Applications of Optical Scanners in an Academic Center
DR. CAROL MOLINARI, Assistant Professor and DR. ROBERT S. TANNENBAUM, Director of Academic Computing Services University of Kentucky Lexington, Ky. As faculty members of a state university that strives to meet the tripartite mission of research, instruction and service, we find that information technology is fast becoming a strategic resource needed to balance these academic demands. In particular, the University of Kentucky's academic computing service group is emerging as a critical technological hub that can direct and keep its faculty on a steady course down the information superhighway. The purpose of this paper is to examine an example of information technology #172;optical scanners #172;by describing how the technology works, as well as its applications in data management and research, development of instructional materials, and providing community services. The university has six scanners, connected to a variety of computers, available to our faculty in two oncampus centers. Basics on Scanners Optical scanners can copy printed text or graphics into an electronic file, which can then be examined and analyzed. Such scanners are fast becoming a key technological resource because they are easy to use; reduce the time and errors associated with adding new information; and are relatively inexpensive. There are three basic types of scanners: OCR, optical character recognition scanners, which "read" printed text from a printed medium; OMR, optical mark readers, which scan marks on specific areas of the page; and Graphic scanners, which copy, enlarge or reduce images from paper or film. Scanners can be used for a wide variety of purposes, limited only by one's needs and imagination. To understand, categorize scanning in two ways—either by the type of input or by the nature of the input medium. The most common types of input are characters, graphic images and marks. The common input media are paper and film, either radiographs or slides. We will consider each of these briefly. Optical Character Recognition (OCR) Often we have a printed copy of a text, but would like to have it in digitized form for computer analysis, editing, formatting, electronic transmission or any of several other uses. Such a text can be typed in using the keyboard, but this may be a laborious, time-consuming process, particularly if one is not a skilled typist or if the text is long or complex. An attractive alternative is to use a scanner and optical character recognition (OCR) software. The scanner digitizes the text image from the page and the OCR software interprets that image, translating the characters into their ASCII equivalents and storing these in a file. There are potential difficulties with using OCR software. The primary problem is accuracy. If the original printed text is clear and crisp, and printed in a standard font, then the resulting ASCII file will probably be quite accurate with only a few errors. Such errors can usually be caught and corrected by visual proofreading or a standard spelling-checking program. If, however, the original is not crisp, or if unusual or multiple fonts are included, the resulting ASCII file will require considerable proofing and correction. Some OCR software contain a provision for "training" the program to recognize special fonts and special characters. If this is necessary for certain original documents, it will add to both software costs and to the time required for the scan. The second possible problem with OCR software is that formatting information contained in the original document is usually lost. For example, line lengths, indentations, pagination, columns, etc. are not normally retained. Some OCR software contain an option to try to reconstruct the original's format in the new ASCII file, but even that is not a sure bet. Of course, as with the proofreading, formatting can be also corrected by hand after the OCR process. Digitizing Graphic Images Perhaps the most common use of a scanner is to convert an image from its current physical medium (paper or film) to a digital representation stored electronically within a computer. Actual contents of the image are irrelevant to the process. What is scanned may contain letters forming words, colored inks creating a picture, photographic chemicals representing an X-ray or any other image. No matter the content, the scanning process subdivides the image into an arbitrary number of tiny dots, called pixels. The number of dots per inch (dpi), is termed the resolution. The higher the resolution, that is, the more dots per inch, the more precise and detailed will be the digital image. However, the higher the resolution, the larger the electronic file produced, increasing both the storage requirements and the processing time. While scanning, most programs allow some preliminary manipulation of the image. You may be able to crop the image to digitize only the portion of interest. You may also be able to enhance the image slightly, by adjusting the scanner in a manner analogous to the way one would focus a camera lens before shooting a picture. And you can often see a preview of the digitized image, at lower resolution, in order to decide if you have made all of the adjustments correctly before proceeding with the final scanning. Optical Mark Reading (OMR) Another important form of scanning is optical mark reading (OMR), sometimes called mark sense reading. OMR is used to digitize choices made to multiple-choice items, such as on a test, survey or questionnaire. In OMR, the unit scans for marks that have been made in specified areas of a page. The exact location is the significant information. What the scanner is actually recording is only a binary piece of information, either the presence or absence of a mark in a particular page location. Software controlling an OMR scan allows the operator to specify the locations on the page to be