the center point such as the center of gravity as the origin. Then, a rectangle circumscribing the pattern of each area is obtained, and the input pattern is
The upper right point p1 (x1, y1) in the first quadrant, the upper left point p2 (x2, y2) in the second quadrant, and the lower right point p3 in the third quadrant
(X3, y3), the lower left point p4 (x4, y
4) Score points. The dictionary pattern includes an upper right point q1 (x1 ′, y1 ′) in the first quadrant and an upper left point q2 (x
2 ', y2'), and the lower right point q3 (x3 ', y
3 '), in the fourth quadrant, a point is set at the lower left point q4 (x4', y4 ').

Alternatively, the points farthest from the origin in each area are defined as p1, p2, p3, p4, q1,
q2, q3, and q4 may be used.

Then, by comparing each point of the (a) input pattern and (b) the dictionary pattern and the transition vector between the points, for example, the following parameters are obtained.

(1) Aspect ratio (| x1-x2 | + | x4-x3 |) / (| y1-y2 |
+ | Y4-y3 |)-(| x1'-x2 '| + | x4'
−x3 ′ |) / (| y1′−y2 ′ | + | y4′−y3 ′
|) (2) inclination y-axis inclination: angle of line segment p1p3−angle of line segment q1q3,
Or ((angle of line segment p1p3−angle of line segment q1q3) + (line segment p
2p4 angle−angle of line segment q2q4)) / 2 X-axis tilt: angle of line segment p2p1−angle of line segment q2q2,
Or ((angle of line segment p2p1−angle of line segment q2q2) + (line segment p
(4p3 angle−line segment q4q3 angle)) / 2 (3) rotation Primary conversion parameter from line segment q4q1 to line segment p4p1 (4) Trapezoid length (q1-q2) / length (q1-q3) < In the case of the threshold value, length (p1-p2) / length (p1-p3) otherwise, 1 or as shown in FIG. 5 and FIG. Each parameter may be obtained by comparing the lengths of the bar and bar elements. For example, in FIG. 5, attention is paid to (b) horizontal bars w1, w2, and w3 having a length equal to or longer than the threshold value of the dictionary pattern, and the corresponding (a) horizontal bars v1, v2, and v3 of the input pattern are respectively angled. In comparison, x-axis tilt: ((angle w1-angle v1) + (angle w2-angle v
2) + (angle w3−angle v3)) / 3 If the x-axis tilt is smaller than the threshold value, and each angle difference
If the absolute values of (angle w1−angle v1) and (angle w3−angle v3) are larger than the threshold, the following is obtained as a trapezoidal parameter.

Trapezoid: y-axis direction change amount (v1 start point−v3 start point) / y-axis direction change amount (v1 end point−v3 end point) Similarly, for the vertical bar as shown in FIG. Can be obtained.

At this time, the dictionary patterns w1, w2, w
In order to find the horizontal bar corresponding to No. 3, matching processing between the dictionary pattern and the input pattern is performed.

The voiced / semi-voiced voice position extraction unit 2-2 converts the character code pattern received from the signal line 101s into a dictionary 2-5.
Then, the dictionary pattern is compared with the input pattern received from the signal line 101s, and the parameters of the deformation of the positions of the voiced / semi-voiced points are extracted as individuality parameters.

FIG. 7 is an example of a character in which the positions of the voiced and semi-voiced characters are unique.

As shown in FIG. 8, matching is performed between (a) the input pattern and (b) the dictionary pattern, and an image of the input pattern corresponding to the voiced / semi-voiced point of the dictionary pattern is extracted. The details of the data structure of the input / dictionary pattern and the matching process are disclosed in Japanese Patent Application Laid-Open No. H10-214312. The positions of the voiced and semi-voiced dots in the dictionary pattern are flagged on the dictionary pattern, or point sequence data indicating the pattern of each image is stored in the dictionary in the order of strokes. It is desirable to be able to search for two images as the corresponding images. FIG. 9 is a diagram for comparing the direction of a cloud point between a dictionary pattern and an input pattern. Similarly to the method described in the description of FIGS. 4 to 6, the inclination and rotation parameters can be extracted based on the coordinates of the start and end points of the turbid point and the transition vector between the start and end points.

