The ‘flatbed’ scanner can take on a wide variety roles in a public library setting. In combination with email software and Internet connectivity, it can function as an alternative to the fax for transmitting material electronically. Combined with a printer, it can serve as a high-quality copier. It makes a surprisingly good still camera for taking digital pictures of relatively flat three-dimensional objects. With optical recognition software (OCR), a scanner can transform a page of text into an editable text file. It can be used to digitally record artwork and photographs for electronic storage or transmission. In fact, the scanner is such a basic tool that its usefulness is mainly limited by one’s imagination.
Your basic flat-bed scanner is a boxy looking thing with a hindged lid, covering a glass ‘bed’ that resembles the top surface of a small photocopier. Inside the box is a fixed-focused camera (usually of the charge-coupled device or CCD breed) and a powerful lamp, both rigged to move across the scan bed in a precise manner.
One way to appreciate how the scanner does its thing is to take a piece of tracing paper, draw 72 parallel lines per inch across it both horizontally and vertically, and then tape this fine grid over a color photograph. Pick out one of the tiny squares of this grid, then from a package of 256 colored pencils choose the color that best matches the part of the photo visible through the square. Fill in the entire square with this color, then repeat the process until the entire grid has been colored. A scanner that has been set for this same resolution (resolution is usually expressed dpi or ‘dots per inch’) and color ‘depth’ will interpret what it sees in a similar way, though the scanner’s job stops once it has assigned a number value for the color of each pixel. Within the capabilities of the scanner's camera, the greater the resolution and color depth, the more closely the recorded data will resemble the original.
A significant advantage of digital imaging is that files can be transmitted many times without sacraficing original image quality. The image may appear different when shown on different monitors or printers, but the original image data is preserved in the file.
Scanning something creates a digital file, which can be thought of as a set of instructions for reconstructing a facsimilie image. One place you can display a reconstructed digital image is on a computer monitor. Since you can move digital files between computers by attaching them to an electronic mail message, sending images by email is an alternative to sending them by fax. Some advantages to using in this method is the cost savings of email vs. long-distance telephone, the option of sending a file attachment at a higher resolution and/or in color, and the fact that they can be sent while a dedicated fax line is busy. Also, many more people own personal computers than fax machines. On the other hand, fax machines are more convenient, because converting an original image into an email attachment is a somewhat involved process. People also tend to check their fax machines more frequently than their email.
A scanner can also serve as a useful companion to a photocopier. Photocopy machines are designed to produce high-contrast copies quickly. Scanners produce high-quality images that can be manipulated in ways that will make them much easier to reproduce. In the example below, a very dark, low-contrast color original was scanned, and the colors were exaggerated to maximize contrast (in this case, green was maximized while red and blue were supressed); the image was then converted to bi-tonal B&W and a negative filter was used to turn the finished image into black lettering on a white background. I know this sounds involved, but the whole procedure took less than five minutes to accomplish.
The combination of scanner and printer can also serve as an economical alternative to a color photocopy machine.
The flatbed scanner has a fixed focus, which means that it is focused to a single preset distance from the camera’s lens. With any lens, there will a minimum and a maximum range at which objects will be in focus; the area between these extremes is known as the ‘depth of field’ of a lens. Rather surprisingly, the field depth on many flatbed scanners is several inches. What this means is that relatively small three-dimensional objects (jewelry, coins, silverware, knives, insignia, watches, etc.) can be conveniently photographed within the library or office. It can also capture details from larger objects, such as manufacturer’s marks. This is a tremendous boon when referring a reference question -- a scanned image of a small object is far clearer than typical 35mm or Polaroid photos. Moreover, this again results in a convenient digital file that can be forwarded to experts by email.
Using a scanner as a still camera is not difficult, but some experimentation will be needed to insure good results. If the scanner's cover doesn't fit over the 3D object very well, try using an oversized piece of off-white poster board positioned a few inches above the print bed (if you try scanning without any cover, the resulting image will be very dark and lacking in contrast.) With the scanners I’ve used in this manner, the original file usually requires adjustments to brightness, contrast and color. Scanners come bundled with relatively simple software capable of making these adjustments. The image shown below was created in under five minutes using a scanner and image manipulation software.
A standard component of scanner software bundles is one or more OCR applications. Depending on the particular program, the scanner first makes fairly low-resolution bi-tonal b&w scan of the text, then runs this through OCR and outputs the results to a text file (usually either ASCII or .rtf). The process is simple and quick from the operator’s perspective, but there are limits to OCR technology that you need to understand. The accuracy of the character recognition process is heavily dependant on the quality of the original image, and to a much lesser extent on the original typeface.
The following two examples illustrate the problems and potential of OCR. In the first example, the original is a photocopy of a carbon paper duplicate of a typescript. By clicking on the image, you can see that the original text appears quite readable; on the other hand, it is a nightmare for the OCR program to interpret because of the many thick and thin letters found throughout the page. It would be more efficient to input this document by hand than to proofread and correct the dozens of errors produced in OCR. By contrast, compare the original and OCR texts on the second example -- the software produced very few errors here.
OCR text, Early S. Daugherty
OCR text, Thread of Life
Scanning and digital archiving has become a standard approach for preserving visual information. Once converted to an uncompressed digital file, a photographic image can be preserved in a very small, convenient format. The current storage format of choice is CD Rom, which is common, cheap and relatively stable; unless something better comes along, DVD Rom (which has backwardscompatability with CD Rom) will probably be the next big thing. The subject of digital preservation will be discussed further in a later article.
Just for grins, here's an image of my latest catscan.
Dave DeLaurant, SJVIS
