TWI231457B - System and method for converting an outline font into a glyph-based font - Google Patents
System and method for converting an outline font into a glyph-based font Download PDFInfo
- Publication number
- TWI231457B TWI231457B TW90129744A TW90129744A TWI231457B TW I231457 B TWI231457 B TW I231457B TW 90129744 A TW90129744 A TW 90129744A TW 90129744 A TW90129744 A TW 90129744A TW I231457 B TWI231457 B TW I231457B
- Authority
- TW
- Taiwan
- Prior art keywords
- pattern
- feature points
- basic
- curve
- points
- Prior art date
Links
Landscapes
- Controls And Circuits For Display Device (AREA)
Abstract
Description
1231457_ #¥日贏明吸) 發明領域 本發明屬於電腦書寫文字或符號的數位表現法,尤指 轉換外圍字型成為圖樣字型的一種系統、方法以及電腦可 讀取之媒介裝置。 發明背景1231457_ # ¥ 日 赢 明 吸) FIELD OF THE INVENTION The present invention belongs to the digital expression method of computer-written characters or symbols, especially a system and method for converting peripheral fonts into pattern fonts, and a computer-readable medium device. Background of the invention
目前已經有很多的字型產生系統,可用於產生亞洲文 字的字型,亞洲字型係指一些表意圖形所組成的文字字 型,亞洲語言可能包含上千個以上的文字,例如中國語言 就包含二萬個以上的文字。At present, there are many font generation systems that can be used to generate Asian fonts. Asian fonts are text fonts composed of ideographic graphics. Asian languages may contain more than a thousand characters. For example, the Chinese language contains More than 20,000 words.
一種傳統的電腦技術用來產生亞洲字型的方法,係採 用字型的外圍邊線來表現文字的形狀,例如(阿斗米系統 公司(Adobe System, Inc.)描述於n PostScript Language Tutorial and Cookbook” (Addi sion-Wesley Publ ishing,1985)的系統就屬於這種技術,其方法係將 文字字型之外圍邊線儲存成直線段及曲線段的集合,這種 技術有如下缺點:(1)由於不同的字型必須定義不同的外 圍形狀,於是需要相對大量的儲存空間,(2)字型外圍形 狀在高解像度空間上所定義的字型資料,僅適用於高解像 度空間上的顯像,並不適用於在相對低解像度空間上做高 品質的顯像用,換言之,字型資料在不同顯像度空間具有 不能相容的特性。A method used by traditional computer technology to generate Asian fonts, which uses the outer edges of the font to represent the shape of the text, such as (described by Adobe System, Inc. in PostScript Language Tutorial and Cookbook "( Addision-Wesley Publ ishing (1985) belongs to this technology. Its method is to store the peripheral edges of the text font as a set of straight and curved segments. This technique has the following disadvantages: (1) due to different Fonts must define different peripheral shapes, so a relatively large amount of storage space is required. (2) Font data defined by font peripheral shapes in high-resolution space is only suitable for development in high-resolution space, and is not applicable. It is used for high-quality imaging in a relatively low resolution space. In other words, font data has incompatible properties in different resolution spaces.
第9頁 i:ift勢換頁_ [(έ)ΕΪPage 9 i: ift page change _ [(έ) ΕΪ
為了克服外圍字型的缺點,本發明人曾經提出採用圖 樣為基礎的字型產生方法並提出專利申請(美國專利申請 案號09/425, 449,申請日期為公元一九九九年十月二十二 曰),在該案中,一個圖樣是指構成亞細亞文字之個別封 閉形狀的單元,而每個亞細亞文字包含一個或多個圖樣, 且每個圖樣是由一個或多個筆晝所組成,在圖樣基礎之字 型裏,每個文字字型是用各別不同組成之基本圖樣來表 現。圖樣基礎字型具有低容量記憶空間以及可在高解像及 低解像空間具有相容之顯像特質,因此圖樣字型非常適合 於不同解像度系統之間供作字型傳訊使用。 因此,在字型傳訊的應用上就可採用圖樣字型適合字 型傳訊的優點,其需求就是要有一套系統,方法,以及電 腦可讀取的媒介裝置,可提供自動轉換文件被擷取之任何 外圍字型成為一圖樣字型,使得該文件可以在不同解像度 之系統裝置間自由傳遞高品質的字型,不致於產生從高解 像度系統裝置傳遞字型資料至低解像度系統裝置時會產生 品質變差之現象,反之亦然。In order to overcome the shortcomings of peripheral fonts, the inventor has proposed to use a pattern-based font generation method and filed a patent application (U.S. Patent Application No. 09/425, 449, the application date is October 2, 1999. (12th), in this case, a pattern refers to the individual closed-shaped units that make up the Asian script, and each Asian script contains one or more patterns, and each pattern is composed of one or more pens In the pattern-based glyph, each text glyph is represented by a basic pattern with a different composition. The pattern basic font has low-capacity memory space and compatible development characteristics in high-resolution and low-resolution spaces. Therefore, the pattern font is very suitable for font communication between different resolution systems. Therefore, in the application of font messaging, the advantages of graphic fonts suitable for font messaging can be adopted. The requirement is to have a system, method, and computer-readable media device that can automatically convert documents to be retrieved. Any peripheral font becomes a pattern font, so that the document can freely transfer high-quality fonts between system devices of different resolutions, and will not produce quality when transferring font data from high-resolution system devices to low-resolution system devices The phenomenon of deterioration and vice versa.
發明總結 本發明提供一套系統、方法,以及電腦可讀取之媒介 裝置,用以定義一組基本圖樣,本發明更進一步地提供一 種方法以及電腦可讀取之媒介裝置具有電腦可執行之指 令,可供自動轉換外圍字型資料成為預先定義之基本圖樣Summary of the Invention The present invention provides a system, method, and computer-readable media device for defining a set of basic patterns. The invention further provides a method and computer-readable media device with computer-executable instructions. , Which can automatically convert peripheral font data into a predefined basic pattern
第10頁Page 10
科奸突貝 的圖樣字型資料。據本發明所轉換出來之圖樣字型資料僅 需相對少之記憶量,適合於在不同解像度之各個不同系統 之間作字型傳訊的應用,不管在高解像度或低解像度之不 同系統的輸出儀上顯像時皆可獲致高品質的顯像字型。 本發明的方法首先定義一組基本圖樣集,然後再用它 們來定義文字的字型集,使得被定義的這些字型適合於在 低或高解像度之輸出裝置上顯像。定義一組基本圖樣集的 方法是從一組文字集中挑選相似形狀的圖樣,然後對每一 被挑選的圖樣逐一定義,每個基本圖樣的定義方式皆一 樣,首先在基本圖樣的外圍邊線定義特徵點,然後再根據 這些被定義的特徵點來定義關鍵點及寬度值。通常,關鍵 點是選在邊緣端點上,或者是形狀彎曲部分的内側或外 侧,或者是主幹中心線通過其他關鍵點之水平或垂直線的 交點上,關鍵點是基本圖樣的第一項外顯參數,其優點是 可藉著移動關鍵點的位置來改變圖樣之外圍形狀。另外, 寬度值也是在基本圖樣之定義裡被採用為另一項的外顯參 數,也具有可藉著改變寬度值來改變圖樣之外圍形狀的優 點。本方法於是再根據這些關鍵點及寬度值反推計算特徵 點位置的方程式,以及再根據特徵點推算曲線形狀之細分 結構裡各個節點上的曲線比例值。曲線比例值乃係根據不 同解像度來分層儲存,使得藉著對不同層級的資料擷取, 可在二個相鄰特徵點間產生不同解像度需求之曲線段,愈 高解像度需求包含愈細密層級的資料,反之,愈低解像度Pattern font information of Kelu Tube. The pattern font data converted according to the present invention only needs a relatively small amount of memory, and is suitable for applications in which fonts are communicated between different systems of different resolutions, regardless of the output instrument of different systems with high or low resolution. You can get high-quality development fonts when developing. The method of the present invention first defines a set of basic patterns, and then uses them to define a font set of text, so that the defined fonts are suitable for displaying on a low or high resolution output device. The method of defining a set of basic patterns is to select similarly shaped patterns from a set of text sets, and then define each selected pattern one by one. Each basic pattern is defined in the same way. First, define features on the outer edges of the basic pattern. Points, and then define keypoints and width values based on these defined feature points. In general, the key points are selected at the edge endpoints, either inside or outside of the curved part of the shape, or at the intersection of the horizontal or vertical line of the trunk center line through other key points. The key points are outside the first item of the basic pattern Display parameters have the advantage that the shape of the periphery of the pattern can be changed by moving the positions of key points. In addition, the width value is also adopted as another explicit parameter in the definition of the basic pattern. It also has the advantage of changing the outer shape of the pattern by changing the width value. The method then inversely calculates the equations for calculating the position of the feature points based on these key points and width values, and then calculates the curve proportion values at each node in the subdivision structure of the curve shape based on the feature points. The curve ratio value is stored in layers according to different resolutions, so that through the acquisition of data at different levels, curve segments with different resolution requirements can be generated between two adjacent feature points. The higher the resolution requirement, the finer the level Data, conversely, the lower the resolution
第11頁 恕^?7 n ' :_ &)L· i 僅需低層級之曲線比例值的資料,本方法藉此機制來產生 與解像度無關之高品質曲線段。本方法反覆使用上述步驟 來定義基本圖樣,使得在文字集的一些相似圖樣,至少皆 能有一個基本圖樣相互對應。Page 11 ^^ 7 n ': _ &) L · i only needs the data of the curve scale value of the lower level, this method uses this mechanism to generate high-quality curve segments that have nothing to do with the resolution. This method repeatedly uses the above steps to define basic patterns, so that at least one basic pattern in the text set can correspond to each other.
根據本發明的一項觀點,本方法在確定每個圖樣外圍 之特徵點的步驟係採自動偵測的程序,該程序先將圖樣固 定在一個極大極小的方框裡,然後確定最左、最右、最 上、最下的切線點為特徵點,決定該些特徵點之後再偵測 其他特徵點的存在,若該些特徵點相鄰之間的曲線段不是 落在相鄰特徵點之連線的同侧,則表示至少有一反曲點的 特徵點需被確定,如果曲線段是在同側則偵查該曲線段上 是否有尖角點的存在可被確定為特徵點,當上述這些特徵 點被決定之後再確定是否該些新的各組相鄰特徵點間之極 大極小方框皆已包含期間之曲線段,若尚未能包含其間曲 線段,則在該曲線段上最左、最右、最上、最下的切線點 處確定特徵點。According to an aspect of the present invention, the step of determining the feature points on the periphery of each pattern of the method adopts a procedure of automatic detection. This procedure first fixes the pattern in a large and small box, and then determines the leftmost, most The right, top, and bottom tangent points are feature points. After determining these feature points, the existence of other feature points is detected. If the curve segments between the feature points are not connected by adjacent feature points, On the same side, it means that at least one characteristic point of the inflection point needs to be determined. If the curve segment is on the same side, it is detected whether there is a sharp corner point on the curve segment. It can be determined as the characteristic point. After the decision is made, it is determined whether the new minimum groups of adjacent feature points have included the period curve segments. If the interval curve segments have not been included, then the curve segment is the leftmost and rightmost. The feature points are determined at the upper, lower, and tangent points.
根據本發明的另一項觀點,基本圖樣係被切割成至少 包含一個單跑段掃瞄線區域,一個單跑段掃瞄線區域係指 在基本圖樣裡的一個沒有空洞的實體部分,該實體部分可 被單跑段掃瞄線填滿,將基本圖樣儲存成單跑段掃瞄線區 域之數據結構的優點,乃是可藉此結構來簡化基本圖樣的 塗黑動作。According to another aspect of the present invention, the basic pattern is cut to include at least one single-run scanning line area. A single-run scanning line area refers to a solid part without holes in the basic pattern. The entity Part of it can be filled by the scanning line of the single run, and the basic pattern can be saved as the data structure of the scanning line area of the single run. The advantage is that the structure can be used to simplify the blacking of the basic pattern.
第12頁Page 12
根據本發明的又一項觀點,用以定義基本圖樣之關鍵 點可以被標示來固定其點矩陣網點區隔上的特定位置,利 用這種標示當作暗示情報,可以微調基本圖樣的細部形 狀,使得字型在低解像度空間上顯像時可避免字型組成圖 樣間推擠在一起的情況發生。According to another aspect of the present invention, the key points used to define the basic pattern can be marked to fix a specific position on the dot matrix dot segment of the dot. Using this marking as hinting information, the detailed shape of the basic pattern can be fine-tuned. When the fonts are displayed on the low-resolution space, the situation that the font composition patterns are pushed together can be avoided.
