CN109035168A - A kind of calculating imaging method and device based on fold mask construction - Google Patents
A kind of calculating imaging method and device based on fold mask construction Download PDFInfo
- Publication number
- CN109035168A CN109035168A CN201810791144.1A CN201810791144A CN109035168A CN 109035168 A CN109035168 A CN 109035168A CN 201810791144 A CN201810791144 A CN 201810791144A CN 109035168 A CN109035168 A CN 109035168A
- Authority
- CN
- China
- Prior art keywords
- mask
- fold
- sequence
- group
- symmetrical
- 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.)
- Granted
Links
- 238000003384 imaging method Methods 0.000 title claims abstract description 54
- 238000010276 construction Methods 0.000 title claims abstract description 33
- 230000000295 complement effect Effects 0.000 claims abstract description 28
- 238000005259 measurement Methods 0.000 claims abstract description 24
- 238000012937 correction Methods 0.000 claims abstract description 20
- 238000005457 optimization Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 48
- 230000002146 bilateral effect Effects 0.000 claims description 45
- 238000000034 method Methods 0.000 claims description 35
- 238000004364 calculation method Methods 0.000 claims description 22
- 238000003860 storage Methods 0.000 claims description 11
- 230000005856 abnormality Effects 0.000 claims description 9
- 230000001052 transient effect Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 abstract description 10
- 206010020751 Hypersensitivity Diseases 0.000 abstract description 2
- 230000003287 optical effect Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 230000003247 decreasing effect Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 230000007306 turnover Effects 0.000 description 2
- 241000287196 Asthenes Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000003325 tomography Methods 0.000 description 1
- 230000004304 visual acuity Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
Abstract
The present invention provides a kind of calculating imaging method and device based on fold mask construction, which comprises by reducing symmetrically with symmetrical anti-fold, generates mask sequence；According to the number of connected region in each mask of the mask sequence, unusual correction is carried out to the mask sequence, obtains optimal mask sequence；The mask that preset number is selected from the optimal mask sequence, carries out complementary differential measurement and image reconstruction.The present invention treats the percentage contribution of imageable target object reconstruction quality according to mask, it is reduced using mask and symmetrically constructs mask sequence with symmetrical anti-fold, and unusual correction is carried out to the sequence of mask by mask connected region number incremental order, reduce the probability that the identical mask of connected region number occurs, reduce the uncertainty of mask, optimize mask construction, image reconstruction is carried out using the mask sequence of optimization, the quality and efficiency for improving image reconstruction have many advantages, such as that super sub-sampling, hypersensitive, sampling dimension is low, high speed is real-time.
Description
Technical field
The invention belongs to calculate technical field of imaging, it is calculated as more particularly, to a kind of based on fold mask construction
As method and device.
Background technique
Optical imagery is that acquisition information is the most universal and effective mode.Traditional imaging technique uses detector array,
That is pixelation sensor or spot scan sampling, have developed relative maturity.But under the image-forming condition of pole dim light and non-visible light
There are still many limitations.
In recent years, scientists have developed calculating imaging technique, using individually the point of spatial resolving power is not visited
Meet the reconstruction image for calculating object, thus also commonly known as " single pixel between the measured value surveyed on device and modulation speckle
Imaging ", and point detector measurement is light intensity summation in entire light field in the modulated pixel picked out.It is good by it
Performance, single pixel imaging technique can be suitable for the imaging of non-visual wave band, such as multi-wavelength imaging, Depth Imaging, polarization very well
Imaging, light spectrum image-forming, ultraviolet imagery, infrared three-dimensional imaging, terahertz imaging etc..The calculating imaging of early stage includes that ghost imaging (is closed
Be unified into picture), calculate illumination imaging, Hadamard imaging, Fourier imaging etc., these overwhelming majority are needed using the orthogonal of full rank
Matrix is directly equal to the image resolution ratio of object to be imaged as calculation matrix, namely measurement number, and which has limited the sampling times.
In order to solve this world-famous puzzle, compressive sensing theory (Compressed Sensing, CS) comes into being.The reason
By pointing out, it is assumed that a signal has sparsity or a compressibility, and natural image can under some base rarefaction representation, by
In the modulation capability of spatial light modulator, it is only necessary to a small amount of non-adaptive linear measurement is carried out to the signal, it just can be to the signal
Perfect reconstruction is carried out, sample rate may be significantly lower than that Nyquist-Shannon rate, and have certain robust to measurement noise
Property.Thus single pixel camera, Magnetic resonance imaging, Tomography, astronomical imaging etc. based on CS have been developed.And
In these applications, priori knowledge is at key.In fact CS technology is to exchange spatial resolution for time of measuring and calculating time
With high pass measurement, in the imaging example of big pixel scale, the disadvantage of CS is completely exposed, namely when needing a large amount of sampling
Between, computation complexity is high, and memory source consumption is huge, and has ignored contribution feature of the feature to image reconstruction of mask, this
The practical development of CS is greatly limited a bit.
