WO1998037495A1 - Procede de representation et de reconstitution d'informations concernant une police de caracteres - Google Patents

Procede de representation et de reconstitution d'informations concernant une police de caracteres Download PDF

Info

Publication number
WO1998037495A1
WO1998037495A1 PCT/US1997/002692 US9702692W WO9837495A1 WO 1998037495 A1 WO1998037495 A1 WO 1998037495A1 US 9702692 W US9702692 W US 9702692W WO 9837495 A1 WO9837495 A1 WO 9837495A1
Authority
WO
WIPO (PCT)
Prior art keywords
stroke
font
information
scaling
width
Prior art date
Application number
PCT/US1997/002692
Other languages
English (en)
Inventor
Zining Fu
Original Assignee
Zining Fu
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zining Fu filed Critical Zining Fu
Priority to AU21329/97A priority Critical patent/AU2132997A/en
Priority to US09/171,780 priority patent/US6288725B1/en
Priority to PCT/US1997/002692 priority patent/WO1998037495A1/fr
Publication of WO1998037495A1 publication Critical patent/WO1998037495A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G06T11/20—Drawing from basic elements, e.g. lines or circles
    • G06T11/203—Drawing of straight lines or curves
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/10—Text processing
    • G06F40/103—Formatting, i.e. changing of presentation of documents
    • G06F40/109—Font handling; Temporal or kinetic typography

