TopWare 18: Liquid

home *** CD-ROM | disk | FTP | other *** search

/ TopWare 18: Liquid / Image.iso / liquid / top2075 / info.doc < prev next >

Wrap

Text File | 1993-04-18 | 36KB | 818 lines

1 MANUAL.DOC for RECOGNITA PLUS "Demo version" (Revision May 28, 1991) You are currently using RECOGNITA PLUS, Version 1.2 Demo, The International OCR Copyright (C) SZKI Recognita Corp., 1991 You may distribute these diskettes free of charge. This MANUAL.DOC consists of four parts. The first one describes the main features of Recognita Plus. The second one tells you what you need to run the program, the third one describes the setup process of RECOGNITA PLUS, the fourth one introduces you to using RECOGNITA PLUS step by step. DEMO VERSION MEANS THAT YOU CAN USE ALL THE FUNCTIONS OF RECOGNITA PLUS, EXCEPT STORING THE RECOGNIZED TEXT IN A TEXT FILE. SO IGNORE ALL REFERENCES TO SAVING OUTPUT TEXT FILES. If you have any comment or suggestion, or intend to buy the product, please address responses to SZKI Recognita Corp. Recognita Corp. of America Budapest, Donati u. 35-45. 1250 Oakmead Parkway, Suite 210 H-1015, Hungary, Europe Sunnyvale, CA 94088-3599 Phone: (361) 201-84-52 Phone: (408) 749-9935 Fax: (361) 201-76-07 or (800) 255-4 OCR Tx: 22-5381 Fax: (408) 730-1180 Recognita Büroautomatisierung GmbH Hainstrasse 17/19 - Aufgang E D-O-7010 Leipzig, Germany Phone/Fax: (37-41) 20 91 93 * * * * * * * * * * * * PART I: ABOUT RECOGNITA PLUS Recognita Plus is an omni-font pre-trained and trainable optical character recognition (OCR) program for the IBM AT, AT/386, AT/486 and PS/2 computers. It offers exceptional speed, both in MS DOS and MS Windows 3.0 environments. It can read in automatic or interactive mode. The latter offers both an on-line editor and an additional learning facility for special tasks. It can also be run from DOS command lines or batch files and so be integrated into application programs. The program supports an exceptionally wide range of desktop scanners. Documents: Recognita Plus can handle typeset documents, also those printed on a letter-quality printer, a near-letter-quality dot-matrix printer, a laser printer, or typed on an electric typewriter. Output: The recognized text can be saved directly to several ASCII formats and also in over 30 widely used word processor formats. 2 Languages: The International Recognita Plus offers twenty-two European languages on-screen, including Greek; in addition it can read almost any language with a Latin-based alphabet, including non-European ones, over 80 in all. It can handle multi-lingual documents, even with a dozen or more languages on one page. Recognita Plus comes in three language implementations: Recognita Plus 1.2 The International OCR Recognita Plus 1.2 The Americas OCR, reads the 4 major languages spoken and written in the Americas (English, French, Spanish and Portuguese) Recognita Plus 1.2 The English OCR, is only for English, like most of the OCR packages on the US market. * * * * * * * * * * * * PART II: WHAT YOU NEED TO RUN RECOGNITA PLUS To use Recognita Plus the following minimum configuration is recommended: - An IBM AT, AT/386, AT/486 or PS/2 computer, with at least 640 kBytes of RAM (in MS DOS environment), or 1 MByte of RAM and memory extension used by MS Windows (in MS Windows environment), 3 MBytes of free hard disk space, a floppy disk drive and one parallel port, which may also be used for a printer. - A CGA, EGA, VGA, Hercules or compatible display card, with a matching colour or monochrome monitor. - One of the supported scanners with at least 300 dpi resolution properly connected to the computer. (Reading from image files is possible without a scanner). - A mouse (recommended). - MS DOS operating system, version 3.0 or later. - MS Windows 3.0 or later. Recognita Plus is able to run in Enhanced Mode, Standard Mode or Real Mode under MS Windows. However, some of the scanners are not supported in Enhanced Mode. Note. Be sure to read the README.DOC file on the RECOGNITA PLUS program disk. README.DOC contains updated information unavailable when this guide was edited. * * * * * * * * * * * * PART III: SETTING UP RECOGNITA PLUS RECOGNITA PLUS must be installed before use. You have a menu driven setup program named RECSETUP on your RECOGNITA PLUS diskette #1. By running it you configure RECOGNITA PLUS to your hardware environment. 