Abstract: SUMMARY This paper presents a statistical approach for font attribute recognition based on features extracted from projection profiles of text lines and using a Bayesian classifier. The presented features allow the discrimination of the font weight, slope and size.

Abstract: A system and method for creating three-dimensional or depth image font text characters using graphic three-dimensional object creation techniques and graphics processors. The text characters can be represented as set descriptions of the characters. The text characters can also be represented as a three-dimensional geometric model including polygons constructed from vertices defined by three-dimensional coordinates. The representations are stored in a font storage and when a user specifies the text characters to be used in a depth image along with the font to be used for the text characters, the geometric representations of the characters are retrieved. If stored as a set, the set is converted into a geometric plot. Appropriate scaling and surface texturing operations are performed as designated by the user to create three-dimensional text character graphic objects. These text character graphic objects are transferred to a graphics processor to be manipulated as desired by the and used to produce an image that can be further processed.

Abstract: The problem of determining font metrics from measurements on images of typeset text is discussed, and least-squares procedures for font metric estimation are developed. When it is shown that kerning is not present, sidebearing estimation involves solving a set of linear equations, called the sidebearing normal equations. More generally, simultaneous sidebearing and kerning term estimation involves an iterative procedure in which a modified set of sidebearing normal equations is solved during each iteration. Character depth estimates are obtained by solving a set of baseline normal equations. In a preliminary evaluation of the proposed procedures on scanned text images in three fonts, the root-mean-square set width estimation error was about 0.2 pixel. An application of font metric estimation to text image editing is discussed. >

Abstract: A technique is specified for the output of fonts on high resolution output devices, such as photosetters in particular. A processor unit optically scales a digitally stored font, by selecting a single master font in a digitized contour coding, in which the contours for each letter are provided with delimiting discrete control points and additionally with instructions, such as those stipulated for the "Intelligent Font Scaling"; by re-scaling the master font with a factor to the desired point size, and by utilizing the instructions in order to steadily vary the stroke thickness of the linearly re-scaled letters in a primary pre-set factor dependency, wherein the stroke thickness is increased for factors smaller than one and is decreased for factors larger than one. The font can also be typographically expanded and condensed which means keeping the widths of the vertical stems during broadening or narrowing, respectively.

Abstract: A computer-implemented apparatus and method for generating an output digital font from a base font and one or more font descriptor files. In one example, the method involves, the steps of retrieving from memory a file containing instructions and data for a generic base font; retrieving from memory a font descriptor file containing specifications for operating upon the base font to produce the desired output font; and then generating the output font by performing operations upon the base font in accordance with the specifications contained in the font descriptor file, to produce a character program for each character in the base font wherein the data representing the output font is the generic font data as transformed in accordance with said specification. Two or more font descriptor files may be combined, such as by using mathematical weighted averaging of the parameter values for different typographic design features, or otherwise, to create font descriptor files for hybrid typefaces.

Abstract: Visual and Technical Aspects of Type gives an introduction to the rules of font design and describes how fonts and their metrics are managed by computers. It is the first book that contains contributions from both type creators and computer scientists. The aim is to provide insights into the production and rendering of digital type and to make traditional type design rules accessible to a wider audience. The first part contains an overview of the evolution of letterforms in their historical and cultural context. The second part is devoted to technical aspects of type; topics covered include character metrics, outline font rasterization techniques, and algorithms for various tasks. Finally, articles by Hans Meier and Fernand Baudin provide an interesting view of the progress of typefaces and page layout, and insight into future developments. This unique book will appeal to graphics designers, computer scientists, typographers and desktop publishers who wish to know more about computer typography.