根據本發明再一項觀點,本發明更進一步地提供一種 方法以及電腦可讀取之媒介裝置具有電腦可執行之指令, 採用一組預先被儲存於基本圖樣資料庫之預先被定義之基 本圖樣集的參數資料,來進行自動轉換外圍字型資料成為 以該些參數定義之圖樣字型資料。如上所述,每個基本圖 樣是被定義成為可產生圖樣形狀輸出的一個程式,其過程 是基於每個基本圖樣是由其外圍形狀上的特徵點為其形狀 輸出的指紋,根據該指紋所確定之特徵點位置,利用預製 預存之劇本句子的解譯來推算關鍵點及寬度值。所以字型 資料之自動轉換的方法係首先去擷取外圍字型的形狀資 料,從擷取字型中逐一以其組合圖樣與該外圍字型的形狀 比對進行將外圍字型資料自動轉換成基本圖樣之參數資 料,首先從外圍字型資料中逐一找出其構成圖樣之特徵點 及其指紋的數據結構,其次從基本圖樣資料庫找尋可匹配 該圖樣指紋之基本圖樣,再根據定義該基本圖樣之參數的 劇本句子解譯如何以特徵點來推算出該圖樣之關鍵點及寬 度值,並將推算出之關鍵點及寬度值儲存於預製之程式According to yet another aspect of the present invention, the present invention further provides a method and a computer-readable medium device having computer-executable instructions using a set of pre-defined basic pattern sets stored in a basic pattern database in advance. Parameter data to automatically convert peripheral font data into pattern font data defined by those parameters. As mentioned above, each basic pattern is defined as a program that can produce a pattern shape output. The process is based on the fingerprint that each basic pattern outputs from its feature points on its peripheral shape for its shape. For the feature point position, the interpretation of pre-stored script sentences is used to estimate the key points and width values. Therefore, the method of automatic conversion of font data is to first extract the shape data of the peripheral fonts, and then automatically convert the peripheral font data to the shape of the combined font and the shape of the peripheral fonts one by one. The basic pattern parameter data, first find out the feature points and fingerprint data structure of the pattern one by one from the peripheral font data, then find the basic pattern that can match the fingerprint of the pattern from the basic pattern database, and then define the basic pattern according to the definition The script sentence interpretation of the pattern parameters explains how to use the feature points to calculate the key points and width values of the pattern, and store the calculated key points and width values in the pre-made program.
第13頁 裡,可藉由該程式之執行來提供該圖樣形狀的輸出,此時 就已經將對應該圖樣之外圍字型資料部分轉換成重新定義 之基本圖樣參數值的圖樣字型資料,反覆上述步驟一直到 該字型資料已經完全轉換成圖樣字型資料為止。 根據本發明之其中一項觀點,當從基本圖樣資料庫搜 尋其中之一基本圖樣來匹配字型資料的圖樣,係指該基本 圖樣之指紋和該字型資料之圖樣形狀的指紋在拓樸結構上 是完全一致的。 根據本發明之又再一項觀點,從事顯像輸出基本圖樣 的程式包含有可從基本圖樣的關鍵點及寬度值決定特徵點 之方程式,該程式更進一步地包含可從每組相鄰兩特徵點 間預製預存於曲線層級表中之曲線比例值來產生該兩特徵 點間的曲線形狀,當該程式在建造某一被選圖樣之顯像 時,該程式包含比對預製預存於基本圖樣之曲線比例值來 匹配該被選圖樣之拓樸形狀。 根據本發明尚且另一項觀點,每個基本圖樣之顯像程 式更進一步地包含一塗黑演算法從事填滿至少一個單跑段 掃瞄線區域,當建造被選圖樣之顯像時,該程式更進一步 地包含擷取其所對應基本圖樣之預製預存塗黑演算法的步 驟0On page 13, the output of the pattern shape can be provided by the execution of the program. At this time, the part of the external font data corresponding to the pattern has been converted into the pattern font data of the redefined basic pattern parameter values. The above steps are continued until the font data has been completely converted into pattern font data. According to one aspect of the present invention, when one of the basic patterns is searched from the basic pattern database to match the pattern of the font data, it means that the fingerprint of the basic pattern and the fingerprint of the shape of the font data are in a topological structure. It is exactly the same. According to yet another aspect of the present invention, a program for developing a basic pattern for development includes an equation for determining feature points from a key point and a width value of the basic pattern. The program further includes two adjacent features from each group. Pre-points are pre-stored in the curve ratio value in the curve level table to generate the curve shape between the two characteristic points. When the program is building an image of a selected pattern, the program contains a comparison of pre-stored pre-stored in the basic pattern. Curve scale value to match the topology of the selected pattern. According to yet another aspect of the present invention, the imaging program for each basic pattern further includes a blacking algorithm to fill at least one single-run scanning line area. When constructing the imaging of the selected pattern, the The program further includes the step of capturing a pre-stored pre-stored blackout algorithm of the corresponding basic pattern.
i231457 IWIC3T ‘明說明—&) 綜上所述,本發明提供一套系統、方法、以及電腦可 讀媒介裝置用以自動轉換外圍字型成為以一組預先定義基 本圖樣為基礎的圖樣字型,本發明之圖樣字型係以圖樣組 合字型,而每個圖樣係由一對應之基本圖樣所定義,因而 大幅降低字型資料的儲存空間,滿足相對少量之儲存記憶 空間的需求,適合於在不同低高解像度之不同輸出裝置間 從事字型傳遞的使用,且由於本發明係在高解像度空間上 定義細密曲線段的層級結構,因此本發明亦可以兼具產生 高品質字型於高解像度及低解像度的字型顯像。 優先實施例之詳細描述 圖一(1 )示出本發明方法可以實現的一個標準計算 環境,其中有一個源頭計算機1,包含一個顯示螢幕2再加 上一個輸出裝置3就可以來實現本發明的方法。源頭計算 機1從一源頭文件4接收到以外圍字型表現的文字資料,源 頭文字4可能經由掃瞄放入源頭計算機1,或者從一磁碟放 入,或者其他適當的裝置方法放入,也可由源頭計算機本 身來產生,源頭計算機1於是將源頭文件4自動轉換成輸出 文件6,輸出文件6係以圖樣字型表現之原文件的文字資 料,輸出文件6適合於有效的文件傳輸使用,例如透過網 路7至一接收計算機8於其輸出裝置9上來顯像輸出,終端 接收計算機8之輸出裝置可能是高解像度的或低解像度 的,它可能和源頭計算機1之顯像裝置2或輸出裝置3具有 不同的解像度,仍然可使得輸出文件6具有高品質外觀的i231457 IWIC3T 'Explanation— &) In summary, the present invention provides a system, method, and computer-readable media device for automatically converting peripheral fonts into pattern fonts based on a set of predefined basic patterns The pattern font of the present invention is a combination of patterns, and each pattern is defined by a corresponding basic pattern. Therefore, the storage space of the font data is greatly reduced, and a relatively small amount of storage memory space is required, which is suitable for The use of font transfer is performed between different output devices with different low and high resolutions, and because the present invention defines a hierarchical structure of fine curve segments on a high resolution space, the present invention can also produce high quality fonts at high resolutions And low-resolution fonts. Detailed description of the preferred embodiment FIG. 1 (1) shows a standard computing environment that can be implemented by the method of the present invention, which includes a source computer 1, including a display screen 2 and an output device 3 to implement the present invention. method. The source computer 1 receives text data in a peripheral font from a source file 4. The source text 4 may be placed in the source computer 1 by scanning, or from a magnetic disk, or other appropriate device method. It can be generated by the source computer itself. The source computer 1 then automatically converts the source file 4 into an output file 6. The output file 6 is the textual data of the original file represented in a pattern font. The output file 6 is suitable for efficient file transmission. Through the network 7 to a receiving computer 8 to develop output on its output device 9, the output device of the terminal receiving computer 8 may be high-resolution or low-resolution, and it may be the same as the display device 2 or output device of the source computer 1. 3 has different resolutions, which can still make output file 6 with high quality appearance
第15頁 s[231457 _ 顯像。 源頭計算機1以及終端接收計算機可能是一個或多個 通用計算裝置,包含個人電腦、工作站、迷你電腦、主機 電腦、多處理器系統、以及其他類似者,網路7可能是直 接聯結式網路、區域網路、廣域網路、或者如網際網路之 全球^網路。P.15 s [231457 _ develop. The source computer 1 and the terminal receiving computer may be one or more general-purpose computing devices, including personal computers, workstations, mini computers, host computers, multi-processor systems, and the like, and the network 7 may be a directly connected network, Local area network, wide area network, or global ^ network such as the Internet.
圖一(2 )示出源頭計算機1之一些主要元件,源頭計 算機1包含一處理器10、一系統記憶體11、一輸入裝置 12、一輸出裝置3以及一外接之介面鍵盤、滑鼠、光筆、 數位板、掃瞄儀或者習知本技藝之可用以控制或登錄數據 於源頭計算機1者。在圖一(1 )所示之輸出裝置3為一印 出裝置,可能是一螢幕顯示器、繪圖儀或者本技藝習知之 其他裝置,外接介面13提供源頭計算機1可藉由它和外界 溝通,它可能是一個或多個調制解調器以及一網路介面 卡。Figure 1 (2) shows some main components of the source computer 1. The source computer 1 includes a processor 10, a system memory 11, an input device 12, an output device 3, and an external interface keyboard, mouse, and light pen. , Tablet, scanner, or those skilled in the art that can be used to control or log data to the source computer 1. The output device 3 shown in FIG. 1 (1) is a printing device, which may be a screen display, a plotter, or other devices known in the art. The external interface 13 provides a source computer 1 which can communicate with the outside world through it. It could be one or more modems and a network interface card.
系統記憶體11可能包含一個或多個只讀記憶體、一隨 機出入記憶體以及如硬碟、磁帶、光碟機等之永久儲存裝 置,系統記憶體14儲存一定義基本圖樣之程式14,被定義 之基本圖樣被儲存於一基本圖樣資料庫15,透過轉換程式 16產生之圖樣基礎字型儲存於一圖樣基礎字型資料庫17, 轉換程式16更進一步耦合至一字典18,該字典儲存將外圍The system memory 11 may include one or more read-only memory, a random access memory and a permanent storage device such as a hard disk, a magnetic tape, an optical disc drive, etc. The system memory 14 stores a program 14 defining a basic pattern, which is defined The basic pattern is stored in a basic pattern database 15, and the pattern basic font generated by the conversion program 16 is stored in a pattern basic font database 17. The conversion program 16 is further coupled to a dictionary 18, which stores the peripheral data.
第16頁 畫负t#頁_P.16 Draw negative t # page_
字型轉換呈圖樣字型的資料,該資料包含每個字型係由那 些基本圖樣所構成,有關這方面的描述在圖四(2 )將做 更詳細的說明。轉換程式16擷取一定義特徵點之程式19, 該程式用以自動確認沿圖樣外圍上之特徵點,其情形將於 圖四(2 )和圖四(3 )詳細說明,特徵點定義程式19亦可 由基本圖樣定義程式14來呼叫,系統記憶體11亦可儲存一 圖樣基礎字型顯像程式20,該程式可在高解像度和低解像 度顯示高品質之輸出字型,並耦合至圖樣基礎字型資料庫 17。程式14、16、19、20可於源頭計算機1上實現成電腦 可執行指令之程式模組,一般而言,程式模組包含一些常 規式、程式、物件、元件、數據結構等用以運作特殊任務 或特殊數據型態,對熟悉本發明技藝的人士而言,系統記 憶體11裡之任何元件亦可以被分散儲存於一個或多個的遠 端電腦,並可由源頭計算機1透過如網路7之傳輸媒介來存 取運作。The font conversion presents the data of the pattern font, which contains the basic patterns of each font system. The description of this aspect will be explained in more detail in Figure 4 (2). The conversion program 16 captures a program 19 defining feature points. This program is used to automatically identify the feature points along the periphery of the pattern. The situation will be described in detail in Figures 4 (2) and 4 (3). Feature Point Definition Program 19 It can also be called by the basic pattern definition program 14. The system memory 11 can also store a pattern basic font display program 20, which can display high-quality output fonts at high and low resolutions, and is coupled to the pattern basic words. Type database 17. Programs 14, 16, 19, and 20 can be implemented on the source computer 1 as program modules for computer-executable instructions. In general, program modules include some conventional formulas, programs, objects, components, data structures, etc. to operate special Tasks or special data types. For those familiar with the technology of the present invention, any element in the system memory 11 can also be stored in one or more remote computers, and can be transmitted by the source computer 1 through the network 7 Media to access operations.