Definitions

  • This invention is in the field of information processing technology and specifically is in the field of character information processing.
  • each variant is a set of new fonts and a set of fonts cannot be enlarged or reduced within an overly wide range . So that, when restoring small sized fonts, it could result in various distortions, such as strokes merging together, uneven stroke separation, inconsistency in stroke width, or loss of stroke tip. And when restoring large fonts, the strokes appear to be thin and the shape of stroke tips is not sufficiently fine sometimes provided for the small, medium and large sized characters respectively for one font type. This has multiplied the font information volume. Sometimes, the use of hint information to avoid the distortion of small sized characters also increases the font volume and makes the design more complicated.
  • This method requires a font representation and restoration method with controllable stroke width to represent the character bases . Its variant consolidation ratio and scaling overlay ratio depends on the character representation and restoration method used.
  • the disadvantage of the method is that when restoring fonts using the space occupation factor proportion method, the binding between some character elements is somewhat too tight or too loose, and this is especially obvious when changing the stroke width.
  • Use of a character element fine adjustment method will increase the font information volume. Meanwhile, because the different stroke skeleton points in the font are related to each other, when doing fine adjustment on one stroke, it is necessary to adjust the other related strokes, which will thus increase the fine adjustment information volume as well as the difficulty of the design.
  • font information is composed of the following four parts: 1. Stroke type identification and the number and coordinates of skeleton points. 2. Stroke width. 3. Start-stop edge angle. 4. Stroke shape coefficient.
  • the outline control point of each stroke is obtained from the above stated four types of data and the control points are connected with each stroke type according to a pre-selected method to obtain the line of outline of each stroke.
  • the advantage is that the stroke width, the start-stop angle and the stroke tip can be controlled separately; that the stroke tip and angle can be changed only by changing the stroke tip data of the type 3 and type 4 related above; that the stroke width can be changed in a proportion which is different from the scaling proportion; and that it can produce fonts of high accuracy.
  • the disadvantage is that the font set needs to be designed character by character; one font set can not produce many sets of fonts with larger stroke width and style differences; horizontal and vertical stroke width ratio can not be controlled separately, and one cannot prevent the strokes from merging together or avoid an uneven stroke gap. Distortion will occur especially when making strokes wider. See Figure 15.
  • Figure 15 shows a font of "Jiu" after its strokes have been simply widened.
  • the dotted lines 1509 in the figure are the edge lines of the reference font with reference stroke width [1501] .
  • 1503 is the frame line of the font.
  • 1507 is the left skeleton point of the upper stroke "Heng" (horizontal stroke) .
  • the Chinese patent CN1105762A (4) (July 26, 1995 publication) explained the representation and restoration method of Chinese character (Hanzi) fonts which divides a character into three-grade structures of character, character element and stroke.
  • the character structural type information is common to all the fonts.
  • position and size information is added to each font.
  • This method has a high compression ratio, however, due to the limited form of each stroke of each font variant (In an example of an actual application, the stroke type of one font is limited within 255) , distortion in stroke representation may occur in the case of describing fonts with high accuracy. Similarly, the limited number of character element forms and the sub-fonts may also cause distortion in representation.
  • the purpose of the present invention is to design a representation method and a corresponding restoration method for font information which has the following: a high compression ratio; a low distortion during restoration; low font production work load; a font set which can be modified in character form or designed with new characteristics easily by users; and compatibility with the process of Chinese patent (1) and having the capacity to improve its disadvantages.
  • the stroke tip control and the stroke widening method of this invention the variant consolidation ratio and the scaling overlay ratio can be increased to a great extent and the font set of many different variants of one primary font can be combined into one font set .
  • stroke information is designed to be comprised of stroke identification information, skeleton point data and identification information of all the stroke tips in a stroke.
  • the reference font information of each primary font is comprised of the above mentioned stroke information of each stroke in every font and a group of stroke shapes control information of this primary font .
  • Composite strokes can also be used to represent the commonly used stroke combinations.
  • the various variants of a font can share the reference font information related above.
  • Each variant includes an information group for stroke shape control.
  • a nonlinear scaling coefficient table of stroke shape control information based on the font scaling proportion and space occupation information is provided.
  • the space occupation information is added to the font information and the space occupation information mentioned here can be a space occupation factor or a dimension.
  • the concavity and convexity information of font edges is also added in the font information.
  • the font restoration process includes a method of obtaining related shape control information of the said stroke of a certain selected variant from the above mentioned stroke shape control information and a method of and the stroke restoration according to the stroke information and the shape control information related above. It also includes a process for the decomposition of composite strokes and a process for the placement of secondary strokes .
  • Nonlinear scaling procedures for stroke width and the procedure adjustment in accordance with the space occupation information can also be added.
  • the structure of the font production process utilizes a method in which a corresponding stroke restoration process is selected based on the stroke type identification, and each primary font has a set of stroke restoration processes, including one restoration process for each type of stroke.
  • a process adjusted based on the concavity- convexity of the font compositing edge is added.
  • the character style selected in the example of actual application is Song Ti (Song style) .
  • the resolution of the reference font selected is 256 x 256.
  • the left bottom corner is taken as the origin of the coordinate system.
  • Figure 1 is the font restoration processing diagram based on the said method of this invention.
  • Figure 2 is an example of font information structure.
  • Figure 3 is a diagram of the Song Ti composite stroke "Ri" .
  • Figure 4a, b, c, d and e are the shape control information and stroke type examples of the Song Ti stroke "Shu" (vertical stroke) .
  • Figure 5a, b and c are stroke widening coefficient tables based on font scaling proportion, font dimension and font space occupation factor, respectively.
  • Figure 6a, b and c describe the concavity-convexity of the font periphery.
  • Figure 7 is an example of stroke shape control information which records a functional relationship.
  • Figure 8 is the shape control information and stroke type example of the Song Ti stroke "Pie" (left falling stroke) .
  • Figure 9 is the diagram of the process to produce the Song Ti composite stroke "Ri” .
  • Figure 10a and Figure 10b are the diagrams of the font concavity-convexity adjustment process for the left and right structure and upper and lower structure respectively.