Abstract: A technique is demonstrated for the output of fonts on high resolution output devices, such as phototypesetters in particular, in which a processor unit optically scales a digitally stored font, expands/condenses and generates a desired weight of the font to be distributed. Two previously generated master fonts with varying weights in a digitized contour coding are used, by which a majority of the contours of each letter are provided with delimiting discrete control points and additionally with instructions, such as those stipulated for "Intelligent Font Scaling". The output font with the desired weight is generated through linear interpolation between the two master fonts. To optically scale to the desired point size, the output font is re-scaled with a factor to the desired point size and the instructions are applied in order to steadily vary the stroke width of the letters, linearly re-scaled with the factor, in a primary pre-set factor dependency whereby the stroke width is increased for factors smaller than 1% and decreased for factors larger than 1%. The font is typographically correctly expanded and condensed, in that the optically scaled output font is brought to the desired width through linear expansion or condensation with a second factor, whereby the instructions are applied in order to hold the stroke width constant during linear expansion or condensation of the letters and in this way to only broaden or narrow the contours.

Abstract: A video controller receives character data, attribute data and font data, each of which are stored in different planes of a video memory. The font data comprises bit maps of at least two character fonts, which may be user fonts or default fonts loaded from a controller BIOS. The video controller retrieves the font data, translates the font data into a page mode, and stores the font data in a hidden font cache in an unused portion of the video memory. The paged font data is divided into a number of pages equal to the number of scan lines per character. Each page contains a number of words, and each word contains at least two bytes. Each byte represents one scan line of a character in a different font. The video controller retrieves the paged fonts in page mode and assembles the scan lines for the characters to be displayed into one video scan line. The use of the page mode increases refresh rate and allows simultaneous display of two fonts.

Abstract: Characters are recognized by a conventional OCR apparatus and converted into outline font form. The system includes a recognition device for optically reading printed characters and recognizing those to obtain information on the recognized characters consisting of text code information and character layout information, an outline font table for retaining outline font data of characters, a character box enlarging function for enlarging enclosing rectangles of the recognized characters obtained by the recognition device by a ratio of an outline font character box to a black pixel component to be drawn in the character box while referring to the outline font table, and modifying the information on the recognized characters by using the enlarged enclosing rectangles as new character boxes of the outline font.

Abstract: SUMMARY The international scope of computing, digital information interchange, and electronic publishing has created a need for world-wide character encoding standards. Unicode is a comprehensive standard designed to meet such a need. To be readable by humans, character codes require fonts that provide visual images — glyphs — corresponding to the codes. The design of a font developed to provide a portion of the Unicode standard is described and discussed.

Abstract: An improved font resolution method is disclosed for a data processing system, which enables the translation of an originally defined font for a document, into available fonts in a local data processing system where the document is to be displayed or printed. The method allows for the selective substitution by a user of alternate Code Page Names and Character Set Names for fonts which may be available on a local data processing system.

Abstract: Kerning information that allows font characters to contact the outlines of immediately preceding characters when the setting of character spaces is zero is set for the data on said font characters. The result of output with the setting of character spaces being zero is such that all font characters contact each other continuously. If the character space is set to a desired value greater than zero, a uniform space is provided between every font character. This insures that in a line of type such as where the letter "A" follows "W", the "A" is automatically kerned into the letter region of "W". As a result, the method of kerning in the processing of documents written in European languages is rationalized to obviate the need to perform calculations for setting letter spaces.

Abstract: A font data output method that applies bandwidth limitation to binary font data using a two-dimensional low-pass filter, and applies gain of one or more to a multilevel-value data group obtained by reduction at a predetermined ratio according to the distribution of values in the multilevel-value data group to generate the gradated font data. Using this method, the loss of line and character density in the font data due to a high reduction ratio can be prevented.

Abstract: SUMMARY In this paper we present a new method of parametrizing PostScript fonts in order to create font families. By changing parameter values one can obtain different weights, condensed or expanded versions, small caps as well as optically scaled fonts. The tool used to parametrize PostScript fonts is D. E. Knuth’s METAFONT program. Instead of designing a font from scratch, METAFONT is used as an extrapolator of existing PostScript fonts: out of the information contained in them we build a meta-font; for every choice of parameter values, special versions of METAFONT allow us to return to PostScript and produce a new PostScript font.

