Page 43 - Digital Analysis of Remotely Sensed Imagery

16 Cha pte r O n e

Rotating
scan mirror

Field
of view Array of detectors

IFOV
H
Scan direction

Swath width

Ground
cell size Direction of
satellite motion
FIGURE 1.7 Relationship among spatial resolution of satellite imagery,
satellite altitude (H), and IFOV (a) of the scanner.

where α is expressed as a radian angle. According to this equation, at
the same altitude a smaller IFOV translates into a smaller pixel size,
and vice versa. At the same IFOV, a lower altitude leads to an image
of a finer spatial resolution, and vice versa.
Spatial resolution denotes the theoretical dimension of ground
features that can be identified from a given remote sensing image. The
finer the spatial resolution, the more detailed the image is. As the pixel
size increases, less detail about the target is preserved in the data (Fig. 1.8).
A small cell size is desirable in those local-scale applications that demand
great details about the target. A fine spatial resolution reduces the number
of mixed pixels, especially if the landscape is highly fragmented and land
cover parcels have an irregular shape. The downside effect of having a
fine spatial resolution is a large image file size. This file size is going to
double or triple if two or three spectral bands are needed. As it is a
common practice to record satellite data in the multispectral mode, an
image file size can reach a few megabits easily. Such a large file is going to
slow down all subsequent analyses. It is thus important to select data
with a spatial resolution appropriate for the needs of an application.
If the digital remote sensing data are obtained through scanning of
existing aerial photographs, their spatial resolution is determined by