IEEE Transactions on Geoscience and Remote Sensing,
Год журнала:
2024,
Номер
62, С. 1 - 12
Опубликована: Янв. 1, 2024
Accurate
estimation
of
building
heights
is
crucial
for
effective
urban
planning
and
resource
management
as
it
provides
essential
geometric
information
about
the
landscape.
Many
end-to-end
deep
learning-based
networks
have
been
proposed
image-to-height
mapping
using
high-resolution
non-optical
optical
remote
sensing
imagery.
In
this
study,
we
develop
a
novel
deep-learning
architecture
that
incorporates
stereo
parallax-based
mathematical
formulation
height
estimation.
We
estimate
parameters
include
differential
parallax
(ΔP)
image,
average
photo-base
(b),
satellite
(h
s
).
The
final
map
computed
by
utilizing
these
in
equation,
thus
combining
closed-form
solutions
within
learning
paradigm.
Moreover,
to
improve
ΔP,
also
introduce
multi-scale
shortcut
connections
module
(MSDSC).
MSDSC
integrates
high-frequency
components
into
lower-resolution
baseline
decoder
features
while
converting
them
features.
To
establish
efficacy
our
network
(Stereollax
Net),
train
evaluate
method
on
densely
populated
cities
China
(42-Cities
dataset)
IEEE
Data
Fusion
Contest
2018
dataset
(DFC2018).
Our
Stereollax
Net
trained
only
with
RGB
imagery
compared
state-of-the-art
methods
utilize
both
panchromatic
multi-spectral
(RGB
Near-infrared)
qualitative
quantitative
results
demonstrate
surpasses
existing
(SOTA)
algorithms,
achieving
superior
performance
fewer
data
training
considerable
margin.
code
will
be
made
publicly
available
via
GitHub
repository.
Remote Sensing,
Год журнала:
2024,
Номер
16(9), С. 1556 - 1556
Опубликована: Апрель 27, 2024
The
traditional
method
for
extracting
the
heights
of
urban
buildings
involves
utilizing
dense
matching
algorithms
on
stereo
images
to
generate
a
digital
surface
model
(DSM).
However,
buildings,
disparity
discontinuity
issue
that
troubles
algorithm
makes
elevations
high-rise
and
surrounding
areas
inaccurate.
occlusion
caused
by
trees
in
greenbelts
it
difficult
accurately
extract
ground
elevation
around
building.
To
tackle
these
problems,
building
height
extraction
from
Gaofen-7
(GF-7)
enhanced
contour
is
presented.
Firstly,
was
proposed
accurate
roof
GF-7
images.
Secondly,
filtering
employed
DSM
(DEM),
can
be
extracted
this
DEM.
difference
between
rooftop
represents
height.
presented
verified
Yingde,
Guangzhou,
Guangdong
Province,
Xi’an,
Shaanxi
Province.
experimental
results
demonstrate
our
outperforms
existing
methods
concerning
accuracy.
Remote Sensing,
Год журнала:
2025,
Номер
17(7), С. 1310 - 1310
Опубликована: Апрель 6, 2025
Volume
of
change
provides
a
comprehensive
and
objective
reflection
land
surface
transformation,
meeting
the
emerging
demand
for
feature
monitoring
in
era
big
data.
However,
existing
methods
often
focus
on
single
dimension,
either
horizontal
or
vertical,
making
it
challenging
to
achieve
quantitative
volumetric
monitoring.
Accurate
measurements
are
indispensable
many
fields,
such
as
open-pit
coal
mines.
Therefore,
main
content
conclusions
this
paper
follows:
(1)
A
method
Automatic
Control
Points
Extraction
from
ICESat-2/ATL08
products
was
developed,
integrating
Land
cover
types
Phenological
information
(ACPELP),
achieving
mean
absolute
error
(MAE)
1.05
m
direction
1.99
vertical
stereo
measurements.
This
helps
correct
image
positioning
errors,
enabling
acquisition
geospatially
aligned
GaoFen-7
(GF-7)
imagery.
(2)
function-based
classification
system
mines
established,
precise
extraction
stereoscopic
region
support
accurate
calculations.
(3)
calculating
mining
stripping
volume
based
GF-7
imagery
is
proposed.
The
utilizes
photogrammetry
extract
elevation
features
combines
spectral
with
data
estimate
volumes,
an
excellent
rate
(ER)
0.26%.
results
indicate
that
our
cost-effective
highly
practical,
filling
gap
changes.
Urban
building
height,
as
a
fundamental
3D
urban
structural
feature,
has
far-reaching
applications.
However,
creating
readily
available
datasets
of
recent
heights
with
fine
spatial
resolutions
and
global
coverage
remains
challenging
task.
Here,
we
provide
150-m
dataset
around
2020
by
combining
the
spaceborne
lidar
(Global
Ecosystem
Dynamics
Investigation,
GEDI),
multi-sourced
data
(Landsat-8,
Sentinel-2,
Sentinel-1),
topographic
data.
The
validation
results
revealed
that
GEDI-estimated
height
samples
were
effective
compared
to
reference
(Pearson's
r
=
0.81,
RMSE
3.58
m).
mapping
product
also
demonstrated
good
performance,
indicated
its
strong
correlation
0.71,
4.73
Compared
currently
existing
datasets,
it
holds
ability
resolution
(150
m)
great
level
inherent
details
about
heterogeneity
flexibility
updating
using
GEDI
inputs.
This
will
boost
future
studies
across
many
fields,
including
environmental,
ecological,
social
sciences.
International Journal of Applied Earth Observation and Geoinformation,
Год журнала:
2024,
Номер
129, С. 103792 - 103792
Опубликована: Март 28, 2024
Most
studies
on
building
height
estimation
using
remote
sensing
imagery
mainly
focus
urban
buildings
with
relatively
regular
shape
and
flat
terrain,
pay
little
attention
to
large
complex
terrain
like
dams.
A
new
dam
measurement
method
was
proposed
in
this
paper,
which
used
shadow
data
metadata
from
multiple
image
sources.
The
not
only
considers
the
geometric
relationship
between
sun,
satellite
dam,
but
also
influence
of
zenith
introduces
correction
factor,
brings
higher
precision.
Two
dams
a
maximum
more
than
200
m
were
studied,
is
estimated
by
considering
topography
around
dam.
experimental
results
show
that
Mean
Relative
Error
(MRE)
our
are
3.1%
4.7%
for
two
study
areas,
while
MRE
traditional
models
13%.
By
doing
so,
we
able
calculate
obtain
information
temporal
changes
during
construction
process,
even
situations
without
Digital
Surface
Model
(DSM).
Therefore,
will
be
propitious
dynamic
supervision
process
effectively.