3 To run RECSETUP you need to know the following: - what kind of computer and display you have - what type of scanner you have. Now we assume that your computer is turned on and the prompt can be seen on the display, e.g. C:> showing that the operating system has been loaded. Insert the RECOGNITA PLUS diskette #1 in drive A, then type in A:RECSETUP <Enter> On the screen you can see the starting screen of the setup pro- gram. Once the setup program is running, follow the instructions on the screen. Within a few seconds you can choose the setup language. Enter the appropriate letter "E" for English, "D" for German, "F" for French, etc. Installation To begin the setup procedure choose 'Installation' by moving the highlight with the direction keys. Then you can select the highlighted item by pressing the <Enter> key. If you intend to work only with import image files (e.g. TIFF files) and there is no scanner in your configuration, choose 'No scanner.' If the scanner interface card or its resident driver is not found, as you leave RECSETUP the screen will list any particular requirements of the scanner you have chosen, e.g. on port addresses, switch settings or interface card needed. Modifications You can make modifications in your configuration at any time by starting the setup program from your installed copy. Type in: RECSETUP <Enter> then move the highlight to 'Modifications' and press the <Enter> key. Make the necessary changes. You may also start RECSETUP from inside MS Windows 3.0 by clicking on its icon 'Recognita Setup'. In this case the modifications will take effect only after closing and re-starting Recognita Plus again. DEVICE DRIVERS Some scanners have a device driver program supplied by the manufacturer. The setup program reminds you to install any device driver needed. If you start Recognita Plus without that, you will see the error message DEVICE DRIVER MISSING or NO SCANNER INSTALLED If you have a mouse in your computer configuration, the resident mouse driver must also be installed in the memory of your PC before running RECOGNITA PLUS. You can use any type of mouse, the 4 only limitation is that its driver should be a Microsoft compatible one. STARTING RECOGNITA PLUS under MS DOS: RECDEMO <Enter> under MS WINDOWS: WIN RECDEMO <Enter> or double-click on the Recognita Plus icon. * * * * * * * * * * * * PART IV: USING RECOGNITA PLUS A/ SOME COMMON TERMS Scanner parameters 'Brightness' and 'Contrast': You can change these parameters to make reading more accurate. For good quality documents use the default values. With a low-quality document, you may have to use a method of trial and error, i.e. to test scan it several times with various settings to see which gives the best results. 'Resolution': There are scanners which can work at differing resolutions. For OCR purposes the best results can be achieved at 300 or 400 dpi. However, many scanners produce 400 dpi only by extrapolation. In this case scanning at 400 dpi will yield worse results than 300 dpi. Reading: You can read text in automatic or interactive modes. Reading in automatic mode means that after activating 'Read' the recognition will automatically be executed, while reading in interactive mode enables on-line correction and teaching during the recognition process. If you wish to abandon reading, press the <Esc> key. Rejection: If the program didn't recognize a character, you will see the rejection symbol (a black filled square by default) on the screen instead of the unrecognized character. Scanning area: The paper length and the margins determine the area of the document to be scanned from scanner. The main purpose of this is to leave out unwanted pieces of a document and to shorten the scanning time. The size range of the paper length accepted by Recognita Plus is from 3 to 14 inches. To type a fraction of an inch or centimetre use a decimal point (e.g. 8.25). If you define margins in the Layout function, be careful not to let the margin frame touch any characters. Reading area: You can define smaller areas (boxes) on the page to read the area inside or outside the boxes. These can overlap one another. Boxes are shown with dotted lines (DOS) or thin red lines (MS Windows). As reading can be performed from areas defined either inside or outside the boxes, you have to decide which way is best. Box settings remain valid until you remove boxes or define new ones. Under MS DOS a box is called a window. 