JPH09230843A - Character generator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a character generator capable of controlling a line width in a slash part and a curve part in a small size. SOLUTION: This character generator enlarges or reduces a character of an outline font data into an arbitrary size by a size conversion means 13, and identifies a horizontal vertical line part and a diagonal curve part respectively by a horizontal vertical line and diagonal curve part identification means 141, and detects start and end points of the diagonal curve part by a diagonal curve detection means 142. This device corrects a line width of the horizontal vertical line part by a line width correction means 143 of a horizontal vertical line part, and calculates a similarity conversion formula equivalent to this change from a position change in the point about each diagonal curve part after correcting the horizontal vertical line part line width by the similarity conversion formula calculation means 144, and similarly converts each diagonal curve part by a diagonal curve part similarity conversion means 145.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character generator applied to a word processor, a personal computer, a desktop top publishing, a computer typesetting machine, a printer, etc., which displays or prints out high quality characters using an outline font. More particularly, the present invention relates to a character generation device having line width correction of a slanted curve portion, which has been conventionally difficult to correct.

[0002]

2. Description of the Related Art Approximately 60 × using an outline font
When outputting a small character of 60 dots or less, since the coordinate value is finally rounded to an integer, a quantization error occurs, resulting in non-uniform line widths and collapsed line intervals, leading to deterioration in character quality.

As a method for solving this problem, line width control called hinting or hint processing has been conventionally performed. Since each company refrains from disclosing these technologies as know-how, there are relatively few prior art examples, but there are some prior applications described later.

Postscript (Adobe, US) and TrueType (US, Apple), which are known as open font technology,
By analogy with the data format disclosed by Microsoft, which holds information such as the line width or line spacing of each horizontal line and each vertical line in the reference data, the above-mentioned hinting is the line width. It can be seen that the process of thinning or thickening is performed with a constant value.

Next, an example of a prior application relating to the line width control technique will be described. (1) In Japanese Patent Laid-Open No. 4-269564, a pair vector forming a line width is detected, the position of one of the vectors is fixed, and the line width of the pair vector before size conversion is multiplied by the size conversion magnification. Move the other vector. This patent does not mention any diagonal curve. (2) In Japanese Unexamined Patent Publication No. 4-188190, after a size conversion, a cumulative total of a series of real number line widths and real number line intervals for each radical and each, and a cumulative total of a series of integer line widths and integer line intervals after quantization. Are compared, and the difference is corrected depending on whether the difference is an odd number or an even number. However, this publication also deals only with the line width and line spacing of the horizontal and vertical line portions and does not make any description regarding the oblique curve portion. (3) Japanese Patent Laid-Open No. 4-362691 also targets only the line width of the horizontal and vertical portions.

As described above, in the related art, in patents for controlling the line width, there is much description about the control of the line width of the horizontal and vertical portions, but there is little about the line width of the oblique curve portion.

However, the character quality of outline fonts is not sufficient with horizontal and vertical line width control alone. This is because most characters include many oblique curve parts, and it is not possible to achieve high quality of the entire character without controlling the line width of the oblique curve parts.

The inventions touching on the control of the line width of the oblique curve portion are as follows. (4) Japanese Unexamined Patent Publication No. 4-104298 suggests that the normal vector method can be used for controlling the line width of the horizontal and vertical line portions in the embodiment as the line width control of the oblique curve portion. It is not a concrete proposal. (5) In Japanese Unexamined Patent Publication No. 4-93992, for the purpose of thickening and thinning characters, the normal line angles at the control points of the horizontal and vertical line parts and the oblique curve parts forming each contour line are calculated, This is realized by moving the control points along each normal vector based on the line width information given in advance. However, the purpose of this proposal is to thicken and thin the decorative characters in a large size, which is insufficient as a so-called hinting function for controlling the line width in a small size.

As described above, the line width control method for the slanted curve portion for hinting has not been established, and it can be said that the conventional line width correction technique relates to the horizontal and vertical line portions. Therefore, this conventional technique will be described below with reference to FIG.

