JP2000347648A - Character face composing and outputting device, and program storage medium for the device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate and output a font of a new face even when data constitutions of original outline fonts are different, by compounding frequency fonts provided by conducting orthogonal transformation of outline fonts corresponding to plural faces. SOLUTION: A CPU 1 transforms outline fonts corresponding to preset plural character faces into frequency fonts of an agrregate of discete frequency components by Fourier transformation respectively. The CPU 1 compounds the respective fellow frequency components of the frequency fonts at a prescribed allocation ratio using as compounding objects the frequency fonts corresponding to the plural faces, and a face font newly generated thereby is inverse- Fourier-transformed to be developmentally output.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character typeface synthesizing output device for synthesizing a plurality of character typefaces to generate and output a new typeface, and a program recording medium therefor.

[0002]

2. Description of the Related Art Conventionally, in a document data processing apparatus, when arranging a character typeface such as a Mincho style or a Gothic style,
There is a technique for making a subtle change by replacing a part of the character with a desired part. Also, a technique for synthesizing a plurality of fonts and generating a new font by completely unifying the data structure of a plurality of outline fonts having different fonts (Japanese Patent Application No. 5-354263, title of invention: character pattern generation Device) is present.

[0003]

However, in the case of the method of exchanging parts, it is necessary to prepare a large number of parts to be exchanged, and there is a problem that the degree of change is small and it takes time. In the case of a method that completely unifies outline data, for example,
There is a problem that the "seriph" at the end of the line existing in the Mincho style must be held in terms of data even though it does not exist in the Gothic style, resulting in an increase in capacity. By the way, the applicant of the present invention has a frequency font generation function of generating a frequency font, which represents the outline of a character by a set of discrete spatial frequency components, using a normal character font, and a character font based on the frequency font. By simply adding a deformed font generation function that generates a deformed font that deforms the outline of a character, you can significantly change the basic outline and thickness of characters, add various expressions, etc. while suppressing the increase in data volume (Japanese Patent Application No. 10-221119, title of the invention: character font deformation output device and its program recording medium) have been proposed. An object of the first invention is to develop a technique previously proposed by the present applicant, and to synthesize a frequency font obtained by orthogonally transforming an outline font corresponding to a plurality of fonts, thereby obtaining an original font. An object of the present invention is to enable a font of a new typeface to be generated and output even if the data configuration of the outline font is different. A second object of the present invention is to generate a variety of new fonts from a single outline font by performing different types of filtering on a frequency font obtained by orthogonally transforming a single outline font. It is to be able to output. A third object of the present invention is to combine a plurality of filter setting values themselves when filtering a frequency font obtained by orthogonally transforming a single outline font, thereby forming a single outline font. Is to be able to generate and output a variety of new fonts.

[0004]

The means of the present invention are as follows. The invention according to claim 1 (first invention) converts an outline font corresponding to a plurality of character fonts specified in advance into frequency fonts represented by a set of discrete spatial frequency components by orthogonal transformation. Font converting means, a font synthesizing means for synthesizing frequency components corresponding to a plurality of typefaces converted by the font converting means, and synthesizing the respective frequency components at a predetermined distribution ratio, and a font synthesizing means. Output means for orthogonally inverting and converting and outputting the new font of the typeface. When the font synthesis means synthesizes frequency fonts corresponding to a plurality of character fonts, if the number of frequency components constituting each frequency font does not match, the number of the frequency components is smaller so that the numbers match. The synthesis process may be performed after complementing the zero data. Further, the font synthesizing means traces the outline of the character constituting the outline font and samples in a predetermined direction, and when converting it into a frequency font by orthogonal transformation, obtains the difference between the lowest frequency components of a plurality of typefaces. The optimum sampling start point may be detected by shifting the sampling start point in the sampling direction until the difference falls within a predetermined allowable range. According to the first aspect of the present invention, the outline font corresponding to a plurality of predetermined character fonts is converted into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, and the converted outline font is converted. The frequency fonts corresponding to a plurality of typefaces are synthesized and their frequency components are synthesized at a predetermined distribution ratio. The generated font of the new typeface is orthogonally inverse-transformed and output. Therefore, by synthesizing frequency fonts obtained by orthogonally transforming outline fonts corresponding to multiple fonts, it is possible to generate and output fonts of new fonts even if the data configuration of the original outline fonts is different. Can be.