It is characterized by the following.

The loop size extracting means 2-3 in FIG. 3 converts the character code pattern received from the signal line 101s into the dictionary 2
-5, the dictionary pattern and the signal line 101
By comparing the input pattern received from s with the input pattern received from s, a parameter of loop size deformation is extracted as an individuality parameter. As for the position of the loop of the dictionary pattern, a flag is added to the dictionary pattern, or an angle change of the pattern is checked, and an image having an angle change equal to or more than a threshold is searched for as an image forming a loop. In the case of the method of adding a flag, point sequence data indicating the pattern of each image is stored in a dictionary in the order of strokes, so that the first image of character data indicates what is a loop shape, or whether each image is a loop or not. Add a flag. Characters including a plurality of loops, such as "@", may be distinguished in the stroke order, such as a third loop / fourth loop.

Then, the areas enclosed by the loops are compared as shown in FIG. 10, and when the ratio is equal to or larger than the threshold value, the ratio is used as a parameter of the deformation of the loop size.

The roundness extracting means 2-4 shown in FIG.
The pattern of the character code received from 1s is read from the dictionary 2-5, the dictionary pattern is compared with the input pattern received from the signal line 101s, and the parameter of the degree of roundness is extracted as an individuality parameter.

For example, when the character pattern received from the signal line 101s is online data from a tablet, the area enclosed by the original handwriting and the approximate line segment is reduced as disclosed in Japanese Patent Application Laid-Open No. H10-214312. Approximation is made so as to be a constant value (FIG. 11A). Similarly, the dictionary pattern read from the dictionary 2-5 is approximated as shown in FIG. Then, in (b), an angle formed by polygonal lines having a length equal to or greater than the threshold value lth and an angle having an angle change greater than or equal to the threshold value ath are searched for a corresponding polygonal line portion in (a) by matching processing. In (a), when a plurality of polygonal lines are formed with an angle change equal to or smaller than the threshold value bth, it is determined that the line is round. Then, a ratio between the angle of the largest angle change in (a) and the angle in (b) is obtained,
This is defined as the degree of roundness. Alternatively, the ratio of the part of the corresponding part of the corner in (b) that is considered to be round may be defined as the degree of roundness.

For details of the input / dictionary pattern matching process, a method disclosed in Japanese Patent Application Laid-Open No. H10-214312 or Japanese Patent Application Laid-Open No. Sho 59-139483 may be used.

FIG. 21 shows the operation procedure of the individuality extracting means 2 described above. First, when a handwritten character written by the user is read in step 2.1, an individuality parameter is extracted by the following processing. First, the entire personality is extracted in step 2.2, the positions of the turbid / semi-voiced points are extracted in step 2.3, the degree of roundness is extracted in step 2.4, and the individuality data is transmitted in step 2.5. Here, the procedures 2.2 to 2.4 need not be performed in this order.

FIG. 22 shows a detailed operation procedure of the procedure 2.2. First, in step 2.2.1, the area is divided into four with the center of the input pattern as the origin. Then, in step 2.2.2, the vertex coordinates of each input area are obtained (see FIG. 4). Similarly, the dictionary patterns are performed in steps 2.2.3 and 2.2.4. Then, in step 2.2.5, it is checked whether there is an aspect ratio, a tilt, a rotation, or a trapezoidal deformation, and if so, the parameters are obtained (see FIGS. 4 to 6).