FIG. 6 is a functional block diagram showing a concrete example of a conventional character generator for correcting the line width of horizontal and vertical line portions.

A CPU (Central Processing Unit) 1 controls the character generator and executes various data processing means 11 to 16 obtained from the program memory 3 in the order of arrows.
The outline font data storage device 2 stores outline font shape data, index data, and related data. The program memory 3 is a data processing means 11 described later.
Stores 16 to 16. The RAM 4 temporarily stores outline font data, and the data processing means 1 described later.
It is used as a work area for 1-16. The display / printing device 5 displays and prints the bitmap image sent from the output means 16 described later on the screen.

Next, the data processing means 11 to 16 in the CPU 1 in FIG. 6 will be described.

The parameter input means 11 receives character specifying information (typeface name / character type / character attribute) and size information specified from outside the apparatus, and the data reading means 12 outlines based on the character specifying information. The corresponding outline font data is read from the font data storage device 2, the size conversion means 13 enlarges or reduces the coordinate value of the outline font data based on the above size information, and the horizontal and vertical line portion line width correction means 143. The quantization error in a small size of the line width of the horizontal and vertical line portions is corrected, the outline data after the correction is converted into the bitmap data by the filling means 15, and the bitmap data obtained by the output means 16 is output in various ways. Output to the device.

[0014]

The first problem with the above-mentioned conventional character generator is that there has been no suitable proposal regarding the line width control of the oblique curve portion.

The reason is that it is more difficult to control the line width of the slanted curve portion than the line width control of the horizontal and vertical portions. This is because it was extremely difficult to automatically determine which part should be the line width and how many values should be set.

The second problem is that the normal vector method in the examples (4) and (5), which touches the line width control of the oblique curve portion described in the section of the prior art, is time-consuming and increases the processing speed. That is not the case.

The reason is that in the normal vector method, the inclination of the tangent line is obtained for each control point, the direction of the normal vector is determined for each point, and the normal vector for each point is decomposed into an X component and a Y component. This is because there is a large amount of calculation processing required. More specifically, the slope of the tangent line at each control point on the contour line of the character is the primary differential value of the function representing the straight line and the curve forming the contour line, and the direction of the normal vector is the direction of the aforementioned contour line. It is in a nested state.

The third problem is that in the normal vector method, when characters are flattened or scaled up, or affine transformation such as rotation or italic or shear is performed, the processing speed for controlling the line width of the oblique curve portion is further slowed. Is to be.

The reason is that the calculation of the differentiated function is first required before the calculation of the differential value mentioned in the second problem.

[0020]

The character generator of the present invention is not limited to a horizontal / vertical line portion whose line width can be easily corrected, but also an oblique curve portion composed of a diagonal portion and a curved portion other than the horizontal / vertical portion. The line width correction of is realized at high speed without using the normal vector method.

More specifically, the character generating apparatus of the present invention comprises a size converting means for enlarging / reducing a character to an arbitrary size using outline font data, and a horizontal / vertical line width and a horizontal / vertical line for identifying a slanted curve portion. Line part / oblique curve part identification means, oblique curve part start / end point detecting means for detecting two points of the oblique curve part start point and end point, horizontal / vertical line part line width correction means, and horizontal / vertical line at the two points mentioned above. Similarity change formula calculating means for detecting a position change of a portion after line width correction and calculating a similarity conversion formula equivalent to this change, and an oblique curve part similarity performing the similar conversion of the oblique curve part using the similarity conversion formula described above. And conversion means.

[0022]

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a functional block diagram showing an embodiment of the character generator of the present invention, and FIG. 2 is a function of the horizontal / vertical line portion / oblique curve portion identifying means 141 of the character generator of FIG. FIG. 3 is a diagram for explaining the function of the oblique line portion start / end point detecting means 142, and FIG. 3 shows the function of the horizontal / vertical line portion line width correcting means 143 of the character generator of FIG. 1 and the character generator of the character generator of FIG. FIG. 4 is a diagram for explaining the function of the similarity conversion formula calculation unit 144, FIG. 4 is a diagram for explaining the function of the similarity conversion formula calculation unit 144 of the character generation device of FIG. 1, and FIG. FIG. 6 is a diagram for explaining the function of the oblique curve portion similarity conversion means 145 of the character generation device, and FIG. 6 is a functional block diagram showing a specific example of a conventional character generation device for correcting the line width of horizontal and vertical line portions. .