scanned and the codes to be recorded in the output file as a result of what the scan "sees" at those locations. For example, a scan might be performed on an answer sheet for a standardized test. This sheet might have 100 or more areas (called fields), corresponding to the 100 or more test items. Each field might have five spaces to be marked. The program translates the presence of a mark into a number or letter code that indicates the respondent's answer. Most programs are sophisticated enough to recognize when more than one choice has been marked for a single item and to record an error code. Some programs can even perform simple test scoring and item analyses, if given the key for the test. The file of data produced from an OMR scan is simply a digital record of the responses recorded on each sheet scanned. Data may represent the results of a survey, answers to a test administered to a class or other categorical data. One then processes this file using standard statistics, test scoring, survey analysis or other appropriate software. Most scanning is done from paper originals. The simplest to scan are single, flat sheets. Most flatbed scanners can accommodate any page up to 11" x 14". Some can handle even larger sheets. For OCR scanning of a manuscript, for example, one normally needs to place paper manually, one sheet at a time, on the scanner. OMR scanners, on the other hand, generally have multiple-page feeding mechanisms to scan an entire set of answer sheets. Books and bound periodicals will present a problem if the entire page cannot be laid flat on the scanner without damaging the binding. If doing so might cause harm, such as to an ancient manuscript, there are scanners that elevate a camera above the page to scan by looking down. These cameras have a depth of focus great enough to compensate for parts of the page that are at different distances from the lens, such as might result from a rippled parchment or a thickly bound book. Input from film that has been printed to photographic paper, such as a personal snapshot, can be scanned directly as you would scan any other piece of paper. Photographic images still on film can also be scanned, with appropriate equipment. Adapters are available to allow one to digitize 35mm slides and radiographs of X-rays or other types of scans, to create digital images. The resulting digital files are analogous to those created from paper images. The input medium d'es not alter the output file; you will still have a file with gray scale or color information encoded for each pixel in the format you have chosen, and you will be able to process the image file in the same ways as any other such file. Scanning Applications in Research For many university faculty, a major issue associated with research activities centers on data collection. Data collection is a very costly component in research. While the supply of external funding remains static, at best, demand continues to escalate. Having an established research record provides a competitive advantage to some. Junior faculty with rigorous research requirements for promotion and tenure, for example, are often at a significant disadvantage when trying to obtain needed funds. One strategy may be to use secondary data for one's research, which circumvents high data-collection costs. An OCR is extremely useful in developing secondary data sets from existing sources. For example, an OCR can scan and append data from different data sources to create and/or customize a new data base. Creating a working research file may require the merging of two or more existing electronic data sets as well as appending information from several printed sources. For example, in my [author Molinari] research on hospital governing boards, I obtained two data sets—one with a national sample of hospital governance information and the other holding financial data for all hospitals in California. By merging them, I obtained a data set containing both governance and financial data for a sample of California hospitals. This resulted in a unique governance data set with a wide array of performance financial outcomes. However, there were several important hospital characteristics lacking in this data set: system affiliation, level of competition and region (urban or rural). These were available in printed form from other sources. Using an OCR, this information was scanned into the data set. This two-pronged scanning approach customized the basic data files, enabling me to pursue research questions and inquiries that were, heretofore, not possible. As a bonus, the customization of this data set helped me attract colleagues for collaboration. For primary data collection and analysis, OMR scanners are efficient tools. For example, if subjects are given a bubblesheet for recording their answers to a written questionnaire or survey, responses can be input quickly and accurately, with minimum time and costs. OMRs also have obvious and well documented applications for testing and evaluation. Using scanners to develop data sets from secondary data can directly or indirectly lead to investigations in new research areas. Such techniques serve the needs of faculty conducting empirical research within the fiscal constraints of the university. Leveraging the investments in basic research is increasingly important in these strict budgetary times. Applications to Enhance Instruction To promote teaching and enhance pre-sentations, several scanners are useful. Developing materials for instruction and presentation (like overhead transparencies) is facilitated by both graphic scanners and OCRs, which incorporate images and text, respectively. For example, pictures can be scanned into an overhead or onto printed notes. A slide scanner, which handles 35mm slides, will especially expedite conversion of text into slides for presentations. OMR scanners, for their part, facilitate grading and analyzing test results, as well as