圖二示出實施本發明之基本步驟,第一個步驟如方塊 70所示係定義有限個「基本圖樣」,如發明背景所述,一 個圖樣係亞細亞文字的一組成單元,一個亞細亞文字包含 至少一個圖樣,每個圖樣至少包含一個筆晝,如圖三 (1 )所示,例如一個文字30就包含了三個圖樣:32,34 以及36,而在圖四(2)的一個文字40包含了三個圖樣: 42,44以及46,在圖四(3 )的一個文字50包含了三個圖 樣:52,54以及56,以及在圖三(4)的一個文字60包含Figure 2 shows the basic steps for implementing the present invention. The first step is to define a limited number of "basic patterns" as shown in block 70. As described in the background of the invention, a pattern is a constituent unit of an Asian script, and an Asian script contains at least A pattern, each pattern contains at least one pen day, as shown in Figure III (1), for example, a text 30 contains three patterns: 32, 34, and 36, and a text 40 in Figure IV (2) contains Three patterns: 42, 44, and 46. A text 50 in Figure 4 (3) contains three patterns: 52, 54, and 56, and a text 60 in Figure 3 (4) contains
第17頁 歡si) 了四個圖樣:6 2,6 4,6 6,以及6 8,可以注意到圖樣3 2, 44,52,以及66 (用白色表現的部分)都有相同的形狀結 構,它們係由一個猶微彎曲的斜筆畫3 2 a以及一個垂直筆 晝32b所構成的圖樣,這四個圖樣32,44,52以及66都可 以用同一個基本圖樣來定義,例如採用圖樣32為基礎來定 義,並使得其他三個圖樣係由基本圖樣的微調產生。據上 例可知亞洲語言的一個特徵是:有限個相似圖樣是一組文 字集所共同擁有,以及相似圖樣具有相同的形狀結構,例 如圖三(1 )〜(4 )所顯示的情況。所以若能藉著圖樣來 定義每一個文字,且相似圖樣可以從一個基本圖樣分別產 生,就可以顯著地降低字型資料量。且每個基本圖樣的參 數值可被一個程式解譯並藉以產生這些基本圖樣的形狀及 其顯像,簡而言之,該程式係根據參數值解釋一些計算的 過程,它包含產生基本圖樣的外圍形狀,以及包含一個演 算法來填滿基本圖樣的内部區域。有關定義基本圖樣的一 些步驟,將在圖四(1 )中詳細描述。 其次的步驟如方塊24所示係涉及採用如方塊22所示預 先定義之基本圖樣集將外圍字型資料自動轉換成圖樣基礎 字型資料,具體而言,每個外圍字型的文字資料係被重新 定義為"個或多個基本圖樣的文字資料’ 種自動轉換的 步驟將在圖四(2)中做更詳盡的描述。 最後如方塊26所示,當所有外圍字型資料完全被自動Page 17 Huansi) four patterns: 6 2, 6, 4, 6, 6, and 6 8. Note that patterns 3 2, 44, 52, and 66 (the parts shown in white) all have the same shape and structure. , They are a pattern composed of a slightly curved oblique stroke 3 2 a and a vertical pen day 32b. These four patterns 32, 44, 52, and 66 can be defined by the same basic pattern, for example, using pattern 32 Define as the basis, and make the other three patterns from the fine-tuning of the basic pattern. According to the above example, one characteristic of Asian languages is that a limited number of similar patterns are commonly owned by a group of character sets, and the similar patterns have the same shape and structure, as shown in Figures 3 (1) ~ (4). Therefore, if each text can be defined by patterns, and similar patterns can be generated separately from a basic pattern, the amount of font data can be significantly reduced. And the parameter value of each basic pattern can be interpreted by a program to generate the shape and display of these basic patterns. In short, the program explains some calculation processes according to the parameter values. It contains the The outer shape, and the inner area that contains an algorithm to fill the base pattern. Some steps for defining the basic pattern will be described in detail in Figure 4 (1). The next step, as shown in block 24, involves automatically converting the peripheral font data into the pattern basic font data using the pre-defined basic pattern set shown in block 22. Specifically, the text data of each peripheral font is Redefining the text data of "one or more basic patterns" This kind of automatic conversion steps will be described in more detail in Figure 4 (2). Finally, as shown in box 26, when all peripheral font data is completely
第18頁 ^231457_ 轉換成圖樣基礎字型資料之後,就可以根據這些被建造出 來的圖樣基礎字型資料產生字型的顯像,而且字型顯像的 品質是和輸出裝置的顯像度無關的,這種全解像度顯示的 步驟將在圖四(6)中加以描述。Page 18 ^ 231457_ After converting to the pattern basic font data, you can generate the font image based on the constructed pattern basic font data, and the quality of the font image is independent of the display device's visibility. The steps for this full resolution display will be described in Figure 4 (6).
圖四(1)示出定義一有限個數之「基本圖樣」的程 序,當作定義每個亞細亞文字之組成圖樣的主要元件,這 種程序主要係依圖一(2)所示之基本圖樣定義程式14來 運作,在運作過程中係藉由人為干預方式之人機介面從事 編柄1以定義每個基本圖樣。Figure 4 (1) shows the procedure for defining a limited number of "basic patterns" as the main element for defining the composition pattern of each Asian character. This procedure is mainly based on the basic pattern shown in Figure 1 (2) The program 14 is defined to operate. In the operation process, the man-machine interface 1 is used to edit the handle 1 to define each basic pattern.
首先參照方塊70,字型設計人員可用掃瞄的文字圖形 檔,或用筆晝基礎的文字圖形檔,當作一組欲被建造成圖 樣基礎的字型集,當然其他方式取得的字模也可以,只要 其影像或者外圍形狀可以被拿來用作圖形的匹配即可。一 組文字集係指相同的字型為主,如果此文字集為中文字, 則不同中文字型的例子有:手寫字體、草體、流體、宋 體、明體等等。字體設計人員可以掃瞄這些文字集,或者 可以使用筆晝基礎之字型集,筆畫基礎字型的定義方法是 描述在美國專利申請案號:08/717, 712 (申請日期:九月 二十日,一九九六年)以及美國專利申請案號: 5, 852, 488,該兩案皆是與本發明相同的讓渡人,簡而言 之,這個方法採用關鍵點及寬度值為其外顯資料,以及採 用特徵點及一個樹狀結構的曲線比例值當作隱含數據。外First refer to block 70. The font designer can use the scanned text graphics file or the pen-based text graphics file as a set of fonts to be built into the pattern. Of course, the fonts obtained in other ways can also be used. , As long as its image or peripheral shape can be used for matching graphics. A group of character sets refers to the same font type. If the character set is Chinese characters, examples of different Chinese character types are: handwritten font, cursive, fluid, Song, Ming, and so on. Type designers can scan these text sets, or they can use the pen-day-based glyph set. The definition of stroke-based glyphs is described in US Patent Application No. 08/717, 712 (application date: September 20 Japan, 1996) and US Patent Application No .: 5, 852, 488, both of which are the same assignees of the present invention. In short, this method uses key points and width values of Explicit data, as well as characteristic points and curve proportion values of a tree structure are used as hidden data. outer
第19頁 1231457Page 19 1231457
顯及隱含數據也將被用來定義在本發明上,所以從現有的 筆畫基礎文字集取得的資料,將更容易被用來定義本發明 的方法,因為我們可以直接利用其中部分相同的外顯及隱 含參數的資料。 當一組文字如方塊70所示被掃瞄進來或產生出來後, 則如方塊72所示,字型設計人員開始確認一組相似的圖樣 形狀,確定相似圖樣可藉由觀察一組被掃瞄進來的文字集 中去擷取,或者透過影像分析的技巧去自動擷取。 其次如方塊74所示,字型設計人員從相似圖樣中選擇 < 其中一個圖樣,當作具有代表性的圖樣,並據以用來定義 ^個基本圖樣’這個基本圖樣最能代表這些相似圖形的產 生,通常最常出現的圖樣被選為最具代表性的圖樣,這種 * 選取的步驟可以透過人工篩選或自動取得。例如在方塊72 所示圖三(1)〜(4)的文字中確定相似圖樣32,44,52 · 以及66,然後選圖樣32為基本圖樣來代表那些在方塊74的 圖樣,同理,可以選圖樣64為基本圖樣來代表相似圖樣42 及6 4的這些圖樣,其次如方塊7 6所示,那個在方塊圖7 4被 選擇的基本圖樣就被指定為第i個基本圖樣,本方法定義 | 第i個基本圖樣的數學表現是如下: G^FiUKj},{Wm},L),i=l,.........,n (整數) ♦ G=基本圖樣的形狀 i =基本圖樣的編號Explicit and implicit data will also be used to define the present invention, so the data obtained from the existing stroke basic text set will be more easily used to define the method of the present invention, because we can directly use some of the same foreign Data for explicit and implicit parameters. When a group of text is scanned in or generated as shown in box 70, then as shown in box 72, the font designer starts to confirm a group of similar pattern shapes, and determines that similar patterns can be scanned by observing a group Retrieve the incoming text set, or use image analysis techniques to automatically retrieve it. Secondly, as shown in block 74, the font designer selects one of the similar patterns as a representative pattern and uses it to define ^ basic patterns. This basic pattern can best represent these similar patterns. The most commonly occurring patterns are selected as the most representative patterns. This * selection step can be obtained manually or automatically. For example, determine the similar patterns 32, 44, 52, and 66 in the text in Figures 3 (1) to (4) shown in block 72, and then select pattern 32 as the basic pattern to represent those patterns in block 74. Similarly, you can Select pattern 64 as the basic pattern to represent these patterns of similar patterns 42 and 64. Secondly, as shown in block 76, the basic pattern selected in block diagram 74 is designated as the i-th basic pattern. This method defines | The mathematical representation of the i-th basic pattern is as follows: G ^ FiUKj}, {Wm}, L), i = l, ........., n (integer) ♦ G = shape i of the basic pattern = Number of basic pattern
第20頁Page 20
Fi =生成第i個基本圖樣形狀的演算法 K=關鍵點的位置; j =在第1個基本圖樣之關鍵點的編號 W=寬度值 m=在第i個基本圖樣之寬度值的編號 L=解像度的層級,L=〇, 1,____ (整數) 於一=每個基ί圖樣(G".........Gn)是被定義及儲存 於一私式,一旦參數值被決定後 (r基本圖「樣的外圍形狀顯示出來,況且 行的' 此項圖樣形狀係有-定的規則可遵 在第 始有方向== 二 義崎的特徵點,這些二用來定 (3 /示出如的處^里技<巧來自動選取特徵點,圖三 認出特徵點,這此從一圖樣外圍形狀中自動確 步驟可再計算機1之特徵點定義程式18Fi = algorithm for generating the shape of the i-th basic pattern K = position of key points; j = number of key points in the first basic pattern W = width value m = number of width values in the i-th basic pattern L = Level of resolution, L = 〇, 1, ____ (integer) Yu Yi = each base pattern (G " ......... Gn) is defined and stored in a private type, once the parameter value After being determined (r basic picture, the outer shape of the pattern is displayed, and the line's shape of this pattern has a -determined rule, which can be followed at the beginning of the direction == the characteristic points of the Yoshizaki, these two are used to determine (3 / As shown in the following example: <Qiaolai automatically selects feature points, Figure 3 recognizes the feature points. This automatically determines the steps from the outer shape of a pattern, and can be redefined by the feature point definition program 18 of the computer 1.
