CN108267945B - A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network - Google Patents
A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network Download PDFInfo
- Publication number
- CN108267945B CN108267945B CN201810038893.7A CN201810038893A CN108267945B CN 108267945 B CN108267945 B CN 108267945B CN 201810038893 A CN201810038893 A CN 201810038893A CN 108267945 B CN108267945 B CN 108267945B
- Authority
- CN
- China
- Prior art keywords
- neural network
- convex lens
- self
- pupil
- organizing map
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 33
- 238000001093 holography Methods 0.000 title claims abstract description 21
- 230000008030 elimination Effects 0.000 title claims description 13
- 238000003379 elimination reaction Methods 0.000 title claims description 13
- 230000006870 function Effects 0.000 claims description 31
- 210000001747 pupil Anatomy 0.000 claims description 31
- 238000012549 training Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 10
- 210000002569 neuron Anatomy 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 3
- 210000004205 output neuron Anatomy 0.000 claims description 3
- 230000008569 process Effects 0.000 description 5
- 238000012546 transfer Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 210000004218 nerve net Anatomy 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 239000000571 coke Substances 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000005622 photoelectricity Effects 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000527 sonication Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G03—PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
- G03H—HOLOGRAPHIC PROCESSES OR APPARATUS
- G03H1/00—Holographic processes or apparatus using light, infrared or ultraviolet waves for obtaining holograms or for obtaining an image from them; Details peculiar thereto
- G03H1/0005—Adaptation of holography to specific applications
-
- G—PHYSICS
- G03—PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
- G03H—HOLOGRAPHIC PROCESSES OR APPARATUS
- G03H1/00—Holographic processes or apparatus using light, infrared or ultraviolet waves for obtaining holograms or for obtaining an image from them; Details peculiar thereto
- G03H1/04—Processes or apparatus for producing holograms
- G03H1/10—Processes or apparatus for producing holograms using modulated reference beam
- G03H1/12—Spatial modulation, e.g. ghost imaging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Abstract
The method for eliminating optical scanner holography defocus noise based on self-organizing map neural network that the invention discloses a kind of, belongs to optical scanner holography field and image denoising field, mainly solves the problems, such as defocus noise in optical scanner holography.The present invention utilize self-organizing map neural network clustering ability, eliminate through accidental enciphering treated rebuild figure defocus noise.The present invention can efficiently solve the defocus noise problem in optical scanner holography, and the method for this removal noise is suitable for every field.
Description
Technical field
The present invention relates to optical scanner holography fields and image denoising field, in particular, being to be related to a kind of be based on from group
It knits the method for the elimination optical scanner holography defocus noise of map neural network and realizes the device of this method.
Background technique
Optical scanner holographic technique, abbreviation OSH are a kind of recording techniques, can be swept by using single 2-D optical heterodyne
It retouches to record 3-D holographic information.Since the technology proposes, applied in many fields, such as 3-D microscope, 3-D mode
The fields such as identification and hologram image encryption.
Eliminating defocus noise is always a hot research problem in optical scanner holography, for the complete of multilayer wall
Breath figure is significant.When rebuilding hologram in a certain strata coke, remainder layer will be as the clear of defocus influence of noise image
Clear degree, therefore, high-resolution reconstruction hologram, domestic and international scientific research personnel have done many related works in order to obtain.
Document " Sectional image reconstruction in optical scanning holography
Using a random-phase pupil " it proposes a kind of realize using random phase plate optical scanner hologram is encrypted
Algorithm, and a kind of method elimination defocus noise using multi-frame mean is proposed, this method can be very good to eliminate defocus noise,
But due to needing multi-frame mean to handle, timeliness is lower.
Document " Defocus noise suppression with combined frame difference and
Connected component methods in optical scanning holography " propose it is a kind of poor based on frame
The method of method and connected domain, this method has preferable timeliness, but the result edge that this method obtains is rough, and compared to
Original image has deformation.
How to guarantee having the problem of faster processing timeliness while eliminating defocus noise well, is being present invention weight
Point one of solves the problems, such as.
In addition, document " U-Net:Convolutional Networks for Biomedical Image
Segmentation " using depth convolutional neural networks divide microbiology cell image, there is shown a kind of U-shaped network structure,
It is designed by the structure in symmetrical network path, makes it have feature few using sample, fireballing.
