| The primary goal for the design of the GeoCam
II optical and imaging systems is to match as closely as possible the results of the
Skylab S190A and S190B camera systems. This includes the selection of a focal length of
lens to match the photographic "footprint" of the S190A cameras and CCD sensors
and filters to duplicate the six spectral bands. The collection of 36,000 photos produced
by EREP, in six registered wavelengths, represents the most complete multispectral film
coverage of the planet ever undertaken. Despite this, the EREP collection remains
tragically unused. GeoCam II intends to operationally re-visit what is a considerable
national asset, and to contribute a comparative update to this remarkable inventory of
planetary data. Because the multispectral bands pioneered on Skylab are very similar to
the first four bands of Landsat, the GeoCam II system becomes a capable research
instrument in its own right. It will give students a powerful system with which to explore
the multispectral images and image processing which is so fundamental to modern remote
sensing. |
|
|
|
|
click on image for closer view
|
The recent proliferation of relatively
inexpensive high resolution CCD chips with color, monochrome and infrared capability makes
it possible to design an affordable camera system which will exactly duplicate the optical
and spectral properties of the six-camera S190A Skylab multispectral camera system. The
high sensitivity of silicon in the infrared means that a CCD-based camera can make an
excellent sensor for the near infrared wavelength range of 750 nm to 1000 nm which present
a difficult problem for film. Six separate digital sensors will simultaneously capture an
image, which will then be downloaded to a central processor using an IEEE 1394 high-speed
interface. The inherently small size of the CCDs will allow the lens/camera systems to
utilize relatively small optical windows. The use of dichoric beam splitters could reduce
the size even more by allowing for multiple wavelengths to be captured through a single
lens system. Such a system could be built now with off-the shelf equipment, and the rapid
development of the digital image field will allow better performance at an even lower cost
in the near future. High-speed architecture and mass storage media will be required to
handle the large amount of data produced and to store it pending downlink opportunities.
Each multispectral image set will produce between 6MB (1k x 1k at 8bit grayscale) and 32MB
(2K x 2K at 12 bits grayscale) of data. Image compression may be used to reduce the data
load, but it is well within the reach of current system capabilities. |
|
The KidSat and EarthKAM
programs have demonstrated the practicality of downloading large digital image files. The
control interface could be modeled on NASA's EarthKAM project, which provides a working
structure for using high-resolution digital Earth images in education. The multispectral
system can serve as a natural extension of the EarthKAM program by allowing the students
to conduct a more advanced research project through the comparative study of twenty-five
years of environmental change. Access to true multispectral images will allow for a richer
learning experience and a chance to explore imagery as a quantitative scientific tool
rather than just for color documentation of geographic morphometry.
|
click on image for complete
diagram |
Because of the
extensive engineering and testing required, a fully operational
prototype system will be built as proof of concept, engineering
test bed, demonstration unit and training tool. The prototype
system will be built from a SONY PCG-F270 Notebook computer
selected because of its built in IEEE-1394 (“Firewire”)
interface used for the operational megapixel resolution cameras
and USB (Universal Serial Bus) interface that will be used in
the lower resolution prototype. Polaris MB-810 infrared and
visible CCD cameras will do the imaging with multi-spectral
filters. Frame capture will use a USB Digital Photo Maker still
image capture unit. |
|
|
|