Abstract: An image data processing apparatus for generating a font with contour lines expressed with gradations comprises a font ROM for storing a binary font data. An output port is provided for reading a target pixel with ambient pixels arranged in 3×3 matrix of the binary font data stored in the font ROM. A shift register produces an address data based on the binary data read from the font ROM. A look-up table has 23×3 elements in which multilevel values in commensurate with the pixel patterns included the 3×3 matrix. One data of the multilevel data is output from the look-up table by designating one element by said address data, and is used as a gradation data of the target pixel. Thus, pixel of font can be expressed with a gradation, the jagged contour of the font is reduced.

Abstract: font-independent character descriptions are important for a systematic approach to automated and semi-automated font design. This is particularly so for large character sets such as Kanji. The paper defines a completely coordinate-independent notation for Kanji, which contains all the necessary information to produce legible character sketches.

Abstract: Character forming method and apparatus are provided. Font data is stored. The conditions to develop the stored font data are stored. An area to develop the font data is reserved in a memory in accordance with the stored conditions. The font data is developed in the area reserved. The font data is vector information. The condition to develop the font data is that the data is either one of the outline data, shadow data, and filler data. The bit map patterns of filler pattern, outline pattern, and shadow pattern are formed in accordance with the developed font data and are individually stored.

Abstract: There are provided a pattern generating method and apparatus for generating a character/symbol pattern. The apparatus comprises a font memory to store outline font data of characters, symbols, and the like, an image memory to store image data based on the outline font data stored in the font memory, storing means for storing the font image based on the outline font data into the image memory, a color data generating circuit to generate color data in accordance with the designated color information, and a painting circuit to paint the inside of the stored font image by the color data. The color data generating circuit executes a masking operation on the basis of the color information which was designated and input and generates the color data. Thus an accurate color image can be reproduced.

Abstract: SUMMARY Many fonts derive from stroke-based ancestry. Pressure applied to the pen or brush provided some variation in the stroke width, which defined a region on each side of a centreline. A simple representation of fonts as variable width strokes is presented in this paper. Advantages include a good first step toward typographic scaling (stroke width scales independently of overall scale factor), and preservation of topology at low resolutions (minimum stroke width can be enforced). A chief disadvantage is the lack of experience designing fonts in this paradigm, or building routines to convert from other paradigms.

Abstract: The method of the invention allows both single-byte character set (SBCS) and double-byte character set (DBCS) fonts in a DBCS code page. The invention stores the SBCS and DBCS text of the document in separate areas. Each area contains the following specific information about the text: the actual text itself, the length in bytes of the text, the horizontal starting position of the text, the font attributes for that text, a flag to indicate that the text is SBCS or DBCS text, and the value which points to the next area containing some text. The font attributes contain information such as the font typeface name, point size, color, weight, width, and the value to indicate whether the font type is an SBCS or a DBCS font type. A document is then set up to use different fonts, SBCS or DBCS, for specific sections of text and alternating back and forth between the fonts as many times as is necessary. The text of the document that uses the different fonts will be in separate areas and each area will contain its own text and font specific information.

Abstract: SUMMARY This paper introduces a novel way to perform dynamic regularisation of outline fonts. In the proposed font representation, the characters are decomposed into the components glyph, contour, knot ,a ndnumber. These components are scaled and mostly rounded before they are

Abstract: A label printing device is disclosed in which font data for a plurality of characters is stored. Characters are derived from the font data for displaying and printing in sizes determined by appropriate scaling factors. In this way, characters derived for printing have the same proportion and appearance as characters derived for displaying. Moreover, a large variety of character sizes can be produced both for printing and displaying.

Abstract: A method for matching fonts in a client-server computer system environment (130, 140) in which an application (134) is being shared by multiple servers (132, 142). Penalty values are computed for font appearance attribute differences. Difference totals (summation of penalty values) are computed for all appearance attributes for a receiving server's (132, 142) candidate font and an application (134) designated font. For each font designated by the application (134), a difference total is computed for each font in a set of candidate fonts in each receiving server (132, 142). For each receiving server (132, 142), and for each application (134) designated font, one receiver font is selected which has the smallest difference total.