Self-organizing map neural network (Self-organizing Feature Maps), abbreviation SOM are based on physiology
The neural network structure proposed with brain science research achievement passes through inherent law and essential attribute in Automatic-searching sample,
Self-organizing adaptively changes network parameter and structure.
The present invention is based on the ingenious clustering characteristics using SOM neural network of the prior art, are incorporated and are made an uproar to defocus
In the problem of sonication, to realize to the defocus noise remove in optical scanner holography.
Summary of the invention
To overcome the above problem in the prior art, the present invention provides that a kind of processing timeliness is higher, eradicating efficacy is preferable
The method of elimination optical scanner holography defocus noise based on self-organizing map neural network, this method utilize SOM neural network,
Without pre-training process, it can be realized and separate defocus noise with object, achieve the purpose that eliminate defocus noise.
To achieve the goals above, The technical solution adopted by the invention is as follows:
A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network, including walk as follows
It is rapid:
(S100) hologram that object obtains encryption is scanned by optical holographic scanning means;
(S200) decryption rebuilds hologram and obtains the reconstruction image with defocus noise;
(S300) training self-organizing map neural network, and clustering is carried out to normalized sample data, disappeared
Except the reconstruction image after defocus noise.
Specifically, the optical holographic scanning means includes laser Laser, and first point corresponding with laser Laser
Beam device BS1, by acousto-optic modulator AOFS, the first reflecting mirror M1, the first pupil p1(x, y) and the first convex lens L1 constitute and with
Corresponding first optical path of first beam splitter BS1, by the second reflecting mirror M2, the second pupil p2(x, y) and the second convex lens L2 are constituted
And corresponding with the first beam splitter BS1 the second optical path, the second beam splitter that the first optical path and the interference of the second optical path are merged
BS2, the scanning light being made of X-Y scanning galvanometer, third convex lens L3 and photoelectric detector PD corresponding with the second beam splitter BS2
Road, wherein the focal length of the first convex lens L1 and the second convex lens L2 are identical, and object to be scanned is placed in the X- of the scanning optical path
Between Y scan galvanometer and third convex lens L3.
Further, in the step (S100), laser Laser tranmitting frequency is ω0Laser, by the first beam splitter
Light wave is divided into two bundles by BS1, wherein a branch of light wave walk the first optical path through acousto-optic modulator AOFS by its frequency by ω0It is promoted to ω0
+ Ω, then successively pass through the first reflecting mirror M1, the first pupil p1(x, y) and the first convex lens L1 is reached at the second beam splitter BS2,
While another Shu Guangbo walks the second optical path and successively passes through the second reflecting mirror M2, the second pupil p2(x, y) and the second convex lens L2,
It reaches at the second beam splitter BS2;
The two-beam wave is interfered at the second beam splitter BS2 forms Fresnel single-slit diffraction, then sweeps through the reflection of X-Y scanning galvanometer
Object is retouched, wherein the light wave through object is converged through third convex lens L3, and is received by photoelectric detector PD, through a series of demodulation
The hologram encrypted.
Wherein, the first pupil p1(x, y) is random phase plate, the second pupil p2(x, y) is 1 function of rectangle.
The optical transfer function of the scanning process are as follows:
Wherein, x and y indicates the position of object under test, and x ' and y ' are integration variable, and z indicates X-Y scanning galvanometer to determinand
The distance of body, λ indicate optical wavelength,Indicate wave number, f is the focal length of the first convex lens and the second convex lens, kxWith
kyIndicate frequency domain coordinates, p1(x, y) and p2(x, y) respectively indicates the first pupil function and the second pupil function,
Herein, since the first pupil is random phase plate, therefore p1(x, y)=exp (j π r (x, y)), due to the second pupil
For 1 function of rectangle, therefore p2(x, y)=1, r (x, y) is the two-dimensional random number of [0,1], then optical transfer function can be expressed as
Wherein, ziIndicate i-th layer of the object under test axial distance with X-Y scanning galvanometer, i=1,2 ..., N, N is determinand
Body indicates p along the axial discrete number of plies, P11The Fourier transformation form of (x, y),*Indicate conjugate operation;
Therefore, the hologram of the encryption is expressed as
Wherein, g (x, y;zi) it is i-th layer of the object under test complex amplitude function being sliced, F and F-1Respectively indicate Fourier transformation
And inverse Fourier transform.