  • Figure 11 is the diagram of the process to determine the stroke shape control information.
  • Figure 12 is the diagram of the process to produce Song Ti stroke "Pie” (left falling stroke) .
  • Figure 13 is the diagram of the font scaling and stroke widening adjustment process.
  • Figure 14 is the diagram of follow-up point coordinate placement process.
  • Figure 15 shows the distortion caused by the existing stroke widening method.
  • Figure 16 is a diagram showing the follow-up characteristics of strokes
  • Figure 2 and Figure 7 respectively show the information structure of one font and the stroke shape control information structure of one character style represented by the method of this invention.
  • the font information base of this character style represented by the method of this invention will be then established.
  • the structure shown in Figure 7 can be added to each font variant of one certain primary character style, and in this way, the font information base of various variants which share the font information of the primary character style will then be established.
  • the structure shown in Figure 2 consists of font code area 201, font information length area 202, space occupation information area 203, concavity-convexity information area 204, stroke number area 205, and the stroke information areas of number m strokes 206, 207, up to 208.
  • the structure of the information area of each stroke is the same as the information area 206 of the number one stroke.
  • the information area 206 of the number one stroke includes the stroke type identification area 209, stroke tip type identification area 210 (composed of the stroke tip type of this stroke type 1 to f#- [1] ) , skeleton point widening type area 211 (composed of the widening type of skeleton point 1 to g#- [1] ) , skeleton point follow- up identification area 212 (composed of the follow-up identification of the skeleton point 1 to g#- [1] of this stroke) and nl number data area 213, 214 up to 215.
  • the space occupation information area 203 can be either a single total space occupation factor or dimension, or the X or Y space occupation factors or dimension of the font. X and Y dimensions 1#- [x] and
  • the skeleton point widening type area 211 records the widening type of every stroke tip.
  • a Y float When the Y coordinate needs to be adjusted vertically according to the width changes of the horizontal stroke of the font, it is called a Y float.
  • X float When the X coordinate needs to be adjusted horizontally according to the width changes of the vertical stroke of the font, it is called an X float.
  • the follow-up identification area 212 identifies the follow-up type of every skeleton point .
  • follow-up refers to the linked movement a certain skeleton point of one stroke makes with one skeleton point of another stroke.
  • the upper end point of stroke 1601 does not move with other strokes
  • the left end X coordinate of stroke 1602 is consistent with the X coordinate of the upper end point 1601
  • the Y coordinate of stroke 1603 is consistent with the Y coordinate of the lower end of stroke 1601
  • the X and Y coordinates of stroke 1605 are consistent corresponding with the X and Y coordinates of the left end of stroke 1604.
  • the follow-up type can be four types corresponding with the four situations in Figure 16: non follow-up, X follow-up, Y follow-up, and X and
  • the follow-up information includes stroke numbers and skeleton numbers, indicating which skeleton point coordinate values of which stroke the said coordinate is consistent with.
  • the stroke 1602 of the font "Shang” in Figure 16 is recorded as "X follow-up" in the point 1 follow-up identification area of the skeleton point follow-up information area 212 in Figure 2, but in the number 1 data area 213, the X coordinate value is no longer recorded and the follow-up information is recorded, in which the stroke number is absolute number 1 or relative number -1.
  • the skeleton number 1 means the said X coordinate is the same as the X coordinate of the first skeleton point of the first stroke of this font.
  • the concavity-convexity information area 204 records the periphery concavity-convexity of the font, which is explained in Figure 6.
  • the N7 to N7 in Figure 6a has defined eight concave-convex directions of a font periphery, with the upper left corner as the starting point. Using the concavity-convexity in the four directions of upper, lower, left and right is also appropriate. If a font has no abutting stroke on the periphery at one certain direction, this character is then considered to be concave in this direction and when it is combined with other characters, the stroke of the other character is allowed to be inserted. If concave is labeled with 1 and the convex is labelled with 0, the periphery concavity-convexity of the character "Shang" can be expressed as "10100001" as shown in Figure 6b.
  • the concavity-convexity of the upper section is the logical expression of the upper right concavity- convexity N2L of the left character and the upper left concavity-convexity NOR of the right character
  • the concavity-convexity of lower section is the logical expression of the lower right concavity- convexity N4L of the left character and the lower left concavity-convexity N6R of the right character.
  • the overall concavity-convexity of the font with an upper and lower structure can be obtained from the concavity-convexity information of the upper character and the lower character using a similar method.
  • Area 207 shows the example of the representation of composite stroke "Ri" in character “Hu” , including stroke identification area 216, stroke tip identification area 217, skeleton point widening type area 218, follow-up identification area 219, every skeleton point data area 220, 221, 222 and 223 (equal to the upper left corner skeleton point coordinate XI and Yl, and the lower right corner skeleton point coordinate XI and Yl) .
  • the stroke tip identification area 217 records the lower left and lower right stroke tip identification of the said composite stroke as type II and type II respectively.
  • the skeleton point widening type area 218 and the follow-up identification area 219 respectively record all the upper left and lower right skeleton points' widening type as "XY float" and the follow-up type as "no follow-up” .
  • Figures 4a, b, and c respectively represent the three types of upper stroke tip of Song Ti stroke "Shu" (vertical stroke) , identified in order as type I, type II and type III.
  • Figure 4d and e respectively show the two types of the lower stroke tip, identified in order as type I and type II.
  • the stroke tip with several skeleton points is determined.
  • there are two stroke tip identifications in the 210 area and three upper tips and two lower tips can be combined into six types of Song Ti "Shu" strokes with different stroke tips.
  • points 401 and point 402 are the upper and lower skeleton points of the vertical stroke
  • points 403 to 412 are the outer/external characteristics points, among which, points 403, 404, 405 and 406 are the main/basic outline points of the stroke, which horizontally move towards the left and the right (Wh/2) from the skeleton points 401 or 402.
  • Point 407 moves horizontally towards the left (Wl) from point 403, point 408 moves vertically downwards (Hu) from point 404, point 409 is located on the extended line of the ligature of points 407 and 408 and has a horizontal distance of Wr from the point 404.
  • the point 410 is located directly below the point 408 and has a vertical distance of Hr from the point 409.
  • the point 411 is located directly below the point 403 and the distance is Hi.
  • the point 412 is located directly above the point 406 and the distance is Hd.
  • the curve from the point 408 to 410 is a quadratic Bezier curve with points 403, 409 and 410 as its characteristic points.
  • the curve from the point 405 to 412 is a quadratic Bezier curve with points 405, 406 and 412 as its characteristic points.
  • the point 401' in Figure 4b is obtained by moving vertically upwards (Hu + Wk/2) from the point 401.