FIG. 23 shows a detailed operation procedure of the procedure 2.3. First, in step 2.3.1, it is determined whether or not the character code received from the signal line 101s is a character having a voiced / semi-voiced point such as [ga, gi, guu, gure, go, da ,,]. Do nothing if there is. If Yes, the following processing is performed. In step 2.3.2, the dictionary pattern of the corresponding character is read, and in step 2.3.2, dictionary / input pattern matching is performed.
This is performed in order to find the positions of the voiced and semi-voiced dots of the input pattern even when the stroke order of the input pattern is incorrectly input. In step 2.3.4, the positions of the elements of the input pattern corresponding to the voiced / semi-voiced points of the dictionary pattern are extracted. And
In step 2.3.5, parameters such as positional deviation, rotation, and long / short length are obtained from the start / end point of the voiced / semi-voiced point of the dictionary / input.

FIG. 24 shows a detailed operation procedure of the procedure 2.4. First, a loop region of an input pattern is searched in step 2.4.1, and the area of the input loop region is obtained in step 2.4.2. Similarly, steps 2.4.3 and 2.4.4 are performed for the dictionary pattern. Then, in step 2.4.5, the area of the dictionary / input is compared to obtain a parameter of the loop size. Prior to the processing, it is determined whether or not the character code received from the signal line 101s is a character having a loop shape such as [A, O, S,,], and if No, nothing is performed. Can be reduced.

FIG. 25 shows a detailed operation procedure of the procedure 2.5. First, in step 2.5.1, search for the corners of the dictionary pattern.
In step 2.5.2, a portion of the input pattern corresponding to the corner of the dictionary pattern is searched. Then, in procedure 2.5.3, it is determined whether the input pattern is a corner or a curve. For example, if the angle difference between the polygonal lines is equal to or less than the threshold value and three or more line segments are continuous, it can be determined that the curve is a curve. If it can be determined as a curve, step 2.
The parameter of the degree of roundness is obtained at 5.4.

FIG. 12 is a detailed block diagram of the personality reflection location extraction means 4. The standard font is read from the signal line 201s, and a part to be deformed to reflect the personality of the standard font is extracted and output to the signal line 202s.

The extraction of the partial individuality may be similar to the extraction of the entire individuality, and may be performed by extracting the parameters of the aspect ratio, inclination, rotation, and trapezoidal deformation as individuality parameters for each drawing.

The font reading means 4-1 designates a character code and reads out a standard font for one character from the standard font, the whole coordinate extracting means 4-2, the cloud point / semi-voice point extracting means 4-2, and the loop extracting means. 4-4, and is passed to the roundness reflection unit extraction means 4-5. This is repeated for necessary characters or for all characters.

FIGS. 14 and 15 show examples of standard fonts. FIG. 14 shows a general outline font, in which some of the methods disclosed in the present invention cannot be applied, such as the inability to reflect the individuality of the size of the loop, but the same individuality, such as reflecting the entire individuality. Personality can be reflected. FIG. 15 shows an outline font in which a portion corresponding to an image of a handwritten character is defined as an independent segment. With this method, it is possible to create a personal font that reflects all the types of personality described in the present invention. In FIG. 14 and FIG. 15, the coordinate point indicated by the square is the on-control point, and the outline of the font passes through this coordinate point. A coordinate point indicated by a triangle is an off-contour point, and a spline curve passing through the preceding and following on-contour points and having the off-contour point as a target point is an outline of the font.

The whole coordinate extracting means 4-2 shown in FIG. 12 extracts the coordinate data of the standard font received from the signal line 201s as a part reflecting the whole personality.

The voiced / semi-voiced point extracting means 4-3 selects a segment corresponding to the voiced / semi-voiced point from the standard font. In order to search for a voiced / semi-voiced segment, a flag indicating this is added to the standard font. Alternatively, the segments of the last two strokes of the font of the voiced character can be searched in the order of the segments since the segments of the last two strokes of the font of the voiced character are the halftone dots. . For example, FIG. 17 shows the font of "Gu", but if the three segments are arranged in the stroke order of the characters, the last two strokes of the character having the voiced dot can be extracted as the voiced portion.