Further, in the step (S200), when the image of r layers of object is rebuild in decryption, setting decoding pupil letter
Number p1d(x, y)=1,Then rebuild zrImage expression is
Wherein, g (x, y;zr) it is in zrPlace rebuilds the image restored, and remainder is noise item on the right side of equation.
Further, the step (S300) specifically includes:
(S310) training self-organizing map neural network,
(S311) establishing input is n-dimensional vector and the two-dimentional SOM neural network with m output node, wherein input
N-dimensional vector is x=[x1,x2,…,xn]T, s-th of connection inputted between neuron node and first of output neuron node
Weight is wsl;
(S312) to wslIt is initialized, is each connection weight wslOne value is randomly selected between [0,1] to make
For initial weight, and keep each initial weight different, the initial value of winning neighborhood is hl,s(x)(0), the initial value of learning rate is η
(0);
(S313) it using the reconstruction image with defocus noise as input sample data, and is normalized;
(S314) input vector is calculated in moment t to the distance of all connection weights:
Wherein, xiIt (t) is value of the input vector in moment t, selection generates DlThe smallest node is used as most matched neuronI.e. winning neuron;
(S315) winning neighborhood function h is definedl,s(x)(t) and learning rate function η (t):
σ (t)=T-k1T ... formula 6
η (t)=η (0)-k2T ... formula 7
Wherein, Rl、Rs(x)It is the position of output node l, s (x) respectively;σ (t) is the radius of neighborhood, and T is the radius of neighbourhood
Initial value, k1For the slope of radius of neighbourhood change curve, 0 < η (t) < 1, k2For the slope of learning rate function curve;
(S316) connection weight of neuron is adjusted by way of stochastic gradient descent,
wsl(t+1)=wsl(t)+η(t)hl,s(x)[xs(t)-wsl(t)] ... formula 8
Wherein, t dullness reduces η (t) at any time, hl,s(x)(t) Gauss neighborhood function is used;
(S317) step (S314) to (S316) is repeated, until learning rate decays to 0, completion is to two dimension SOM nerve net
The training of network;
(S320) m class data set is obtained after the completion of training, wherein last a kind of data set is the reconstruction image after denoising.
Compared with prior art, the invention has the following advantages:
(1) present invention uses self-organizing map neural network and optical scanner holographic technique, utilizes self-organizing map neural
The clustering ability of network, can be very good by the reconstruction image with defocus noise defocus noise and object divided
Class, to realize the defocus noise for eliminating the hologram after accidental enciphering.
(2) method that the present invention uses self-organizing map neural network, the hologram after efficiently solving accidental enciphering
With defocus noise problem.
(3) this method does not need pre-training process compared to other deep learning methods, and the algorithm training time is short.
(4) not only implementation is simple, is convenient for operation by the present invention, while having very strong usability.
Detailed description of the invention
Fig. 1 is the basic structure schematic diagram of optical holographic scanning means in the present invention.
Fig. 2 is that one group of subject image of scanning used in the embodiment of the present invention is illustrated.
Fig. 3 is the reconstruction image that the embodiment of the present invention must have defocus noise, and (a) and (b) respectively corresponds corresponding in Fig. 2
Subject image.
Fig. 4 is the training flow diagram of SOM neural network used in the embodiment of the present invention.
Fig. 5 is SOM neural network basic model used in the embodiment of the present invention.
Fig. 6 is the denoising result of the embodiment of the present invention, and (a) and (b) respectively corresponds the reconstruction with defocus noise in Fig. 3
Image.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples, and embodiments of the present invention include but is not limited to
The following example.
Embodiment
It as shown in Figures 1 to 6, should be based on the side of the elimination optical scanner holography defocus noise of self-organizing map neural network
Method, the specific steps are as follows:
(S100) hologram that object obtains encryption is scanned by optical holographic scanning means:
The basic structure of its used optical holographic scanning means is as shown in Figure 1, comprising: laser Laser, with laser
The corresponding first beam splitter BS1 of device Laser, by acousto-optic modulator AOFS, the first reflecting mirror M1, the first pupil p1(x, y) and
First optical path that one convex lens L1 is constituted and corresponding with the first beam splitter BS1, by the second reflecting mirror M2, the second pupil p2(x,
Y) and the second convex lens L2 is constituting and corresponding with the first beam splitter BS1 the second optical path, and the first optical path and the second optical path are done
The second beam splitter BS2 of merging is related to, it is corresponding with the second beam splitter BS2 to be visited by X-Y scanning galvanometer, third convex lens L3 and photoelectricity
Survey the scanning optical path that device PD is constituted, wherein object to be scanned is placed in the X-Y scanning galvanometer and third convex lens of the scanning optical path
Between mirror L3;Wavelength X=the 632.8nm for the light wave that laser issues, the focal length phase of the first convex lens L1 and the second convex lens L2
It together, is all 400mm, the distance of X-Y scanning galvanometer to bilayer slice is respectively z1=12mm, z2=13mm, scanned object is as schemed
Shown in 2, for the slice of use having a size of 1mm × 1mm, sampling pixel points are 256 × 256.