A CPU (Central Processing Unit) 1 in FIG. 1 controls the character generator and executes various data processing means 11 to 16 obtained from the program memory 3. The outline font data storage device 2 stores outline font shape data, index data, and related data. The program memory 3 includes the data processing means 11 to 11.
16 are stored. The RAM 4 temporarily stores outline font data and is used as a work area for the data processing means 11-16. The display / printing device 5 displays and prints the bitmap image sent from the output means 16 on the screen.

The data processing means 11 to 16 have been described in detail in the description of the prior art example. Here, the description is omitted except for the line width correcting means 14 which replaces the horizontal and vertical line portion line width correcting means 143 of the prior art. .

The data processing means 141 of the line width correction means 14
~ 145 will be described in detail.

The horizontal / vertical line portion / oblique curve portion identifying means 141 identifies the horizontal / vertical line portion and the oblique curve portion, respectively, and the oblique curve portion start / end point detecting means 142 identifies two points of the starting point and the ending point of the oblique curve portion. The horizontal / vertical line portion line width correction means 143 corrects the line width of the horizontal / vertical line portion, and the similarity conversion formula calculation means 144 corrects the horizontal / vertical line portion line width to determine the start and end points of each oblique curve portion. From the position change, a similarity transformation equation equivalent to this change is calculated. Finally, the oblique curve part similarity conversion means 145 performs similarity conversion on each of the oblique curve parts. As a result, it is possible to realize the slanted curve portion line width correction corresponding to the horizontal and vertical line portion line width correction.

Next, the horizontal / vertical line portion / oblique curve portion identifying means 1
The function of 41 will be described with reference to FIG.

Here, FIG. 2 is a diagram showing a concrete example of an outline font formed from a straight line vector and a curved line vector.

The horizontal and vertical line portions are formed by only straight line vectors, and the slanted curved line portions are formed by straight line vectors and curved line vectors.

The horizontal / vertical line portion / oblique curve portion discriminating means 141 classifies many straight line vectors and curve vectors forming such an outline font into straight line vectors forming horizontal / vertical line portions and others. A method of identifying horizontal and vertical line portions is known. For example, there is a method of detecting a pair vector disclosed in Japanese Patent Laid-Open No. 4-269564. In this embodiment, this method is used to identify the horizontal and vertical line portions, and all the vectors that do not form the horizontal and vertical line portions are regarded as the vectors forming the oblique curve portion.

As specific examples, the slanted curve portions in FIGS. 2A to 2D are listed. In FIG. 2 (a), arc a ₁ a ₂ and arc a
₃ a ₄ , the arc b ₂ b ₃ and the arc b ₃ b ₄ in FIG.
In FIG. 2C, the arcs c ₅ c ₆ , arc c ₆ c ₇ , arc c ₇ c ₈ and arc c ₈ c ₉ , and in FIG. 2D, the arc d ₅ d ₆ , arc d ₆ d ₇ and arc d ₇ d ₈ , the arc d ₈ d ₉ , the arc d ₉ d ₁₀ , the arc d ₁₀ d _11, and the arc d ₁₆ d ₁₇ and the arc d ₁₇ d ₁₈ are oblique curve portions.

Further, the function of the oblique curve portion start / end point detecting means 142 will be described with reference to FIG.