5 Graphics separation: Recognita Plus automatically separates text from graphics. However, this takes more time, so to speed up the text recognition on a page with large size graphics, photos or images, it's a good idea to define margins and/or boxes to cut out these parts of the page. Image file: The image file created by the program can be stored in compressed PackBits or uncompressed TIFF format. An uncompressed TIFF file of an A4 size page at 300 dpi needs over 1 Mbyte of free space on your hard disk, while in compressed form it needs only 20-60% of an uncompressed one. Supported Input Image File Formats: TIFF (uncompressed, compressed CCITT G3, CCITT G4, compressed PackBits) and PCX. Text formats: In addition to the ASCII formats, Recognita Plus offers over 30 word processor formats. Recognita Plus enables you to transmit the original character sizing to the output text file, i.e. transmit to the selected word processor all character sizes, line spacing, tabulator positions from the original document, provided a word processor format was chosen and 'Original' was set. Character set: The delivery default character set is IBM ASCII (PC8). Further possibilities are Roman-8, ANSI, Greek-ELOT and Greek-MEMOTEK. You can see the relevant character code tables on- screen by clicking on 'Show Char. Set...'. Other sets: In addition Recognita Plus is supplied with thirteen modified IBM ASCII sets for reading less common languages. B/ WORKING UNDER MS DOS READING THE TEST.TIF file supplied with Recognita Plus and containing multi-lingual text Steps for reading from image file 1. Select the parameter FROM FILE: *.TIF. under IMAGE SOURCE in the Main Menu. 2. Type in the name of the desired image file (TEST.TIF) or leave *.TIF to call up and use a file selection box. 3. Activate the READ function. 4. When Recognita Plus indicates 'Ready', press the <Enter> key or click on O.K. Our test file has only four languages. So, in order to improve the program's accuracy rate, select only the languages found in the test file: French, German and Spanish. The OPTIONS function enables you to define the languages. To select a language, de- select INTERNATIONAL, then click on the language required. A tick will appear in front of the name. To de-select it click on it again. Steps for defining the languages for the TEST.TIF file 1. Activate OPTIONS. 2. De-select INTERNATIONAL and select the appropriate languages (German, French and Spanish). 3. Activate O.K. 6 Let's suppose you intend to read only some paragraphs of the test file. The LAYOUT fuction is used to display the layout of a page and define the parts to be read. Steps for reading a part of an image file by windows 1. Check that FROM FILE is selected under IMAGE SOURCE. 2. Check that CORRECT AND LEARN and LANDSCAPE in the Main Menu are de-selected. 3. Name the image file to be read (TEST.TIF). 4. Activate the LAYOUT function in the Main Menu. 5. Activate SCAN IMAGE to scan the page to display its layout. 6. Decide which to select: TEXT INSIDE WINDOWS or TEXT OUTSIDE WINDOWS. 7. Define window(s) for the required parts of the page. Move the cursor (a cross) to one corner of the window to be created. Press the left mouse button, drag it to the opposite corner, then release. 8. If the window definition is not satisfactory, activate REMOVE WINDOW and click in one window or REMOVE ALL and define window(s) once more. 9. Return to the Main Menu by activating BACK TO MAIN MENU. 10. Activate the READ function in the Main Menu to read the text area defined. 11. Press the <Enter> key or click on O.K. to return to the Main Menu when the 'Ready' message appears. SETTING SCANNER PARAMETERS Steps for setting scanner parameters 1. Insert a page in the scanner. 2. Activate the TEST SCAN function. 3. Enlarge some characters (left mouse button). 4. Click on TEST SCAN repeatedly, enlarge other characters and examine other parts of the page. Use the right mouse button to scroll the image horizontally. 5. Activate the EJECT function. 6. Adjust scanner parameters in the Main Menu. 7. Activate TEST SCAN again to examine the quality of the scanned image produced with the new settings. 8. Repeat steps 5 to 7 until the result is satisfactory. 9. Return to the Main Menu (use EJECT if necessary). 10. Read the page with your optimised scanner settings. SETTING OPTIONS Steps for setting options 1. Activate OPTIONS. 2. Select the appropriate text or word processing format. 3. Select RETAIN ORIGINAL SIZES if you wish to transfer the original character sizes to the output file. 4. Select the appropriate language(s), or INTERNATIONAL for a highly multi-lingual document. 5. If you want to read a document with the most common Greek symbols in it, select PLUS GREEK. 7 6. If your document contains nearly all numerical information, select NUMBERS ONLY,PLUS and define characters additionally needed. 