The loop extracting means 4-4 selects a segment having a loop shape from standard fonts. For this purpose, a flag is set for the loop segment of the standard font, or the segment is held in the stroke order, and the number of strokes (the number of segment strokes) is passed as data from the loop size extracting means 2-3 in FIG. You may get it. Alternatively, in the standard font, attention is paid to the line segment only in the right direction and the downward direction, and when the sum of these angle changes is equal to or more than the threshold value, it is extracted as a loop-shaped portion.

The roundness reflection part extracting means 4-5 in FIG. 12 selects a segment having a square shape from the standard font. For this purpose, a flag is added to a segment having a square shape of a standard font, or the segment is held in the stroke order, and the number of strokes (the number of segment strokes) is passed as data from the roundness extraction means 2-4 in FIG. You may be asked to do so. Alternatively, in a standard font, attention is paid to a line segment having only the rightward and downward directions and having a length equal to or greater than the threshold value Lth. It is extracted as a portion having a square shape.

As described above, the individuality parameters extracted by the global coordinate extracting means 4-2, the cloud point / semi-voiced point extracting means 4-3, the loop extracting means 4-4, and the roundness reflecting section extracting means 4-5 are reflected. The data at the desired position is indicated by a signal line 202.
s.

FIG. 13 is a detailed block diagram of the individuality reflecting means 5. Data indicating which part of the standard font reflects the personality parameter is received from the signal line 202s, and a personal font is output to the signal line 200s in response to the personality parameter from the signal line 100s.

The overall individuality reflecting means 5-1 is connected to the signal line 202.
The coordinate data of the standard font received from s is converted according to the individuality parameter. The whole individuality extracting means 2 of FIG.
Inverse transformation is performed from the parameters described in the description of 1. FIG. 16 shows an example in which the standard font is converted to reflect the overall individuality.

The cloud point / semi-voice point reflecting means 5-2 is connected to the signal line 2
The coordinate data of the segment corresponding to the cloudy point / semi-voiced point of the standard font received from 02s is converted according to the individuality parameter. For example, the segment corresponding to the cloud point / semi-voice point in FIG. 17 is translated or rotated according to the personality parameter such as upward / downward.

The loop individuality reflecting means 5-3 in FIG. 13 converts the coordinate data of the segment having the loop shape of the standard font received from the signal line 202s according to the individuality parameter. For example, if the individuality parameter of the loop is “150%
If "large", b / a * 100 = 1 as shown in FIG.
In order to satisfy 50 and b * a = 1, the non-loop-shaped segment is a = 0.8 times, and the loop-shaped segment is b = 1.2 times. To make the multiple multiple, the coordinate point is made to be a multiple of the designated number after the parallel movement with the center of gravity of each segment as the origin. The synthesis is performed after the translation in the opposite direction to the beginning of returning the center of gravity of each segment to the original center of gravity. When making it a multiple, it is advisable to perform processing that does not change the font line width. The processing in which the line width of the font does not change is disclosed in detail in JP-A-9-292873 and JP-A-9-90933. Further, it is better to adjust the position and the size after the composition so that the position and the size are the same as those of the standard font, that is, within the same character frame.

The roundness reflection means 5-4 shown in FIG.
The coordinate data of the segment having the angular shape of the standard font received from 02s is converted according to the individuality parameter.

For example, in the case where the standard font is a Gothic font as shown in FIG. 19, the on-control point at the vertex of the corner is set as the off-control, and the on-control points are set before and after the vertex of the corner. The distance between the corner vertex and the on-contour point increases as the degree of roundness increases.

In the case where the standard font is a Mincho font, a corner decoration is provided at the vertex of the corner in order to express the pressing by brush writing as shown in FIG. ), Two line segments where the extension lines intersect at 90 degrees with a polygonal line having a length equal to or greater than the threshold value are searched, and after the corner decoration is removed, the roundness is reflected similarly to the Gothic body.

The font registering means 5-5 shown in FIG. 13 outputs a personal font to the signal line 200s.