Abstract: Today’s font-scalers generate screenfonts with acceptable quality on-the-fly from a generic font representation. However, as closed systems they discourage the integration of separate solutions to different aspects of font-scaling. This paper illustrates an object-oriented approach that allows for both contour and rendering independence. Refined solutions can be packaged separately into intelligent contour and rendering objects. The approach results in a small and efficient font-scaling system that masters complexity by concept rather than industriousness.

Abstract: Preface How to Use this Book Assumptions Related Documents Font Conventions Used in This Book Request for Comments Bulk Sales Information Acknowledgments Chapter 1 An Introduction to X Administration 1.1 The Design of X11 1.1.1 Display Servers 1.1.2 Clients and Resources 1.1.3 Toolkits and GUIs 1.2 X Administration 1.2.1 Installing X 1.2.2 Supporting Users 1.2.3 Maintaining Software 1.2.4 Maintaining Multiple Machines 1.2.5 A Philosophy of X Administration Chapter 2 The X User Environment 2.1 The Configured X Session 2.1.1 The Twilight Zone 2.2 Components of the X Environment 2.2.1 Window Managers 2.2.2 Customizing Clients 2.2.2.1 The -fn Command-line Option 2.2.2.2 The -geometry Command-line Option 2.2.2.3 Specifying Colors 2.2.2.4 Using Resources 2.2.3 The Startup Script 2.2.3.1 The Foreground Process 2.3 The Shell Environment 2.3.1 Setting the DISPLAY Variable 2.3.1.1 Complications with Display Names 2.3.2 Redefining the Search Path 2.3.2.1 Setting the Search Path for OpenWindows Support 2.3.2.2 Setting the Search Path for Mixed Environments 2.3.3 xterm Issues 2.3.3.1 xterm and Terminal Emulation 2.3.3.2 The Resize Client 2.3.3.3 xterm and the Login Shell (C Shell) 2.3.4 Starting Remote Clients 2.3.4.1 Starting a Remote Client with rsh 2.4 Startup Methods 2.4.1 xinit and startx 2.4.2 Differences Between .xinitrc and .xsession 2.5 Related Documentation Chapter 3 The X Display Manager 3.1 xdm Concepts 3.2 xdm Configuration Files 3.3 xdm the Easy Way 3.4 Troubleshooting xdm 3.5 Customizing xdm 3.5.1 The Master Configuration File (xdm-config) 3.5.2 Listing X Servers (the Xservers File) 3.5.2.1 Xservers Syntax 3.5.3 xdm Host Access Control: the Xaccess File (R5 Only) 3.5.3.1 Direct and Broadcast Access 3.5.3.2 Indirect Access and the Chooser 3.5.3.3 Using Macros 3.5.3.4 Advantages and Disadvantages of the Chooser 3.5.4 The Xresources File 3.5.4.1 Configuring the Login Box 3.5.4.2 The xconsole Client 3.5.5 Starting Up Individual X Sessions (the Xsession File) 3.5.5.1 No Home Directory? (R5) 3.5.6 Display Classes 3.6 Testing Your xdm Setup 3.6.1 Resetting the Keyboard 3.6.2 Restarting xdm Using xdm-pid (R4 and Later) 3.6.3 Rereading XDM Configuration Files (R3) 3.7 Permanent Installation of xdm 3.8 Related Documentation Chapter 4 Security 4.1 Host-based Access Control 4.1.1 The /etc/Xn.hosts File 4.1.2 The xhost Client 4.1.3 Problems with Host-based Access Control 4.2 Access Control with MIT-MAGIC-COOKIE-1 4.2.1 Using MIT-MAGIC-COOKIE-1 with xdm 4.2.2 The xauth Program 4.2.3 Using MIT-MAGIC-COOKIE with xinit 