|
|
|
Education/Training
|
The Impact of New Technologies on Remote Sensing Education
Kurt T. Rudahl and Sally E. Goldin (USA)
Goldin-Rudahl Systems Inc.
6 University Drive #213, Amherst, MA, USA
E-mail : 0003086210@mcimail.com
The need for Education in Remote Sensing
Here at the 15th Asian Conference on Remote Sensing, one hardly needs two argue the importance of education in remote sensing and related disciplines. The conference program has included session on educational topics since its early years. Many conference attendees are actively involved in education or training, either in formal academic programs or in applied, professionally oriented settings.
In fact, the importance of effective wide, spread education concerning remote sensing and related topics had never been greater. This is primarily due to two factors: the multi faceted environmental degradation currently threatening the earth and its people, and the economic significance to each country of intelligently monitoring and administering their natural resources.
Few people dispute that the plane is or soon will be in a state of environmental crisis. Pollution, population growth exceeding the support capability of the land, loss of biodiversity and global climate change are only few of the problems that face today’s and tomorrow’s generations. Remote sensing and related technologies can contribute to our understanding of these problems as well as tot eh implementation of practical solutions.
Meanwhile countries are becoming increasingly aware that long term economic viability requires a balanced, informed strategy for the exploitation of natural resources forecasting agricultural yields, and developing new energy sources.
In order to meet these challenges, we need professional who are familiar with the capabilities and limitations of the geospatial technologies. This includes researchers who can develop new methodologies and applications, as well as technicians and practitioners who can apply proven techniques to specify problems.
Even as our need for trained personnel increases, the technological bases for these disciplines have also been expanding rapidly. Remote sensing began such as visual interpretation of aerial photography. Although significant time and training were required to become a skilled photo interpreter, the instruments and tools involved were for the most part accessible and familiar.
Today, visual interpretation skills are still important but to extract the vast amount of information contained in modern remotely sensed data, an array of other knowledge is needed: knowledge of the characteristics, advantages and limitations of the characteristics, advantages and limitations of the many different data sources available; familiarity with digital image processing techniques and algorithms; basic competence in mathematics and statistical analysis; at least some knowledge of computer hardware and software; and of course and understanding of the physical or biological processes under examination. Even the letter has become more complex, as we find an increasing need to understand interactions among natural processes and systems, rather than isolated phenomena.
Thus, it is more important than ever that remote sensing educational programs provide an adequate exposure to modern technologies, as well as solid grounding in remote sensing theory and principles.
Traditional Obstacles
In the Author’s ten years of involvement with remote sensing education, we have seen many changes and improvements. Until recently, however, efforts to provide comprehensive education or training in remote sensing faced a variety of serious obstacles:
- Adequate powerful computer hardware was too expensive and too complex for educational organizations to maintain particularly since effective hands on training requires a low student to compute ratio. Affordable hardware did not offer sufficient power to provide a realistic introduction to the relevant technologies. Furthermore, the available computers were too bulky and delicate to permit use on site in field exercises or even (in some cases)
- Remote sensing/GIS software was complex cryptic, and difficult to use, particularly for individuals with only modest computer experience and for individuals whose native language was not English. In addition, software was not available for the affordable platforms most appropriate to educational settings.
- Digital imagery was expensive and difficult to acquire. Even if imagery were made available through donation partnership arrangements, capacity limitations of small affordable computers precluded using the data effectively.
- Remote sensing education was difficult to justify and fund. The importance of learning how to analysis pictures from earth orbiting satellites was not always obvious to non-technical government decision makers strugging to meet the needs of their people with limited financial resources.
New (or Newly Affordable) Technologies
None of these problems has disappeared. During the last few years, however, there have been developments in the computer industry and in the remote sensing / GIS community that are beginning to have a significant positive impact, and which we expect to become even more important in the future.
Table 1
Problem:
Need affordable, adequately-powerful hardware
New Solutions:
Full-color graphics boards for $200 and under
Personal computer price/performance (speed, storage, memory capacity)
Problem:
Need accessible, educationally-appropriate siftware
New Solutions
Remote sensing and GIS software designed for personal computers (eg. Dragon/ips)
Internationalization technologies
Graphical user interfaces
Problem:
Need affordable digital imagery & ways to manage it.