FIG. 32 shows an example of a personal font reflecting the whole personality and the partial personality. (A) (a) For the standard font, (b) the one reflecting the overall individuality: x-axis inclination of 15 degrees, and b = 1.2 times only the image data with the loop flag set (C) shows the result obtained by multiplying the other values by a = 0.8, and (d) shows the result reduced to the vertical and horizontal sizes of the standard font. Personal fonts can be similarly created for other personalities.

The individual can freely specify whether the standard font should be Mincho font or Gothic font. Alternatively, when the individual roundness parameter is large, the Gothic font may be automatically selected as the standard font.

In the above embodiment, a case has been described in which a personal font is used by one PC or form information terminal.

Next, an embodiment in which the present invention is applied to a system for providing a personal font on a network will be described with reference to FIGS.

FIG. 26 shows a system configuration of a system for providing a personal font on a network. On the network 160, as a client terminal, a pen input personal computer (Pen P
C) 100, personal computer (PC) 110,
There are 120. On the other hand, the server includes a personal font server 130, a print server 140, and a greeting card server 150. These systems work in conjunction with each other. Necessary information operates while being exchanged between servers. For example, the print server 140 and the greeting card server 150
Downloads a program and a personality parameter dictionary required for generation from the personal font server 130 when the conversion of the personal font occurs, and outputs the characters of the corresponding document in the personal font.

A specific example will be described below.

First, it is necessary to create a personality parameter dictionary of a user in the personal font server in advance. FIG. 27 shows a system configuration for performing this registration.
The PenPC 100 can collect handwritten characters with a pen. The handwritten characters input on the screen 100c are ink management 1
00b, ink data (coordinate data) is collected,
This data, along with the personal ID, is stored in the network management (N
This is sent to the personal font server 130 connected on the network via the (ET management) 100a. In the embodiment of the present invention, the ink data handwritten by the Pen PC 100 is transmitted to the personal font server 130. However, the personality parameter may be extracted by reading the handwritten characters by a scanner and transmitting the image data to the server. The procedure for creating this personality parameter dictionary in advance is the same as that described with reference to FIG.

On the other hand, the personal font server 130
Then, via the network management (NET management) 130a, the ink data is transferred to the personality parameter extraction service means.
The data is received at 130f, the personality parameter extraction process is performed, and the result data is registered as a personality parameter dictionary 130g. Here, the personality parameter extraction service means 130f
It refers to the individuality extracting means 2 in FIG. The personality parameter dictionary 130g is stored in association with a personal ID in order to respond to requests from other servers and client terminals.

The personal font server 130
Is a personal font generation program service means 130e that services a personal font generation program. It is typically served as a web browser applet. That is, if the function does not exist in the required system, the function is automatically downloaded from the server. It may be downloaded as a plug-in by the FTP function.

Further, it has a recognition service function for recognizing handwritten characters represented by ink data and converting them into character codes. This function includes a recognition dictionary 130d, a recognition engine 130c, and a recognition engine service unit 130b. When there is a recognition request from the client terminal,
The recognition engine service means 130b sends the input ink data to the recognition engine 130c. Recognition engine 13
0c performs recognition with reference to the recognition dictionary 130d, and sends it back to the client terminal via the recognition engine service means 130b.

Next, a system configuration for sending the document 110b from the personal ID 110c and the document 110b from the client terminal PC 110 to the print server 140 and printing out the document 110b will be described with reference to FIG. First, the personality parameter dictionary 140c corresponding to the personal ID 110c
Is not in the print server 140, and if there is no personal font generation program 140e, the personal print service means 140g detects this and downloads the personality parameter dictionary 140c and the personal font generation program 140e from the personal font server 130. I do. Here, in the print server 130, the personal font generation program 130e, after being downloaded, remains in the system as it is. However, the personality parameter dictionary 14
Although 0c exists only for the cache memory, basically, it is basically downloaded for each personal ID 110c.

Personal font generation program 140
e is the standard font 140d and the personality parameter dictionary 14
Using 0c, a personal font 140f is generated. This personal font 140f is developed according to the character code of the read document 110b and is used as a cache font memory. The font typeface is also Mincho P
F1, Gothic PF2, and Brush PF3 are developed for each font format. If memory size is allowed,
The font set of all characters may be expanded.