7. If your document is mostly text, LANGUAGE SET should be selected; define any special characters needed in its text box if desired. 8. Define the CHARACTER SET used by your Word Processor. 9. Define a character for the rejection symbol if the default is not satisfactory. 10. Activate O.K. to confirm settings and return to the Main Menu. READING TEXT FROM SCANNER Steps for reading a single page 1. Insert the page into the scanner. Check that the text is not skewed. 2. Set scanner parameters. 3. Set options (languages, character set, word processor, etc). 4. Check that MULTIPLE PAGES, LANDSCAPE and CORRECT AND LEARN in the Main Menu are de-selected. 5. TEXT RESULT should be selected; type in a suitable name for the output file. 6. Activate the READ function (or press <Ctrl> + <r>). 7. Wait for the 'Ready' message to appear. 8. Check that the box in front of TO FILE is selected. 9. Press the <Enter> key or click on O.K. to save your text and return to the Main Menu. MULTIPLE PAGES Steps for reading multiple pages from ADF 1. Prepare the document to be read and insert it in the scanner's ADF. 2. Select MULTIPLE PAGES and set other parameters. 3. Specify a suitable output file name. 4. Activate the READ function. 5. Press the Enter key or click on O.K. to return to the Main Menu after getting the message: 'All the pages are read'. Steps for reading multiple pages without ADF 1. Prepare the document to be read and insert the first page in the scanner. 2. Select MULTIPLE PAGES and set other parameters. 3. Specify a suitable output file name now or under (7). 4. Activate the READ function. 5. Feed the next page when the message: 'Feed the next page, or click on NO MORE' appears. 6. After feeding or placing each new page press the <Enter> key or click on O.K. 7. Click on NO MORE to return to the Main Menu after reading all the pages. PAGE ORIENTATION Steps for reading a page in landscape format 1. Insert the page in the scanner. 8 2. Select LANDSCAPE DOCUMENT in the Main Menu. 3. Remove any remaining margins or windows. 4. Set all other options and parameters for the page. 5. Activate READ. Don't forget to de-select LANDSCAPE DOCUMENT when you have finished reading such documents to go back to PORTRAIT orientation! DEFINING THE SCANNING AND READING AREA The LAYOUT function allows you to define the paper length, margins and windows on the page. Defining a margin frame: Activate SCAN IMAGE on the LAYOUT screen to scan in the page to see its structure. To define a margin frame, move the cursor into the layout display - the cursor will change its shape to a cross. Move it to the top left corner of the text. Press the <Shift> key, hold it down and at the same time click with the left mouse button and drag the cursor to the bottom right corner of the desired text. A margin frame displayed by continuous lines will appear. Be careful not to let the frame touch any characters. The resulting values for the margins measured from the edges of the page will be seen in the margin boxes (left, right, top and bottom). Deleting a margin frame: If your margin frame does touch some characters or isn't correct, simply define a new frame by dragging the mouse while holding down the <Shift> key again. The previous margin frame will automatically be replaced by the new one. All four margins settings can be reset to zero in one step by clicking on RESET (to the right of the margin boxes). Enlarging the layout display: Activate SCAN IMAGE again without changing the margin frame. Notice that the resulting layout window contains only the area defined by the margins and that the image is enlarged. This gives a better view of the layout and makes it easier to define windows accurately. Note: When reading from an image file PAPER LENGTH cannot be changed. Windows can be defined both for images coming from a scanner and from image files. You can define up to 16 windows and these can overlap one another. Windows are shown with dotted lines. Reading can be executed from inside or outside windows. READING TEXT IN INTERACTIVE MODE Reading text in interactive mode can be done either from scanner or from an image file scanned in earlier. Steps for on-line correcting 1. Start the program. 2. Select the parameter CORRECT AND LEARN in the Main Menu. 3. Select image source FROM SCANNER or FROM FILE. If from image file, specify the correct name. 9 4. Check that the currently loaded CHARACTER SET is the most useful one for your current document. If not load another. 5. Check all other parameters are set correctly. 