Optics Express,
Год журнала:
2021,
Номер
29(25), С. 41865 - 41865
Опубликована: Ноя. 24, 2021
The
veiling
effect
caused
by
the
scattering
and
absorption
of
suspending
particles
is
a
critical
challenge
underwater
imaging.
It
possible
to
combine
image
formation
model
(IFM)
with
optical
polarization
characteristics
effectively
remove
recover
clear
image.
performance
such
methods,
great
extent,
depends
on
settings
global
parameters
in
application
scenarios.
Meanwhile,
learning-based
methods
can
fit
information
degradation
process
nonlinearly
restore
images
from
scattering.
Here,
we
propose
for
first
time
method
full
scene
imaging
that
synergistically
makes
use
an
untrained
network
By
mounting
Stokes
mask
polarizer
CMOS
camera,
simultaneously
obtain
different
states
IFM
calculation
optimize
automatically
without
requiring
extra
training
data.
This
nonlinear
fitting
ability
neural
corrects
undesirable
imperfect
parameter
classical
scenes
.
shows
good
removing
impact
water
preserving
object
information,
making
it
achieve
PLoS ONE,
Год журнала:
2024,
Номер
19(4), С. e0301182 - e0301182
Опубликована: Апрель 26, 2024
The
three-dimensional
swimming
tracks
of
motile
microorganisms
can
be
used
to
identify
their
species,
which
holds
promise
for
the
rapid
identification
bacterial
pathogens.
also
provide
detailed
information
on
cells’
responses
external
stimuli
such
as
chemical
gradients
and
physical
objects.
Digital
holographic
microscopy
(DHM)
is
a
well-established,
but
computationally
intensive
method
obtaining
cell
from
video
data.
We
demonstrate
that
common
neural
network
(NN)
accelerates
analysis
data
by
an
order
magnitude,
enabling
its
use
single-board
computers
in
real
time.
establish
heuristic
relationship
between
distance
focal
plane
size
bounding
box
assigned
it
NN,
allowing
us
rapidly
localise
cells
three
dimensions
they
swim.
This
technique
opens
possibility
providing
real-time
feedback
experiments,
example
monitoring
adapting
supply
nutrients
microbial
bioreactor
response
changes
phenotype
microbes,
or
pathogens
drinking
water
clinical
samples.
Engineering Research Express,
Год журнала:
2023,
Номер
5(3), С. 032005 - 032005
Опубликована: Сен. 1, 2023
Abstract
Holography
is
a
technique
to
record
and
reconstruct
three
dimensional
(3D)
information
without
mandating
lenses.
Digital
holography
(DH)
provides
direct
access
the
complex
amplitude
of
reconstructed
wavefront.
This
feature
differentiates
DH
from
other
imaging
techniques
enables
it
provide
quantitative
object
under
investigation.
Advancements
in
technologies
digital
image
sensors,
coherent
sources,
computation
algorithms
hardware,
has
paved
way
holographic
systems
for
industrial
applications.
work
presents
an
overview
scientific
applications
where
can
play
important
role.
Few
areas
including
microscopy,
non-destructive
testing,
displays,
environment,
cloud
ocean
studies
are
discussed.
Abstract
We
introduce
a
digital
inline
holography
(DIH)
method
combined
with
deep
learning
(DL)
for
real-time
detection
and
analysis
of
bacteria
in
liquid
suspension.
Specifically,
we
designed
prototype
that
integrates
DIH
fluorescence
imaging
to
efficiently
capture
holograms
flowing
microfluidic
channel,
utilizing
the
fluorescent
signal
manually
identify
ground
truths
validation.
process
using
tailored
DL
framework
includes
preprocessing,
detection,
classification
stages
involving
three
specific
models
trained
on
an
extensive
dataset
included
generic
particles
present
sterile
five
bacterial
species
featuring
distinct
morphologies,
Gram
stain
attributes,
viability.
Our
approach,
validated
through
experiments
synthetic
data
spiked
different
bacteria,
accurately
distinguishes
between
particles,
live
dead
Gram-positive
negative
similar
morphology,
all
while
minimizing
false
positives.
The
study
highlights
potential
combining
as
transformative
tool
rapid
clinical
industrial
settings,
extension
other
applications
including
pharmaceutical
screening,
environmental
monitoring,
disease
diagnostics.
Optics Express,
Год журнала:
2021,
Номер
29(25), С. 41865 - 41865
Опубликована: Ноя. 24, 2021
The
veiling
effect
caused
by
the
scattering
and
absorption
of
suspending
particles
is
a
critical
challenge
underwater
imaging.
It
possible
to
combine
image
formation
model
(IFM)
with
optical
polarization
characteristics
effectively
remove
recover
clear
image.
performance
such
methods,
great
extent,
depends
on
settings
global
parameters
in
application
scenarios.
Meanwhile,
learning-based
methods
can
fit
information
degradation
process
nonlinearly
restore
images
from
scattering.
Here,
we
propose
for
first
time
method
full
scene
imaging
that
synergistically
makes
use
an
untrained
network
By
mounting
Stokes
mask
polarizer
CMOS
camera,
simultaneously
obtain
different
states
IFM
calculation
optimize
automatically
without
requiring
extra
training
data.
This
nonlinear
fitting
ability
neural
corrects
undesirable
imperfect
parameter
classical
scenes
.
shows
good
removing
impact
water
preserving
object
information,
making
it
achieve