According to a fourth aspect of the present invention, there is provided a font conversion means for converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, A font generating means for generating a plurality of typeface frequency fonts by respectively applying different types of filter processing to the frequency fonts generated by the font converting means; and a plurality of typeface frequency fonts generated by the font generating means. It is provided with a font synthesizing means for synthesizing a font of a new typeface and an output means for developing and outputting a font of the new typeface generated by the font synthesizing means. In addition, there is provided inverse conversion means for converting the frequency fonts of the plurality of fonts into coordinate sequence data by performing orthogonal inverse conversion on the frequency fonts of the plurality of fonts generated by the font generation means, and the font synthesis means comprises A font of a new font may be generated by synthesizing coordinate sequence data of fonts corresponding to a plurality of fonts converted by the conversion unit at a predetermined distribution ratio. According to a fourth aspect of the present invention, a single outline font is converted into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, and different types of filter processing are performed on the frequency font. As a result, a frequency font of a plurality of fonts is generated, and a new font is synthesized based on the frequency fonts of the plurality of fonts generated and output.
Therefore, by applying different types of filter processing to the frequency font obtained by orthogonally transforming a single outline font, it is possible to newly generate and output various fonts from the single outline font.

The invention according to claim 6 (third invention) is a font conversion means for converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, When performing a filtering process on the frequency font converted by the font converting unit, a plurality of filter setting values are synthesized at a predetermined distribution ratio, and then the filtering process is performed to generate a new typeface font. Means and an output means for orthogonally transforming the font of the new typeface obtained by the synthesizing means and developing and outputting the result. Claim 6
The described invention converts a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, and performs a filtering process on the frequency font. A font of a new typeface is generated by synthesizing the filter set values at a predetermined distribution rate and then performing a filtering process, and the new typeface is orthogonally inverse-transformed and developed and output. Therefore, when filtering a frequency font obtained by orthogonally transforming a single outline font, a variety of fonts are newly created from a single outline font by synthesizing multiple filter setting values themselves. Can be generated and output.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) A first embodiment of the present invention will be described below with reference to FIGS. FIG. 1A is a block diagram showing the entire configuration of the document data processing device. The CPU 1 is a central processing unit that controls the overall operation of the document data processing device according to various programs. The storage device 2 has a storage medium 3 in which an operating system, various application programs, data files, character font data and the like are stored in advance, and a drive system thereof. The storage medium 3 is fixedly provided or removably mountable, and includes a magnetic / optical storage medium such as a floppy (registered trademark) disk, a hard disk, an optical disk, and a RAM card, and a semiconductor memory. Have been. Also, programs and data in the storage medium 3 may be stored in a CP
The data is loaded into the RAM 4 under the control of U1. Further, C
The PU 1 receives a program and data transmitted from another device via a communication line or the like and stores the program and data in the storage medium 3 or a program stored in a storage medium provided in the other device. And data can be used via a communication line or the like. An input device 5, a display device 6, and a printing device 7, which are input / output peripheral devices, are connected to the CPU 1 via a bus line. The CPU 1 controls these operations according to an input / output program.

The input device 5 has a keyboard, a pointing device such as a mouse, and an image reader for inputting character string data and the like and various commands.
Here, when the document data is input from the input device 5 at the time of document creation, the document data is displayed on the text screen of the display device 6 and the determined character string determined by the kana-kanji conversion is stored in the RAM 4. Note that the display device 6 is a liquid crystal display device, a CRT display device, a plasma display device, or the like that performs multicolor display, and the printing device 7 is a full-color printer device, which is a non-impact printer or an impact printer such as thermal transfer or inkjet. Output document data in color.

Here, when printing a document, if it is instructed to print the document with handwritten characters such as written with a brush, the CPU 1 uses an orthogonal transform, that is, a discrete Fourier transform (DFT) expansion system. Therefore, the outline of the existing character font (outline font) is converted into a data structure represented by a set of discrete spatial frequency components. Here, each part constituting a character can be represented by a continuous function having a cycle of one round of the contour (closed curve).
The basic frequency of the FT is converted into a data structure of an aggregate in which the basic frequency and a frequency that is an integral multiple of the basic frequency are combined. Note that the maximum order of the conversion result obtained by the DFT processing corresponds to the sampling theorem.

FIG. 1B shows the main structure of the RAM 4, and various memory areas are allocated to the RAM 4. The character string storage unit 4-1 temporarily stores a character string to be subjected to the DFT processing. The CPU 1 reads out each character code constituting the character string from the head thereof and reads the character code from the normal font storage unit in the storage device 2. 2-1 is searched, and the corresponding character font is fetched. Normal font storage 2
-1 stores an outline font in association with various characters.
The DFT process is performed on the character font read out from No. 1 for each part. At this time, the frequency conversion work unit 4-2 is used. The frequency conversion work unit 4-2 is a work area for temporarily storing a conversion result for each part, and the CPU 1 deforms the contour data of one part in the frequency conversion work unit 4-2 that has been subjected to the DFT processing. It is stored in the font storage unit 4-3. The deformed font storage unit 4-3 temporarily stores a newly generated synthesized font by synthesizing a plurality of typefaces, and the CPU 1 performs an inverse Fourier transform of the outline data constituting the synthesized font to obtain the original coordinate sequence. After the data is generated, the coordinate sequence data is developed in the print buffer 4-4.