The oblique curve portion start / end point detecting means 142 detects the start / end and the end point of the oblique curve portion where the oblique curve portion is in contact with the horizontal / vertical line portion. As a method of detection, the oblique curve part up to the appearance of the straight line vector forming the horizontal and vertical line parts in the continuation of the vector forming the outline font is collectively considered as one oblique curve part, and the two points of the start point and the end point are considered. To detect. Which is the start point or the end point depends on the direction of the contour line. Specific examples of the start and end points of the oblique curve portion are listed in FIGS. 2 points a ₁ a ₂ and 2 points a ₃ a ₄ in FIG. 2A, 2 points b ₂ b ₄ in FIG. 2B, and 2 points c ₅ c in FIG. 2C.
_9, FIG. 2 (d) in 2-point d ₅ d ₁₁ and two points d ₁₆ d ₁₈ are start and end points of the oblique curve section.

Next, the function of the horizontal / vertical line width correction means 143 will be described with reference to FIG.

The horizontal / vertical line portion line width correcting means 143 corrects the line width of the horizontal / vertical line portion, which is important in forming an outline font, so that the line width unevenness generated in a small size of about 60 × 60 dots or less is generated. Prevents line spacing from being collapsed. As a method of correcting the line width of the horizontal and vertical line parts,
(1) JP-A-4-269564, (3) JP-A-4-362691, and (4) JP-A-4-1042
There is a method of forcibly aligning line width nonuniformity due to a quantization error to a certain fixed value as disclosed in Japanese Patent Publication No. 98, for example.

As a concrete example of the result of the line width correction of the horizontal and vertical line portions, the line width correction examples in FIGS. 3A to 3D will be listed.

FIG. 3A is a correction example in which a horizontal line of a circular Gothic body is evenly thinned, FIG. 3B is a correction example in which a vertical line is evenly thickened, and FIG. An example of correction in which vertical and horizontal lines are thickened in the right direction and upward direction respectively, and FIG. 3D is an example of line width correction in which only the vertical line of the character "sentence" is evenly thinned. Depending on the line width correction method, whether the correction is performed evenly or horizontally or vertically is different, but in the present embodiment, which of the above-mentioned known examples (1), (3), and (4) is horizontal and vertical. It can also be used as a line width correction method for lines. However, only the line width correction of the horizontal and vertical line portions in the above-described known example causes distortion at the joint between the horizontal and vertical line portions and the oblique curve portion. As a specific example, FIG.
In (a), 2 points a ₁ a ′ ₁ and 2 points a ₂ a ′ ₂ and 2 points a ₃
a ′ ₃ and 2 points a ₄ a ′ ₄ and 2 points b ₂ b ′ in FIG. 3 (b).
₂ and 2 points b ₄ b ′ ₄ , 1 point c ₉ c ′ ₉ in FIG. 3 (c),
In FIG. 3D, two points d ₅ d ′ ₅ and two points d ₁₁ d ′ ₁₁ and 2
There is a disagreement between the start and end points of the straight line vector of the horizontal and vertical line portions and the start and end points of the vector of the slanted curve portion as in the points d ₁₆ d' ₁₆ and 2 points d ₁₈ d' ₁₈ . One of the objects of the present embodiment is to eliminate the distortion of the joining portion of the horizontal and vertical line portions whose line width has been corrected and the oblique curve portion to obtain matching.

Next, the function of the similarity transformation formula calculating means 144 will be described with reference to FIGS. 3 and 4.

In order to eliminate the distortion of the horizontal and vertical line portions due to the line width correction, for example, the arc a ₁ a _{2 in} FIG. 3A.
The a or arc a ₃ a ₄ arc a _'1 a' ₂ may be similarity transformations arc a _'3 a' _4. These two similar transformation formulas are the line segments a ₁
a ₂ to line segment a ′ ₁ a ′ ₂ and line segment a ₃ a ₄ to line segment a ′
_This is the same as the formula for similarity transformation to ₃ a ′ ₄ . Figure 4
FIG. 4 is a diagram for explaining a calculation for performing a similarity transformation on a certain line segment pq into an arbitrary line segment rs. In the transformation formula for transforming a line segment into an arbitrary line segment, (1) change in position is represented by parallel movement, (2) change in elevation angle is represented by rotation, and (3) change in line segment length is represented by It can be expressed in scalar magnification. 4, the elevation angle of the line segment pq is α, the elevation angle of the line segment rs is β,
The coordinates of the point p are (X _p , Y _p ), and the coordinates of the point q are (X _q , Y
_q ), the coordinates of the point r are (X _r , Y _r ), and the coordinates of the point s are (X
_s , Y _s ). First, (1) the point p is moved in parallel to the origin o, (2) it is rotated by (β-α) so that the elevation angle α becomes the elevation angle β, and (3) in order to adjust the length of the line segment (the line segment rs Length)
/ (Scalar length of line segment pq) is multiplied, and (4) point p is translated to the position of point r. This conversion matrix is represented by the following formula.

