AgriScience and Technology,
Journal Year:
2024,
Volume and Issue:
unknown, P. 317 - 325
Published: Jan. 1, 2024
Winter
wheat
constitutes
a
significant
share
of
Armenia’s
grain
production,
but
can
be
severely
affected
by
the
spread
fungal
diseases.
According
to
recent
scientific
studies,
methods
based
on
dynamics
spectral
signatures
crops
allow
for
timely
detection
early
signs
crop
diseases,
prevention
epidemiological
threats,
and
minimization
economic
losses.
The
paper
addresses
problem
determining
optimal
monitoring
periods
phytopathogens
across
republic
Armenia
using
these
methods.
study
covers
various
regions
Armenia,
taking
into
account
their
climatic
agronomic
characteristics,
as
Normalized
Difference
Vegetation
Index
(NDVI),
weather
conditions
digital
elevation
model.
data
show
potential
timing
key
phenological
stages
in
winter
development,
which
is
essential
this
type
monitoring.
Cartographic
materials
have
been
created
indicate
recommended
start
dates
monitoring,
average
from
2018-2024.
also
outlines
necessary
steps
improve
outcomes
proposed
methodology,
well
actions
implementing
approach
practice.
Remote Sensing,
Journal Year:
2025,
Volume and Issue:
17(5), P. 756 - 756
Published: Feb. 22, 2025
Assessing
land
use/land
cover
changes
currently
represents
an
important
avenue
for
achieving
a
better
understanding
of
the
urbanization
phenomenon.
Various
free
datasets
based
on
satellite
imagery
are
available,
but
user
should
decide
which
one
is
most
suitable
their
study
area.
The
aim
present
paper
to
perform
accuracy
assessment
built-up
areas
using
four
datasets:
Corine
Land
Cover
Backbone
(CLC
Backbone),
High
Resolution
Layers
(HRL)–Imperviousness,
Esri
and
Dynamic
World.
case
represented
by
12
major
metropolitan
(MAs)
in
Romania
have
dynamic
economic
development
urban
expansion.
Confusion
matrices
were
created,
following
metrics
been
computed:
overall
(OA),
kappa
coefficient
(k)
(UA).
analysis
was
performed
three
levels:
entire
surface
MAs
separately
rural
sides.
results
at
level
show
that
even
though
CLC
2018
extracting
built
(0.85
accuracy),
HRL
could
also
be
used,
as
they
share
same
values
(0.67).
Significant
differences
exist
between
areas.
(0.87)
than
(0.84).
other
recorded
variations
has
second
greatest
(0.81),
while
accurate,
after
Backbone,
assessing
In
conclusion,
best
performance
all
levels
analysis.
significance
lies
datasets,
levels.
This
aims
help
researchers
decision
makers
choose
dataset
use
changes.
Additionally,
having
reliable
may
compute
indicators
used
monitor
Sustainable
Development
Goals
(SDGs).
Sensors,
Journal Year:
2025,
Volume and Issue:
25(10), P. 3092 - 3092
Published: May 14, 2025
Conventional
remote
sensing
classification
approaches
based
on
single-source
data
exhibit
inherent
limitations,
driving
significant
research
interest
in
improved
multimodal
fusion
techniques.
Although
deep
learning
methods
convolutional
neural
networks
(CNNs),
transformers,
and
graph
(GCNs)
have
demonstrated
promising
results
fusing
complementary
multi-source
data,
existing
methodologies
demonstrate
limited
efficacy
capturing
the
intricate
higher-order
spatial–spectral
dependencies
among
pixels.
To
overcome
these
we
propose
HGCN-HL,
a
novel
framework
that
integrates
hypergraph
(HGCNs)
with
lightweight
CNNs.
Specifically,
an
adaptive
weight
mechanism
is
first
designed
to
preliminarily
fuse
spectral
features
of
hyperspectral
imaging
(HSI)
Light
Detection
Ranging
(LiDAR),
enhancing
feature
representation
ability.
Then,
superpixel-based
dynamic
hyperedge
construction
enables
joint
characterization
homogeneous
regions
across
both
modalities,
significantly
boosting
large-scale
object
recognition
accuracy.
Finally,
local
detail
are
captured
through
parallel
CNN
branch,
complementing
global
relationship
modeling
HGCN.
Comprehensive
experiments
conducted
three
benchmark
datasets
superior
performance
our
method
compared
state-of-the-art
approaches.
Notably,
proposed
achieves
improvements
training
efficiency
inference
speed
while
maintaining
competitive
Current Landscape Ecology Reports,
Journal Year:
2024,
Volume and Issue:
10(1)
Published: Dec. 16, 2024
Abstract
Purpose
of
Review
Landscapes
can
be
defined
as
mosaics
different
land
covers,
habitats,
ecosystems,
or
land-use
systems.
The
link
between
spatial
heterogeneous
patterns
and
ecological
processes
is
the
core
concept
in
research
field
landscape
ecology.
Nowadays,
advanced
computational
methods
are
essential
to
due
its
cross-disciplinary
nature,
increasing
availability
data,
complexity
Recent
Findings
This
review
provides
an
overview
recent
developments
that
have
We
focus
on
key
topics
such
patterns,
connectivity,
genetics,
sampling,
simulations
modeling,
planning.
Summary
highlights
innovations,
challenges,
potential
future
directions
field,
emphasizing
role
addressing
complex
questions.
Remote Sensing,
Journal Year:
2024,
Volume and Issue:
16(20), P. 3866 - 3866
Published: Oct. 18, 2024
The
escalating
human
pressures
on
natural
ecosystems
necessitate
urgent
and
effective
conservation
strategies
to
safeguard
biodiversity
ecosystem
functions.
This
review
explored
current
techniques
for
mapping
pressure,
with
a
particular
focus
their
application
in
nature
conservation,
especially
within
protected
areas
(PAs).
Specifically,
we
analyzed
the
impacts
of
seven
major
types
PAs.
Additionally,
discussed
four
key
methods
including
land
use
intensity,
footprint,
digital
other
proxies,
examining
distinct
characteristics
respective
advantages
disadvantages.
our
research
pressure
assessing
its
suitability
applications
delineating
directions
future
work.
These
insights
contributed
better
support
management
AgriScience and Technology,
Journal Year:
2024,
Volume and Issue:
unknown, P. 317 - 325
Published: Jan. 1, 2024
Winter
wheat
constitutes
a
significant
share
of
Armenia’s
grain
production,
but
can
be
severely
affected
by
the
spread
fungal
diseases.
According
to
recent
scientific
studies,
methods
based
on
dynamics
spectral
signatures
crops
allow
for
timely
detection
early
signs
crop
diseases,
prevention
epidemiological
threats,
and
minimization
economic
losses.
The
paper
addresses
problem
determining
optimal
monitoring
periods
phytopathogens
across
republic
Armenia
using
these
methods.
study
covers
various
regions
Armenia,
taking
into
account
their
climatic
agronomic
characteristics,
as
Normalized
Difference
Vegetation
Index
(NDVI),
weather
conditions
digital
elevation
model.
data
show
potential
timing
key
phenological
stages
in
winter
development,
which
is
essential
this
type
monitoring.
Cartographic
materials
have
been
created
indicate
recommended
start
dates
monitoring,
average
from
2018-2024.
also
outlines
necessary
steps
improve
outcomes
proposed
methodology,
well
actions
implementing
approach
practice.