4.2.4 xauth vs. xhost 4.3 The XDM-AUTHORIZATION-1 Mechanism (R5) 4.4 The SUN-DES-1 Mechanism (R5) 4.4.1 Public Key Encryption 4.4.2 Prerequisites for Using SUN-DES-1 4.4.3 Using SUN-DES-1 with xdm 4.4.4 Using SUN-DES-1 with xinit 4.4.5 Adding Another User with SUN-DES-1 4.4.6 xterm and SUN-DES-1 4.4.7 Troubleshooting SUN-DES-1 4.5 xterm and Secure Keyboard 4.6 Other Security Issues 4.6.1 The Console xterm (R4 and Earlier) 4.6.2 The Console and xdm (R5) 4.6.3 Hanging the Server Remotely (R3) 4.6.4 Reading the Framebuffer (Sun Workstations) 4.6.5 Removing Files in /tmp 4.6.6 The Network Design 4.7 Related Documentation Chapter 5 Font Management 5.1 Fonts on the X Window System 5.1.1 xlsfonts 5.1.2 xfd 5.1.3 xfontsel 5.1.4 The Font Path 5.1.5 The Font Directory File 5.1.6 The fonts.scale File (R5 only) 5.1.7 Wildcards 5.1.8 Aliases 5.1.8.1 The FILE_NAMES_ALIAS Alias 5.2 All About Fonts 5.2.1 Bitmap Versus Outline Fonts 5.2.2 Font Formats 5.2.3 Format Conversion Tools 5.3 Adding New Fonts 5.3.1 Adding a Single Font 5.3.2 Adding Multiple Fonts 5.3.2.1 Multiple Font Example 5.3.3 Problems with Running Vendor-specific Clients 5.3.4 DECWindows Examples 5.3.4.1 Aliasing 5.3.4.2 DECWindows Conversion 5.3.5 AIXWindows Example 5.3.6 OpenWindows Example 5.3.6.1 Aliasing 5.3.6.2 OpenWindows Conversion 5.3.6.3 Converting from X11/NeWS to PCF or SNF 5.3.6.4 More Conversions 5.4 Providing Fonts Over the Network 5.5 The R5 Font Server 5.5.1 The Configuration File 5.5.2 Installing the Font Server 5.5.2.1 Testing By Hand 5.5.2.2 Changing BSD Boot Files 5.5.2.3 Changing System V Boot Files 5.5.2.4 Changing AIX Boot Files 5.5.3 Font Server Name Syntax 5.5.4 Debugging the Font Server 5.5.5 Font Server Clients 5.5.6 The Font Path and the Font Server 5.5.7 Hostname Aliases 5.5.8 A Font Server Example 5.6 Related Documentation Chapter 6 Color 6.1 Color Specification in Release 4 and Earlier 6.1.1 RGB Color Names 6.1.2 Numeric Color Values 6.1.3 Adding Your Own Color Names (RGB) 6.1.4 Fixing a Corrupted Color Database 6.2 Color Specification in Release 5 (Xcms) 6.2.1 Xcms Color Names 6.2.2 Adding Your Own Color Names in Xcms 6.2.3 Xcms Database Example 6.2.4 Device Profiles 6.3 Related Documentation Chapter 7 X Terminals 7.1 Buying an X Terminal: What's What 7.1.1 Monitors 7.1.1.1 Screen Size 7.1.1.2 Resolution 7.1.1.3 Depth 7.1.1.4 Refresh Rate 7.1.2 Keyboard and Mouse 7.1.3 X Server Software 7.1.4 Special Features 7.1.5 Memory Configuration 7.1.6 Network Interface 7.2 X Terminal Setup 7.3 Network Setup 7.3.1 Getting the IP Address Using RARP 7.3.2 Getting Information Using BOOTP 7.3.3 Trivial File Transfer Protocol (TFTP) 7.3.4 Setting Up the Network on the X Terminal 7.3.5 Debugging Hints 7.3.5.1 Error Messages 7.3.5.2 Updating the arp Table 7.3.5.3 Name Server Problems 7.4 Fonts on X Terminals 7.4.1 Font Formats 7.4.2 The Font Server (R5) 7.4.3 Choosing TFTP or NFS for Font Access 7.4.3.1 Reading Fonts Using TFTP 7.4.3.2 Reading Fonts Using NFS 7.5 Configuring for the X Display Manager 7.5.1 Configuring the X Terminal for xdm 7.5.2 Configuring an R5 Host 7.5.3 Configuring an R4 Host 7.5.4 Configuring xdm