[0041]

In the above equation, (x, y) is the control point of the oblique curve portion before similarity conversion, and (X, Y) is the control point of the oblique curve portion after similarity conversion.

Therefore, the calculated similarity conversion formula is as follows:
In FIG. 3A, two similar transformation formulas for transforming the line segment a ₁ a ₂ into the line segment a ′ ₁ a ′ ₂ and the line segment a ₃ a ₄ into the line segment a ′ ₃ a ′ ₄ and FIG. In (b), the line segment b ₂ b _{4 is changed} to the line segment b ′.
One similarity transformation formula for converting into ₂ b ′ ₄ and one similarity transformation formula for converting the line segment c ₅ c ₉ into the line segment c ′ ₅ c ′ ₉ in FIG. Then the line segment d ₅ d ₁₁ is converted to the line segment d ′ ₅
Two similar conversion formulas for converting the line segment d ₁₆ d ₁₈ into the line segment d ′ ₁₆ d ′ ₁₈ are obtained for d ′ ₁₁ .

Next, the function of the oblique curve portion similarity conversion means 145 will be described with reference to FIG.

Here, FIG. 5 shows the effect that the joint portion with the corrected horizontal and vertical line portions is smoothly connected by the similarity transformation of the oblique curve portion according to the present invention. The similarity conversion formula obtained by the similarity conversion formula calculation means 144 is the arc a ₁ a _{2 in} FIG.
And the arc a ₃ a ₄ and the figure b ₂ b ₃ b _{4 in} FIG.
To the figure c ₅ c ₆ c ₇ c _{8 of} FIG. 5C, and the figure d ₅ d ₆ d ₇ d ₈ d ₉ d ₁₁ and the figure d ₁₆ d ₁₇ d _{18 of} FIG. 5D. By adapting to, it is possible to smoothly connect the joint portion of the horizontal and vertical line portions after the line width correction and the oblique curve portion without distortion. That is, the start and end points a ₁ a ₂ and a ₃ a ₄ of the arc a ₁ a ₂ and the arc a ₃ a ₄ of the oblique curve portion in FIG. 5A are the start and end points a ′ ₁ a ′ _{2 of the} horizontal and vertical line portions. a ′ ₃ a ′ ₄ in the same position as in FIG.
The start and end points b ₂ b ₄ of the figure b ₂ b ₃ b ₄ in the oblique curve portion of (b)
To the start and end points b _'2 b' ₄ at the same position of the horizontal and vertical line portion, start and end points c of the swash curved portion of the graphic c ₅ c ₆ c ₇ c ₈ shown in FIG. 5 (c)
To ₅ c ₉ is start and end points c _'5 c' ₉ and the same position of the horizontal and vertical line portion, FIG shapes oblique curve section of _{(d) d 5 d 6 d} 7 d 8 d 9
d ₁₁ and the start and end points d ₅ d ₁₂ of figure _{d 16 d 17 d 18 d 16} d 18 is the start and end points d _'5 d' ₁₂ and d _'16 d' ₁₈ at the same position of the horizontal and vertical line portion.

In addition, when the characters are enlarged or reduced in a flat shape or subjected to affine transformation such as rotation, italics or shearing,
It suffices to simply obtain the above conversion formula and perform similarity conversion. In contrast to this, in the normal vector method, (1) the differentiated function is calculated after affine transformation, (2) the primary differential value that is the slope of the tangent line is obtained for each control point, and (3) the normal line for each point. The direction of the vector is determined, and (4) each control point is set to the X of the normal vector for each point.
It is necessary to move in the direction and the Y direction. Therefore, it is clear that the present invention allows faster computation.