(圖一(2 ))裡來執行,該程式包含一組確認特徵點的 規則,如圖四(3 )方塊150所示,基本圖樣的圖形座標是 被設定在一個極小方框理,例如在圖六(1 )的一個圖樣 151是用一個極小方框152來逼近,其次在圖四(3)的方 塊153,本方法於最左、最下、最右、及最上之極小方框 的邊緣上確認四個主要特徵點,在圖六(1 )的四個主要 特徵點Pi〜p4就是這樣被確定的。 其次如方塊154所示,對於每一組相鄰的主要特徵 點,本方法再決定是否其間的曲線段皆在它們連線的同一 邊,參考方塊15 6,例如在圖六(1 ),我們可以看出在兩 個曲線段157a及157b並不位於連接點匕到匕線段的同一 邊,這時必須如方塊158所示,確定轉折點為特徵點,例 如在圖六(1 )的特徵點S4就是一個轉折點。 再如方塊156所示,如果所有曲線段都已經在兩個相 鄰特徵點連線的同一邊,或者已經如方塊158所示,所有 轉折點都已確認,則如方塊160所示,再決定它們之間是 否還有其他尖角點的存在,如圖六(1 )的$就是一個尖角 點,因此如方塊162所示,$就被選為是一個特徵點,最 後,在圖六(1 )的情形就會得到六個特徵點: (Pp $,P2, S4, P3,P4) 如方塊1 60所示,如果已經知道沒有任何尖角點存(Figure 1 (2)), the program contains a set of rules for identifying feature points, as shown in box 150 of Figure 4 (3). The graphic coordinates of the basic pattern are set in a minimal frame, such as in A pattern 151 in FIG. 6 (1) is approximated by a small box 152, followed by box 153 in FIG. 4 (3). The method is applied to the edges of the leftmost, bottom, right, and uppermost small boxes. The four main feature points are identified above, and the four main feature points Pi ~ p4 in Figure 6 (1) are thus determined. Secondly, as shown in block 154, for each set of adjacent main feature points, the method then determines whether the curve segments in between are on the same side where they are connected. Refer to block 15 6. For example, in Figure 6 (1), we It can be seen that the two curve segments 157a and 157b are not located on the same side of the connection point dagger to the dagger line segment. At this time, as shown in block 158, the turning point must be determined as the characteristic point. For example, the characteristic point S4 in FIG. 6 (1) is A turning point. As shown in block 156, if all curve segments are already on the same side where two adjacent feature points are connected, or if all the turning points have been confirmed as shown in block 158, then as shown in block 160, then determine them. Is there any other sharp point between them? As shown in Figure 6 (1), $ is a sharp point, so as shown in block 162, $ is selected as a feature point. Finally, in Figure 6 (1 ) Will get six feature points: (Pp $, P2, S4, P3, P4) As shown in box 1 60, if you already know that there are no sharp corner points stored
第22頁 在,€ 角點已被確定,則再如方塊163所示,做出以 兩個相鄰特徵點為對角端點的方框,再如方塊164所示,Page 22 Now, if the corner point has been determined, then as shown in block 163, make a box with two adjacent feature points as diagonal endpoints, and then as shown in block 164.
察看是否有任何其間的曲線段超出了此方框,果真有的 話,則如方塊166所示,擴展方框來完全包容所有的曲線 段,並確定此新擴展方框的最左、最下、最右、及最上邊 線可能相切的點為特徵點,如圖六(1 )的例子,方框 168a係由兩個相鄰特徵點S4以及Ρ3所組成,這時有部分曲線 段超出了方框168a,因此需要擴展方框168a至新的方框 168b,才可完全包容所有在此兩個特徵點間的曲線段,所 以就在其上確定了特徵點$,它係在方框168的最左邊線 上,其他特徵點S和33亦是透過相同的程序被確定出來 的。 方塊170是個可有可無的步驟,該步驟將在基本圖樣 外圍邊線上示出所有被確定的特徵點,如方塊172所示, 這時,字型設計人員就可觀察其正確性,如前所述,圖樣 的特徵點應當被定義在其外型有曲率變化處,如果有特徵 點沒有被適當地定義,則如方塊174所示,字型設計人員 可直接補增該特徵點。Check to see if there are any curve segments beyond this box. If so, expand the box as shown in box 166 to fully contain all the curve segments, and determine the left and bottom of the new extended box. The points that may be tangent to the rightmost, rightmost, and uppermost edges are the feature points. As shown in the example of Figure 6 (1), box 168a is composed of two adjacent feature points S4 and P3. At this time, some curve segments exceed the square. Box 168a, so you need to expand Box 168a to a new Box 168b to fully encompass all the curve segments between these two feature points, so the feature point $ is determined on it, which is in the box 168 On the far left line, other characteristic points S and 33 are also determined through the same procedure. Block 170 is an optional step. This step will show all identified feature points on the peripheral edge of the basic pattern, as shown in block 172. At this time, the font designer can observe its correctness, as previously described. It is stated that the feature points of the pattern should be defined where the shape has curvature changes. If any feature points are not properly defined, as shown in block 174, the font designer can directly add the feature points.
一旦在圖樣外框中的所有特徵點皆被適當地確認後, 就如方塊176所示,沿著外框的順序儲存所有被確定的特 徵點,例如在圖六(1 )中,本方法可能從h開始追跡,沿 著圖樣外框逆時針方向逐一儲存特徵點直至P4為止。這Once all the feature points in the outer frame of the pattern are properly identified, as shown in block 176, all identified feature points are stored in the order of the outer frame. For example, in Figure 6 (1), this method may Start tracking from h and store the feature points one by one counterclockwise along the outer frame of the pattern until P4. This
第23頁Page 23
時,在相鄰特徵點Pi和^之間就形成曲線段11 ;在相鄰特 徵點Si和82之間就形成曲線段12 ;其餘曲線段亦是這樣形 成的,當然連接任何一個圖樣外框的特徵點可依順時針或 逆時針方向來排列,圖六(3 )係以樹狀結構示出在圖六 (2)中的特徵點,這種樹狀結構的排列順序就構成該基 本圖樣外形的『圖樣指紋』,任何一個基本圖樣表示的所 有相似圖樣,皆和其對應的基本圖樣具有相同的圖樣指 紋,換言之,圖樣指紋確定基本圖樣及其所代表之相似圖 樣的外形。In this case, a curve segment 11 is formed between the adjacent feature points Pi and ^; a curve segment 12 is formed between the adjacent feature points Si and 82; the rest of the curve segments are also formed in this way, of course, connecting any frame of the pattern The feature points of can be arranged in a clockwise or counterclockwise direction. Figure 6 (3) shows the feature points in Figure 6 (2) as a tree structure. The arrangement order of this tree structure constitutes the basic pattern. The "pattern fingerprint" of the shape. All similar patterns represented by any one basic pattern have the same pattern fingerprint as the corresponding basic pattern. In other words, the pattern fingerprint determines the appearance of the basic pattern and the similar pattern it represents.
如圖四(1 )方塊79所示,每個基本圖樣之圖樣指紋 係被儲存起來,並提供進一步地定義基本圖樣使用。As shown in block 79 of Figure Four (1), the pattern fingerprint of each basic pattern is stored and provides further definition of basic pattern use.
其次如方塊8 0所示,本方法定義一些曲線比例值,用 以造出相鄰特徵點間的曲線段,曲線比例值的結構表示了 相鄰特徵點間的曲線段是如何在不同層級解像度的空間裡 被建造出來的。在本發明的具體化表現中,係採用貝茲曲 線產生的演算法來表現曲線段,以及採用曲線層級表格對 照不同顯像度之曲線比例值的列表來分辨曲線段的細密程 度。 圖七(1 )示出一個二次貝茲曲線的幾何,二次貝茲 曲線的產生,可由兩個特徵點h *F2及一個可移動之中間 控制點h所構成的三角形來表現,移動中間控制點込可以Secondly, as shown in block 80, this method defines some curve ratio values to create curve segments between adjacent feature points. The structure of the curve ratio values indicates how the curve segments between adjacent feature points are resolved at different levels. Built out of space. In the embodiment of the present invention, the algorithm generated by the Bezier curve is used to represent the curve segments, and the curve level table is used to distinguish the fineness of the curve segments using a list of curve ratio values for different visualizations. Figure 7 (1) shows the geometry of a quadratic Bezier curve. The generation of a quadratic Bezier curve can be represented by a triangle formed by two characteristic points h * F2 and a movable intermediate control point h. Control point
第24頁Page 24
取得最逼近欲得到的曲線,若Pi和1^的終點連線形成L2線 段,且L2的中點為A,則二次貝茲曲線是一條和Pi Fi以及 Pi Fi+1分別相切於Fi和Fi+1以及通過A的二次曲線’此處的一 條曲線段可用C(r)表示,其中nr”表示一種特定的曲線比 例值,其定義如下·· C(r)二(Cr (X),Cr (y) 其中(^(xXJxj/LJx)以及Get the curve that most closely approximates. If the end points of Pi and 1 ^ are connected to form the L2 line segment, and the midpoint of L2 is A, then the quadratic Bezier curve is a tangent to Pi Fi and Pi Fi + 1 respectively. And Fi + 1 and a quadratic curve through A. Here a curve segment can be represented by C (r), where nr ”represents a specific curve proportion value, which is defined as follows: C (r) 二 (Cr (X ), Cr (y) where (^ (xXJxj / LJx) and
Cr(y )二 LJyVLi (y) 因此,再參照圖七(3 ),那麼圖七(1 )所示的曲線 段就被定義為(;(1),係屬於在特徵點巧和Fi+1間之第0層級 解像度(L=0)的第一條曲線,其值於是被儲存在對應一個 樹狀結構之曲線層級列表的第π Οπ層級裡。 當第一條曲線段需要再細分,以求得更精細的逼近曲 線形狀,可在高解像度輸出裝置上取得更高品質的顯像 時,那就如圖七(2 )所示,將它再細分成二條線段,每 條曲線段也是用貝茲曲線來定義。如圖七(2 )的情形, 在Fi+i,Α以及可移動之中間控制點Ρ2所形成三角形裡,產 生一條被細分的二次貝茲曲線段,同理,此曲線段會和 Ρ2Α以及P2Fi+1*別相切,而且會通過P2*AFi+1線段終點的連 線終點B,這條曲線段就被定義為(;(2),A和匕間的曲線段 被定義為Cr(3),這二條曲線段Cr(2)和Cr(3)就被儲存在解Cr (y) two LJyVLi (y) Therefore, referring to Fig. 7 (3), the curve segment shown in Fig. 7 (1) is defined as (; (1), which belongs to the feature point Qiao and Fi + 1 The first curve of the 0th level resolution (L = 0) is then stored in the π Οππ level of the curve level list corresponding to a tree structure. When the first curve segment needs to be subdivided, When obtaining a more detailed approximation curve shape, when a higher-quality image can be obtained on a high-resolution output device, as shown in Figure 7 (2), it is subdivided into two line segments, and each curve segment is also used. Bezier curve is defined. As shown in Fig. 7 (2), in the triangle formed by Fi + i, Α and the movable middle control point P2, a subdivided quadratic Bezier curve segment is generated. Similarly, this The curve segment will be tangent to P2A and P2Fi + 1 *, and will pass the end point B of the end of the P2 * AFi + 1 line segment. This curve segment is defined as (; (2), the curve between A and D The segment is defined as Cr (3), and the two curve segments Cr (2) and Cr (3) are stored in the solution
第25頁 顆 _ 像度層級為n l’’的列表裡,如圖七(3 )所示,曲線段 Cr(2)可能在被細分成二條曲線段Cr(4)和Cr(5),Cr(4)和 CF(5)就被儲存在解像度層級為Π2Π的列表裡,因此,如圖 七(3 )所示,每條曲線段都會和其所屬解像度層級(L = 0, 1,2)發生關連。In the list of _ image level n l '' on page 25, as shown in Figure VII (3), the curve segment Cr (2) may be subdivided into two curve segments Cr (4) and Cr (5). Cr (4) and CF (5) are stored in a list with a resolution level of Π2Π. Therefore, as shown in Figure 7 (3), each curve segment will be associated with its resolution level (L = 0, 1, 2 ) Related.
藉著對不同解像度層級來定義曲線段的方式,使得本 發明在不必特別在意低或高解像度之輸出裝置的情況下, 皆能得到高品質的字型輸出。字型設計人員可以在兩個相 鄰特徵點間,細分成很多曲線段來定義該兩點間曲線段的 形成,愈細密的曲線段組織愈能提供高解像度的顯像用, 反之,愈上層組織的曲線段愈適合用來提供低解像度的顯 像。例如,參看圖七(3 )的情形,若解像度層級數L 是最上層),則僅是Cr(l)的曲線段會被擷取,若L 是ΠΓ’(中間層),則Cr(2)和Cr(3)的曲線段會被擷取,若 L是Π2Π(最下層),則不同被細分的Cr(3)和Cr(4)以及Cr(5) 的曲線段會被擷取。By defining the curve segments for different resolution levels, the present invention can obtain high-quality font output without having to pay particular attention to low or high resolution output devices. The font designer can subdivide into many curve segments between two adjacent feature points to define the formation of the curve segment between the two points. The finer the curve segment organization, the more it can provide high-resolution imaging, otherwise, the higher the layer The more curved sections of the tissue are more suitable for providing low-resolution imaging. For example, referring to the situation of Fig. 7 (3), if the resolution level L is the uppermost layer), only the curve segment of Cr (l) will be captured. If L is ΠΓ '(the middle layer), then Cr (2 ) And Cr (3) curve segments will be retrieved. If L is Π2Π (the lowest layer), then the different subdivided curve segments of Cr (3) and Cr (4) and Cr (5) will be retrieved.