New Solutions:
Affordable scanners frame grabbers, digital cameras
CD-ROM technology
Write-your-own CD-ROMs and Photo-CD
Problem:
Need to justify and fund remote sensing
New Solutions:
Public awareness of remote sensing
Geographic Information Systems
In the sections that follow, we examine each of these developments in more detail, clarifying how each can make remote sensing education more effective.
The Incredibly Cheap PC
Most readers are aware that remarkably powerful micrometer hardware is now available at very low prices and we do not wish to belabor this point. However, many may not have considered the specific implications of these market facts for education.
Assuming a fixed budget for purchasing a computer (and ignoring the influence of inflation). A small computer such as the 1986 model shows, provided adequate resources for book keeping and work processing. The early versions of our DRAGON/ips software demonstrated that a computer of this type could even be used for some aspects of remote sensing education. However, realistic remote sensing analyses on data sets significant demand on a computer’s speed, memory and storage capacities, and graphical display.
Table 2 - US$ 2000 worth of PC hardware |
| | 1986 | 1990 | 1994 |
| Processor | 80286 | 80386 | 80486 |
| Speed | 8Mhz | 16Mhz | 66Mhz |
| Co-processor | extra cost | extra cost | included |
| Memory | 640 KByte | 2 MByte | 8 MByte |
| Hard Disk | 10 MByte | 80 Mbyte | 500 Mbyte |
| Graphics Res. | 320 x 200 | 640 x 480 | 1024 x 700 |
| Graphics Colors | 16 or 256 | 256 | 16 million |
|
Although the 1994 model does not show a qualitative change from earlier available hardware, the quantitative changes are such as to change impractical tasks to simple ones.
Two of many possible examples will suffice :
- A maximum likelihood classification (4 bands, 512 x 512 image size) performed using DRAGON on the 1990 machine shown above would have required several hours to complete in 20 minutes or less (and, if desired, the image size could be increased to 1024 x 10240).
- A multi band composite image display combines the data from three bands into one natural color display: potentially 8 bits per band or 24 total significant bits of information. On a 16 color display (4 bits of co called display depth), only 1 significant bit per band is available. This is essentially useless. Even on a 256 color display, however, only 2-1/2 bits per band are available. This corresponds to 6 shades per primary color. Only with the 32,000 or 16 million color displays of today’s computer can be user see the full range of data being represented.
A full colour Pc graphics adapter currently costs only about US$200. It is interesting to note these color capabilities better than what is typically supplied with much more expensive Unix workstations.
At the same time that basic computer hardware has fallen price, peripherals important for remote sensing have also become significantly more available. The primary categories include hardcopy input devices color printers, and the range of CD ROM technology which is only beginning to be explored.
Another area where more affordable hardware impacts remote sensing education is portable computing. Today’s laptop and notebook computers are sufficiently powerful, portable, and robust, that doing image processing in the field as become quite practical. While a portable computer may be as such as 1005 more expensive than an equivalent desktop configuration, process are still well within the range of many educational organizations, particularly since one or two units will usually suffice.
Finally, the expansion of low cost computer technology, particularly in the are of graphics has had a side effect which is potentially very useful for remote sensing educators : the proliferation of computer graphics/desktop publishing service bureaus in may urban areas.
These commercial enterprises can provide high quality output at reasonable cost, for presentation and publication. This removes the need for educational organizations to purchase expensive hardcopy output devices; the money saved can instead be invested in a larger number of student workstations, in software, or in data. While these service bureaus may not yet be available in all locations, they are rapidly growing economic sector which should be attractive to entrepreneurs in almost every major city.
Table 3 Important New Remote Sensing Peripherals
| |
Low-end price in USA
(approximate) |
| Color flatbed scanner |
$1000 |
| Video france grabber |
$1000 |
| Color 35-mm flim scanner |
$1000 |
| Color inkjet printer |
$350 |
| Color printer with PostScript
software |
$600 |
| Color thermal tranfer printer |
$2000 |
| CD-ROM drive |
$200 |
| Write-your-own CD-ROM disk |
$125 |
| 8-mm tape unit |
$1500 |
| Removable 150 MByte disk drive |
$500 |
|
|
|
|
|
|
|
