| 4.1 Introduction |
|
When the specialized photointerpreter investigates a part of
the natural earth surface or a phenomenon/fact/event in the scene by the use of
photointerpretation and remote sensing methods and techniques, he must have a good
understanding of the overall Photointerpretation methodology he applies, closely depending
on the basic principles of the scientific field underlying the investigation. In all
cases, the following steps should be followed, in a holistic, systematic and integrated
way:
|
1. Examine the basic elements constituting the position of the specific problem,
including all their characteristics and their relations, interdependencies and
interactions with the surrounding environment.
2. Collect, evaluate and use the existing general and specific informative material on the
area under investigation.
3. Select the
appropriate specific methodology or combination of methodologies for the problem under
investigation.
4. Select the appropriate remote sensing
technique.
5. Select the appropriate equipment for data processing.
|
So, the first step in applying remote sensing methods and techniques is to:
1.1 Define the nature of the problem and identify the scientific and technical
possibilities for dealing with it, referring to:
|
1.1.1 the level of the basic human knowledge concerning the specific
problem
1.1.2 the level of the relevant traditional methods which have been developed, and which
could perhaps be supported by remote sensing methodologies
1.1.3 the level of the relevant available technology that can be applied.
|
| 1.2
Identify the limitations and constraints imposed by the specific problem concerning: |
1.2.1 the results of methods that have been applied up to now under
real natural and socioeconomic conditions
1.2.2 the margins resulting from the cost effectiveness and promptness of the techniques
used
1.2.3 the available material/equipment to be used. |
1.3 Evaluate the scientific/technical/technological personnel involved in the application
of the appropriate photointerpretation and remote sensing methods and techniques.
1.4 Evaluate all the available spatial and qualitative information (literature) for the
greater area under investigation, in order to support this specific study (e.g.
cartographic, statistical, bibliographical, climatic etc data, thematic maps, aerial
photographs, remotely sensed images etc). |
| |
| 4.2 Specific interpretation characteristics of remotely sensed
images |
TOP |
The knowledge of the specific interpretation characteristics of each kind of aerial
photographs or remotely sensed images is essential for the selection of the most
convenient photointerpretation methods.Below, we present, the interpretation
characteristics of:
a. Conventional aerial photographs.
b. Thermal infrared imagery.
c. LANDSAT images.
d. Radar images.
4.2.1 Specific interpretation
characteristics of conventional aerial photographs
4.2.1.1 Panchromatic aerial photographs
Panchromatic aerial photographs are usually acquired with a yellow filter and are
sensitive to visible light. Their reduced sensitivity in the green portion limits their
specific ability to discriminate objects with different colours in the relevant
application fields (forestry: reduced discrimination ability of different tree
categories).
Panchromatic aerial photographs are used for most photointerpretation applications,
because not only the appearance of objects’ shadows impede photointerpretation, but also
trees, forestry roads, boundaries of a recently arable field and the general area’s
morphology are clearly defined.
The panchromatic photointerpretation characteristics yield the best results in application
fields such as urban
planning, geography, geology, civil engineering projects, with photograph acquisitions
during spring and summer seasons.
4.2.1.2 Infrared aerial photographs
They are especially sensitive in the portion of blue, mauve and reflected infrared
radiation and are usually acquired with a red or a dark filter. In infrared aerial
photographs, the surfaces of water concentrations (without admixture, mud etc.), absorb
the infrared radiation, so they are presented in extremely dark tones (black) and this
characteristic facilitates the easier detection of tributaries, river-beds with current
flow, networks of natural surface drainage, swamps, irrigation or drainage canals, tidal
lines and coastlines.
Infrared aerial photographs represent reality more clearly than the panchromatic ones in
cases of fog and suspension of dust particles in the air.
However, this advantage becomes less important in the case of a thick fog or of an
extremely humid atmosphere with heavy cloud cover.
Infrared aerial photographs can also present better differences between arable and
non-arable land and between broad-leaved trees (which have high reflectance and are
therefore presented in bright tones) and coniferous ones (which absorb the infrared
radiance and are presented in dark tones). This characteristic of infrared aerial
photograph interpretation helps the identification of different timber types and also, the
detection of green colour “camouflage” (green colour doesn’t reflect infrared
radiance) or of “cut” (dead) vegetation.
It is obvious, that the above-mentioned characteristics are disadvantages for some
photointerpretation applications, because:
|
a. The water (dark) image, objectively eliminates some tone differences and consequently,
reefs, obstacles in straights or canals, shallow marshes etc. can not be distinguished.
b. The intense (dark) shadows of infrared aerial photographs hide useful information.
c. The use of red filter helps discriminate better different vegetation types (because of
increased tone differences), but at the same time it reduces the sharpness of the image.
d. The non-conventional tone correspondence (depending on the photointerpreter’s
conception criteria), can cause confusion during the mono-image identification, for
example of light-coloured objects (such as dusty roads/broad – leafed low trees). |
As a result of their specific photointerpretation characteristics, infrared aerial
photographs are more suitable for applications like natural resource inventories and
monitoring especially in forestry, ecology and flora and fauna management, with
photographic acquisition early in the spring and late in the summer.