圖四(4)示出如上所述之定義曲線段的流程,如方塊 180所示,本方法定義一條或多條二次貝茲曲線段,再如 方塊182所示,本方法指定每條曲線段所屬的曲線層級, 然後如方塊184所示,本方法決定是否被定義的曲線段已 經吻合了基本圖樣外形,如果沒有則如方塊186所示,本 方法繼續細分沒有吻合部分的曲線段,再回到方塊180用Figure 4 (4) shows the process of defining curve segments as described above. As shown in block 180, this method defines one or more quadratic Bezier curve segments. As shown in block 182, this method specifies each curve. The curve level to which the segment belongs. Then, as shown in block 184, the method determines whether the defined curve segment already conforms to the basic pattern shape. If not, as shown in block 186, the method continues to subdivide the curve segment that does not have a matching part, and then Back to block 180
第26頁 s羽頁_ 二次貝茲曲線技巧來定義被細分的曲線段,以及到方塊 184查看是否新建的細分曲線段可以吻合該部分外形,反 覆使用這樣的程序直到曲線段已能吻合基本圖樣外形。再 如方塊18 8所示,本方法儲存代表曲線段的所有曲線比例 值以及它們分別所屬的曲線層級於一個樹狀結構之曲線層 級列表裡,如方塊190所示,本方法再指定點矩陣尺寸的 範圍給每個曲線層級,代表每個層級數所管轄之點矩陣尺 寸的範圍,此觀點將在下面加以敘述。Page 26 s feather page _ Quadratic Bézier curve technique to define the subdivided curve segment, and go to block 184 to see if the newly created subdivision curve segment can fit the shape of the part, use this procedure repeatedly until the curve segment can match the basic Pattern appearance. As shown in block 18-8, the method stores all curve proportion values representing the curve segments and the curve levels to which they belong respectively in a tree-level curve level list. As shown in block 190, the method specifies the size of the point matrix. The range of is given to each curve level, and represents the range of the size of the point matrix governed by the number of levels. This point will be described below.
一旦所有的曲線段皆被定義後,就回到圖四(1 ), 如方塊87所述,本方法將基本圖樣切成一個或多個單跑段 掃瞄線區域,一個單跑段掃瞄區域是指在基本圖樣理的一 個實體區塊,因為它不含有空心,所以它可被單跑段掃瞄 線填滿。將基本圖樣切成單跑段掃瞄線區域可以簡化塗黑 填滿基本圖樣的工作,例如在圖八裡,就示出一個基本圖 樣的樣本194被切成四個單跑段掃瞄線區域,則區域1獨立 形成一個單跑段掃瞄線區域,它可以被箭頭196所示的單 跑段掃瞄線塗黑填滿,同樣地,圖樣中間部分1 94分成2和 3二個區域,分別可以被箭頭198 a及198 b的單跑段掃瞄 線塗黑填滿,因此,本方法僅需對每個被定義出來的單跑 段掃瞄線區域做出塗黑填滿的演算即可。 如圖四(5)方塊199所示,有關對每個單跑段掃瞄線 區域定義塗黑演算法,本方法首先如方塊200所示,指定Once all the curve segments have been defined, return to Figure Four (1). As described in block 87, this method cuts the basic pattern into one or more single-run scan line regions, and a single-run scan The area is a solid block in the basic pattern. Because it does not contain hollows, it can be filled by the single-scanning line. Cutting the basic pattern into the single-scanning scanning line area can simplify the work of filling the basic pattern with black. For example, in Figure 8, a sample of the basic pattern 194 is cut into four single-scanning scanning line areas. Then, area 1 independently forms a single-run scan line area, which can be filled with black by the single-run scan line shown by arrow 196. Similarly, the middle part of the pattern 1 94 is divided into 2 and 3 areas. It can be filled with black lines by the single-scanning lines of arrows 198 a and 198 b respectively. Therefore, this method only needs to make a black-filling calculation for each defined single-scanning line area. can. As shown in block 199 in Figure 4 (5), regarding the definition of the blackout algorithm for the scanning line area of each single run, the method firstly specifies the block as shown in block 200, specifying
第27頁Page 27
一個單跑段碼給每個區域,再如方塊202所示,本方法確 定區域左右二邊曲線段的號碼’例如在圖八的情形’區 域’’1”是由{左曲線段=20 3&,右曲線段=2031>}所定義,其中 曲線段230a和203b係如前述之圖四(1)方塊80所定義, 同理,區域’’2”是由{左曲線段=2 04&,右曲線段=2041)}所定 義,以此類推。藉著這樣的建構,則塗黑圖樣1 94來顯像 的工作就比較容易,因為這時只要依序來分別塗黑填滿每 個單跑段掃瞄線區域即可。A single run segment code is given to each area, and as shown by block 202, this method determines the number of the left and right curve segments of the area 'for example, in the case of FIG. ;, The right curve segment = 2031 >}, where the curve segments 230a and 203b are defined as the box 80 in Figure 4 (1) above. Similarly, the area `` 2 '' is defined by {left curve segment = 2 04 & , Right curve segment = 2041)}, and so on. With this construction, the work of blackening the pattern 194 for development is easier, because at this time, it is only necessary to blacken and fill each scanning line area in order.
其次,方塊2 0 5是個可有可無的步驟,本方法將各個 單跑段掃瞄線區域以及它們的左右曲線段顯示在螢幕上, 讓字型設計人員可以驗證,如方塊206所示,字型設計人 員決定某個區域是否適當地被定義,如果沒有則如方塊 208所示,字型設計人員透過輔助工具適當地將此區域定 義,然後如方塊210所示,本方法於是儲存這個區域的左 右曲線段資料,俾供塗黑演算法使用,如方塊211所示,Secondly, block 205 is an optional step. This method displays the scanning line area of each single running segment and their left and right curve segments on the screen for font designers to verify, as shown in block 206. The font designer determines whether an area is properly defined. If not, it is shown in block 208. The font designer appropriately defines this area through auxiliary tools, and then shown in block 210. The method then saves this area. The left and right curve segment data is used for the blackout algorithm, as shown in block 211.
本方法決定是否所有的單跑段掃瞄線區域皆已定義,如果 沒有則跳到方塊19 9,繼續定義另一個尚未被定義的單跑 段掃瞄線區域,提供塗黑演算法使用,如果已如方塊2 11 所示,所有的單跑段掃瞄線區域皆已被定義,本方法於是 回到圖四(1)的方塊81。 其次如圖四(1 )方塊82所示,本方法定義及儲存第i 個基本圖樣之「關鍵點」,參照圖五以知曉關鍵點&至1(4This method determines whether all the scan line areas of the single run have been defined. If not, skip to box 19 9 and continue to define another scan line area of the single run that has not yet been defined to provide a blackout algorithm. If As shown in block 2 11, all the scanning line areas of the single run have been defined. The method then returns to block 81 in FIG. 4 (1). Secondly, as shown in block 82 of FIG. 4 (1), the method defines and stores the “key points” of the i-th basic pattern. Refer to FIG. 5 to know the key points & to 1 (4
第28頁 -//r ,ΐ4ρ。'::_Page 28-// r, ΐ4ρ. ':: _
是怎樣被定義的。關鍵點是用來定位以及從事局部變化基 本圖樣之形狀用,因此關鍵點的選取是依照基本圖樣之外 觀,以及能表現相似圖樣的特點來定義,關鍵點可由字型 設計人員透過輔助工具來選取,或者可依據語言描述的軟 體工具加予定義。關鍵點可選在基本圖樣的内部、邊線上 面、或其外部,使得可藉著移動關鍵點來控制基本圖樣的 局部形狀,一般而言,關鍵點通常被選在基本圖樣的起點 或終點上,或者基本圖樣突然改變方向的局部位置上,使 得改變關鍵點的位置,可以造成增長、拉大或縮小,或者 彎曲基本圖樣之部分區域的形狀,藉此特點可以修正基本 圖樣的形狀來逼近吻合文字理得各個相似圖樣,因此一個 圖樣至少需要兩個關鍵點,有關於移動關鍵點來改變基本 圖樣之形狀的方法將在下面有更完整的敘述。 其次如方塊82所示,當某個關鍵點被標示為『暗示情 報』時,則該點會被強迫座落在網格的特定位置上,並作 為重新計算特徵點位置及圖樣之外圍形狀的基礎,簡而言 之,標示關鍵點為『暗示情報』具有避免文字字型扭曲在 一起的優點,有關暗示情報的詳細說明將在以下敘述。How it is defined. The key points are used to locate and engage in the shape of the basic pattern of local changes. Therefore, the selection of key points is defined according to the appearance of the basic pattern and the characteristics that can represent similar patterns. The key points can be selected by font designers through auxiliary tools. , Or it can be defined based on the software tools described in the language. The key points can be selected inside, above the border, or outside the basic pattern, so that the local shape of the basic pattern can be controlled by moving the key points. Generally speaking, the key points are usually selected at the starting point or end point of the basic pattern. Or the basic pattern suddenly changes the direction of the local position, so that changing the position of key points can cause growth, enlargement or reduction, or bend the shape of a part of the basic pattern. With this feature, the shape of the basic pattern can be modified to approximate the text. Each similar pattern is processed, so a pattern needs at least two key points. The method of moving the key points to change the shape of the basic pattern will be described more fully below. Secondly, as shown in block 82, when a key point is marked as "suggested intelligence", the point will be forced to be located at a specific position on the grid and used as a recalculation of the feature point position and the outer shape of the pattern. Basically, in short, marking the key point as "suggested intelligence" has the advantage of avoiding distorted text fonts. A detailed description of hinted information will be described below.
圖九(1 )和(2 )分別出示沒有標示及標示關鍵點的 差異,在圖九(1 )和(2 )的實現部分示出圖樣外形,而 斜線填充的部分示出圖形網格在螢幕顯像的部分,通常圖 形塗黑的部分若佔有網格50 %或以上者,此網格就被填Figures 9 (1) and (2) show the differences between the unmarked and marked key points, respectively. The implementation part of Figures 9 (1) and (2) shows the shape of the pattern, and the hatched part shows the graphic grid on the screen. The imaged part, usually the black part of the figure, if the grid is 50% or more, the grid is filled.
第29頁 成眷5¾百_ —_并,禅) ,在圖九(1 )的141a及141b是二個沒有標示的關鍵 點,這時中間部分會佔有二個網格而造成變粗的現象,這 使品質變差,因為其他部分皆只佔一個網格的粗細,乃形 成不協調的現象。 反之,如圖九(2 )所示,當141a和141b被標示後,On the 29th page, it is dependent on five hundred and fifty _ —_ and, Zen). In Figure 9 (1), 141a and 141b are two unmarked key points. At this time, the middle part will occupy two grids and cause a thickening phenomenon. This makes the quality worse, because the other parts only occupy the thickness of a grid, which is an inconsistent phenomenon. On the contrary, as shown in Figure 9 (2), when 141a and 141b are marked,
它們會自動地移動至其網格的中心點,使得中間部分只會 有一個網格的顯像,因此不會使品質變差。當然,上例將 標示關鍵點移至網格中心位置僅是使用暗示情報的一例, 其他選擇的使用亦可以,例如強迫關鍵點座落於網格邊緣 線的中點等。 字型設計人員可以指定從那個解像度層級以上才需要 處理暗示情報的動作,因為超過該解像度層級後是屬於高 解像度之顯像使用,就沒有必要再去處理微調動作。例如 若字型設計人員指定層級1是啟動暗示情報的層級,則等 於或低於該層級是屬於低解像度之顯像,必須去移動被標 示之關鍵點來產生暗示情報的處理以及顯現其效果。They will automatically move to the center point of their grid, so that only one grid will be visible in the middle, so the quality will not be deteriorated. Of course, moving the marked key points to the center of the grid in the above example is just one example of using hinted information. Other options can also be used, such as forcing the key point to be located at the midpoint of the edge of the grid. The font designer can specify that the action of hinting information is only needed from that resolution level onwards, because beyond this resolution level, it is used for high-resolution development, so there is no need to deal with fine-tuning. For example, if the font designer specifies that level 1 is the level to initiate hinting information, then if the level is lower than that level, it is a low-resolution display. The key points marked must be moved to generate the hinting information processing and show its effect.
