)R
gical
shery
cean
rface
phyll
pe of
their
hms.
vards
'd on
ttern
1ated
PFZs
ound
h for
] for
n of
rdon
al of
from
iS. of
)01a)
d for
pixel
rface
iture.
' the
; and
give
phyll
ating
mber
The
r the
nt in
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India, 2002
The texture has typically been a qualitative measure, it can be
enhanced using mathematical algorithms. Feature extraction was
attempted using texture analysis algorithms. For this, a subset of
the chlorophyll image was selected to apply different types of
texture analysis algorithms. Three texture analysis algorithms i.e.
mean euclidean distance (1* order), variance (2" order) and
skewness (3 order) were used for extraction of oceanic features.
These algorithms are shown below,
S [Xe -xijy]^
(1). Mean Euclidean distance: -—-—-————7——--— ----- (1)
n-1
Y(xuij -M)?
(2), Variance z . e Late (2)
n—l
Z(xij - M”)
(5) SRewness = + || | m (3)
(n-1) (V^
Where, xij = DN value of pixel ij
xc = DN value of a windows central pixel
n = number of pixel in in the window
M = mean of moving window
»xij
mean = -------
n
V = variance
3. RESULTS AND DISCUSSION
Surface profile allows us to visualize in three dimensions the
concentration/reflection spectrum of rectangular area of data file
values an image. Spatial profile allows us to visualize the
concentration/reflectance spectrum of poly line of data file values
in an image. Surface profile gives an idea of structural patterns of
the features as well as the morphology. Spatial profile helps in
understanding the pattern of increase/decrease in concentration of
chlorophyll concentration within the given feature spatially.
Figure 1 indicates important features for exploring fishery
resources (Solanki et. al. 2001b).
3.1 Patterns of different features
The patterns surface and spatial profiles of oceanic features have
been shown in figure 2. The description of each feature types, its
importance for fisheries and the profiles interpretation have been
discussed here.
Rings: Rings are derivatives of meanders and eddy. It can be easy
to identify on an image as a circle shape. Rings are productive and
already localized developed eco-system. These features ensure the
secondary and tertiary production. The rings are in the circular
fashion observed with higher chlorophyll concentrations in the
periphery and low concentration in the center in surface profile.
There is sharp increase in concentration and decline in
465
Figure 1: Oceanic features utilised to study the patterns
concentration in the areas of center is clearly seen in spatial
profile. This type of features can be easily identified through
surface profile as well as spatial profile.
Eddies: The current of water often on the side of main current,
especially one moving in circle. Easy to monitor in space and
time using remotely sensed data. Rotating water masses causes
deep mixing and hence nutrient enrichment occured leading high
production. Persistence is relatively longer duration. The visual
predators like tunas prefer periphery of eddies and steamer.
Circulating features with high chlorophyll concentration is clearly
seen in surface profile. The increasing chlorophyll concentration
with circulating pattern is seen in three dimension plots. The
appearance of eddy is clearly observed in the surface profile.
Similarly the picks and declines observed in spatial profiles.
Fronts: Fronts are the boundary between the two water masses
with different property. It can be easily detected as breaks ocean
colour (chlorophyll concentration). High chlorophyll is an
indicator of bio-mass production. Hence, resource sustained for
longer period. The chances of development of local eco-system
are more, which enables benthos exploration. The front indicates
the gradient of chlorophyll concentration. There is abrupt
increasing pattern in concentration is observed in surface profile.
Similarly abrupt increase in concentration is seen in spatial profile
as distance increases.
Meander: Meandering pattern is turn or windings of the current
that may be detached form the main stream. Easily detected
through curvatures in the images. This enables large area extends
in relatively smaller area. So that even if feature shift potential
area may not shift totally. This also helps in delayed fishing. In
the smaller area large amount of phytoplankton concentration is
available as compared to linear feature. It forms enclosed pocket,
hence confine the resources. Some time it terns into rings, which