FIG. 2 is a diagram conceptually showing a state in which an existing font is converted into a set of frequency components for each part by DFT conversion. One part of the font to be processed is drawn, and the outline of this bitmap is plotted. Here, the number of points for plotting one part is determined depending on how finely the conversion is desired. In the case of the discrete Fourier transform, the number is selected from a multiplier of “2”, for example, “512”. The contours of the parts are plotted, XY coordinate values are fetched for each of the points P0, P1, P2,..., And the coordinate values are subjected to DFT processing. Note that the transform result by DFT is reversible unless the calculation accuracy is strictly determined, and can be returned to the original coordinate value by Fourier transform. In the case of a normal outline font, the number of parts is defined and attributes such as a reference point coordinate value, a straight line, and a Bezier curve are defined for each part (each image). In the case of the frequency font in the embodiment, the coordinates of the outline font and the description part of the attribute are a table of the number of spatial frequency components. In other words, the DFT processing result for each part in association with the character code, the number of parts, and the part number is a numerical table describing a DC component, a fundamental frequency component, and a frequency component obtained by multiplying the DC component, the fundamental frequency component, and the like. Here, the processing result by the DFT is expressed as a complex number, and its size is discretely obtained for a real axis part and an imaginary axis part on a complex plane. In the figure, "Xrl" is a complex plane. Above, the value of the real axis portion in the X-axis direction is shown, and “Xim” indicates the value of the imaginary axis portion. Similarly, “Yrl” and “Yim” are the value of the real axis part and the value of the imaginary axis part in the Y-axis direction. In the figure, “F0” is a DC component, and the center part coordinates of the contour are described in the real part thereof. Note that a value of “zero” is always described in the imaginary axis part of “F0”.
“F1” is a fundamental frequency component representing a frequency that vibrates once when making a round of the contour of the part, and the size and direction of the part are governed by this value. “F2” is a frequency component (secondary high-frequency component) that vibrates twice when making one round of the contour of the part, and the shape of the part is governed by this value, and this value increases in a part that is greatly bent. The subsequent "F3", "F4",... Are third- and fourth-order high-frequency components, and the maximum order thereof is half the number of coordinates when sampling the contour of the conversion target part. This is because the value of the second half can be easily derived only by performing the inversion of the positive / negative and the reverse order from the value of the first half.

FIG. 3 shows a state in which the character "sum" is decomposed for each part and subjected to DFT processing, and the vertically long parts forming the "nogi bias" and the "fresh" are formed. A table is shown in which values of a real part and values of an imaginary part obtained by performing DFT processing for each part are discretely distributed on a frequency axis by paying attention to a largely bent part. In this frequency distribution, the characteristics of the vertically long parts are exhibited at the fundamental frequency F1, and the characteristics of the parts which are greatly bent are exhibited at the next frequency F2 in addition to the fundamental frequency F1. FIG. 4 shows specific numerical values of a real axis portion, an imaginary axis portion, a real axis portion, and an imaginary axis portion in the X-axis direction obtained by performing the DFT processing on the character "sum". FIG. 8 is a diagram showing a correspondence relationship between each part constituting the “sum” and frequency data of each part. Here, as shown by "1" to "7" in the figure, this character "Sum" is composed of seven parts, and a set of spatial frequency components corresponding thereto is data of seven parts.

By using such a frequency font, the outline of a character can be freely changed. FIG. 5A shows a data structure in which all high frequency components of the spatial frequency components obtained by the DFT processing are cut by a digital filter (low-pass filter) to make all the frequency components after F3 zero.
A data structure in a case where a specific high-frequency component, for example, F56 is added and its value is set to a constant value to perform deformation by finely oscillating the contour line is shown. FIG. 5B shows the deformed state, in which the approximate shape of the character is dominated by the low-frequency components of F0, F1, and F2, and its outline is finely vibrated by the high-frequency component of F56. ,
The number and intensity of the vibrations can be arbitrarily controlled. To obtain a large vibration, the value may be increased, and to obtain a fine vibration, the order of the high frequency component may be increased.