6. Activate the READ function. 7. When Recognita Plus stops on a character and its suggestion is correct, click on ACCEPT. 8. Recognita Plus then moves on to the next problem character. 9. When Recognita Plus gives an incorrect suggestion or says NONE, give a correction, either from the keyboard, from the on-screen character set or with a code. 10. Check your result, then click on CORRECT. 11. Click on ACCEPT/CORRECT with CORRECT: empty to place a rejection symbol if you cannot give a suitable correction. 12. Click on DROP GARBAGE to remove unwanted characters. 13. Press the <Enter> key, or click on O.K. to go back to the Main Menu after the 'Ready' message appears. SAVING IMAGE FILES Steps for saving an A4 page as an image file 1. Start Recognita Plus. 2. Insert the page into the scanner, making sure the text is not skewed. 3. Activate the LAYOUT function in the Main Menu. 4. Set the margins to zero and leave page length at its default value (11.75 inches/29.7cm) 5. Select the parameter TO FILE: *.TIF. 6. Set the parameter COMPRESSED if you wish. 7. Type in a suitable image file name, or leave *.TIF to call up a dialog box. 8. Activate SCAN IMAGE to scan the page, display it and save it as an image file. 9. Activate BACK TO MAIN MENU. QUITTING When you decide to end your session with Recognita Plus, activate the QUIT function. This saves the current settings in the file RECOGN.STS, quits the program and returns control to the operating system. C/ WORKING UNDER MS WINDOWS READING THE TEST.TIF FILE supplied with Recognita Plus and containing a multi-lingual text (English, German, French, Spanish). To read an image file (TEST.TIF) 1. Set the text parameters Steps for setting text parameters: Select 'Options', then 'Text Parameters...'. Select your word processor or the best format for you. Select 'Original' or 'General' character sizing according to your needs. 10 De-select 'International' and use 'Languages...' to select only the languages required, i.e. German, French and Spanish, then click on 'OK' to close the Languages dialog box. Click on 'OK' to close the Text Parameter dialog box. 2. Select 'File', then 'Open...'. 3. Select TEST.TIF in the Open Image Source dialog box 4. Click on 'Read'. 5. Click on 'OK' or press <Enter> when the message 'Recognition is done' appears. Displaying layout, reading from bitmap The layout of the whole scanned area will be displayed in the 'Overview' window and at the same time its original bitmap either in full-size (100%) or reduced to 50% or 25% can be seen in the 'Layout' window. A coloured area (the so-called positioning rectangle) in the 'Overview' window indicates which part of the whole image is being currently displayed in the 'Layout' window. The boxes cursor is the default cursor shape in the 'Layout' window. If you point with the cursor on the marked area in the 'Overview' window it will change its shape into a four-headed arrow. Use this cursor to move the marked area, in order to see a different part of the bitmap in the 'Layout' window. To see the bitmap of the marked area at full-size or reduced to 50% or 25%, use the appropriate tool or select the desired percentage in the 'Show' menu. To display the TEST.TIF file and read from the bitmap 1. Set the text parameters (languages, word processor, etc.) 2. Select 'File', then 'Open...'. 3. Select TEST.TIF in the Open Image Source dialog box 4. Click on 'Open As Image' in the Open Image Source dialog box to display the image file on screen. 5. Define the reading area with boxes, if you wish to read only a part of the image file, select Text Inside or Text Outside Boxes in the 'Boxes' menu. 6. Select 'CurrentBitmap', then 'Read'. 7. Click on 'OK' or press <Enter> when the message 'Recognition is done' appears. 8. Select 'File', then 'New' to clear the screen. 9. Select 'Remove All' in the 'Boxes' menu of the 'Layout' window to remove box(es) and close the 'Layout' window. TO READ A PAGE FROM SCANNER 1. Place the page into the scanner in the position required by the scanner model you use (e.g. sheet-feed or flat-bed). Check that the text is not skewed. 