  • Figure 3 is an example of the composite stroke "Ri” .
  • a composite stroke is a group of strokes which is composed of two or more other strokes and has certain relations within all the strokes.
  • this specific “Ri” is consisted of stroke “Zuo Shu” (left vertical stroke) , “Heng Zhe Shu”, “Zhong Heng” (center horizontal stroke) and “Xia.Heng” (bottom horizontal stroke)
  • the “Heng Zhe Shu” stroke is further consisted of a horizontal stroke and a vertical stroke and the distances between the center horizontal stroke and the bottom and the upper vertical strokes are same .
  • the stroke shape control parameter includes the length (Hh) , which extends downwards correspondingly to the type II stroke tip of "Ri" stroke, and the widths of the present vertical and the horizontal strokes (Wk and Wh) (Shown in Figure 7) .
  • the dotted lines at the left and right lower parts are respectively the outlines of the left lower stroke tip of type II and the right lower stroke tip of type II.
  • the solid lines above the dotted lines are the outlines of the corresponding lower left stroke tip of type I and the lower right stroke tip of Type I .
  • the dotted line at the right upper corner is the outline of "Heng Shu Zhe" stroke when its right upper stroke tip is type I, and the solid line is the outline of its type II stroke tip.
  • PI an P2 are the skeleton points of stroke "Ri" at the left upper corner and right lower corner.
  • P3 and P4 or P'4 (corresponding to the situations when the left lower stroke tip is type
  • P5 or P'5 correspondent with the situations when the left lower stroke tip is type I and II respectively
  • P6 and P7, and P8 and P2 are the left and right skeleton points of the center and bottom horizontal strokes, respectively.
  • the stroke number of the composite stroke "Ri" is 2 and the skeleton point number is also 2.
  • the coordinate value of the skeleton points of each sub-stroke of this stroke "Ri" can be calculated according to the following equation:
  • the use of the composite stroke can reduce stroke redundancy. With this composite stroke, the distortion such as 1505 shown in figure 15 can also be eliminated and the equidistance of the isometric type composite stroke can be maintained.
  • the basic stroke type is determined and then the composite strokes composed of other strokes are determined and labelled as stroke identification by order.
  • the stroke compositions with frequent appearance, more coordinate repeating information or the isometric type strokes are determined as composite strokes.
  • Each stroke has the same skeleton point number and same stroke tip number.
  • the stroke tip type of each tip point and turning point with different shapes of each stroke are determined and labelled separately as identifications.
  • each font is processed according to the font code order.
  • the font code, the concavity- convexity information and the stroke number of each font are placed in areas 201, 204 and 205.
  • step 101 is to obtain the font starting position from the font information table based on the font code.
  • step 102 is to execute the determination process of the stroke shape control information shown in Figure 11.
  • step 103 is to execute the font scaling and the stroke widening adjustment process shown in Figure 13.
  • Step 104 is to execute the locating process of the follow-up point coordinates shown in Figure 14.
  • Step 105 is to obtain the present font stroke number m and set the present stroke number s to 1.
  • Step 106 is to select and execute one corresponding process from 107, 108 and so forth based on the primary character style to which the present character style belongs.
  • Step 107 is to select and execute one process from 109, 110 and all the following Song style strokes' production processes based on the present stroke type identification.
  • Step 109 is to execute the production process for Song style stroke "Pie” shown in Figure 12.
  • Step 110 is to execute the production process for Song style composite stroke "Ri” shown in Figure 9, etc.
  • Step 108 is to select and execute one process from 111, 112 and all the following Hei style strokes' production processes based on the present stroke type identification.
  • Steps 111 and 112 are, respectively, the production processes of the Hei style stroke "Pie” and the composite stroke "Ri", etc.
  • Step 113 is to add 1 to the present stroke number s .
  • Step 114 is to determine if s>m. If the condition is untenable, it will execute step 106, or else it is the end of the process .
  • the determination process of the stroke shape control information shown in Figure 11 is used in step 102 shown in Figure 1.
  • Step 1100 is to select the stroke shape control information of a variant of the primary character style as the present control information according to the present character style .
  • Step 1101 is to search the table in Figure 5 using the present X and Y scaling proportion Kx and Ky respectively, and to obtain the width scaling proportion KWh and KWk of the vertical and horizontal strokes under the present scaling proportion.
  • Step 1102 is to calculate KWk * Wk and KWh * Wh, and obtain the reference stroke widths Wkl and Whl of the horizontal and vertical strokes under the present scaling proportion.
  • Step 1103 is to obtain the dimensions lx and ly of the present font from 203 in Figure 2, then multiply them respectively by Wk and Wh and further obtain the X and Y space occupation coefficients WX and WY of the font. And then, the left and right sections of the table in Figure 5c are searched using WX and WY to obtain the stroke width adjustment coefficients CX and CY of the horizontal and vertical strokes.
  • Step 1104 is to calculate CX * Wkl and CY * Whl to obtain the actual horizontal and vertical stroke widths W'k and W'h, respectively, of the present font.
  • Step 1105 is to calculate other coefficients using Wk, W'h and the functional relations recorded in Figure 7, and then form the shape control coefficient structure of the present font shown in Figure 2 to be used in the present font restoration steps 103 to 114 in Figure 1, and at last return.
  • it may also use lx and ly directly in step 1103 to search the left and right sections, respectively, shown in Figure 5b to obtain CX and CY.
  • Another plan is to use the following method in step 1102 and 1104: Use lx and ly to search the left and right sections shown in Figure 5b respectively, obtain the adjusted reference horizontal and vertical stroke widths from the 3rd and 6th columns, and multiply them respectively by KWh and KWk, and then similarly, W'k and W'h are obtained.
  • Steps 1104 and 1105 ensure that the stroke width and each coefficient value are equal or greater than 1 and the stroke tip will not get lost even without using hint information.
  • the effects obtained are: 1.
  • the stroke narrowing proportion is smaller than the font reduction proportion (or the opposite when the font is enlarged)
  • Increase in the width of vertical strokes' Wh or horizontal strokes' Wk of the characters with fewer vertical or horizontal strokes (or the opposite with the characters with more horizontal or vertical strokes) and at the same time, other stroke shape control coefficients are also adjusted correspondingly. Therefore, although there is no width information in each stroke, the proper font stroke width is guaranteed.
  • An independent set of tables as shown in Figure 5a, b and c can also be set up for each primary character style.
  • 1701 is the side frame line of the reference font
  • 1702 is obtained from reduction in 1701 when the width of both horizontal and vertical strokes increase compared with the reference font
  • 1703 is obtained from enlargement of 1701 when the width of both horizontal and vertical strokes diminish compared with the reference font .
  • 1704 and 1705 are the centerline of the reference font
  • 1706 and 1707 are obtained by moving 1702 and 1703 respectively in the horizontal direction.
  • Zh and Zw are the height and width of the reference font, respectively.
  • Y'i Yi * ny + ⁇ Wk/2 (Eqation 2)
  • Equation 1 and Equation 2 are used in steps 1307 and Equation 2 are used in steps 1307 and Equation 1 and Equation 2 are used in steps 1307 and Equation 1 and Equation 2 are used in steps 1307 and