When the development of the personal font 140f is completed, the personal print service means 140g
The expanded personal font 140f is sent to the printer 14
Output to 0h.

As described above, document printing of a personal font can be performed in accordance with the personal ID.

Next, a system configuration for providing a greeting card such as a Christmas card or a New Year's postcard will be described with reference to FIG. 29, in the client terminal PC 110, the greeting card document 110d and the mailing address 110e are stored in the print server 14 shown in FIG.
0b. Greeting Card Document 11
0d is a greeting card composed of a picture, a greeting, and a text, created in, for example, an HTML style. A personal font can be called for a character font in a document. In FIG. 29, in the greeting card server 150, the greeting card service means 150g manages the entire server. The created greeting card provides a service of attaching the greeting card to a mail server (not shown) and transferring a personal font based on the information of the mailing address 110e designated by the client.

Note that when a facsimile server is constructed,
With the same configuration as the greeting card server 150, the greeting card service means 150g is faxed.
As a service means, and a greeting card document 1
It can be constructed by using 10d as a FAX document. The facsimile service unit calls a personal font for the character font in the facsimile and converts it into image data for facsimile.

Next, the system configuration and data flow of a mail document will be described with reference to FIG. FIG.
Shows a case where a mail document is received by the other client terminal PC 110 from the PC 120 as the client terminal. In this figure, the only difference from the print server 140 and the greeting card server 150 is the mail management means 110g, and the rest is the same. That is, a necessary personality parameter dictionary is specified from the mail address and the personal ID, a personal font is developed according to the character code of the mail document, and displayed and output on the screen 110h. Thereby, the contents of the mail can be displayed in a rich expression.

FIG. 31 shows an operation procedure for providing a personal font on the network described above. First,
In step 3.1, a pattern is fetched, and in step 3.2, individuality is extracted to extract individuality parameters. Next, in step 3.3, a personality reflection location is extracted from the standard font, and in step 3.4, the personality is generated by reflecting the personality in the location. Then, in order to determine whether the generated personal font is used in each service, it is checked in step 3.5 whether the creator of the display / print document source is the owner of the personal font. If the owner of the personal font does not want to be used by others without permission, prepare a personal font owner password,
Attach a password to the display / print document source. Alternatively, it is confirmed that the creator of the display / print document source is the owner of the personal font by the personal authentication means other than the above. In steps 3.6 to 3.8, printing and display are performed using the personal font in each service.

As described above, in the embodiment of the present invention, the system is constructed based on the server on the network 160. However, if it is limited to a specific application, the element functions of the personal font server are implemented in the client terminal. You can still email and print.

[0094]

According to the present invention, there is provided a function for creating a personal font using an outline font.

Also, the present invention provides a function for creating a beautiful personal font that reflects personality even for a person who cannot write beautifully.

In the embodiment of the present invention, since a system for providing a personal font on a network is constructed, a document to be converted into a personal ID and a personal font is transmitted from a client terminal to a print server or a greeting server. Only with this, there is an effect that it is possible to easily output a personal font. Furthermore, since a method for transferring only the personality parameters to the network has been realized,
There is an effect that a personal font can be displayed on a mail document or the like in real time.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram of an embodiment.

FIG. 2 is an example of an individuality parameter of the embodiment.

FIG. 3 is a partial configuration diagram of an embodiment.

FIG. 4 is an operation example of the embodiment.

FIG. 5 is an operation example of the embodiment.

FIG. 6 is an operation example of the embodiment.

FIG. 7 is an operation example of the embodiment.

FIG. 8 is an operation example of the embodiment.

FIG. 9 is an operation example of the embodiment.

FIG. 10 is an operation example of the embodiment.

FIG. 11 is an operation example of the embodiment.

FIG. 12 is a partial configuration diagram of an embodiment.

FIG. 13 is a partial configuration diagram of an embodiment.

FIG. 14 is an example of a font used in the embodiment.