其次如方塊83所示,本方法定義並儲存至少一個基本 圖樣的寬度值,例如圖五有四個寬度值W1至W4被定義出 來。定義寬度值可由字型設計人員透過輔助工具來設定, 或者可依據語言描述的軟體工具加以定義。同理,寬度值 是用來控制基本圖樣的局部形狀,因此也可以藉由其值的Secondly, as shown in block 83, the method defines and stores the width value of at least one basic pattern. For example, in Figure 5, four width values W1 to W4 are defined. The definition width value can be set by font designers through auxiliary tools, or can be defined according to software tools described by the language. Similarly, the width value is used to control the local shape of the basic pattern, so you can also use its value
第30頁Page 30
截释日咖:0 $Si 改變來逼近相似圖樣的局部形狀,有關控制基本圖樣之局 部形狀的方法將在以下有更詳細的說明。 一旦所有的關鍵點及寬度值都已經被定義後,就如方 塊84所示,本方法推導一些方程式,使得特徵點(在方塊 78定義)可由關鍵點及寬度值(在方塊82和方塊83分別定 義)推算出來,參照圖五,令^(丨& |,丨&丨)表示一個 =徵點在X-Y軸的座標值;κ =(丨&丨,丨&丨)表示一個關 的特徵點可藉由以下的方程式推算 出 來: 〜”工1 7 1 lx = 丨 K1 lx-l/2Wi ; IFi I y: 1 K! U (2) If2 Ix= IVU ; IF2 丨y=l K2 U (3) 丨 f3 1 x= 1 K4 lx-l/2W2; IF3 ly= 1 K4 |y + l/2W3 (4) 丨 F4 1 x = 1 K3 lx~l/2W2; IF4 ly= 1 K3 ly + l/3W4 (5) 1 F5 I X = 丨 K3 lx- 1/2W2; |F5 ly- 1 K3 ly-1/3W4 (6) 丨 f6 1 x= 1 K4 lx~l/2W2; If6 ly= 1 K4 ly-l/2W3 (7) IF7 lx= 丨 K1 lx- 1/2W2; IF7 ly=l u (8)Intercepted Japanese coffee: 0 $ Si changes to approximate the local shapes of similar patterns. The method of controlling the local shapes of basic patterns will be explained in more detail below. Once all keypoints and width values have been defined, as shown in block 84, this method derives some equations so that feature points (defined in block 78) can be determined by keypoints and width values (in blocks 82 and 83, respectively). Definition) Calculated, referring to Figure 5, let ^ (丨 & |, 丨 & 丨) represent the coordinate value of an = sign point on the XY axis; κ = (丨 & 丨, 丨 & 丨) represents a relationship The characteristic points of can be calculated by the following equations: ~ "Work 1 7 1 lx = 丨 K1 lx-l / 2Wi; IFi I y: 1 K! U (2) If2 Ix = IVU; IF2 丨 y = l K2 U (3) 丨 f3 1 x = 1 K4 lx-l / 2W2; IF3 ly = 1 K4 | y + l / 2W3 (4) 丨 F4 1 x = 1 K3 lx ~ l / 2W2; IF4 ly = 1 K3 ly + l / 3W4 (5) 1 F5 IX = 丨 K3 lx- 1 / 2W2; | F5 ly- 1 K3 ly-1 / 3W4 (6) 丨 f6 1 x = 1 K4 lx ~ l / 2W2; If6 ly = 1 K4 ly-l / 2W3 (7) IF7 lx = 丨 K1 lx- 1 / 2W2; IF7 ly = lu (8)
如方塊8 6所示「劇太a, jTL ^ . =如;=徵點來計算得到的==== 以用以示關鍵點1^以及寬度咖可 (9)1 lx'1/2( lFl lx+ lF? lx); |K! Iy= IFi U or |F7 |y 丨 K2 丨 χ= 1 F2 | χ ; 1 K2 1 y= 1F2 |y (10) 丨 K3 |x = l/2( 1 F4 1 丨 y- 丨 F5 1 y) x+ |F5 U); |K2 Iy= (11) 1 F4 | y-2/3( | F4 K4=交(垂線(1 κ3 1 x),線(f5, f6)) (12) W1= 1 p7 lx- 1 Fx |x (13) w2= 1 f5 lx- 1 f4 Ix (14) W3= 1 f3 ly- 1 F6 Iy (15) W4= I F4 y- IF5 |y (16) 其中垂線是通過括號内之點的一條垂直線,例如 通過丨As shown in box 86, `` Drama a, jTL ^. = Such as; = sign points calculated ==== to show the key point 1 ^ and the width of coffee (9) 1 lx'1 / 2 ( lFl lx + lF? lx); | K! Iy = IFi U or | F7 | y 丨 K2 丨 χ = 1 F2 | χ; 1 K2 1 y = 1F2 | y (10) 丨 K3 | x = l / 2 (1 F4 1 丨 y- 丨 F5 1 y) x + | F5 U); | K2 Iy = (11) 1 F4 | y-2 / 3 (| F4 K4 = cross (perpendicular line (1 κ3 1 x), line (f5, f6)) (12) W1 = 1 p7 lx- 1 Fxx (13) w2 = 1 f5 lx- 1 f4 Ix (14) W3 = 1 f3 ly- 1 F6 Iy (15) W4 = I F4 y- IF5 | y (16) where the vertical line is a vertical line passing through the point in parentheses, such as through 丨
第32頁Page 32
31457 : : _ 麟賴 以及 點之>垂直線,線是通過括號内之二點的一條直線 交是括號内二條直線的交點。 如方塊88所示,一旦所有的基本圖樣皆已被完整地 定義,則本方法將各個基本圖樣的定義資料儲存於基本圖 樣資料庫15 (圖一(2 )),例如第i個基本圖樣的號碼 (ID),其關鍵點、寬度值,從關鍵點及寬度值計算特徵點 的方程式、曲線比例值儲存成一個曲線層級列表,以及提 供輸出顯像之塗黑演算法的程序,就是那些定義第i個基 本圖樣的資料’而且每個基本圖樣的顯不程式(Gi )’基本 圖樣的圖樣指紋(儲存於圖四(1 )之方塊79 )以及劇本 句子(方塊86 )皆儲存於基本圖樣資料庫15裡。 其次如方塊9 0所示,本方法決定是否所有的基本圖樣 已經被完整定義並適當地被儲存成程式可執行的樣子,換 言之,本方法決定是否文字集裏所有的圖樣皆至少有一個 基本圖樣來代表其形狀,如果沒有則本方法跳回方塊72, 繼續如上所述的方式去定義另外的圖樣。 一旦所有的基本圖樣都已經被完整定義後,此時基本 圖樣資料庫包含了一組程式集可用以顯示所有基本圖樣的 形狀,其中每個基本圖樣的顯示程式是和其特殊之圖樣指 级以及劇本句子是關連在一起的,亦即基本圖樣資料庫包 含一組預先定義之基本圖樣集,可用於將外圍字型資料自31457 : _ Linlai and point > vertical line, the line is a straight line passing through the two points in the brackets and the intersection is the intersection of the two straight lines in the brackets. As shown in block 88, once all the basic patterns have been completely defined, the method stores the definition data of each basic pattern in the basic pattern database 15 (Figure 1 (2)), such as the i-th basic pattern. Number (ID), its key points, width values, equations for calculating feature points from key points and width values, curve ratio values are stored as a curve level list, and programs that provide blackout algorithms for output visualization are those definitions The data of the i-th basic pattern 'and the display program (Gi) of each basic pattern's pattern fingerprint (stored in box 79 of figure four (1)) and the script sentence (box 86) are stored in the basic pattern Library 15. Secondly, as shown in block 90, this method determines whether all the basic patterns have been completely defined and appropriately stored as executable programs. In other words, this method determines whether all the patterns in the text set have at least one basic pattern. To represent its shape, if not, the method jumps back to block 72 and continues to define another pattern in the manner described above. Once all the basic patterns have been fully defined, the basic pattern database contains a set of programs that can be used to display the shapes of all basic patterns. The display pattern of each basic pattern is related to its special pattern reference and The script sentences are related, that is, the basic pattern database contains a set of pre-defined basic pattern sets, which can be used to convert the external font data from
第33頁 T231457_ gif說日科26) 動轉換k圖樣基礎字型資料的方法上,此乃本發明揭露的 重點。Page 33 T231457_ gif said Japanese Section 26) This method is the key point of the present disclosure in the method of dynamically converting k-pattern basic font data.
圖四(2)示出圖一(2)之轉換程式16用於將外圍字 型資料轉換成圖樣基礎字型資料的一般步驟。首先如方塊 130所示,擷取定義外圍字型資料之以直線段以及曲線段 組成的外圍線段,這些被計算機1擷取之外圍字型資料可 以整批或分批載錄至計算機的系統記憶體11,再如方塊 132所示,從被擷取之字型檔中選其中一字逐一轉換,如 方塊134所示,比對該字典18 (圖一(2 ))的資訊,字典 18儲存的資訊包含字碼、字體、製造商、每字對應的基本 圖樣數目、每個圖樣之相對座標等,從該資訊中挑選對應 該字知其中一個基本圖樣來逐一和該字的部分形狀作比 對。Fig. 4 (2) shows the general steps of the conversion program 16 of Fig. 1 (2) for converting the peripheral font data into the pattern basic font data. First, as shown in block 130, the peripheral line segments composed of straight line segments and curved line segments that define the peripheral font data are retrieved. These peripheral font data retrieved by the computer 1 can be recorded into the computer's system memory in batches or in batches. Body 11, and then as shown in block 132, select one of the words from the extracted font file to convert one by one. As shown in block 134, compare the information of the dictionary 18 (Figure 1 (2)) with the dictionary 18 to store The information includes the code, font, manufacturer, the number of basic patterns corresponding to each character, the relative coordinates of each pattern, etc. From this information, select one of the basic patterns corresponding to the character to compare one by one with the partial shape of the character .
如方塊136所示,對於該被挑選之圖樣而言,就如圖 四(3)所示之特徵點定義程式19,首先被呼叫出來從事 確認該外圍字型部分形狀的特徵點,並將之轉換成圖樣的 特徵點及它們在圖樣指紋的結構,如前所述,特徵點定義 程式19包含一組規則,提供決定特徵點之取得方法、步驟 以及順序等運算規則,換言之,如圖四(3 )方塊170、 172以及174的步驟是由該組規則來自動運作的。 其次如方塊138所示,在方塊136被確認之特徵點於是As shown in block 136, for the selected pattern, the feature point definition program 19 shown in Figure 4 (3) is first called to confirm the feature points of the shape of the peripheral font, and The feature points converted into a pattern and their structure on the pattern fingerprint. As mentioned earlier, the feature point definition program 19 contains a set of rules that provide calculation rules that determine the method, steps, and order of obtaining feature points, in other words, as shown in Figure 4 ( 3) The steps of blocks 170, 172, and 174 are automated by the set of rules. Next, as indicated by block 138, the feature point identified at block 136 is then
第34頁Page 34
依其圖樣指紋之定義被建造起來,以提供進一步地運算該 圖樣在其對應基本圖樣之其他參數值的填補工作。如方塊 140所示,該圖樣指紋必須要吻合其對應基本圖樣所定義 之圖樣指紋,才能進行下一步驟之運算程序。It is constructed according to the definition of the fingerprint of the pattern to provide further operation for filling the pattern with other parameter values corresponding to the basic pattern. As shown in block 140, the pattern fingerprint must match the pattern fingerprint defined by its corresponding basic pattern before the next step of the calculation process.
如方塊142所示,一旦該基本圖樣指紋被確認之後, 該基本圖樣預先定義的劇本句子就被擷取出來,再採用如 方塊13 6所示之圖樣指紋的特徵點位置來計算關鍵點位置 以及寬度值,一旦該基本圖樣之所有被預先定義之關鍵點 及寬度值皆依預存之劇本句子被推算出來之後,其次就如 方塊144所示,該基本圖樣之編碼就指定給該被選圖樣,As shown in block 142, once the fingerprint of the basic pattern is confirmed, the pre-defined script sentence of the basic pattern is extracted, and then the feature point positions of the pattern fingerprint shown in block 13 6 are used to calculate the key point positions and Width value. Once all the pre-defined key points and width values of the basic pattern are calculated from the pre-stored script sentences, the next step is as shown in block 144. The encoding of the basic pattern is assigned to the selected pattern.