2. Set the scanner parameters Steps for setting scanner parameters: Select 'Scan', then 'Test' to examine the typographical quality of the document Select 'ReadMore!' to scan the next slice of the bitmap Enlarge some characters by moving the cursor on a character and clicking, or by dragging the mouse 11 Examine the typographical quality (line thickness, filling of holes, touching or broken characters, etc.) of the character enlarged in the 'Zoom In' window Select 'ScannerParams...' Adjust 'Brightness' and 'Contrast' If the character shapes are very thin, or even broken, adjust 'Brigthness' to the left on the scale in order to 'darken' it. Conversely, if the characters are too thick, so they run into each other, or some holes inside them are filled, adjust the parameter in the opposite direction, to 'lighten' it. Select 'ReScan!' Repeat the steps from 'ReadMore!' to see if the new settings have improved the result Select 'Eject!' to finish the testing process Close the 'Test Scan' window 3. Set the text parameters (languages, word processor, orientation, etc.) 4. Select 'Recognition' 5. Select 'Read' 6. Edit the text on screen if you wish (provided you have chosen one of the ASCII formats) Insert and/or delete characters from the keyboard using the direction and text editing keys Select 'Search', then select one of the editing functions according to your needs Type in the characters you want to find or change from and to in the resulting dialog box, then click on the appropriate function key ('OK', 'Change', 'Change All') TO DISPLAY THE LAYOUT OF A PAGE FROM SCANNER 1. Select 'Scan', 'Layout', 'ScanImage', 'DoScan' 2. Use the four-headed cursor shape to move the marked area in the 'Overview' window. 3. Select the desired percentage to see the bitmap of the marked area. To define the scanning area (paper length, margins) from the Layout window: 1. Use the Margins tool to get the margins cursor and define a margin frame in the 'Overview' or 'Layout' window. The margins will be displayed with thick blue lines. 2. Drag the side or the corner of the margin frame to resize it in the 'Layout' window, or define a new one, if necessary, or 1. Select 'ScanImage', then 'Scanning Area...' 2. Enter the new values for paper length, margins in the dialog box. 3. Select 'ScanImage' and 'DoScan' to scan the paper again. Note: Margins are valid only for reading from scanner. 12 To define the reading area (inside or outside boxes) 1. Use the Boxes tool to get a boxes cursor 2. Drag boxes on the scanned bitmap, resize them, if necessary To resize boxes, drag the appropriate side of the box or its corner. 3. Use Eraser or Remove All tool to remove boxes, or Select 'Remove Box' or 'Remove All' in the 'Boxes' menu To delete one box move the box eraser cursor inside the box to be deleted, then click 4. Use the alternative Text In/Out tool, or Checkmark 'Text Inside Boxes' or 'Text Outside Boxes' in 'Boxes' menu according to your needs Note: Boxes can be defined both for images coming from a scanner and from image files. TO READ A MULTI-LINGUAL DOCUMENT Select 'International' in 'Language Set' of the Text Parameters dialog box or 1. De-select 'International' 2. Use 'Languages...' to select only languages required. 3. Use the text box of 'Language Set' to enter additionally up to 25 accented characters to include them in the subset of recognizable characters. TO READ FINANCIAL TABLES, BALANCE SHEETS, etc. 1. Select 'Numbers Only' in 'Enabled Characters' of the Text Parameters dialog box 2. Use the text box to include up to 16 additional characters TO READ DOCUMENTS IN INTERACTIVE MODE 1. Set text parameters before reading (text format, character set, languages, orientation, etc. If you read from scanner, set scanner parameters, too.) 2. Select 'Recognition', then 'Read and Correct/Learn' to start interactive reading from scanner. 3. In the 'Read and Correct/Learn' dialog box: Click on 'Accept' if the suggestion is correct. Click on 'Zoom In' to enlarge a problem character. Click on 'Show Char. Set...' to transfer character(s) from the on-screen character set to the text box, or enter the correct character(s) in the text box, if there is no suggestion or the suggestion is wrong, then click on 'Correct'. Click on 'Reject' to insert the rejection symbol. Click on 'Drop Garbage' to delete unwanted character(s). 13 Note: Reading text in interactive mode can be done either from scanner or from an image file scanned in earlier, or from the currently scanned bitmap. To end your session with Recognita Plus 1. Select 'File' 2. Select 'Exit' Note. When leaving Recognita Plus, it will automatically save the current parameter settings into the RECOGN.PAR file, provided some parameter has been changed during your work. The RECOGN.PAR file loads these settings when you start Recognita Plus next time. If the file RECOGN.PAR does not exist when you start the program, the delivery default settings will be loaded. * * * * * * * * * * * * End of MANUAL.DOC.