In this embodiment, outline fonts corresponding to a plurality of character fonts arbitrarily designated in advance are respectively converted into frequency fonts by DFT conversion, and frequency fonts corresponding to the plurality of fonts are set as synthesis targets. Are synthesized at a predetermined distribution ratio, and the font of the newly generated typeface is inversely transformed by the DFT and developed and output. Figure 6 shows the outline font data structure and D
It shows the relationship with the frequency font data structure after FT conversion, where (A) shows before conversion and (B) shows after conversion. Here, the coordinate data of the outline font is represented by two components, an X component and a Y component.
One stroke of the outline is sampled for each stroke constituting a character. In this case, the outline outline data for one stroke is 1
Even if it is expressed by a Bezier curve of a book or by 100 straight lines, it is still the same as the coordinate sequence data, so that the DFT conversion can be performed irrespective of the data structure of the outline font. However, when synthesizing a plurality of fonts as a condition for DFT conversion from an outline font, the same image of each font is sequentially specified and processed, so the number of strokes of the multiple fonts must match. Become. Further, it is required that the starting point of the sampling (for example, outline data) and the sampling direction are also unified. Then, when an outline font of a plurality of fonts is DFT-converted under such conditions, the conversion result is a value on the real axis and a value on the imaginary axis in the X and Y directions. , Ie, line symmetric (real part) and point symmetric (imaginary part) data. Also, the number of samples is determined by how finely the coordinates are sampled, and when one image is large, the contour is long and the number of samples is large, and when it is small, the number of samples is small. By the way, as shown in FIG. 5, even if all the data after a certain frequency component are set to zero as shown in FIG. 5, the shape of the frequency font is not rounded but the whole shape is collapsed. You can keep the shape. Therefore, even if the number of data of the frequency fonts of the multiple fonts to be synthesized is different, if the data numbers of both fonts are matched by adding zero, it becomes possible to synthesize the fonts. DFT conversion is possible without being affected. For example, the frequency font of the first typeface to be synthesized has 8 samples,
Assuming that the number of samples of the frequency font of the second font is 16, it is sufficient to add eight zeros after the frequency font of the first font to match the number of samples of the second font. By averaging the values of the respective components of the frequency font of each font in which the number of samples (the number of data) is matched in this manner, a frequency font of a new font obtained by combining two fonts is obtained.

Next, the operation of the document data processing apparatus will be described with reference to FIGS.
This will be described with reference to the flowchart shown in FIG. Here, a program for realizing each function described in these flowcharts is stored in the storage medium 3 in the form of a program code readable by the CPU 1, and the CPU 1 performs an operation according to the program code. Execute sequentially. This is the same in other embodiments described later. FIG. 7 is a flowchart showing the overall operation when printing a character string. That is, the normal character font (outline font) is changed to 1
In addition to converting to a frequency font for each part, a transformed font is generated by processing the data that constitutes the frequency font, and the transformed font is returned to the original coordinate data by inverse Fourier transform. 3 is a flowchart showing an operation of performing character expansion and printing out. First, after the line pointer is initialized (step A1), the character string storage unit 4-1 is accessed with this pointer value to read the character string code for one line (step A2). Next, after the digit pointer is initialized (step A3), the 1 designated by the digit pointer
The character codes for the digits are read (step A4). Then, one-character printing processing described later is performed (step A5),
When printing of one character is completed, the value of the digit pointer is updated and the next character is designated (step A6). Thereafter, the process returns to step A5 until the end of printing of one line is detected in step A7, and printing of one character is repeated. When printing of one line is completed, the value of the line pointer is updated (step A8), and the process returns to step A2 until the end of all lines is detected in step A9, and the above operation is repeated.

FIG. 8 is a flowchart detailing step A5 (one-character printing process) in FIG. First, the character 1
An image pointer is initialized to designate an image (step B1). Then, data of one stroke indicated by the stroke pointer is read from the first font outline font from the normal font storage unit 2-1 out of the plurality of fonts designated in advance as a combination target (step B2), and is read by the DFT.
Conversion (step B3). That is, the number of parts (the number of strokes) constituting the font to be processed is acquired, and sampling is performed by tracing the contour line of the part, and the coordinate value of each sampled point is subjected to DFT transformation. Similarly the second
Is obtained and DFT-transformed (steps B4 and B5). Then, a synthesizing process for synthesizing the DFT conversion results of the first and second fonts is performed (step B6).

FIG. 9 is a flow chart showing the synthesizing process. Each frequency component of a plurality of fonts is sequentially designated by the value "n" of the F pointer, and if there is a frequency component corresponding to each font, the frequency The components are combined at a predetermined distribution ratio, but if one of the fonts reaches the upper limit quickly and there is no data thereafter, zero data is forcibly added. That is, in step C1,
Initialize the F pointer. Then, check whether the frequency component indicated by the value "n" of the F pointer exceeds the upper limit frequency component S1nmax of the first typeface (step C).
2) Since the frequency component F0 of the first font is specified at first, the process proceeds to step C4, and the value "n" of the F pointer is similarly set to the upper limit frequency component S2nmax of the second font for the second font. Find out if you have crossed. Also in this case, since the frequency component F0 of the second typeface is specified, it is detected that the frequency component is less than the upper limit. Then, the next step C
From 5 to C8, the frequency components of the first font and the second font designated by the value "n" of the F pointer are combined at a predetermined distribution ratio.

