Data in Brief,
Год журнала:
2024,
Номер
57, С. 111016 - 111016
Опубликована: Окт. 10, 2024
The
fish
are
incorporated
with
ice
to
preserve
their
freshness
when
sold
on
the
market.
Ordinary
people
can
only
detect
its
some
basic
knowledge.
Therefore,
non-destructive
inspection
is
an
innovative
solution
help.
This
dataset
provides
a
medium
develop
system
for
detection
of
freshness.
There
three
data
variations:
sensor
data,
images,
and
organoleptic
examination.
includes
species:
mackerel,
tilapia,
tuna,
using
21
each
species.
Data
generation
was
carried
out
11
days,
where
800
MQ
(Metal
Oxide)
135
TGS
(Taguchi
Gas
Sensor)
2602
80
images
were
generated
every
day.
Organoleptic
examinations
Indonesian
National
Standard
(SNI)
2729-2013
six
parameters:
eyes,
gills,
body
surface
mucus,
meat,
smell,
textures.
be
used
system,
regression
modeling
estimate
deterioration
in
freshness,
standard
grouping
classes.
Comprehensive Reviews in Food Science and Food Safety,
Год журнала:
2024,
Номер
23(6)
Опубликована: Ноя. 1, 2024
Food
fraud
undermines
consumer
trust,
creates
economic
risk,
and
jeopardizes
human
health.
Therefore,
it
is
essential
to
develop
efficient
technologies
for
rapid
reliable
analysis
of
food
quality
safety
authentication.
Machine
vision-based
methods
have
emerged
as
promising
solutions
the
nondestructive
authenticity
quality.
The
Industry
4.0
revolution
has
introduced
new
trends
in
this
field,
including
use
deep
learning
(DL),
a
subset
artificial
intelligence,
which
demonstrates
robust
performance
generalization
capabilities,
effectively
extracting
features,
processing
extensive
data.
This
paper
reviews
recent
advances
machine
vision
various
DL-based
algorithms
authentication,
DL
lightweight
DL,
used
such
adulteration
identification,
variety
freshness
detection,
identification
by
combining
them
with
system
or
smartphones
portable
devices.
review
explores
limitations
challenges
include
overfitting,
interpretability,
accessibility,
data
privacy,
algorithmic
bias,
design
deployment
DLs,
miniaturization
sensing
Finally,
future
developments
field
are
discussed,
development
real-time
detection
systems
that
incorporate
combination
expansion
databases.
Overall,
techniques
expected
enable
faster,
more
affordable,
accurate
authentication
methods.
