slicing are those which provide for maximum flexibility in selecting
the density ranges which are to be displayed as specific colors, as
opposed to those which are limited to the use of equal density ranges
for each display color.
Image Analysis Systems
LANDSATS I and II have provided earth resource scientists the op-
portunity to acquire multispectral data repetitively over large regions.
These capabilities provide a mechanism for monitoring change in earth
resources, mapping resources over extensive regions, and inventorying
specific resources through use of appropriate statistical sampling pro-
cedures. Successful application of these data to resource management
problems necessitates development of interpretive methodologies which
allow quick, consistent, and accurate extraction of pertinent infor-
mation in a cost éffective manner. The capabilities developed during
the 1960's and early 1970's and the availability of LANDSAT data has
provided the impetus for engineering and development of new, improved,
low cost image analysis equipment.
There has been an increasing trend toward more interactive systems
providing the user greater flexibility in analyzing multispectral data
at a faster rate than previously available. Virtually all image anal-
ysis systems are based on the assumption that earth resource features
exhibit characteristic reflectance patterns and through processing of
multispectral data, an analyst can efficiently and accurately classify,
map, and inventory the areal extent and quantity of specific resource
parameters. Systems now being developed utilize both hardware and soft-
ware functions and vary considerably in their design concept, method of
operation, and capabilities afforded the user. Generally, the hardware
associated with image analysis systems include data input devices (e.g.,
tape drives, vidicon cameras, graphic digitizers), a computer with ap-
propriate peripherals, communication and graphics terminal, image dis-
play device, and an output display device. Detailed descriptions of
several advanced image processing systems can be found in Erikson (1975);
Schaller and Towles (1975); Johnson (1975); Quinn (1975); Whitley (1975);
Henze and DeZur (1975); Bernstein and Ferneyhough (1975); and Goetz, et.
al. (1975).
The most common method of inputting data to an image analysis system
is through digital tapes. LANDSAT computer compatible tapes (CCT's),
SKYLAB S192 tapes, and aircraft multispectral scanner tapes can be anal-
yzed on the variety of systems available. Black and white multiband and
color photography can also be input to these systems through digitization
with scanning digitizers or vidicon cameras with appropriate color sep-
aration filters. The most common method of displaying an image is on a
color monitor. The size of an area displayed is a function of the color
monitor and the user defined image scale. Monitors capable of display-
ing a 512 x 512 or 1024 x 1024 array of picture elements are common.
Such monitors allow displaying an image as a color composite and provides
the user flexibility in adjusting color saturation, brightness, and scene
contrast on the display. Since an entire LANDSAT scene includes over 7.5
million picture elements, an entire scene cannot be displayed unless the
data are sampled. 0n a 512 x 512 color monitor, approximately 37 of the