The
work
is
devoted
to
the
process
of
research
and
development
an
object
selection
tracking
in
images.
article
offers
a
solution
problem
crossing
bounding
frames
that
arises
operation
YOLO
R-CNN
type
neural
networks
by
developing
method
for
estimating
optimal
size
search
area,
which
will
allow
finding
appropriate
its
frame,
offer
generalized
approach
visualization
process,
visually
represent
overlap
facilitate
object.
To
confirm
effectiveness
proposed
method,
there
conduct
experiments
on
data
set.
Drones,
Год журнала:
2024,
Номер
8(10), С. 579 - 579
Опубликована: Окт. 13, 2024
Ocean
beaches
are
a
major
recreational
attraction
in
many
coastal
cities,
requiring
accurate
visitor
counts
for
infrastructure
planning
and
value
estimation.
We
developed
novel
method
to
assess
beach
usage
on
the
Gold
Coast,
Australia,
using
507
drone
surveys
across
24
beaches.
The
covered
30
km
of
coastline,
accounting
different
seasons,
times
day,
environmental
conditions.
Two
AI
models
were
employed:
one
counting
people
land
water
(91–95%
accuracy),
another
identifying
types
(85–92%
accuracy).
Using
data,
we
estimated
annual
at
34
million
2022/23,
with
55%
45%
water—approximately
double
most
recent
estimate
from
lifeguard
counts,
which
spatially
limited
prone
human
error.
When
applying
similar
restrictions
as
surveys,
data
15
visits,
aligning
closely
(within
9%).
Temporal
(time
day
week,
season)
spatial
(beach
location)
factors
strongest
predictors
usage,
additional
patterns
explained
by
weather
variables.
Our
method,
combining
drones
AI,
enhances
coverage,
accuracy,
granularity
monitoring,
offering
scalable,
cost-effective
solution
long-term
assessment.
LATAM Revista Latinoamericana de Ciencias Sociales y Humanidades,
Год журнала:
2023,
Номер
4(2)
Опубликована: Авг. 2, 2023
Este
artículo
ofrece
un
análisis
de
la
literatura
internacional
producido
entre
el
año
2012
y
2022
sobre
abordaje
del
Cambio
Cobertura
Uso
Terreno
(CCUT),
a
través
una
metodología
mixta.
El
fenómeno
involucra
múltiples
interacciones
con
los
factores
físicos,
sociales,
económicos,
políticos
culturales.
Por
esta
razón,
se
requiere
comprender
manera
integral
las
causas,
efectos
procesos
que
inducen
estos
cambios
nivel
local
regional.
La
mixta
perspectiva
combina
datos
cuantitativos
(teledetección)
cualitativos
(percepción
social).
objetivo
investigación
fue
revisión
estudios
científicos
utilizan
aplicada
al
CCUT,
para
conocer
estado
actual
conocimiento
enfoques
teóricos,
metodológicos,
tendencias
en
este
campo
investigación.
Se
encontró
limitada
revistas
acceso
abierto
aborden
problemática
enfoque,
diversidad
término
“metodología
mixta”,
no
existe
información
explícita
método
artículos
revisados.
identificaron
herramientas
técnicas
más
empleadas
cuantitativa
cualitativa,
estrategias,
alcances
limitaciones
han
reportado
diversos
estudios.
destaca
capacidad
enfoque
metodológico
obtener
comprensión
problemática,
debido
proporciona
completa
causas
conducen
CCUT.
También
identificó
limitación
importante
como
establecer
vínculos
previos
confianza
actores
locales
puedan
colaborar
diseño
ser
parte
fundamental
The
work
is
devoted
to
the
process
of
research
and
development
an
object
selection
tracking
in
images.
article
offers
a
solution
problem
crossing
bounding
frames
that
arises
operation
YOLO
R-CNN
type
neural
networks
by
developing
method
for
estimating
optimal
size
search
area,
which
will
allow
finding
appropriate
its
frame,
offer
generalized
approach
visualization
process,
visually
represent
overlap
facilitate
object.
To
confirm
effectiveness
proposed
method,
there
conduct
experiments
on
data
set.