Summary of the invention
Above-mentioned existing calculating imaging method sample rate is high, computation complexity is high and memory consumption is big to overcome the problems, such as or
Person at least is partially solved the above problem, and the present invention provides a kind of calculating imaging method and device based on fold mask construction.
According to the first aspect of the invention, a kind of calculating imaging method based on fold mask construction is provided, comprising:
By reducing symmetrically with symmetrical anti-fold, mask sequence is generated；
According to the number of connected region in each mask of the mask sequence, unusual correction is carried out to the mask sequence,
Obtain optimal mask sequence；
According to the optimal mask sequence, image reconstruction is carried out.
A kind of calculating imaging device based on fold mask construction is provided according to a second aspect of the present invention, comprising:
Generation module, for generating mask sequence by reducing symmetrically with symmetrical anti-fold；
Optimization module, for the number of connected region in each mask according to the mask sequence, to the mask sequence
Unusual correction is carried out, optimal mask sequence is obtained；
Reconstructed module, for carrying out image reconstruction according to the optimal mask sequence.
According to the third aspect of the invention we, a kind of electronic equipment is provided, comprising:
At least one processor, at least one processor and bus；Wherein,
The processor and memory complete mutual communication by the bus；
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to refer to
Order is able to carry out foregoing method.
According to the fourth aspect of the invention, a kind of non-transient computer readable storage medium, the non-transient calculating are provided
Machine readable storage medium storing program for executing stores computer instruction, and the computer instruction makes the computer execute foregoing method.
The present invention provides a kind of calculating imaging method and device based on fold mask construction, and this method is treated according to mask
The percentage contribution of imageable target object reconstruction quality is reduced using mask and symmetrically constructs mask sequence with symmetrical anti-fold, and pressed
Mask connected region number incremental order carries out the sequence to mask and carries out unusual correction, reduces that connected region number is identical to be covered
The probability that mould occurs, reduces the uncertainty of mask, optimizes mask construction, so that important mask is in spatial light modulator
Preferential display carries out image reconstruction using the mask sequence of optimization, and introduces the mode of operator in imaging algorithm, so that big number
According to the calculation matrix of amount without storing repeatedly in calculating process, largely simplification matrix multiplication and inversion operation, are reduced
Computation complexity and calculating memory consumption, and then realize the overdelicate supercomputing imaging of the super sub-sampling of high pixel resolution,
Improve the quality and efficiency of image reconstruction.
Detailed description of the invention
Fig. 1 is the calculating imaging method overall flow schematic diagram provided in an embodiment of the present invention based on fold mask construction；
Fig. 2 is first group of mask signal in the calculating imaging method provided in an embodiment of the present invention based on fold mask construction
Figure；
Fig. 3 is second group of mask signal in the calculating imaging method provided in an embodiment of the present invention based on fold mask construction
Figure；
Fig. 4 is first in calculating imaging method second group of mask provided in an embodiment of the present invention based on fold mask construction
The generation schematic diagram of a mask；
Fig. 5 is 256 rank mask sequences in the calculating imaging method provided in an embodiment of the present invention based on fold mask construction
In unusual mask set position view；
Fig. 6 is the calculating imaging device overall structure diagram provided in an embodiment of the present invention based on fold mask construction；
Fig. 7 is that overall system architecture schematic diagram is imaged in the calculating provided in an embodiment of the present invention based on fold mask construction；
Fig. 8 be another embodiment of the present invention provides based on fold mask construction calculating be imaged overall system architecture signal
Figure；
Fig. 9 is electronic equipment overall structure diagram provided in an embodiment of the present invention.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
A kind of calculating imaging method based on fold mask construction is provided in one embodiment of the invention, and Fig. 1 is this
The calculating imaging method overall flow schematic diagram based on fold mask construction that inventive embodiments provide, this method comprises: S101,
By reducing symmetrically with symmetrical anti-fold, mask sequence is generated；
Wherein, mask is 0 and 1 matrix constituted.If the Pixel Dimensions of target object to be imaged are p × q, wherein p and q
It is even number, then the size of each mask is also p × q.If initial mask, i.e., first mask is monodromy matrix, such as all values
For 1 matrix.It carries out reducing symmetrical and symmetrical anti-fold since initial mask, new mask is generated, then to new mask
It carries out reducing symmetrical and symmetrical anti-fold again, new mask is generated, until the mask count of generation reaches preset sum.