使得該圖樣字型之外顯資料為基本圖樣編碼關鍵點位置, 以及寬度值所組成。如方塊146所示,檢查是否所有組成 該外圍字型形狀之圖樣皆已如前所述步驟處理過,如果沒 有則再如上述步驟反覆啟動圖樣定義程式,直到所有外圍 資料皆已經被轉換成圖樣字型之外顯資料以及其隱含參數 值後才停止該程式,隱含參數值包含如何從關鍵點及寬度 值推算出特徵點之參數值,以及如何從特徵點再推算出它 們之間曲線形狀的曲線比例值。一旦所有這些組合圖樣的 外顯和隱含參數值都被決定後,就將這些資料儲存在圖樣 基礎字型資料庫17,在圖一(2 )裡當作該被選外圍字型 之文字的新字型資料。如方塊148所示,決定是否所有外 圍字型資料皆已被自動轉換成圖樣基礎字型資料,若尚未 完成則繼續對未被選外圍字型反覆做上述步驟,直到所有Make the pattern font external data composed of the key position of the basic pattern coding and the width value. As shown in block 146, check whether all the patterns that make up the peripheral font shape have been processed as described above. If not, then start the pattern definition program repeatedly as described above until all the peripheral data have been converted into patterns. The program is stopped only after the font's explicit data and its implicit parameter values. The implicit parameter values include how to calculate the parameter values of the feature points from the key points and width values, and how to calculate the curve between them from the feature points. The curve scale value of the shape. Once the explicit and implicit parameter values of all these combined patterns are determined, these data are stored in the pattern base font database 17 and used as the text of the selected peripheral font in Figure 1 (2). New font information. As shown in block 148, determine whether all the surrounding font data has been automatically converted into the pattern basic font data. If it has not been completed, continue to repeat the above steps for the unselected surrounding fonts until all
第35頁 外圍字型皆被自動轉換之處理後才停止轉換的程序。 一旦所有文字已經被自動轉換後,則圖樣基礎字型資 料17儲存了所有文字被轉換過的新字型資料,使得每個文 字可以透過一組預製預存之圖樣程式來顯示其形狀或者輸 出其塗黑的點矩陣。Page 35 The process of stopping the conversion after the peripheral fonts are automatically converted. Once all the text has been automatically converted, the pattern basic font data 17 stores the new font data that all text has been converted, so that each text can display its shape or output its paint through a set of pre-stored pattern programs. Black dot matrix.
圖四(6)示出依本發明之圖樣基礎字型資料產生文 字顯像的方法,字型顯像的過程正是上述自動轉換之建造 技術的逆向處理步驟。圖四(6 )的步驟係在圖樣基礎字 型顯像程式20裏來完成,該程式係在計算機1 (圖一 (2))裡採用儲存於圖樣基礎字型資料庫之外顯及隱含 參數值,逐一解譯適合於低或高解像度之高品質的顯像輸 出。如方塊300所示,本方法在一般應用程式,例如文字 處理的程式中,擷取其文字來顯像或印出時,就如方塊 302所示,本方法首先決定顯像裝置或印表機的解像度層 級,再如方塊304所示,從每個組成圖樣預存於曲線層級列 表中擷取對應之曲線比例值,它們剛好適合於該解像度層 級之印表機或顯像裝置的使用。Fig. 4 (6) shows a method for generating a character image according to the basic font data of the pattern of the present invention. The process of character image development is exactly the reverse processing step of the above-mentioned automatic conversion construction technology. The steps in Figure 4 (6) are completed in the graphics basic font development program 20, which is used in computer 1 (Figure 1 (2)) to store the implicit and explicit font database outside the graphics basic font database. Parameter values, one by one, are suitable for high-quality imaging output with low or high resolution. As shown in block 300, when the method captures the text for development or printing in a general application, such as a word processing program, as shown in block 302, the method first determines the developing device or printer As shown in block 304, the corresponding curve scale values are pre-stored in the curve level list of each component pattern, which are just suitable for the use of the printer or the display device of the resolution level.
如方塊3 0 6所示,本系統決定是否印表機或顯像裝置 的特殊顯像層級需要去啟動暗示情報的動作,如果不要去 啟動暗示動作,則如方塊308所示,圖樣之所有的關鍵點 乃依實際計算在顯像空間位置佈置各個組成圖樣,反之則As shown in block 3 06, the system determines whether the special development level of the printer or developing device needs to initiate the action of hinting information. If it is not necessary to initiate the action of hinting, as shown in block 308, all of the patterns The key point is to arrange each component pattern in the development space position according to the actual calculation, and vice versa
第36頁Page 36
如方塊31 0所示,被標示的關鍵點被強制移動至點矩陣網 格之特定位置上,而未標示的關鍵點則不被移動,因此如 方塊312所示,圖樣的顯像是依據這些最後得到之關鍵點 及寬度值來計算,計算的過程是從關鍵點及寬度值求出特 徵點的位置,再依據從曲線層級列表擷取的曲線比例值計 算相鄰特徵點間的曲線段,該曲線層級列表係依據特殊輸 出裝置的解像度製作,最後使用塗黑演算法將圖樣内部區 域填滿’以產生點矩陣的顯像^因為母個圖樣是被切割成 多個單跑段掃瞄線區域,塗黑演算法係將每個單跑段掃瞄 線區域做簡單填滿,例如某依點矩陣的網格被50 %以上的 單跑段掃瞄線區域蓋住,那麼此網格的點矩陣就被塗黑, 或者若此網格的中心點係落入單跑段掃瞄線區域内,那麼 此網格的點矩陣也會被塗黑。As shown in block 3 10, the marked key points are forcibly moved to a specific position in the point matrix grid, and the unmarked key points are not moved. Therefore, as shown in block 312, the pattern is developed according to these The key points and width values obtained at the end are calculated. The calculation process is to find the position of the feature points from the key points and the width values, and then calculate the curve segment between adjacent feature points according to the curve proportion value extracted from the curve level list. The curve level list is made based on the resolution of the special output device. Finally, a blackout algorithm is used to fill the inner area of the pattern 'to produce a dot matrix image. ^ Because the mother pattern is cut into multiple single-run scanning lines. Area, the blackout algorithm simply fills the scan line area of each single run. For example, the grid of a point matrix is covered by more than 50% of the scan line area of a single run. The point matrix is blackened, or if the center point of the grid falls within the scan line area of a single run, the point matrix of the grid is also blacked out.
本發明可藉由圖形人機介面(GUI)的電腦輔助設計 (CAD)工具來貫徹實施,以下描述視窗作業系統下貫徹本 發明之GUI工具的操作,但需瞭解本發明亦可用其他作業 系統來貫徹,並不僅限於一種作業系統來實施,而且本方 法藉由CAD工具來貫徹的方法,亦可藉由影像分析技術來 自動實施。 圖十示出本方法採用一種GUI工具定義特徵點,在圖 十中,一種GUI之CAD工具400包含一個標題、主欄、按鈕 以及二維顯示區402有一個圖樣406被顯示出來,字型設計The present invention can be implemented by a computer aided design (CAD) tool of a graphical human machine interface (GUI). The following describes the operation of the GUI tool of the present invention under a Windows operating system, but it should be understood that the present invention can also be implemented by other operating systems. The implementation is not limited to an operation system, and the method is implemented by a CAD tool, and can also be automatically implemented by image analysis technology. Figure 10 shows that this method uses a GUI tool to define feature points. In Figure 10, a GUI CAD tool 400 includes a title, main bar, buttons, and a two-dimensional display area 402. A pattern 406 is displayed. The font design
第37頁Page 37
人員從視窗系統的主搁選1’ Add Point” 408下達命令,並移 動十字游標以按鈕點在特徵點上,每個被按鈕點上的位置 就定義一個特徵點,例如顯示區402以”+’’表示的位置就是 被定義的特徵點。The person selects 1 'Add Point ”408 from the main window system, and moves the cross cursor to the button point on the feature point. Each position on the button point defines a feature point, for example, the display area 402 starts with“ + The position indicated by '' is the defined feature point.
圖十一(1)至(3)示出本方法在CAD工具上定義關鍵點 及寬度值,在圖十一(1)中,字型設計人員從CAD工具的主 搁選n Add Key11 410下達命令,並移動十字游標至各個欲被 定義為關鍵點的位置上,再以按钮確定該位置來定義一個 關鍵點,例如顯示區402裡以π+π表示的位置就是被定義的 關鍵點。 同理,在圖十一(2)中,字型設計人員選擇n Add Thicknessn412下達命令,產生一個如圖十一(3)所示的視 窗414,然後字型設計人員在顯示區402移動十字游標取二 點418 及420,貝丨J>^lfAdd Thi ckness” 視窗裡n New Thickness”欄盒41 6裡會顯示它們之間的距離,若按 下” 0K”則以此距離值定義寬度值。Figures 11 (1) to (3) show that the method defines key points and width values on the CAD tool. In Figure 11 (1), the font designer selects Add Key11 410 from the main selection of the CAD tool Command, and move the cross cursor to each position to be defined as a key point, and then use the button to determine the position to define a key point. For example, the position indicated by π + π in the display area 402 is the defined key point. Similarly, in Figure 11 (2), the font designer selects n Add Thicknessn 412 to issue a command to generate a window 414 as shown in Figure 11 (3), and then the font designer moves the cross cursor in the display area 402. Take the two points 418 and 420. In the “New Thickness” window in the “New Addness” box, the distance between them will be displayed. If you press “0K”, the width value will be defined by this distance value.
圖十二(1)至(5)示出字型設計人員如何在特徵點之間 定義曲線比例值,如圖十二(1)所示,字型設計人員先從 視窗42 0的1’ Curve Levelπ輸入曲線層級lf 0Π在曲線層級碼 的方塊422裡,然後如圖十二(2)所示,在特徵點ΡΑ和1\之 間產生一條曲線段及決定其形狀的中間控制點己,如圖所Figures twelve (1) to (5) show how the font designer defines the curve proportion value between the feature points. As shown in Figure twelve (1), the font designer first starts from the 1 'Curve of the window 42 0 Levelπ enters the curve level lf 0Π in block 422 of the curve level code, and then as shown in figure twelve (2), a curve segment is generated between the characteristic points PA and 1 \ and the intermediate control point that determines its shape, such as Map
第38頁Page 38
示,剛開始時其形狀是一直線段,因為PA&PB連線是在共 線上,因此π 0’’的起始設定讓PA和匕之間的曲線段是一條直 線段,此後,字型設計人員開始移動Pi來逼近PA和1\之間的 曲線,產生一條逼近想要得到之曲線形狀的一條二次曲 線,並記下如圖七(1)所示的 (XrXL^xVL^x),L^yVLJy))為層級π0’’ 的曲線比例 值。 如圖十二(3)所示,字型設計人員在曲線層級碼的方 塊422輸入層級’’ Γ’,開始定義層級’’ Γ’的曲線比例值,如 圖十二(4 )所示,想要得到的原始曲線形狀被分成二段,ΡΑ 和Pe以及Pe和匕二段,再如上所述在其間分別產生中間控制 點匕及已,並依上述步驟分別移動P2和?3用它們所決定的二 次曲線段分別逼近原始曲線形狀,並分別記下如圖七(2) 所示Cr(2)及Cr(3)為層級丨f 1Π的曲線比例值,依此反覆步 驟類推,於是產生愈高層級愈細密分段的各層級曲線比例 值。 如圖十二(5)所示,字型設計人員將每個層級所對應的解 像度範圍填入,字型設計人員使用n Add Level”視窗420裡 的解像度範圍顯示區432,在顯示區432列出曲線層級數以 及對應的解像度範圍,例如在圖十二(5)裡,字型設計人 員設定24x24點矩陣及其下尺寸係屬於曲線層級0,25x25 至64x64 尺寸之間的解像度屬於曲線層級1,通常,字型In the beginning, the shape is a straight line segment, because the PA & PB connection is on a collinear line, so the initial setting of π 0 '' makes the curve segment between PA and dagger a straight line segment. After that, the font design The person starts to move Pi to approach the curve between PA and 1 \, and generates a quadratic curve that approximates the shape of the curve that he wants, and writes down (XrXL ^ xVL ^ x) as shown in Figure 7 (1), L ^ yVLJy)) is the curve scale value of the level π0 ''. As shown in Fig. 12 (3), the font designer enters the level "Γ '" in block 422 of the curve level code, and starts to define the curve proportion value of the level "Γ', as shown in Fig. 12 (4). The original curve shape you want to get is divided into two sections, PA and Pe, and Pe and dagger. Then, as described above, the intermediate control points dagger and dagger are generated respectively, and P2 and? Are moved respectively according to the above steps. 3 Use the quadratic curve segments determined by them to approximate the original curve shape, respectively, and note down the curve proportion values of Cr (2) and Cr (3) as shown in Figure 7 (2), and repeat them accordingly. The steps are analogized, and then the curve proportion values of each level are divided into more detailed levels. As shown in Figure 12 (5), the font designer fills in the resolution range corresponding to each level. The font designer uses the resolution range display area 432 in the n Add Level window 420, and there are 432 columns in the display area. Show the number of curve levels and the corresponding resolution range. For example, in Figure 12 (5), the font designer sets the 24x24 point matrix and its lower dimensions to belong to the curve level 0. The resolution between the 25x25 to 64x64 dimensions belongs to the curve level 1. , Usually, font
第39頁Page 39
MllSl; ; _ 設計人員係根據其觀察曲線細分的資訊,決定如何設定曲 線層級數及對應的解像度尺寸範圍,例如若曲線段的細分 在低解像度上顯示所產生的效果一樣,則細分前的上一層 曲線資訊就已足夠來產生該解像度的顯示。 綜上所述,本發明提供一種方法與系統來定義一組基 本圖樣集,以及一種方法與電腦可讀取媒介裝置具有電腦 可執行指令採用預先定義之基本圖樣集將外圍字型資料自 動轉換成圖樣基礎字型資料。圖樣字型僅需相對少量數據 的技藝空間,非常適合於在各種不同解像度之裝置間做字 型傳遞的使用,由於本方法定義及產生曲級段係根據解像 度層級來擷取適當的數據使用,因此本方法不管在高和低 解像度裝置上使用,皆能產生高品質的字型輸出。 一旦本發明之優先實施例被揭露及詳細敘述後,很明 顯地,熟悉此項技藝的人士就不難根據本發明的精神及範 圍做出各項改變,來達成與本發明一樣的目的。MllSl;; _ The designer decides how to set the number of curve levels and the corresponding resolution size range according to the information of the observation curve subdivision. For example, if the subdivision of the curve segment shows the same effect on the low resolution display, the One layer of curve information is sufficient to produce a display of that resolution. In summary, the present invention provides a method and system for defining a set of basic patterns, and a method and a computer-readable medium device having computer-executable instructions for automatically converting peripheral font data into a predefined basic pattern set. Pattern basic font information. The pattern font requires only a relatively small amount of data in the technical space, which is very suitable for the use of font transfer between devices with various resolutions. Because the method defines and generates a curved segment, the appropriate data is retrieved according to the resolution level. Therefore, no matter it is used on high and low resolution devices, this method can produce high quality font output. Once the preferred embodiment of the present invention is disclosed and described in detail, it will be apparent that those skilled in the art will not have to make various changes in accordance with the spirit and scope of the present invention to achieve the same purpose as the present invention.