  • equation 1 the above related effect is actually being reached by reducing the font in width with the amount of ⁇ Wh and then moving it in the horizontal direction with the amount of ⁇ Wh/2 (only change the X coordinate of the points in the skeleton point widening type identification area 211 in Figure 2, which are identified as necessary to be widened in the X or XY direction) .
  • the case of vertical strokes being narrowed is just the opposite.
  • the method for stroke widening adjustment in the vertical direction using the equation 2 is similar to this .
  • the effect reached by this method is that when the stroke width is changed within the same side frame line, no distortion like 1505 will occur to the font and at the same time the font structure is kept even, which has overcome a shortcoming of the Chinese patent (3) .
  • step 103 is to obtain the stroke number m and set the present stroke number v as 1.
  • step 1302 is to obtain the present stroke skeleton point number ts from the stroke shape control information and set the present stroke skeleton point number v as 1.
  • Step 1303 is to obtain the widening type of the present skeleton point and the follow-up identification from 211 and 212, respectively, in Figure 2.
  • Step 1310 is to determine if the present point's Y coordinate is independent. If it is independent, step 1311 is skipped. Step 1311 is to carry out widening adjustment to the Y coordinate of the present point according to the equation 2.
  • Step 1312 is to add 1 to the present skeleton point number v.
  • Step 1313 is to determine if v is greater than ts . If the condition is untenable, step 1303 and the following steps are executed, or else step 1314 is executed. Step 1314 is to add 1 to the present stroke number s.
  • Step 1315 is to determine if s is greater than m. If the condition is untenable, step 1302 and the following steps are executed again, or else the process is ended and returns to start.
  • the follow-up point coordinate locating process shown in Figure 14 is used at 104 in Figure 1. Steps 1401 and 1402 are the same as steps 1301 and 1302 in Figure 13, respectively. Step 1403 is to obtain the X,
  • Step 1404 is to determine if the X coordinate of the present point is follow-up. If it is not follow-up, step 1405 is skipped (At this point, the X coordinate value obtained by step 1403 is actually the follow-up information) .
  • Step 1405 is to obtain the order number of the stroke which is followed up and the order number of the point which is followed up from the X follow-up information of the point, obtain the X coordinate value of this point and give the X coordinate position of the present point.
  • Step 1406 is to determine if the Y coordinate of the present point is follow-up. If it is not follow-up, step 1407 is skipped.
  • Step 1407 is to obtain the order number of the stroke which is followed up and the order number of the point which is followed up from the Y follow-up information obtained by step 1403, obtain the Y coordinate value of this point and give the Y coordinate position of the present point.
  • Step 1408 is to determine if v equals or is greater than ts. If the condition is tenable, step 1410 is executed, or else the process will continue from step 1403 after step 1409 is executed. Step 1409 is to add 1 to the present skeleton point number v. Step 1410 is to determine if s is equal to or greater than m. If the condition is tenable, the process returns, otherwise it continues from step 1402 after step 1411 is executed. Step 1411 is to add a to the present stroke number s.
  • the production process of the Song style stroke "Pie” in Figure 12 is used in step 109 in Figure 1.
  • the produced stroke is as shown in Figure 8.
  • the entrance coefficient of this process includes stroke tip type identification, skeleton point coordinates and the present font stroke shape control information.
  • the stroke "pie” shown in Figure 8 has three skeleton points 801, 802 and 803. At the upper end there are three stroke tip types of types I, II and III.
  • the points 805, 806, 807, 808, 809, 810 and 803 are the basic shape control points of this stroke. Therefore, when the stroke "pie” is represented using the area 210 in Figure 2, the stroke identification number fl is 1, and the skeleton point number gl is 3.
  • the shape control coefficient of the stroke "pie” includes the upper end width Wx, the middle point width Wz, the lower pointing tip radius WR, the right upper tip outer point distance Wt and the right upper tip height Wy. Together with the stroke identification number and the skeleton point number, they are all given in the stroke shape control information shown in Figure 7.
  • the skeleton line passing through the skeleton points 801, 802 and 803 is a quadratic Bezier curve.
  • the points 805 and 806 are on the normal of the skeleton line which passes through the point 801, and both points have a distance from the point 801 of Wx/2.
  • the points 807 and 808 are on the normal of the skeleton line which passes through the point 802, and both points have a distance from the point 802 of Wz/2.
  • the point 804 is on the skeleton line and has a distance from the point 803 of WR.
  • the points 809 and 810 are on the normal of the skeleton line which passes through the point 804, and both points have a distance from the point 804 of WR.
  • the left outline which passes through the points 805, 807 and 809, and the right outline which passes through 806, 808 and 810 are the quadratic Bezier curves passing through the above related two sets of points, respectively.
  • the curve which passes through the points 810, 803 and 809 is a half circular arc passing through these three points.
  • the point 811 is on the extension of the line passing through the points 805 and 806, and has a distance from the point 806 of Wt .
  • step 1201 is to determine the skeleton line passing through the skeleton points 801, 802 and 803.
  • step 1202 is first to determine the point 804, and then determine the basic control points 805 up to 810.
  • step 1203 is to determine the curves that pass through the points 805, 807, 809 and points 806, 808 and 810.
  • steps 1201 and 1203 the unknown coefficient t of the Bezier curve at the center point is given as 0.5 and thus the coefficient of the parametric equation is obtained.
  • Step 1204 is to calculate other stroke shape control points according to the description in Figure 8.
  • Step 1205 is to select according to the upper stroke tip type and execute steps 1206, 1207 and 1208 respectively for the stroke tips of type I, II and III. Steps 1206, 1207 and 1208 produce the outline curves of the stroke "Pie" of three types.
  • Step 1209 is a known blackening process which fills the area within the closed outline produced by the last step, and then returns .
  • Step 901 is to calculate the coordinate of each point according to the description in Figure 3.
  • Step 902 is to determine if X2 - XI ⁇ 2 (If the left and right vertical strokes merge together when the vertical width is 1) . If the condition is untenable, step 903 is skipped.
  • Step 903 is to set X2 as XI + 2.
  • Step 904 is to determine if Yl - Y2 ⁇ 4 (If three horizontal strokes merge together when the horizontal width is 1) . If the condition is untenable, step 905 is skipped.
  • Step 905 is to set Y2 as Yl - 4.
  • Step 906 is to determine if three horizontal strokes merge together, namely to determine if Yl - Y2 ⁇ 2 * Wk + 2. If there is no merging together, step 907 is skipped. Step 907 is to set the horizontal width W'k of the present composite stroke "Ri" as (Yl - Y2 - 2)/2. Step 908 is to determine if the left and right vertical strokes merge together, namely to determine if X2 - XI ⁇ Wh + 1. If not, step 909 is skipped. Step 909 is to set the vertical width W'h of the present composite stroke "Ri" as X2 - XI - 1.
  • Step 907 and 909 are to recalculate each shape control coefficient and produce the stroke shape control coefficient of the present stroke according to Figure 7.
  • Step 910 is to set the upper stroke tip type as type II and the lower stroke tip type as type II.