That is, the first value indicated by the value of the F pointer
And in the frequency component of the second font, the distribution ratio P (0 ≦ P ≦) is assigned to the real part S1FnXr in the X direction of the first font.
1) multiplied by the X-direction real part S2F of the second typeface
By adding nXr to a value obtained by multiplying (1-P), a real part FnXr in the X direction of the frequency component of the synthesized font is obtained (step C5). Similarly, for the imaginary part in the X direction, FnXi = S1FnXi *
The imaginary part FnX of the frequency component in the X direction in the synthesized font is calculated by the formula of P + S2FnXr * (1-P).
i is obtained (step C6). Further, similarly in the Y direction, FnYr = S1FnYr * P + S2FnYi *
While executing the calculation of (1-P) (step C7),
FnYi = S1FnYi * P + S2FnYi * (1-
The calculation of P) is executed (step C8). Then, the value of the F pointer is advanced by one and updated (step C).
9) It is checked whether the same processing as described above has been executed for all frequency components (step C10). That is, the first and second
The process returns to step C2 until the processing is completed for all the frequency components having the higher upper limit of the typeface. Here, when it is detected that the value of the F pointer has exceeded the upper limit frequency component of one of the first and second fonts, step C3 or step C4 is executed. Here, when it is detected that the value of the F1 register exceeds the upper limit frequency component of the first typeface, every time it is detected, the nth new frequency component indicated by the F register in the first typeface is detected. As "0"
Is added (step C3). Similarly, for the second typeface, when it is detected that the frequency component exceeds the upper limit frequency component of the second typeface, the first type is detected each time it is detected.
Is added as the nth new frequency component indicated by the F register to the typeface (step C).
4). The newly added "0" data and the frequency components of other fonts are combined at a predetermined distribution ratio (steps C5 to C8). Hereinafter, the same operation as described above is repeated until the end of all frequency components.

After the completion of the synthesizing process, the process proceeds to step B7 in FIG. 8, and the Fourier inverse transform is performed on the font of one part of the new typeface generated by the synthesizing process. In this case, if the contour data of one part is inversely transformed by DFT, the coordinates return to the coordinates at the time of plotting, so that an outline font using this as a short vector is obtained. Then, a rasterizing process for painting the inside of the outline is performed on the print buffer 4-4 (step B8). Then, the next image is designated (step B9), and it is checked whether the entire image for one character is completed.
Then, the data of the first font and the second font are read out and generated for each stroke, and are developed and output. When printing of one character is completed, the printing is performed for all characters and all lines according to the flowchart of FIG.

As described above, in the first embodiment, outline fonts corresponding to a plurality of character fonts designated in advance are respectively converted into frequency fonts which are represented by a set of discrete spatial frequency components by orthogonal transformation. In addition to the conversion, the frequency fonts corresponding to the plurality of converted fonts are combined and the respective frequency components are combined at a predetermined distribution ratio, and the font of the newly generated font is orthogonally transformed and expanded. Since the data is output, a font having a new format can be generated and output even if the data configuration of the original outline font is different. In this case, if the number of frequency font data corresponding to a plurality of typefaces to be combined does not match, zero is added to the smaller number so that the number of data matches, and then the combining process is performed. As a result, it is possible to combine fonts having greatly different character shapes and sizes.
As described above, merely installing at least two types of outline fonts makes it possible to generate a variety of fonts and to obtain a low-cost and expressive font. FIG. 10 shows a printing example in this case. In this example, the pop font and the brush are combined, but by changing the distribution ratio at the time of the combination, FIG.
A variety of fonts as shown in (A) and (B) can be obtained, and synthesis of such a typeface in which the shape and size of one stroke are completely different cannot be realized by conventional techniques.
It can be easily realized.

In the first embodiment described above,
This is based on the condition that the sampling start point at the time of DFT conversion is identical between the fonts to be combined. After the outline data of the first typeface is DFT-transformed,
When the outline data of the typeface is subjected to DRT conversion, the sampling start point may be searched for the optimum point by trial and error. FIG. 11 shows the one-character printing process in this case. Steps D1 to D5 are the same processes corresponding to steps B1 to B5 in FIG. 8 described above, and after the first stroke of the first font has been DFT-converted. DF for the first stroke of the second typeface
At the time of T-conversion, it is determined in step D6 whether the difference between the lowest frequency components of the two fonts is within a predetermined allowable range.

FIG. 12 is a flowchart showing this allowable range determination processing. The X-axis real number part S2F1Xr of the lowest frequency component of the second font is subtracted from the X-direction real number part S1F1Xr of the lowest frequency component of the first font.
A raised value Xrd (positive number) is obtained (step E1).
Similarly, for the imaginary part in the X direction, Xid = (S1
F1Xi-S2F1Xi) By the calculation of ² , the difference X
The id is obtained (step E2). Further, similarly in the Y direction, Yrd = (S1F1Yr-S2F1Yr) ²
, The difference Yrd is obtained (step E
3), Yid = (S1F1Yi-S2F1Yi) ² is calculated to find the difference Yid (step E4).
Each value rXd, Xid, Yrd, Y thus obtained
id are all added, and the cumulative difference d is obtained (step E).
5) The accumulated difference d is compared with a predetermined judgment value (step E6). As a result, if d <determination value, it is determined to be within the allowable range, but if d ≧ determination value, it is determined to be outside the allowable range.