Without XDMCP 7.5.5 Setting Up Server Access Control 7.6 Remote Configuration of X Terminals 7.6.1 Remote Configuration on NCD Terminals 7.6.2 Remote Configuration on Visual Terminals 7.6.3 Remote Configuration on Tektronix Terminals 7.7 Reconfiguring the Host 7.7.1 Increasing the Number of Processes 7.7.2 Increasing the Number of Pseudo-ttys 7.7.3 Increasing the Amount of Swap Space 7.7.3.1 Swapping to a File 7.7.3.2 Swapping to a Disk 7.8 Related Documentation Chapter 8 Building the X Window System 8.1 Installation Issues 8.1.1 Should You Use MIT Source? 8.1.2 Types of Vendor-supplied X Distributions 8.1.2.1 X from Your OS Vendor 8.1.2.2 X from a Third Party 8.1.3 X Source Code from MIT 8.1.4 Complete or Client-only Distribution? 8.1.5 Installing Multiple X Releases 8.2 Source Preparation 8.2.1 Do You Have Enough Disk Space? 8.2.2 Is Your Platform Supported? 8.2.3 Applying OS Patches 8.2.4 Applying X Patches 8.2.5 Creating a Link Tree (Optional) 8.3 Simplest Case Build 8.4 Host Problems 8.4.1 Disk Space 8.4.1.1 Changing the tmp Directory Using TMPDIR (Ultrix and HP-UX) 8.4.1.2 Changing the tmp Directory Using -temp (SunOS) 8.4.2 Shared Library Installation (SunOS) 8.4.3 NFS Installation 8.4.3.1 NFS Installation Without Root Access 8.4.3.2 Installation Over the Network (rdist) 8.4.4 Installing the termcap or terminfo Definition for xterm 8.5 Simple Configuration 8.5.1 Configuration Parameters 8.5.1.1 site.def 8.5.1.2 The ProjectRoot Flag 8.5.1.3 The Platform Configuration File (platform.cf) 8.5.2 Configuration Example 1 8.5.3 Configuration Example 2 8.5.4 Configuration Example 3 8.5.5 Configuration Example 4 8.5.6 Configuration Example 5 8.5.7 Other Build Flags 8.5.7.1 Xterm Build Flags 8.6 Building Programs After X Is Installed 8.6.1 xmkmf 8.6.2 Include Files 8.6.3 Libraries 8.7 More About imake 8.7.1 The make Program 8.7.2 The C Preprocessor 8.7.3 Imake Syntax 8.7.3.1 Comments in imake 8.7.3.2 Multi-line Macros (

Abstract: Automated text recognition is a difficult but important problem. It can be summarized as: how to enable a computer to recognize letters and digits from a predefined alphabet, possibly using contextual information. Various attempts at solving this problem, using different selections of features and classifiers, have been made. Human performance has been achieved in accuracy by automated text recognition systems, and has been bypassed in speed for the case of single size, single font, high quality, known layout, known background, text. When one or more of the above parameters are changed, the problem becomes increasingly difficult. In particular, attaining human performance in recognizing cursive script of varying size, varying style, unknown layout, unknown background is far from the reach of today's algorithms, despite the continuous research effort for almost four decades. In this report, we analyze the problem in detail, present the associated difficulties, and propose a coherent framework for addressing automated text recognition. (Copies available exclusively from MIT Libraries, Rm. 14-0551, Cambridge, MA 02139-4307. Ph 617-253-5668; Fax 617-253-1690.)