  • Step 911 is to determine the left lower stroke tip type, and execute step 912 with type I stroke tip and step 913 with type II stroke tip.
  • Step 912 or 913 is to execute the production process of the stroke "Shu" using the upper and lower skeleton points PI, P4 or PI, P'4 and the stroke tip type set by step 910.
  • Step 914 is to set the left, left upper and the lower stroke tip type of the stroke "Heng Zhe Shu" as type I, type II and type II respectively.
  • Step 915 is to determine the right lower tip type of the stroke "Ri' , and execute steps 916 and 917 with the type I and type II strokes respectively.
  • Step 916 or 917 is to execute the production process of the stroke "Heng Zhe Shu” using the left upper and right lower skeleton points PI and P5 or PI and P'5 and the stroke tip type set by step 915.
  • Step 918 and 919 are to set both left and right stroke tip of the center horizontal stroke and the lower horizontal stroke as type I and execute the production process of the vertical strokes using the left and right skeleton points P6, P7 or P8 , P2 and the pre-set stroke tip type. At last, the process returns.
  • the production process of sub-strokes are executed using the stroke shape control coefficient of the present stroke.
  • the distortion shown by 1505 in Figure 15 can be eliminated by using the composite strokes.
  • equal-distance type composite strokes can be prevented from becoming unevenly distanced or merging together when using the small font sizes.
  • the stroke shape control information of this character style is then formed.
  • the stroke shape control information volume of different character styles may be different.
  • the production process of each basic stroke can be determined by the following method:
  • the entrance coefficient includes the position of each skeleton point of this stroke and the stroke tip type of each end point and turning point.
  • First is the determination step of each skeleton point and turning point (including step of selecting corresponding process step for this stroke tip type, and the determination step of each control point of each stroke tip type) .
  • the last is step of executing a filling process and step of returning.
  • the entrance coefficients are the same as those of the basic strokes and the its production process is as follows: First, calculate the coordinate of each skeleton point, then determine the skeleton point and the stroke tip of each component stroke according to the skeleton and the stroke tip, and at last, using the skeleton point and the stroke tip information of each component stroke as coefficients, call the production process of each component stroke.
  • steps of equal-distance adjustment and the stroke width adjustment which is to avoid the stroke merging together, are added.
  • the shape of each stroke is controlled by its specific restoration process based on the shape control information.
  • Different primary character styles have different restoration processes and the different variants of the same primary character style have different shape control information.
  • This type of design has transferred the complicated stroke shape control to the specific stroke restoration process, thus simplified the font information and attained the effect of easy design and flexible control.
  • the stroke identification and the relevant control information By changing the position of the skeleton points, the stroke identification and the relevant control information, many different styles of one stroke can be produced through one production process of this stroke.
  • the stroke shape can be precisely controlled by the stroke restoration process and the connections can be made smooth without increasing the stroke information volume.
  • the present invention can increase the variant consolidation and the scaling overlay compression ratio by a great amount.
  • FIG. 1 and the following related representation method, and Figure 6 and the following related restoration method in the said patent are taken as examples.
  • Each font base in the font information area 107 of the font base described in the said patent has used the structure shown in Figure 2 of the present invention.
  • the various variants of each primary character style share a font base area which uses the method of the present invention, and each variant has a separate stroke shape control information area.
  • the font production process 602 of the said patent is replaced with the font production process in Figure 1 of the present invention.
  • step 1001 is to obtain the concavity-convexity information of the left and right font respectively.
  • Step 1002 is to determine if the opposed implantation is possible.
  • step 1003 is executed. If it is possible, step 1005 is then executed. Step 1003 is to determine if the implantation can be done in one direction of left or right. If yes, it will turn to step 1006, or else step 1004 is executed. Step 1004 is to reduce the two sub- fonts by a certain proportion and move the right font towards the right. Step 1005 is to enlarge the two sub-fonts by a certain proportion, move the right font towards the left and make the two fonts implanted with each other. Step 1006 is to adjust the left and right fonts according to the width difference of the vertical strokes. Step 1007 is to obtain the overall concavity-convexity according to the method shown in Figure 6.
  • Step 1008 is to return to the overall concavity-convexity.
  • Steps 1004 and 1006 respectively keep an extra space of ⁇ Wh and ⁇ Wh/2 between the left and right sub-fonts.
  • the ⁇ Wh is 0, it is the concavity-convexity adjustment process of the left and right fonts without considering the influence from the strokes' widening.
  • step 1006 can be skipped.
  • general technical people in the field can easily carry out the concavity-convexity adjustment process shown in Figure 10b for the fonts of upper and lower structure types. This method has rectified the shortcomings in the standard font combining process of the Chinese patent (1) .
  • the font information which are generated by the font information demonstration method stated in the present invention can be saved in an information storage device.
  • the restoration method stated above can be realized by an information processing program.
  • This program can also be saved in an information storage device and can be run in this device to make products including the invention method.
  • the said information processing device usually has the capabilities of information storage, processing, input, and output, such as a computer.
  • the present invention can be used independently, or used in combination with the previously related Chinese patent (1) as well. It can be either used for the Chinese fonts or the European fonts. It can also be used as an auxiliary tool for font design.
  • the combined use of the present invention with the Chinese patent (1) can make full use of the advantages of the Chinese patent (1) , such as high compression ratio, high independence of the standard font information and the low design work load of the standard font combination, and the advantages of the present invention, such as high font representation accuracy, high variant consolidation and compression ratio, great stroke widening range and the high scaling overlay compression ratio, and is one of the best practical application plans.
  • the total storage volume of one set of a primary character style (including its different kinds of variants) is approximately 300KB, with an approximate increase of 150KB per additional primary character style. Therefore, the total volume of ten sets of primary character styles (may include up to 50-100 sets of variants, namely 50-100 sets of fonts as generally called) does not exceed 2000KB, which is equivalent to the volume of a 20,000- character font set of the same accuracy. Its overall compression ratio is far higher than any of the present technology and the restoration accuracy is maintained high at the same time. Because the standard font takes more than 98% of the 60,000 characters, the font design work load is very low.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Controls And Circuits For Display Device (AREA)