FIG. 15 is an example of a font used in the embodiment.

FIG. 16 is an operation example of the embodiment.

FIG. 17 is an example of a font used in the embodiment.

FIG. 18 is an operation example of the embodiment.

FIG. 19 is an example of a font used in the embodiment.

FIG. 20 is an example of a font used in the embodiment.

FIG. 21 is an operation procedure of the embodiment.

FIG. 22 is an operation procedure of the embodiment.

FIG. 23 shows an operation procedure of the embodiment.

FIG. 24 is an operation procedure of the embodiment.

FIG. 25 is an operation procedure of the embodiment.

FIG. 26 is a system configuration diagram when the present invention is constructed on a network.

FIG. 27 is a diagram illustrating a configuration diagram of a personal font server and a method of registering a personality parameter dictionary.

FIG. 28 is a system configuration diagram of a print server.

FIG. 29 is a system configuration diagram of a greeting card server.

FIG. 30 is a diagram showing a system configuration in the case of transferring a mail.

FIG. 31 shows an operation procedure of a system using a personal font.

FIG. 32 is an example of a font used in the embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Pattern taking-in means, 2 ... Individuality extraction means, 3 ... Standard font, 4 ... Individuality reflection location extraction means, 5 ... Individuality reflection means, 6 ... Personal font, 130 ... Personal font server, 140 ... Print server, 150 ...
Greeting card server.

Continued on the front page (72) Inventor Keiko Abe 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture Inside the Hitachi Research Laboratory, Hitachi, Ltd. (72) Inventor Toshiaki Nakamura 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture Inside Hitachi Research Laboratory, Hitachi, Ltd. (72) Inventor Akihiro Katsura 7-1-1, Omikacho, Hitachi City, Ibaraki Prefecture Inside Hitachi Research Laboratory, Hitachi, Ltd. (72) Inventor Yasushi Fukunaga 7, Omikamachi, Hitachi City, Ibaraki Prefecture No. 1-1 F-term in Hitachi Research Laboratory, Hitachi, Ltd. (Reference) 5B009 RA05 5B021 JJ05 JJ06 5C082 AA01 AA17 AA25 AA27 BA02 BB01 BB34 BB49 CA31 CA54 CA82 CB06 DA32 DA87 DA89 MM09 MM10

Claims (12)