The process that the optical holographic scanning means executes are as follows: laser Laser tranmitting frequency is ω0Laser, by first
Light wave is divided into two bundles by beam splitter BS1, wherein a branch of light wave walk the first optical path through acousto-optic modulator AOFS by its frequency by ω0It mentions
It is upgraded to ω0+ Ω, then successively pass through the first reflecting mirror M1, the first pupil p1(x, y) and the first convex lens L1 reaches the second beam splitter
At BS2, while another Shu Guangbo walks the second optical path and successively passes through the second reflecting mirror M2, the second pupil p2(x, y) and the second convex lens
Mirror L2 is also reached at the second beam splitter BS2, wherein the first pupil p1(x, y) is random phase plate, the second pupil p2
(x, y) is 1 function of rectangle;
The two-beam wave is interfered at the second beam splitter BS2 forms Fresnel single-slit diffraction, then sweeps through the reflection of X-Y scanning galvanometer
Object is retouched, wherein the light wave through object is converged through third convex lens L3, and is received by photoelectric detector PD, through a series of demodulation
The hologram encrypted.
The optical transfer function of the process are as follows:
Wherein, x and y indicates the position of object under test, and x ' and y ' are integration variable, and z indicates X-Y scanning galvanometer to determinand
The distance of body, λ indicate optical wavelength,Indicate wave number, f is the focal length of the first convex lens and the second convex lens, kxWith
kyIndicate frequency domain coordinates, p1(x, y) and p2(x, y) respectively indicates the first pupil function and the second pupil function,
Herein, since the first pupil is random phase plate, therefore p1(x, y)=exp (j π r (x, y)), due to the second pupil
For 1 function of rectangle, therefore p2(x, y)=1, r (x, y) is the two-dimensional random number of [0,1], then optical transfer function can be expressed as
Wherein, ziIndicate i-th layer of the object under test axial distance with X-Y scanning galvanometer, i=1,2 ..., N, N is determinand
Body indicates p along the axial discrete number of plies, P11The Fourier transformation form of (x, y),*Indicate conjugate operation;
Therefore, the hologram of the encryption is expressed as
Wherein, g (x, y;zi) it is i-th layer of the object under test complex amplitude function being sliced, F and F-1Respectively indicate Fourier transformation
And inverse Fourier transform.
(S200) decryption rebuilds hologram and obtains the reconstruction image with defocus noise, as shown in Figure 4;
In order to rebuild the image of r layers of object, setting decoding pupil function p1d(x, y)=1,Then rebuild zrImage can be expressed as:
Wherein, g (x, y;zr) it is in zrPlace rebuilds the image restored, and remainder is noise item on the right side of equation.
