Geoscience Reference
In-Depth Information
Figure 3-1.
The tranquil beauty of the bog is apparent in this early morning, foggy view at Nigula, southwestern
Estonia. Photo courtesy of E. Karofeld.
As Tiner (1997) emphasized, aerial photogra-
phy continues to be essential for wetland studies.
With the launch of the i rst Earth Resources
Technology Satellite in 1972 (later renamed
Landsat I), applications were immediately rec-
ognized for wetlands and related water resources
(e.g. Carter, McGinness and Anderson 1976;
Higer, Cordes and Coker 1976; Krinsley 1976).
From the frog's-eye view on the ground
(Fig. 3-1) to the vantage from outer space, we
now have the ability to observe wetlands from
many heights and resolutions (Fig. 3-2).
pixel can rarely be determined by itself. For
positive identii cation, usually a group of four
to nine pixels is necessary, which leads to a
general rule of thumb (Hall 1997):
Positive recognition of objects in remotely sensed
images requires a GSD at least three times smaller
than the object size.
The spatial resolution (GSD) of an aerial photo-
graph or satellite image is only one factor that
comes into play for image interpretation. Other
basic aspects of resolution are temporal and
spectral characteristics for different kinds of
remote sensing. Additional factors include the
following features (based on Avery and Berlin
1992; Teng et al. 1997; Jensen 2007):
3.2.1 Image resolution and
interpretability
The term “resolution” is used in widely different
senses in various disciplines. In the general
context of remote sensing, it refers to the small-
est object or feature that can be identii ed by
an experienced image interpreter. For digital
imagery, the smallest object is a single cell or
picture element (pixel) in the scene. Each pixel
corresponds to an actual spot on the ground,
and the linear size of this spot is known as
ground sample distance (GSD), which is an
appropriate measure for image scale (Comer
et al. 1998). However, the identity of a single
• Tone or color - Gray tone (b/w) and color
are how we recognize and distinguish objects.
The tone or color of an object helps to sepa-
rate it from other features in the scene, espe-
cially for features with high contrast.
• Shape - Natural shapes tend to follow the
lay of the land. Cultural (human) shapes, on
the other hand, are often geometric in nature
with straight lines, sharp angles, and regular
forms. In single images only plan shape can
