In general, the consensus on digital maps is one of value with only minor drawbacks. For instance, the Oversized Color Images Project at Columbia University arrived at the conclusion, after experimenting with several different scanning resolutions, that "the highest quality printouts show that fine details can be captured successfully from [paper original, microfiche, and transparency]." Their results also showed a color shift "in all printouts and in all images as viewed at a variety of monitors." So, the original and its scanned offspring differed in what they communicated. A second example will illustrate a common concern with digital maps. A review of the World Atlas and United States Atlas by The Software Toolworks, Inc., an electronic atlas, found that colors displayed on the screen were "illogical." However, the hard-copy printouts were of better quality. Again, the improper image communicated inaccurate information.
The Oversized Color Images Project, which intimately investigated acceptable preservation and digital access techniques for dealing with oversized color images associated with text, arrived at conclusions that are both positive but also disconcerting. The positive findings continue with the claim that on the maps scanned, the 1 mm elevation marks are legible at high resolution, and all color pattern codes can be distinguished. The investigators claimed a useable product. However, in reading their results, one finds clues that difficulties are present in actually using the images.
Firstly, the use of different equipment results in different images. Several vendors provided images to be inspected, and some were more in focus than others. If differences are possible, is this really communicating properly or being true to the original? Secondly, the size of an image in terms of storage size results in slow retrieval time using a Web browser and lengthy waits are common while transferring files across the Internet. Thirdly, returning to the color difference among equipment, the researchers suggest a solution: calibrate the equipment with standard colors and bring the scanned image into line with the original. As the sophistication of software packages increases, the color issue will be less of a problem. What is still a problem is file size and resulting display time. The project, more than simply for preservation purposes, also sought to provide access over the Internet to scholars. In feed back comments, users noted the slow retrieval time of such large images. A fifth issue, even though the principal investigators claimed adequate success, most maps, if in high enough detail, are not viewable all at once on a computer screen. One is forced to scroll around the image.
The purpose of a map is to communicate information about some spatial data. Much study and research has been invested in user needs and usage of maps. Viewing the map in its entirety, with peripheral stimulus, makes for the best use of the information presented. However, needing to pan around the screen in order to view the map causes a new sensation in perception and effectiveness of the map. The alternatives are making the file size smaller, thereby reducing the quality of the image and making it unusable because of illegible cartographic marks. Others have suggested "mosaic-ing" a map into segments. This reduces the load time, but breaks apart the map, which was intended to be viewed as one unit.
When file sizes are too large or a map image needs touch up, tools are available to alter the image. Ideally, a reduction in file size would not have a subsequent result in reduced image quality. Quality here is defined as legibility of text, distinguishing colors, and legibility of fine lines. Compression options, such as lossless GIF and lossy JPEG, were used by the Oversized Color Images Project with varying success. Color images remained quite large. The greater reduction by compression or the reduction of dots per inch (dpi), again, make a slightly inferior image.
A second processing tool are arithmetic algorithms which actually reduce the amount of data. Using a GIS, the operator can generalize the map by using a data point reduction algorithm. Eliminating data points is necessary when the scan produces a rough line instead of the desired clean line. Dropping out data points does save file space, because the data base has less information in it, but this process is at the expense of the original features. In clear terms, GIS specialist Zhilin Li says, "data point reduction algorithms should not be used for generalization purposes."
Displaying spatial data on a computer screen, using a GIS, takes the form of vector or raster files. For communication purposes, raster images, which are series of dots in a matrix pattern, provide a superior product than vector images, which are straight lines. Now, depending on the intended audience and the data being mapped, a vector presentation might be adequate. But generally speaking, a raster image, especially an air photo, grabs the user because of recognizable features on the earth, instead of graphic representations in the form of lines. With proper sophistication and computing power, raster data can actually be integrated with vector data to produce a highly effective map. As expected, there is a price. Raster data requires more storage. On the positive side, printers and plotters favor the dot pattern over the line pattern in vector data. The problems of fitting an image onto one screen remains, but printouts on quality plotters are clean, clear, and readable. However, surveys do show that GIS users most commonly produce a map on-screen, rather than as a hard copy printout.
Although the majority of digital maps are viewed and manipulated on a computer screen, users do need a hard copy printout for publication or other presentation purposes. Just as colors, text, and lines have varying degrees of communicative success on the screen, the printouts possess similar drawbacks. Basic to images, maps displayed on a screen are achieved with the additive colors of Red, Green, Blue (RGB), but printed with the subtractive colors Cyan, Magenta, Yellow, and Black (CYMK) using device drivers to translate colors from RGB to CYMK. It is in this translation process that colors represented on the screen do not correspond to the hard copy output.
Even though digital maps earn the Wow! factor from users, I am wary, at this point, of claiming the electronic versions to be superior to the traditional paper format. The limitation placed on communicating information resulting from small computer screens reduces the effectiveness of a map. Users value viewing the entire map at once. Even though they do not focus on the whole map at once, but just one section, users can reference what they are looking at based on the surrounding map. Reading a digital map limits the user to one screen at a time. The 21 inch monitors available are a great improvement over 14 inch monitors, but until technological advancements can offer an inexpensive, large monitor to the general public, reading digital maps will continue to be inadequate.
Another barrier preventing digital map images from being truly effective is the file size. The file size issue is slowly becoming moot, but accessing such large files, for now, is cumbersome for a user. In the author's own experience with map images, text is most often illegible. Alternatively, choropleth and other thematic maps are useable, but attempting to read anything with small font sizes only results in frustration.
Digital maps are here to stay, this is not in doubt. What is questioned, though, is the quality of maps at present. Scanning a paper map for the purpose of viewing on a computer screen is not, in the author's view, acceptable for communicating spatial information. As file sizes decrease and screen sizes increase, then digital maps will proliferate and be widely accepted as quality communicators of accurate information.
Authored by Randall b.
Kemp : email@example.com
For ILS 603: Image Database