Abstract

Selon l'invention, les informations concernant une police de caractères comprennent des informations sur le type de trait, les points de squelette et le type d'extrémité de trait (types I, II et III), communes à différentes variantes de polices de caractères. Chaque variante comprend en outre des informations sur la largeur des traits horizontaux et verticaux, ainsi que des informations de commande de la forme des traits (803 et 805-810) exprimées en tant que fonctions desdites variantes. Chaque variante peut comprendre également des informations sur l'élargissement (Wx, Wy, Wz, Wr et Wt) des points de squelette du trait et des informations sur les liaisons entre les points de squelette, ainsi que des informations de mise à l'échelle non linéaire de la largeur des traits et des informations sur des traits composites du style de caractère primaire, fondées sur des informations concernant les rapports de mise à l'échelle et l'occupation de l'espace. Le procédé de reconstitution consiste à déterminer les informations de commande de la forme des traits et les informations de commande des contours, à ajuster la largeur des points de squelette, à déterminer les points de liaison entre les points de squelette, à ajuster la largeur des traits présents, à ajuster de façon isométrique les traits composites et à réajuster la largeur des traits, afin d'éviter la fusion de traits.
PCT/US1997/002692 1997-02-24 1997-02-24 Procede de representation et de reconstitution d'informations concernant une police de caracteres WO1998037495A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
AU21329/97A AU2132997A (en) 1997-02-24 1997-02-24 Representation and restoration method of font information
US09/171,780 US6288725B1 (en) 1997-02-24 1997-02-24 Representation and restoration method of font information
PCT/US1997/002692 WO1998037495A1 (fr) 1997-02-24 1997-02-24 Procede de representation et de reconstitution d'informations concernant une police de caracteres

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/US1997/002692 WO1998037495A1 (fr) 1997-02-24 1997-02-24 Procede de representation et de reconstitution d'informations concernant une police de caracteres

Publications (1)

Publication Number Publication Date
WO1998037495A1 true WO1998037495A1 (fr) 1998-08-27