[0047]

The first effect is to provide a line width correcting means for a slanted curve portion other than the horizontal and vertical line portions, which has been difficult to correct the line width in the past.

The reason is that by calculating a similarity conversion formula that can cope with changes in the start and end points of the slanted curve portion after the correction of the line width of the horizontal and vertical line portions, and by carrying out the similarity conversion of this slanted curve portion, horizontal smoothing can be achieved without distortion. This is because the vertical line portion and the oblique curve portion can be connected.

The second effect is that it is possible to provide a line width correcting means for the oblique curve portion which is faster than the conventional normal vector method.

The reason is that the calculation amount is small and the calculation is simple.

A third effect is that it is possible to provide a line width correction means for a slanted curve portion which is fast even after affine transformation.

The reason is also due to the small amount of calculation and the simplicity of calculation.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing an embodiment of a character generation device of the present invention.

2 is a diagram for explaining a function of a horizontal / vertical line portion / oblique curve portion identifying means 141 and a function of an oblique curve portion start / end point detecting means 142 of the character generator shown in FIG.

3 is a diagram for explaining a function of a horizontal / vertical line part line width correction means 143 of the character generation device of FIG. 1 and a function of a similarity conversion formula calculation means 144 of the character generation device of the character generation device of FIG. 1; .

FIG. 4 is a diagram for explaining the function of a similarity conversion formula calculating unit 144 of the character generator of the character generator of FIG.

5 is an oblique curve portion similarity conversion means 1 of the character generator of FIG.
It is a figure for demonstrating the function of 45.

FIG. 6 is a functional block diagram showing a specific example of a conventional character generator for performing line width correction of horizontal and vertical line portions.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 CPU 2 Outline font data storage device 3 Program memory 4 RAM 5 Display / printing device 11 Parameter input means 12 Data reading means 13 Size changing means 14 Line width correction means 141 Horizontal / vertical line / oblique curve part identification means 142 Oblique curve part Start / end point detection means 143 Horizontal / vertical line width correction means 144 Similarity conversion formula calculation means 145 Oblique curve part similarity conversion means

Claims (3)

[Claims] 1. A size conversion means for enlarging or reducing a character to an arbitrary size using outline font data, and a line width correction means for a horizontal and vertical line portion for arbitrarily thickening or narrowing the line width of the horizontal and vertical line portions. In the character generator having the above, not only the horizontal / vertical line portion whose line width correction has been easy conventionally, but also the slanted line portion and the curved portion other than the horizontal / vertical portion corresponding to the line width correction result of this horizontal / vertical line portion are formed. A character generator characterized by optimally correcting the line width of a slanted curve portion for each character size. 2. The character generating device according to claim 2, wherein the line width-corrected horizontal and vertical line portions and the oblique curve portions are joined together by eliminating distortion. 3. A size conversion means for enlarging or reducing a character to an arbitrary size by using outline font data, and a line width correction means for a horizontal and vertical line portion for arbitrarily thickening or narrowing the line width of the horizontal and vertical line portions. In the character generation device having the above, the line width of the horizontal and vertical line portions, which has been easier to correct the line width than before, and the slanted curve portion composed of the slanted line portion and the curved portion, which are difficult to correct the line width other than the horizontal and vertical portions, are identified. Horizontal / vertical line portion / oblique curve portion identifying means, oblique curve portion start / end point detecting means for detecting two points of the starting point and the ending point of the oblique curve portion, and the line width of the horizontal / vertical line portion can be arbitrarily thickened or narrowed. A horizontal and vertical line portion line width correction means, a similarity conversion formula calculation means for detecting a position change of the horizontal and vertical line portions after the line width correction at the two points, and calculating a similarity conversion formula equivalent to this change; Use a conversion formula to change the diagonal curve A character generation device, comprising: a slanted curve portion similarity conversion unit that converts the character.