If it is determined that the value is outside the allowable range, the process proceeds to step D7 in FIG. 11, and the sampling start point of the second font is shifted by one in the sampling direction. Then, the process returns to step D5, and the second font is again subjected to DFT conversion using this point as a sampling start point. Thereafter, trial and error are performed while shifting the sampling start point until it is detected that the second font is within the allowable range. As a result, when it is detected that the sampling is within the allowable range, the sampling start point of the second font is determined to be optimal for the first font, and the fonts are combined with each other (step D8). .
Thereafter, similarly to Steps B7 to B9 in FIG. 8, the combined result is subjected to the inverse DFT transform and developed and output (Steps D9 to D9).
11). Then, the process returns to step D2 until the end of all the images is detected in step D12, and the above operation is repeated.

The method for finding the sampling start point of the second typeface is not limited to the method shown in FIG.
When performing DFT conversion on the second typeface, a large number of DFT conversions are performed while shifting the sampling start point in advance, and the difference between the first and second lowest frequency components is obtained. The sampling start point of the two fonts can also be used. By doing so, it is possible to automatically match the sampling start points, and as a result, it is possible to increase the degree of freedom in the data configuration of the original outline font. (Second Embodiment) Hereinafter, a second embodiment of the present invention will be described with reference to FIGS. In the first embodiment, outline fonts of two types of fonts arbitrarily designated in advance are FDT-converted and synthesized. In the second embodiment, however, a single outline font is converted to a DFT. By applying different types of filter processing to the converted frequency font, a plurality of typeface frequency fonts are generated.
That is, a new font can be generated by operating each component by the filtering process as shown in FIG. 5, so that in the second embodiment, a single outline font is obtained by DRT conversion. Frequency filter with two types of filtering
And the frequency font of the second font.

FIG. 13 is a flowchart showing a one-character printing process according to the second embodiment. Step F1
Steps F3 to F3 are the same as steps B1 to B3 in FIG. 8 described above. In this case, however, the data of one stroke of the arbitrarily designated single font outline font is read and subjected to DFT conversion. Then, the conversion process is stored in the memory, and the filtering process is performed (steps F4 and F5). in this case,
The filter processing includes a low-pass filter that cuts a certain frequency component or more, an addition filter that adds a specific frequency, and the like, and deforms the outline and size of the character without breaking the basic character shape. The first typeface thus transformed is subjected to DFT inverse transformation, and the result of this inverse transformation is stored in a memory (steps F6 and F7). Next, the result of the DFT conversion stored in step F4 is read out and another type of filter processing is performed (step F8) to obtain a second typeface, which is subjected to inverse DFT conversion (step F8).
9).

After the first typeface and the second typeface are obtained by the two types of filter processing and inversely converted as described above, in the next step F10, the inverse conversion result of the first typeface and the inverse conversion result of the second typeface are obtained. Is synthesized. That is, in the above-described first embodiment, the frequency fonts subjected to the DFT conversion are combined, but in the second embodiment, the result of the inverse DFT conversion is combined in order to increase the processing efficiency. Here, the first and second typefaces obtained by the two types of filter processing have the same situation when a single outline font is sampled, so that the sampling number and the sampling start point are the same. Further, since the data configuration does not change even after performing the DFT inverse transform, it is possible to combine the data after the inverse transform.

FIG. 14 is a flowchart showing a synthesizing process for synthesizing the inverse conversion results of the first font and the second font.
Since the synthesis in this case is synthesis of coordinate data, it is only necessary to allocate the X coordinate and the Y coordinate at the allocation ratio P (1 ≦ P ≦ 0). That is, it is only necessary to find the subdivision point between the coordinate points. That is, the point pointer is initialized (step G).
1), the X coordinate S1Xn of the first font represented by this value is P
And the value obtained by multiplying the X coordinate S2Xn of the second typeface by (1-P)
Is added to the value obtained by multiplying by X to obtain the internal division point X of the X coordinate.
n is obtained (step G2). Similarly, for the Y coordinate, the subdivision point Yn of the Y coordinate is obtained by calculating Yn = S1Yn * P + S2Yn * (1-P) (step G3). Then, the point pointer is advanced by one to update the value (step G4), and thereafter, the process returns from step G5 to step G2 until all the sampling numbers have been completely processed, and the above operation is performed for each point. When the combining process is completed, the process proceeds to step F11 in FIG. 13, where the rasterizing process is performed and the next image is designated (step F12). Then, the process returns to step F2 until the end of all the images, and the above operation is repeated.

