|
Vol. 5, No 3 May/June 1997 |
Subscribe to AT Messenger Download PDF Viewer |
Making Computers Read
Ed Salisbury
Kent County ATRC
The computer can be an invaluable tool for giving people with visual impairments, as well as those with reading difficulties, access to printed material. The computer, with the addition of software and a few peripherals, can read books, magazines, newspaper articles, and even the daily mail using computer generated speech. Although computerized reading systems have been available for many years, recent improvements in technology have made them smaller, faster, more accurate, and less expensive. Because of the variety of products available and a number of different options to consider, the purchase of a reading system can be somewhat confusing.
All computerized reading systems consist of five major components, the first being a computer. Although an IBM or compatible computer is most often used for this type of application, a Macintosh can be used as well. Memory and processor requirements will vary depending on the complexity of the reading system that is chosen.
The second component is a scanner. A scanner, somewhat resembling a small photocopier, is used to import (scan) an image into the computer. This image can be a photograph, drawing, or a page from a book or magazine. Once the image is scanned, it is visible on the computer screen. This image can then be printed or manipulated using a software program. Unfortunately, a page of scanned text is acquired as an image—a picture of the page—thus making it inaccessible to a person with a visual impairment. Because it is only a picture, it cannot be directly read to the operator, nor can it be added to or edited by the operator who is using a word processing program. Common scanners include a variety of models from Hewlett Packard, Epson, Umax, Microtek, and others. Scanners range in price from around $300 to $2000 depending on speed, resolution, bit depth, and color accuracy. For text scanning, any of the entry level scanners are sufficient.
The third component is used to change the image of the page into text that is recognizable by a word processor as well as a screen reading program. Screen reading programs will be discussed later. This component is referred to as optical character recognition (OCR) software. OCR software converts the image to editable text. If the text on the page is of an ornate type style, the OCR software may get confused and recognize an “I” as an “L”, a “c” as an “o”, an “m” as an “n”, etc. The clarity of the original copy will also affect the quality of the recognition. Stray marks on the original may be interpreted as letters or punctuation. Common OCR software programs include TextBridge from Xerox and OmniPage from Caere. OCR programs range in price from $80 to $500. Many scanners and fax modems include a limited version of OCR software.
The fourth component is a speech synthesizer. A speech synthesizer is a combination of hardware and software that enables the computer to speak. Many inexpensive speech synthesizers sound very robotic or lack accuracy in pronunciation. Better synthesizers closely resemble human voice in accuracy, clarity, and inflection. Common speech synthesizers include Accent, Doubletalk, Keynote, and DECTalk. They range in price from around $200 to over $1000.
The fifth component acts as a liaison between the OCR software and the speech synthesizer. This software, called screen reading (or screen review) software reads the converted text, then sends instructions to the speech synthesizer to have the text spoken. The operator is able to navigate through a document using the mouse or keyboard, having only those parts of the document read that he or shewants. Screen reading programs also read text in spreadsheets, word processors, databases, and any other software that presents information to the operator in a text format. Because screen reading programs are specifically designed to be used by people with visual impairments, they also read menu selections, dialog boxes, and error messages. Common screen reading programs include JAWS, OutSpoken, VocalEyes, and IBM Screen Reader. Screen reading software can be purchased for as little as $25; fully featured packages sell for $700 or more.
If all of this sounds rather complicated, there are alternatives to piecing together a system. Xerox, Kurzweil, and Arkenstone are three companies marketing reading systems that work with an IBM or compatible computer. In these systems, OCR and screen reading software are combined. They are generally sold with a choice of scanners and other adaptive aids.and speech synthesizers. The Omni 1000 and Omni 3000 from Kurzweil Educational Systems also include a software based speech synthesizer that works with the computer’s sound card. All of these systems include a seamless interface designed for people with visual impairments. Several companies, including Xerox and Arkenstone, produce self-contained models. These models include a computer processor, keypad, scanner, OCR and screen reading software, and speech synthesizer. These systems are somewhat portable and highly specialized for reading printed material. In other words, it would be difficult to play games or install other software on one of the self-contained systems. Although generally more expensive than piecing together a reading system, self-contained systems offer a easy to use interface and guaranteed compatibility among components.
With improvements in technology, computerized reading systems have made the transition from large, expensive, robotic-sounding machines to a natural, affordable, and easy-to-use tool for many people. To find out more about any of the reading systems listed above, or to see a demonstration, contact your local ATRC.