(S300) training self-organizing map neural network, and clustering is carried out to normalized sample data, disappeared
Except the reconstruction image after defocus noise:
(S310) training self-organizing map neural network,
(S311) establishing input is n-dimensional vector and the two-dimentional SOM neural network with m output node, wherein the n inputted
Dimensional vector is x=[x1,x2,…,xn]T, s-th of connection weight inputted between neuron node and first of output neuron node
Value is wsl;N=65536, m=18 are selected in the present embodiment;
(S312) to wslIt is initialized, is each connection weight wslOne value is randomly selected between [0,1] to make
For initial weight, and keep each initial weight different, the initial value of winning neighborhood is hl,s(x)(0), the initial value of learning rate is η
(0);
(S313) using the reconstruction image with defocus noise as input sample data, since input sample image is
256 × 256, matrix-vector is become to 65536 × 1 vector, is then normalized, comply with input n tie up to
The format of amount;
(S314) input vector is calculated in moment t to the distance of all connection weights:
Wherein, xsIt (t) is value of the input vector in moment t, selection generates DlThe smallest node is used as most matched neuronI.e. winning neuron;
(S315) winning neighborhood function h is definedl,s(x)(t) and learning rate function η (t):
σ (t)=T-k1T ... formula 6
η (t)=η (0)-k2T ... formula 7
Wherein, Rl、Rs(x)It is the position of output node l, s (x) respectively;σ (t) is the radius of neighborhood, and T is the radius of neighbourhood
Initial value, k1For the slope of radius of neighbourhood change curve, 0 < η (t) < 1, k2For the slope of learning rate function curve;According to aforementioned ginseng
It counts, in the present embodiment, T=3, k1=0.1, η (0)=0.92, k2=0.04;
(S316) connection weight of neuron is adjusted by way of stochastic gradient descent,
wsl(t+1)=wsl(t)+η(t)hl,s(x)[xs(t)-wsl(t)] ... formula 8
Wherein, t dullness reduces η (t) at any time, hl,s(x)(t) Gauss neighborhood function is used;
(S317) step (S314) to (S316) is repeated, until learning rate decays to 0, completion is to two dimension SOM nerve net
The training of network;
(S320) m class data set is obtained after the completion of training, wherein last a kind of data set is the reconstruction image after denoising.
As shown in Figure 6, it can be seen that after neural network, eliminate most of defocus noise.
Above-described embodiment is merely a preferred embodiment of the present invention, and it is not intended to limit the protection scope of the present invention, as long as using
Design principle of the invention, and the non-creative variation worked and made is carried out on this basis, it should belong to of the invention
Within protection scope.
Claims (6)
1. a kind of method of the elimination optical scanner holography defocus noise based on self-organizing map neural network, which is characterized in that
Include the following steps:
(S100) hologram that object obtains encryption is scanned by optical holographic scanning means;
(S200) decryption rebuilds hologram and obtains the reconstruction image with defocus noise;
(S300) training self-organizing map neural network, and to normalized sample data carry out clustering, be eliminated from
Reconstruction image after burnt noise:
(S310) training self-organizing map neural network,
(S311) establishing input is n-dimensional vector and the two-dimentional SOM neural network with m output node, wherein s-th of input
Connection weight between neuron node and first of output neuron node is wsl;
(S312) to wslIt is initialized, is each connection weight wslOne value is randomly selected between [0,1] as just
Beginning weight, and keep each initial weight different, the initial value of winning neighborhood is hl,s(x)(0), the initial value of learning rate is η (0);
(S313) it using the reconstruction image with defocus noise as input sample data, and is normalized;
(S314) input vector is calculated in moment t to the distance of all connection weights:
Wherein, xiIt (t) is value of the input vector in moment t, selection generates DlThe smallest node is used as most matched neuronI.e. winning neuron;
(S315) winning neighborhood function h is definedl,s(x)(t) and learning rate function η (t):
σ (t)=T-k1T ... formula 6
η (t)=η (0)-k2T ... formula 7
Wherein, Rl、Rs(x)It is the position of output node l, s (x) respectively;σ (t) is the radius of neighborhood, and T is the initial value of the radius of neighbourhood,
k1For the slope of radius of neighbourhood change curve, 0 < η (t) < 1, k2For the slope of learning rate function curve;
(S316) connection weight of neuron is adjusted by way of stochastic gradient descent,
wsl(t+1)=wsl(t)+η(t)hl,s(x)[xs(t)-wsl(t)] ... formula 8
Wherein, t dullness reduces η (t) at any time, hl,s(x)(t) Gauss neighborhood function is used;
(S317) step (S314) to (S316) is repeated, until learning rate decays to 0, completion is to two dimension SOM neural network
Training;
(S320) m class data set is obtained after the completion of training, wherein last a kind of data set is the reconstruction image after denoising.
2. the side of the elimination optical scanner holography defocus noise according to claim 1 based on self-organizing map neural network
Method, which is characterized in that the optical holographic scanning means includes laser Laser, the first beam splitting corresponding with laser Laser
Device BS1, by acousto-optic modulator AOFS, the first reflecting mirror M1, the first pupil p1It is that (x, y) and the first convex lens L1 are constituted and with the
Corresponding first optical path of one beam splitter BS1, by the second reflecting mirror M2, the second pupil p2What (x, y) and the second convex lens L2 were constituted
And the second optical path corresponding with the first beam splitter BS1, the second beam splitter BS2 that the first optical path and the interference of the second optical path are merged,
The scanning optical path being made of X-Y scanning galvanometer, third convex lens L3 and photoelectric detector PD corresponding with the second beam splitter BS2,
Wherein, the focal length of the first convex lens L1 and the second convex lens L2 are identical, and the X-Y that object to be scanned is placed in the scanning optical path is swept
It retouches between galvanometer and third convex lens L3.