As described above, the second embodiment has the same effects as the first embodiment described above, and has only one outline font as an original.
There is no misalignment due to the composition, and high-quality composition can be performed. FIGS. 15A to 15E illustrate synthetic fonts in this case, and the fonts are different in thickness and length, so that a highly expressive typeface can be obtained. In this case, the first typeface and the second typeface are inversely converted and then synthesized, so that a beautiful finish can be obtained by a simple operation. In other words, when two typefaces having many straight lines are combined, the frequency font is likely to fluctuate due to a calculation error, but the result after DFT inverse transformation having a lot of fluctuates is a cleaner finish, and the calculation is simpler. (Third Embodiment) Hereinafter, a third embodiment of the present invention will be described with reference to FIGS. In the above-described second embodiment, the first filter obtained by the filtering process is used.
In the third embodiment, the first and second fonts are not generated, but a single outline font is subjected to DFT conversion and filter processing. The filter set values (parameters) are combined at a predetermined distribution ratio. That is, FIG. 16 shows a one-character printing process in this case. After the image pointer is initialized, data of one stroke is read from the outline font and DFT-converted (steps H1 to H).
3). Then, a predetermined filter process is performed on the conversion result (step H4), a font of a new typeface is generated, and the DFT is inversely converted and rasterized (steps H5 and H6). When the printing process for one image is completed, the next image is designated (step H).
7), the process returns to step H2 until the end of all images is detected in step H8, and the above operation is repeated.

FIG. 17 is a flowchart showing step H4 in FIG.
9 is a flowchart illustrating (filter processing). This filtering process performs a predetermined filtering process on the result of the DFT transformation. For example, in addition to a low-pass filter that cuts a certain frequency component or more, a partial addition filter that adds a specific frequency component, A wide variety of transformations are performed using a total integration or partial integration filter that integrates values into real and imaginary values without breaking the basic character shape of the character. For example, in the case of a low-pass filter, a high-frequency component in a set of frequency components is deleted, and a contour ta having a roundness consisting of only low-frequency components that govern the general shape of a character is generated. Further, in the case of the total integration filter, the value of the real axis part and the value of the imaginary axis part of each frequency component in the X and Y directions are multiplied by a constant.
Generates outline data in which the thickness and length of characters are changed. In the case of the partial integration filter, the imaginary axis portion of a predetermined frequency component such as the lowest frequency component F1 in the set of frequency components is left as it is, and the real axis portion is multiplied by a constant to change the momentum of the character. Generate a contour. In the case of the partial addition filter, by performing an addition process of changing a value of a predetermined frequency in a set of frequency components, a contour of a character is vibrated or a contour with a partially changed thickness is generated.

First, after performing a parameter synthesizing process for changing a filter set value (step J1), a low-pass filtering process (step J2), an overall integrating filtering process (step J3), and a partial integrating filtering process (step J1).
4), a partial addition filter process (step J5) is executed, and these filter processes are to deform the outline shape of the character without breaking the shape of the character as described above. The arbitrarily designated filter processing is executed. Here, the parameter synthesizing process is shown in FIG.
8 is executed according to the flowchart shown in FIG. First, a set value of a filter set in advance as a processing target is obtained and copied and stored (step K).
1). Then, a filter process for synthesizing the set value for each filter type is performed (steps K2 to K5). In this case, a plurality of filter set values preset for each type of filter are distributed at a predetermined distribution ratio, thereby obtaining these internally dividing points, and setting these as new filter set values.

As described above, in the third embodiment, when performing a filtering process on a frequency font obtained by performing DFT conversion on a single outline font, a plurality of filter setting values are allocated in a predetermined distribution. Since a font of a new typeface is generated by performing a filtering process after synthesizing at a rate, and the orthogonally-inverted transform is performed to develop and output the font, the third embodiment also has the above-described first embodiment. In addition to having the same effect as, only the filter parameters need to be synthesized, the amount of calculation is minimized,
The processing speed is the fastest than the other embodiments described above, and high-speed character synthesis is possible.

[0032]

According to the first aspect of the invention, a frequency font obtained by orthogonally transforming outline fonts corresponding to a plurality of typefaces is synthesized, so that even if the data configuration of the original outline font is different. , A new typeface font can be generated and output. According to the second invention, a variety of fonts are newly generated from a single outline font by performing different types of filter processing on a frequency font obtained by orthogonally transforming a single outline font. Can be output. According to the third aspect, when filtering is performed on a frequency font obtained by orthogonally transforming a single outline font, a plurality of filter setting values themselves are combined to form a single outline font. Can generate and output various fonts.

[Brief description of the drawings]

FIG. 1A is a block diagram showing an overall configuration of a document data processing device, and FIG. 1B is a diagram showing a main part of a RAM 4.

FIG. 2 is a diagram illustrating a state in which a frequency font is generated by sampling a normal character font for each part and performing DFT processing.

[Fig. 3] Decomposes a normal character font into parts for each DF
FIG. 7 is a diagram illustrating a state of T processing and a frequency distribution diagram in which values of a real part and an imaginary part obtained by performing DFT processing are discretely distributed on a frequency axis.

FIG. 4 is a diagram showing specific numerical values forming a frequency font and showing a correspondence relationship between which part corresponds to which data.

5A is a diagram for explaining a data structure of a frequency font, and FIG. 5B is a diagram showing an output state of a character represented by the frequency font having such a data structure.

6A and 6B are diagrams showing a relationship between a data structure of an outline font before DFT conversion and a data structure of a frequency font after DFT conversion, wherein FIG.