student evaluation and feedback. By facilitating easy development of visual materials, scanners go a long way toward enhancing instructional quality. Applications That Extend Service Telemedicine is a relevant example of how scanning technology facilitates the university's community service to under-served areas around the state. Graphic scanners enable remote, rural populations to access specialized diagnostic and therapeutic services. In conjunction with interactive video, scanners can assist in transmitting medical data over long distances. In fact, scanners can provide more information to medical personnel because of their ability to magnify or crop parts of X-rays, thus enhancing the quality of care. For example, radiography scans X-rays from patients living in outlying areas, which are then read by radiologists at the University of Kentucky's Medical Center in Lexington. By helping provide access to quality medical care, scanners significantly contribute to this university's commitment to serve communities (especially in rural areas) throughout the state. On Balance As with all emerging forms of technology, there are cost limitations and legal constraints associated with optical scanners. Equipment and personnel costs need to be assessed. Luckily, scanners are not a fad and, once bought, will offer years of productivity. Longevity helps amortize the units' initial cost. Similarly, the skills required of personnel to use scanning technology are relatively quick and easy to acquire. Lastly, both the hardware and software are becoming more user-friendly in design and operation. It should be noted that legal questions in terms of copyright restrictions, raised by copying information from other sources, have yet to be resolved. While universities should have some guidelines regarding copyright abuse, common sense and a conservative approach should suffice in most cases. The bottom line is simple. University faculty are prudent to tap these technological resources of their academic computing center by learning how to use and apply different scanners in their teaching, research and service activities. Carol Molinari is an assistant professor in health administration at the University of Kentucky in Lexington. E-mail:
[email protected] Robert Tannenbaum is director of Academic Computing Services at the University of Kentucky. Reminder: Safety Considerations for Using Computers Computer use by all students, from elementary school through college, continues to grow. As the hours that students spend on computers increase, so do the possibilities of physical problems associated with prolonged use. n Physical Perils It's a fact that today's typical computer game can take up to 100 hours to complete. Add that time to the amount children spend working, not playing, on the computer—completing homework assignments, writing letters or designing artwork—and you have many hours of intense visual concentration, rigid body postures and potential problems with one's back, wrists, legs or other muscular aches and pains. Over time, there's the possibility that these aches and pains can become chronic. In addition to school, home computers are becoming a standard in many families. In fact, if current trends continue, the number of American homes with PCs could increase to well over 26 million by the year 2000. And 44% of the current computer market is made up of married couples with children under the age of 18—students. n What to Look for in Furniture What kinds of issues should teachers be aware of as pupils in their classroom work or play on the computer each day? "Computers are frequently part of a work station," says Nancy Osterman, ergonomics specialist at Quill Corp., the nation's largest independent direct marketer of office products. "That's why it's important to choose computer accessories that are adjustable for every child who will be using the computer." If computer work stations are adjusted for adults, children may find themselves perched on the edge of a chair that's too large for them, legs dangling. In selecting a chair for a work station that will be used by children, choose one that adjusts quickly to the height and body size of the child. Look for chairs marked "ergonomic" and teach your pupils how to readjust the chair for proper fit each time they sit there. Also consider the surface holding the keyboard and mouse. Ideally, its height will adjust to the user. The work station layout should allow children to move around a bit while seated, and be organized to reduce excessive reaching or awkward postures. In addition, look for work stations with rounded edges. They will reduce the contact pressures that sharp edges can put on a child's arms and hands. n Minimizing Eye Fatigue Another good tip from Osterman is to center the monitor directly in front of and at least an arm's distance away from the child. Select a monitor that is adjustable in height and angle for each child's requirements. If possible, adjust the screen's background colors to be consistent with the walls' colors or other features of the room. Following all of these measures will minimize eye fatigue. Teachers should also encourage their students to take short, but frequent, "mini breaks" from the computer. Even getting up and walking to another part of the room or out for a drink of water gives eyes time to rest and recover. "It's important for teachers and parents to help children practice these routines each time they sit down at the computer to play or work," says Osterman. "Using the computer 'safely' not only will increase the benefits to kids, but make their bodies thankful as well." For more on working safely at the computer, send for a free booklet entitled "Easy 8-Step Ergonomics Guide." Although written for the office environment, many of its tips are easily applied to other settings. Write to Quill Corp. Public Relations Dept., 100 Schelter Rd., Lincolnshire, IL 60069-3621.