第40頁Page 40
TFT mx 圖一(1 )係運作本發明之一般環境的示意圖 圖一(2)係實現本發明之計算機内一些基本元件的功能方 塊圖。 〔圖號說明〕 1---源頭計算機; 2---顯像螢幕; 3, 9---廣義印表機(輸出裝置); 4,6---外圍字型之文字文件; 5---廣義記憶裝置;7---網路; 8——接收端計算機 圖二係本發明方法之三個基本步驟的流程圖。 圖三(1)〜(4)示出一些中文文字的樣本,其中每個文字包 含複數個圖樣以及每個圖樣包含複數個筆晝。 〔圖號說明〕 30--_中文字;32,34,36---中文字30的組成圖樣 32a,32b 筆晝部分;42, 44, 46——中文字40的組成圖 樣;52, 54, 56---中文字50的組成圖樣; 62, 64, 66, 68 中文字60的組成圖樣 圖四(1)〜(5)是本發明在定義基本圖樣之方法的流程圖以 及根據被預先定義之基本圖樣來從事將外圍字型資料自動 轉換成圖樣基礎字型資料之方法的流程圖。TFT mx Figure 1 (1) is a schematic diagram of the general environment in which the present invention operates. Figure 1 (2) is a functional block diagram of some basic elements in a computer that implements the present invention. [Illustration of drawing number] 1 --- source computer; 2--developing screen; 3, 9 --- generalized printer (output device); 4,6 --- text file of peripheral font; 5- -Generalized memory device; 7 --- Network; 8-The computer diagram at the receiving end is a flowchart of the three basic steps of the method of the present invention. Figures three (1) to (4) show some samples of Chinese characters, where each character contains multiple patterns and each pattern contains multiple pen days. [Illustration of Drawing Numbers] 30 --_ Chinese characters; 32,34,36 --- Composition patterns of Chinese characters 30a, 32b Day part; 42, 44, 46--Composition patterns of Chinese characters 40; 52, 54 , 56 --- the composition pattern of Chinese character 50; 62, 64, 66, 68 The composition pattern of Chinese character 60. Figures 4 (1) to (5) are flowcharts of the method for defining the basic pattern of the present invention and are The defined basic pattern is a flowchart of a method for automatically converting peripheral font data into pattern basic font data.
第41頁Page 41
圖四(6)是圖樣基礎字型之全解像度(低、高解像度)_ 像方法的流程圖。 圖五示出包含關鍵點、寬度值以及特徵點之—個圖樣的範 例0 〔圖號說明〕 32---圖樣範例 圖六(1)〜(3)示範本發明在確認圖樣指紋之特徵點的方 〔圖號說明〕 151---圖樣;152-—極大極小方框 157a, 157b---含反曲點之曲線段; 168a---以相鄰特徵點為對角之方框 168b---擴展之方框 方法 圖七(1)〜(3)示出本發明在定義曲線段的Figure 4 (6) is a flowchart of the full resolution (low, high resolution) image method of the basic font of the pattern. Figure 5 shows an example of a pattern containing key points, width values, and feature points. 0 [Illustration of drawing number] 32 --- Example of patterns Figures 6 (1) to (3) demonstrate the feature points of the present invention in confirming pattern fingerprints Square [Illustration of drawing number] 151 --- pattern; 152 --- maximum and minimum box 157a, 157b --- curve section with inflection points; 168a --- box 168b with adjacent feature points as diagonal --- Expanded Box Method Figures 7 (1) ~ (3) show the invention in defining curve segments.
義單跑段掃瞒區域的方法,藉以 〔圖號說明〕 194--簡單基本圖樣形狀樣本;196___掃瞎線;The method of sweeping the concealed area in the Yidan section, by means of [illustration of the drawing number] 194--simple basic pattern shape sample; 196___ sweeping line;
第42頁 圖式簡單說胡 1 98a, 1 98b---分開的二段掃瞒線; 203a, 203b——單跑段掃瞄線區域1之左右兩旁分別的線段; 204a, 204b---单跑段掃瞎線區域2之左右兩旁分別的線段 圖九(1)〜(2)示出未標示及標示關鍵點經處理後之字型顯 像的樣子。 〔圖號說明〕 141a,141b---關鍵點; ❿ 圖十是電腦輔助工具用以定義特徵點的螢幕晝面 〔圖號說明〕 400 圖形人機介面工具; 402---顯不區 406 圖樣;408 指令菜單之一道指令 圖十一(1)〜(3)是電腦輔助工具用以定義關鍵點的螢幕晝 面。 〔圖號說明〕 410---指令菜單之一道指令;416---寬度值顯不棚 418, 42 0——決定寬度值之胖瘦點 圖十二(1)〜(5)是電腦輔助工具用以定義不同解像度層級The diagram on page 42 simply says Hu 1 98a, 1 98b --- two separate sweep lines; 203a, 203b-single-line sweep lines on the left and right sides of area 1; 204a, 204b --- Figures 9 (1) ~ (2) of the left and right sides of the single-run sweeping line area 2 respectively show the zigzag appearance of the unmarked and marked key points after processing. [Illustration of drawing number] 141a, 141b --- key points; ❿ Figure 10 is the screen day screen of computer-aided tools to define characteristic points [Illustration of drawing number] 400 graphical human-machine interface tool; 402 --- display area 406 Patterns; one of the 408 command menus. Instruction diagrams eleven (1) to (3) are computer-aided tools used to define the key points of the screen. [Illustration of drawing number] 410 --- an instruction in the instruction menu; 416 --- the width value is not displayed 418, 42 0—the fat and thin points that determine the width value Figure 12 (1) ~ (5) are computer-aided Tools to define different resolution levels
第43頁 ί231457 之曲線比例值的螢幕晝面。 〔圖號說明〕 420 曲線層級視窗; 422 曲線層級碼棚 430---視窗; 432---層級設定視窗 j^· 第44頁Page 43 The screen of the curve scale value of 231457. [Illustration of drawing number] 420 curve level window; 422 curve level code booth 430 --- window; 432 --- level setting window j ^ · page 44
Claims (1)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW90129744A TWI231457B (en) | 2001-11-30 | 2001-11-30 | System and method for converting an outline font into a glyph-based font |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW90129744A TWI231457B (en) | 2001-11-30 | 2001-11-30 | System and method for converting an outline font into a glyph-based font |
Publications (1)
Publication Number | Publication Date |
---|---|
TWI231457B true TWI231457B (en) | 2005-04-21 |
Family
ID=36121952
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW90129744A TWI231457B (en) | 2001-11-30 | 2001-11-30 | System and method for converting an outline font into a glyph-based font |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI231457B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103425627A (en) * | 2012-05-21 | 2013-12-04 | 文鼎科技开发股份有限公司 | Method for generating stroke and method for generating font |
TWI482101B (en) * | 2012-04-10 | 2015-04-21 | Arphic Technology Co Ltd | Method of generating strokes |
TWI571751B (en) * | 2016-05-17 | 2017-02-21 | 亞太戲網股份有限公司 | Acceleration method for showing asian language fonts in pages, font server, and browsing end |
US10346034B2 (en) | 2015-10-19 | 2019-07-09 | Dynacomware Taiwan Inc. | Dynamically generating characters with personalized handwriting character font characteristics method and system thereof |
-
2001
- 2001-11-30 TW TW90129744A patent/TWI231457B/en not_active IP Right Cessation
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI482101B (en) * | 2012-04-10 | 2015-04-21 | Arphic Technology Co Ltd | Method of generating strokes |
CN103425627A (en) * | 2012-05-21 | 2013-12-04 | 文鼎科技开发股份有限公司 | Method for generating stroke and method for generating font |
CN103425627B (en) * | 2012-05-21 | 2016-05-11 | 文鼎科技开发股份有限公司 | Produce the method for stroke and the method for generation font |
US10346034B2 (en) | 2015-10-19 | 2019-07-09 | Dynacomware Taiwan Inc. | Dynamically generating characters with personalized handwriting character font characteristics method and system thereof |
TWI571751B (en) * | 2016-05-17 | 2017-02-21 | 亞太戲網股份有限公司 | Acceleration method for showing asian language fonts in pages, font server, and browsing end |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TW486642B (en) | Glyph-based outline font generation independent of resolution | |
JP6956555B2 (en) | How to detect objects in an image and object detection system | |
US10437466B2 (en) | Formula inputting method and apparatus | |
US7576753B2 (en) | Method and apparatus to convert bitmapped images for use in a structured text/graphics editor | |
US8875016B2 (en) | Method and apparatus to convert digital ink images for use in a structured text/graphics editor | |
US20020064308A1 (en) | System and methods for spacing, storing and recognizing electronic representations of handwriting printing and drawings | |
US6661417B1 (en) | System and method for converting an outline font into a glyph-based font | |
JP5774558B2 (en) | Handwritten document processing apparatus, method and program | |
JP2003298837A (en) | Graphical input and display system for creating and manipulating electronic image, method for editing source electronic image inputted to display device to meaningful image object, and product including computer usable medium with computer readable program code incorporated | |
JP2013061660A (en) | Method for generating stroke-based font character for low-resolution display | |
EP1901233A2 (en) | Techniques for image segment accumulation in document rendering | |
CN115812221A (en) | Image generation and coloring method and device | |
US6771266B2 (en) | Method and apparatus for improving the appearance of digitally represented handwriting | |
JP4162670B2 (en) | How to generate a silhouetted gray font | |
JP2022066321A (en) | Information processing device and program | |
TWI231457B (en) | System and method for converting an outline font into a glyph-based font | |
JPH06131436A (en) | Method and device for plotting processing | |
US20230334223A1 (en) | Font customization based on stroke properties | |
US7961945B2 (en) | System and method for on-the-fly segmentations for image deformations | |
TW434493B (en) | Stroke-based glyph-outline font generation in low/high resolution space | |
TWI282947B (en) | Silhouette oriented gray font generation | |
JP4148867B2 (en) | Handwriting processor | |
JP7324305B2 (en) | Electronic form creation device, electronic form creation method, and program | |
Notowidigdo | User-directed sketch interpretation | |
JP5361315B2 (en) | Information processing apparatus and information processing method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
MK4A | Expiration of patent term of an invention patent |