Family

ID=22260405

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1997/002692 WO1998037495A1 (fr) 1997-02-24 1997-02-24 Procede de representation et de reconstitution d'informations concernant une police de caracteres

Country Status (2)

Country Link
AU (1) AU2132997A (fr)
WO (1) WO1998037495A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1517231A1 (fr) * 2002-03-29 2005-03-23 Enmedia System COPR. Banque de caracteres chinois a courbes integrees decrites sur la base de l'axe de reference

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4931953A (en) * 1987-04-30 1990-06-05 Hitachi, Ltd. Character pattern generation method and system therefor
US4933866A (en) * 1988-11-17 1990-06-12 Vital Lasertype, Inc. Method and apparatus for generating aesthetically alterable character designs
US5018217A (en) * 1988-09-26 1991-05-21 Brother Kogyo Kabushiki Kaisha Data converting apparatus having means for changing ornamental stroke end data of character outline
US5305433A (en) * 1989-09-29 1994-04-19 Seiko Epson Corporation Method and system for generating characters with compact character data
US5398311A (en) * 1987-02-25 1995-03-14 Canon Kabushiki Kaisha Character processing apparatus and method for processing character data as an array of coordinate points of contour lines
US5416852A (en) * 1990-08-24 1995-05-16 Fuji Xerox Co., Ltd. Pattern processing system and method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5398311A (en) * 1987-02-25 1995-03-14 Canon Kabushiki Kaisha Character processing apparatus and method for processing character data as an array of coordinate points of contour lines
US4931953A (en) * 1987-04-30 1990-06-05 Hitachi, Ltd. Character pattern generation method and system therefor
US5018217A (en) * 1988-09-26 1991-05-21 Brother Kogyo Kabushiki Kaisha Data converting apparatus having means for changing ornamental stroke end data of character outline
US4933866A (en) * 1988-11-17 1990-06-12 Vital Lasertype, Inc. Method and apparatus for generating aesthetically alterable character designs
US5305433A (en) * 1989-09-29 1994-04-19 Seiko Epson Corporation Method and system for generating characters with compact character data
US5416852A (en) * 1990-08-24 1995-05-16 Fuji Xerox Co., Ltd. Pattern processing system and method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1517231A1 (fr) * 2002-03-29 2005-03-23 Enmedia System COPR. Banque de caracteres chinois a courbes integrees decrites sur la base de l'axe de reference
EP1517231A4 (fr) * 2002-03-29 2010-04-14 Enmedia System Corp Banque de caracteres chinois a courbes integrees decrites sur la base de l'axe de reference

Also Published As

Publication number Publication date
AU2132997A (en) 1998-09-09

Similar Documents

Publication Publication Date Title
US6288725B1 (en) Representation and restoration method of font information
EP1762983B1 (fr) Ajustement de glyphes pour le rendu d'images tramées à haute définition
JP4358472B2 (ja) イメージデータの非対称スーパーサンプリングラスタ化のための方法およびシステム
KR100324879B1 (ko) 문자 표시 장치, 문자 표시 방법 및 기록 매체
EP0255833A1 (fr) Generation automatique de caracteres en mode point a partir de contours
JPH0652472B2 (ja) イメージ処理方法
EP1762984A2 (fr) Réglage du contour de glyphes lors de la restitution
JPS63271290A (ja) 文字パタ−ン生成方式
WO1998037495A1 (fr) Procede de representation et de reconstitution d'informations concernant une police de caracteres
DE4124554C2 (de) Datenumwandlungseinrichtung und Zeichenausgabeeinrichtung
JPS59218493A (ja) 図形表示情報記憶方法
JPS62272295A (ja) 線幅制御機能を持つ文字パタ−ンの拡大・縮小方式
DE2700828A1 (de) Ablenksystem fuer eine farbfernsehbildroehre
WO2004032060A1 (fr) Procede d'analyse et de modification d'une empreinte
JP3391822B2 (ja) アウトラインフォントファイル圧縮方法
JP2669539B2 (ja) 文字出力方法および文字出力装置
JP2912638B2 (ja) 漢字フォントの階層的ベクトル化方式
JP2004029598A (ja) 文字作成方法および文字作成プログラム
JP3224138B2 (ja) 漢字フォント生成方式
JPH01270094A (ja) 文字フォント生成方法
JPS63187331A (ja) マルチウインドウによるデ−タ表示方式
JPS60242492A (ja) 文字発生装置
JP2939159B2 (ja) 手書き文字の位置揃え処理装置および方法
JP4059438B2 (ja) 図形または文字の表示装置、表示方法、表示プログラムおよび表示プログラムを記録した記録媒体
JPH07106644B2 (ja) 文字発生装置

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AL AM AT AU AZ BA BB BG BR BY CA CH CN CU CZ DE DK EE ES FI GB GE HU IL IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MD MG MK MN MW MX NO NZ PL PT RO RU SD SE SG SI SK TJ TM TR TT UA UG US UZ VN

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): KE LS MW SD SZ UG AM AZ BY KG KZ MD RU TJ TM AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN ML MR NE SN TD TG

WWE Wipo information: entry into national phase

Ref document number: 09171780

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

NENP Non-entry into the national phase

Ref country code: CA

122 Ep: pct application non-entry in european phase