3. a kind of elimination optical scanner holography defocus noise based on self-organizing map neural network according to claim 2
Method, which is characterized in that in the step (S100), laser Laser tranmitting frequency be ω0Laser, by the first beam splitting
Light wave is divided into two bundles by device BS1, wherein a branch of light wave walk the first optical path through acousto-optic modulator AOFS by its frequency by ω0It is promoted to
ω0+ Ω, then successively pass through the first reflecting mirror M1, the first pupil p1(x, y) and the first convex lens L1 reaches the second beam splitter BS2
Place, while another Shu Guangbo walks the second optical path and successively passes through the second reflecting mirror M2, the second pupil p2(x, y) and the second convex lens
L2 is also reached at the second beam splitter BS2;
The two-beam wave is interfered at the second beam splitter BS2 forms Fresnel single-slit diffraction, then reflects scanning object through X-Y scanning galvanometer
Body wherein the light wave through object is converged through third convex lens L3, and is received by photoelectric detector PD, is obtained through a series of demodulation
The hologram of encryption.
4. the side of the elimination optical scanner holography defocus noise according to claim 3 based on self-organizing map neural network
Method, which is characterized in that the first pupil p1(x, y) is random phase plate, the second pupil p2(x, y) is 1 function of rectangle.
5. the elimination optical scanner holography according to any one of claims 1 to 4 based on self-organizing map neural network from
The method of burnt noise, which is characterized in that the hologram of the encryption is expressed as
Wherein, i=1,2 ..., N, N are object under test along the axial discrete number of plies, g (x, y;zi) it is i-th layer of object under test slice
Complex amplitude function, F and F-1Fourier transformation and inverse Fourier transform are respectively indicated, x and y indicate the position of object under test, z table
Show X-Y scanning galvanometer to object under test distance,Indicate wave number, λ indicate optical wavelength, f be the first convex lens and
The focal length of second convex lens, kxAnd kyIndicate frequency domain coordinates, P1Indicate the first pupil function p1(x, y)=exp (j π r (x, y))
Fourier transformation form,*Indicate that conjugate operation, r (x, y) are the two-dimensional random number of [0,1],
6. the side of the elimination optical scanner holography defocus noise according to claim 5 based on self-organizing map neural network
Method, which is characterized in that in the step (S200), when the image of r layers of object is rebuild in decryption, setting decoding pupil function p1d
(x, y)=1,Then rebuild zrImage expression is
Wherein, g (x, y;zr) it is in zrPlace rebuilds the image restored, and remainder is noise item on the right side of equation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810038893.7A CN108267945B (en) | 2018-01-16 | 2018-01-16 | A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810038893.7A CN108267945B (en) | 2018-01-16 | 2018-01-16 | A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108267945A CN108267945A (en) | 2018-07-10 |
CN108267945B true CN108267945B (en) | 2019-08-09 |
Family
ID=62775606
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810038893.7A Expired - Fee Related CN108267945B (en) | 2018-01-16 | 2018-01-16 | A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108267945B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109884869A (en) * | 2019-04-22 | 2019-06-14 | 重庆邮电大学 | One kind calculating holographic imaging systems and method based on dicoria heterodyne |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6445470B1 (en) * | 1990-04-06 | 2002-09-03 | University Of Southern California | Apparatus for simultaneous spatial modulation of incoherent/coherent angulary multiplexed optical beams |
WO2002086565A1 (en) * | 2001-04-20 | 2002-10-31 | Bookham Technology Plc | Photonic band gap device |
CN101221662A (en) * | 2008-01-31 | 2008-07-16 | 复旦大学 | Remote sensing image mixed image element decomposition method based on self-organizing mapping neural network |
CN104159094A (en) * | 2014-07-09 | 2014-11-19 | 四川大学 | Method for improving optical scanning holographic tomography effect |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006270656A (en) * | 2005-03-24 | 2006-10-05 | Fuji Xerox Co Ltd | Apparatus and method for image processing |