FIG. 7 is a flowchart illustrating an overall operation when a character string is printed out.

FIG. 8 is a flowchart detailing step A5 (one-character printing process) in FIG. 7;

FIG. 9 is a flowchart detailing step B6 (synthesis processing) in FIG. 8;

FIGS. 10A and 10B are diagrams showing a printing example of the first embodiment.

FIG. 11 is a flowchart for explaining a modified application example of the one-character printing process in the first embodiment.

12 is a step D6 of FIG. 11 (permissible range determination processing);
5 is a flowchart detailing FIG.

FIG. 13 is a flowchart illustrating a one-character printing process according to the second embodiment.

FIG. 14 is a flowchart detailing step F10 (synthesis processing) in FIG. 13;

FIGS. 15A and 15B are diagrams illustrating a printing example according to the second embodiment.

FIG. 16 is a flowchart illustrating a one-character printing process according to the third embodiment.

FIG. 17 is a flowchart detailing step H4 (filter processing) in FIG. 16;

FIG. 18 is a flowchart detailing step J1 (parameter combining processing) in FIG. 17;

[Explanation of symbols]

 Reference Signs List 1 CPU 2 Storage device 3 Storage medium 4 RAM 5 Input device 6 Display device 7 Printing device 4-1 Character string storage unit 4-2 Frequency conversion work unit 4-3 Conversion font storage unit 4-4 Print buffer

Claims (9)

[Claims] 1. Font conversion means for converting outline fonts corresponding to a plurality of character fonts designated in advance into frequency fonts represented by a set of discrete spatial frequency components by orthogonal transformation. Means for synthesizing frequency fonts corresponding to a plurality of typefaces converted by the means and synthesizing the respective frequency components at a predetermined distribution ratio; and orthogonalizing a font of a new typeface generated by the font synthesizing means. An output unit for performing inverse conversion and developing and outputting the output data. 2. The frequency synthesizer according to claim 1, wherein when synthesizing the frequency fonts corresponding to the plurality of character fonts, if the number of frequency components constituting each frequency font does not match, the number of the frequency components is adjusted so that the numbers match. 2. The character typeface synthesis output device according to claim 1, wherein the synthesis process is performed after complementing zero data to a smaller number. 3. The font synthesizing means traces the outline of a character constituting an outline font and samples it in a predetermined direction. When converting it into a frequency font by orthogonal transformation, the font synthesizing means calculates the difference between the lowest frequency components of a plurality of typefaces. 2. The character font according to claim 1, wherein an optimum sampling start point is detected by shifting the sampling start point in the sampling direction until the difference falls within a predetermined allowable range. Synthetic output device. 4. A font conversion means for converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation, and a frequency font generated by the font conversion means. Font generating means for generating a plurality of typeface frequency fonts by respectively applying different types of filter processing; and font synthesizing means for synthesizing a new typeface font based on the plurality of typeface frequency fonts generated by the font generating means. Output means for developing and outputting a font of a new typeface generated by the font synthesizing means. 5. A font synthesizing unit, comprising: an inverse transform unit for transforming a frequency font of a plurality of fonts into coordinate sequence data by performing an orthogonal inverse transform on the frequency font of the plurality of fonts generated by the font generating unit; 5. The font according to claim 4, wherein a font of a new font is generated by synthesizing the coordinate sequence data of the fonts corresponding to the plurality of fonts converted by the inverse conversion means at a predetermined distribution rate. Character font synthesis output device. 6. A font converting means for converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transform, and a frequency font converted by the font converting means. Means for synthesizing a plurality of filter setting values at a predetermined distribution ratio when performing filter processing, and then performing filter processing to generate a new typeface font; and a new font obtained by the synthesis means. Output means for orthogonally and inversely transforming the font of the font and developing and outputting the font. 7. A recording medium having a program code read by a computer, wherein an outline font corresponding to a plurality of character fonts specified in advance is represented by a set of discrete spatial frequency components by orthogonal transformation. A function to convert to each frequency font, a function to synthesize frequency fonts corresponding to a plurality of converted fonts at a specified distribution rate, and a function to orthogonally transform the font of a new typeface A recording medium having a program code for implementing a function of performing inverse conversion and developing and outputting. 8. A recording medium having a program code read by a computer, comprising: a function of converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation; A function to generate a frequency font of a plurality of fonts by applying different types of filter processing to the generated frequency font, a function of synthesizing a font of a new font based on the frequency font of the generated fonts, and A recording medium having a program code for realizing a function of developing and outputting a new typeface font. 9. A recording medium having a program code read by a computer, comprising: a function of converting a single outline font into a frequency font represented by a set of discrete spatial frequency components by orthogonal transformation; A function of generating a new font by performing a filtering process after combining a plurality of filter setting values at a predetermined distribution ratio when performing a filtering process on the converted frequency font; and A recording medium having a program code for realizing a function of orthogonally inversely transforming a font and developing and outputting the font.