4.2.1.3 Coloured aerial photographs
In order to take satisfactory coloured aerial photographs (sensitive to visible light) not
only the right lighting and acquisition geometry disposal conditions are required, but
also the appropriate use of filters.
coloured aerial photographs are more reliable than panchromatic and infrared ones on the
presentation of objects colour gradation, because they can “understand” and present a
wider scale of colour shade than, for example panchromatic photographs in the case of
grey’s gradation.
They are therefore valuable in the identification of different soil categories, surface
rocks, coast types, hydrographical controls and exploration of shallow water concentration
bottoms because of the possibility of water-transparency compared to the panchromatic
photographs.
4.2.1.4 Coloured infrared aerial photographs
The need for photointerpretation to discriminate between “healthy” vegetation and
leafage camouflages absorbing the infrared radiance has lead the development of
coloured-infrared aerial photograph applications and techniques.
The specific photointerpretation characteristics of coloured infrared aerial photographs
mainly guarantee: the possibility of identifying objects during the comparative
photointerpretation process with coloured aerial photographs of the same areas.
They also permit: |
 The detection of plant
diseases on the beginning or on the first stage of evolution and expansion and forest
damages, caused by the same insect categories.
|
The mapping of different tree types, depending on the infrared
radiance reflectance of leafage.
|
| The discrimination between evergreen and deciduous trees of the same
colour, because the “healthy” deciduous trees reflect the infrared radiance much more
than the evergreen ones during the period in which they have leaves . |
4.2.1.5 Multi-spectral aerial photographs
Multi-spectral aerial photographs
are acquired with special photogrammetric cameras carrying four or more lenses. They look
like coloured infrared ones with the help of appropriate filters and they offer us the
possibility to reveal interesting natural, chemical and biological characteristics, which
for certain reasons are presented in a different way on the relevant images when we change
the colour, the density and the shades of colour with a specific viewer . |
Today, there are a lot of applications for multi-spectral aerial
photographs in forestry, agriculture, geomorphology, natural resource inventories,
environmental pollution monitoring etc.
4.2.2 Specific interpretation characteristics of thermal infrared images
The thermal infrared (optico-mechanical) scanners are remote sensors, sensitive to changes
of the reflected object’s temperature, which are recorded as tone gradation.
So, in a positive print of thermal infrared image, the brighter tones indicate objects
with high absolute temperature.
The thermal infrared images are not at all and in any case corresponding to the
conventional aerial photographs, because they are based on the surfaces emissivity and not
on the surfaces sensitivity of reflected light.
The basic source of the surfaces reflected or emitted thermal radiance is the solar
radiance. So, during daytime, surfaces of high absorbance store a great quantity of
temperature, although the exactly opposite happens to surfaces with high reflectance.
Consequently, the thermal infrared image acquisition time influences the quality and the
quantity of data recorded.
4.2.3 Specific interpretation characteristics of LANDSAT images
There is a basic structural difference between conventional aerial photographs and LANDSAT
images (stereoscopic vs monoscopic): in the first case, in spite of rays' deviation caused
by the relief, which permits the stereoscopic vision of the overlap of two successive
aerial photographs, the scanning process of multi-spectral scanners (MSS) permits the
one-dimensional conception of relief only to the scanning direction.
In the second case, we obtain an image of the scanning line image with the sensor
perpendicularly projected on the image’s center, so, the base to the axis of image
acquisition is practically zero and consequently, LANDSAT images stereovision is
impossible at this direction.
At the equator sane overlap is 14% whereas in latitude 80o it becomes 85%. The
distance between two successive scanning centerlines is 159 Km.
4.2.4 Specific interpretation characteristics of Radar images
The Radar (Radio Detection and Ranging) is a microwave active remote sensing system
working in the area 0.3 – 100 cm of the electromagnetic spectrum, which functions in all
lighting and weather conditions existing in the area under investigation.
The SLAR (Side Looking Airborne Radar) system, in its typical form, consists of a system
of a transmitter–antenna and a receiver of electromagnetic energy scattered on the
ground and recorded by the system.
Depending on the object’s orientation to the SLAR’s antenna direction, its texture,
shape, form and electric properties, we either have strong return signals (which means
clear tones on the image) or weaker signals (different tone gradations) or weak to zero
return signals (dark or black tone).
So, a hill oriented parallel to the SLAR’s flight axis direction is represented on a
radar image by a clear tone and so does (but for different reasons) a storehouse metallic
roof.
Reference:.Rokos, D. “Photointerpretation and Remote
Sensing”, NTUA, 1979
|