[Claims] 1. A pattern capturing means for capturing a character pattern; a personality extracting means for extracting a personality parameter of a handwritten character; a standard font; and a personality reflection location for extracting a portion to be transformed from the standard font to reflect the personality. A font creation device and a storage medium for font creation, comprising extraction means and personality reflection means for transforming a personality reflection location of a standard font according to a personality parameter. 2. A pattern capturing means for capturing a character pattern, a personality extracting means for decomposing a handwritten character into segments and extracting respective personality parameters, a standard font, and a modification for reflecting the personality from the standard font. A font creation device and a storage medium for font creation, comprising: a personality reflection location extracting means for extracting a portion; and a personality reflection means for transforming a personality reflection location of a standard font according to a personality parameter. 3. The method according to claim 1, wherein the personality extracting means has a dictionary for holding a pair of a character coat and its standard pattern, and includes a dictionary pattern (hereinafter, referred to as a dictionary pattern).
The reference point of the dictionary pattern) is compared with the reference point of the pattern (hereinafter referred to as “input pattern”) fetched from the pattern fetching means.
A font creation device and a storage medium for font creation, wherein parameters of roundness, aspect ratio, inclination, rotation, and trapezoidal deformation are extracted as individuality parameters. 4. The standard font according to claim 1, wherein the standard font is composed of a set of segments represented by coordinate points, and the individuality reflection point extracting means converts the segment of the standard font into a clouded point / semi-voiced point, an L-shaped shape. The personality reflection unit receives the parameters of the position of the stagnation point / semi-voiced point, the size of the loop, and the deformation of the roundness as individuality parameters, and determines the corresponding segment according to these. A font creation apparatus and a font creation storage medium, which are deformed and receive parameters of aspect ratio, inclination, rotation, and trapezoidal deformation as individuality parameters and convert entire coordinate points. 5. The font creating apparatus according to claim 1, wherein said individuality extracting means uses a predetermined value for the individuality parameter when the individuality extracting means cannot extract the individuality parameter. Storage medium. 6. A handwritten character personality extracting means and a handwritten character personality extracting storage medium, comprising: a pattern capturing means for capturing a character pattern; and a personality extracting means for extracting a personality parameter of the handwritten character. 7. The segment associating means according to claim 6, wherein said individuality extracting means performs a corresponding correspondence between a standard pattern, a segment of said input pattern and a segment of said standard pattern. And a personality extraction unit for handwritten characters, comprising: a difference calculation unit that uses a difference obtained by comparing a segment of the input pattern and a segment of the standard pattern associated by the association unit as an individuality parameter. A storage medium for extracting personality of handwritten characters. 8. A standard font, a personality reflecting portion extracting means for extracting a portion to be transformed from the standard font to reflect the personality, and a personality reflecting portion for transforming the personality reflecting portion of the standard font according to the personality parameter. A font creation device and a font creation storage medium characterized by the above-mentioned features. 9. A personal font service system in which a client terminal and a personal font server for providing a personal font service are connected on a network, a handwritten character is read from the client terminal, and a personality parameter of the handwritten character is extracted and serviced. Personality parameter extraction service means; personal font generation program service means for providing a program for generating a personal font from a standard font which the system has as standard and the personality parameter dictionary extracted by said personality parameter extraction service means; And a network management means for managing the service on a network in response to the request of (1). 10. A personal font service system in which a client terminal and a personal font server for providing a personal font service and a print server for providing a print service are connected on a network, wherein a handwritten character is read from the client terminal and the personality of the handwritten character is read. Personality parameter extraction service means for extracting and servicing parameters, and personal font generation for providing a program for generating a personal font from a standard font which the system has as standard and the personality parameter dictionary extracted by the personality parameter extraction service means Personal phone comprising program service means and network management means for managing the service on a network in response to a request from a client Client server, and input the print document and personal ID from the client terminal, and receive the personality parameter dictionary and personal font generation program services from the personal font server,
Further, a personal font service system is connected to a print server which receives a print request from a client, converts a print code document by a personal font generation program, and outputs the converted print code document to a printer. 11. A personal font service system to which a personal font server for providing a personal font service with a client terminal on a network, a mail server for transferring mail, and a greeting server for providing a greeting card service, are connected with a handwritten character. Is read from the client terminal, the personality parameter extraction service means for extracting and providing the personality parameters of the handwritten characters, and a personal font based on a standard font which the system has as standard and the personality parameter dictionary extracted by the personality parameter extraction service means. A personal font generation program service means for providing a font generation program; and a network for managing the service on a network in response to a request from a client. A personal font server composed of network management means, a print document, a personal ID, and a mailing address are input from a client terminal, and the personal font server receives services of a personality parameter dictionary and a personal font generation program from the client terminal. A personal font service system, wherein the personal font service system is connected to a greeting server which receives a transfer request from the server and converts the greeting code document with a personal font generating program and transfers the converted document to a mail server. 12. A personal font service system in which a client terminal and a personal font server for providing a personal font service and a facsimile server for performing a facsimile transfer service are connected on a network. Personality parameter extraction service means for extracting and providing a personality parameter of the personal computer; and a personal service providing a program for generating a personal font from a standard font which the system has as standard and the personality parameter dictionary extracted by the personality parameter extraction service means. A personal font comprising a font generation program service means and a network management means for managing the service on a network in response to a request from a client. Client server, a print document, a personal ID, and a telephone number are input from the client terminal, the personal font server receives the service of the personality parameter dictionary and the personal font generation program, and further receives the transfer request from the client and faxes. A personal font service system, wherein a facsimile server for converting a code document by a personal font generation program and performing facsimile transfer is connected.