JP4961182B2 (en) * | 2005-10-18 | 2012-06-27 | 株式会社リコー | Noise removal apparatus, noise removal method, noise removal program, and recording medium |
US8270732B2 (en) * | 2009-08-31 | 2012-09-18 | Behavioral Recognition Systems, Inc. | Clustering nodes in a self-organizing map using an adaptive resonance theory network |
CN103488662A (en) * | 2013-04-01 | 2014-01-01 | 哈尔滨工业大学深圳研究生院 | Clustering method and system of parallelized self-organizing mapping neural network based on graphic processing unit |
CN105204311B (en) * | 2015-07-06 | 2018-05-18 | 电子科技大学 | A kind of optical scanner holography edge detection method based on Gauss apodization |
CN107085838B (en) * | 2017-06-02 | 2021-02-26 | 新疆大学 | Method and device for removing hologram noise |
-
2018
- 2018-01-16 CN CN201810038893.7A patent/CN108267945B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6445470B1 (en) * | 1990-04-06 | 2002-09-03 | University Of Southern California | Apparatus for simultaneous spatial modulation of incoherent/coherent angulary multiplexed optical beams |
WO2002086565A1 (en) * | 2001-04-20 | 2002-10-31 | Bookham Technology Plc | Photonic band gap device |
CN101221662A (en) * | 2008-01-31 | 2008-07-16 | 复旦大学 | Remote sensing image mixed image element decomposition method based on self-organizing mapping neural network |
CN104159094A (en) * | 2014-07-09 | 2014-11-19 | 四川大学 | Method for improving optical scanning holographic tomography effect |
Also Published As
Publication number | Publication date |
---|---|
CN108267945A (en) | 2018-07-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108089425B (en) | A method of the elimination optical scanner holography defocus noise based on deep learning | |
Ramos et al. | Real-time, multiframe, blind deconvolution of solar images | |
Kang et al. | Phase extraction neural network (PhENN) with coherent modulation imaging (CMI) for phase retrieval at low photon counts | |
Pellizzari et al. | Imaging through distributed-volume aberrations using single-shot digital holography | |
Yang et al. | A hybrid method for multi-focus image fusion based on fast discrete curvelet transform | |
Mohammed et al. | Quality assessment of refocus criteria for particle imaging in digital off-axis holography | |
Tang et al. | RestoreNet: a deep learning framework for image restoration in optical synthetic aperture imaging system | |
Leo et al. | Multilevel bidimensional empirical mode decomposition: a new speckle reduction method in digital holography | |
CN104008536A (en) | Multi-focus noise image fusion method based on CS-CHMT and IDPCNN | |
Leihong et al. | Restoration of single pixel imaging in atmospheric turbulence by Fourier filter and CGAN | |
Chen et al. | DH-GAN: a physics-driven untrained generative adversarial network for holographic imaging | |
CN108267945B (en) | A method of the elimination optical scanner holography defocus noise based on self-organizing map neural network | |
Zhang et al. | Learning synthetic aperture radar image despeckling without clean data | |
Saponara et al. | Reconstruct fingerprint images using deep learning and sparse autoencoder algorithms | |
Su et al. | Explainable analysis of deep learning methods for sar image classification | |
Balasubramani et al. | Influence of noise-reduction techniques in sparse-data sample rotation tomographic imaging | |
EP3535623B1 (en) | System and method for reconstruction of holographic lens-free images by multi-depth sparse phase recovery | |
Li et al. | Deep adversarial network for super stimulated emission depletion imaging | |
Liu et al. | Joint artifact correction and super-resolution of image slicing and mapping system via a convolutional neural network | |
Sulaiman et al. | Subspace wavefront estimation using image sharpening and predictive dynamic digital holography | |
CN107835074A (en) | A kind of method for eliminating accidental enciphering optical scanner holography defocus noise | |
Wang et al. | NSTMR: Super resolution of Sentinel-2 images using nonlocal nonconvex surrogate of tensor multirank | |
Latorre-Carmona et al. | Three-dimensional imaging with multiple degrees of freedom using data fusion | |
CN108153132B (en) | A kind of optical scanner holography self-focusing method based on mean value gradient function | |
Wu et al. | Hyperspectral image fusion based on L-NMF |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190809 Termination date: 20200116 |