Using all masks of generation as a mask sequence.Mask reduce and symmetrically refers to the length of mask and wide reduces respectively
To original half, then using the mask of diminution as the upper left of new mask, newly covered according to the generation of the mask of diminution
Mould, new mask top and the bottom are symmetrical, and left-right parts are symmetrical.Symmetrical anti-fold includes anti-fold symmetrical above and below, bilateral symmetry is counter turns over
Folding and anti-fold symmetrical above and below and the anti-fold of bilateral symmetry.Anti- fold symmetrical above and below refers to that the upper half block of holding is constant, makes lower half
Block and upper half block took originally the inverse value of plate pixel value down at axial symmetry, then by the pixel value of lower half block.Bilateral symmetry is counter to be turned over
Folding, which refers to, keeps left half block constant, and right half block and left half block is made to take original right panel block at axial symmetry, then by the pixel value of right half block
The inverse value of pixel value.Anti- fold symmetrical above and below refers to the anti-fold of bilateral symmetry carries out a left side for the anti-result turned down symmetrical above and below
Right symmetrical anti-fold；Alternatively, the symmetrically anti-result turned down is carried out anti-fold symmetrical above and below.
S102 carries out unusual correction to mask sequence, obtains according to the number of connected region in each mask of mask sequence
Optimal mask sequence；
Wherein, connected region refer to by mask with same pixel value and position it is adjacent pixel composition image district
Domain.Fewer in view of the connected region number of mask, the reconstruction quality contribution that mask treats imageable target object is bigger.To mask
Symmetrical anti-fold is carried out to generate in new mask process, it, will not when the pixel value to mask some regional area takes inverse value
Change the number of connected region in this regional area, but the part-area edge and other region joining parts may be because of taking
The variation of value and the variation that connected domain number occurs, carry out needing when anti-fold symmetrical above and below and the anti-fold of bilateral symmetry at the same time through
Cross that two sub-symmetries are counter to be turned down, the contribution for treating imageable target object reconstruction quality is smaller.And when carrying out reducing symmetrical, symmetrically
The pixel value of axis two sides is centainly identical, this make the part-area edge in the case of this connect caused by the variation of connected domain number
Smaller, i.e., it is larger to the contribution of object reconstruction quality.It is only carrying out anti-fold symmetrical above and below and is only carrying out symmetrically counter turn over
It can successively decrease there is a situation where connected domain number when folding, that is, occur unusual；For this purpose, unusual correction need to be carried out.After abnormality is corrected
Mask sequence is as optimal mask sequence.
S103 carries out image reconstruction according to optimal mask sequence.
Image reconstruction is carried out to target object to be imaged according to optimal mask sequence.The present embodiment is not limited to image reconstruction
Method.
The present embodiment treats the percentage contribution of imageable target object reconstruction quality according to mask, using mask reduce symmetrically with
Symmetrical anti-fold construction mask sequence, and carry out the progress abnormality of the sequence to mask by mask connected region number incremental order and entangle
Just, the probability that the identical mask of connected region number occurs is reduced, the uncertainty of mask is reduced, optimizes mask construction, make
Image reconstruction is carried out with the mask sequence of optimization, improves the quality and efficiency of image reconstruction.
On the basis of the above embodiments, mask sequence is generated by reducing symmetrically with symmetrical anti-fold in the present embodiment
The step of specifically include: carry out anti-fold symmetrical above and below, the anti-fold of bilateral symmetry and symmetrical above and below anti-respectively to initial mask
Fold and the anti-fold of bilateral symmetry, using three kinds of symmetrical anti-fold results of initial mask and initial mask as first group of mask；
Wherein, the anti-result turned down symmetrical above and below is carried out as second mask in first group of mask for initial mask, it will be to first
Beginning mask carries out the anti-result turned down of bilateral symmetry as the third mask in first group of mask；Alternatively, will be to initial mask
The anti-result turned down of bilateral symmetry is carried out as second mask in first group of mask, initial mask will be carried out symmetrical above and below
The anti-result turned down is as the third mask in first group of mask；It is symmetrically and symmetrical anti-by reducing according to first group of mask
Fold generates mask sequence；Wherein, the step of carrying out anti-fold symmetrical above and below to initial mask specifically includes: holding is initially covered
The top half of mould is constant, and the lower half portion of initial mask is taken inverse value；Wherein, the top half of initial mask with initially cover
The lower half portion of mould is at axial symmetry；The step of bilateral symmetry instead turns down is carried out to initial mask to specifically include: keeping initial mask
Left-half it is constant, the right half part of initial mask is taken into inverse value；Wherein, the left-half and initial mask of initial mask
Right half part at axial symmetry；The step of carrying out anti-fold symmetrical above and below and symmetrically anti-fold to initial mask is specifically wrapped
It includes: the initial mask anti-result turned down symmetrical above and below is subjected to the anti-fold of bilateral symmetry；Alternatively, initial mask bilateral symmetry is anti-
The result of fold carries out anti-fold symmetrical above and below.
As shown in Fig. 2, 4 masks that Fig. 2 is first group.Wherein, the 1st mask is preset initial mask, just
Beginning mask is monodromy matrix, might as well be set as complete 1 matrix.2nd mask be by the 1st mask by it is symmetrical above and below it is anti-fold and
, the 3rd mask is to be obtained by the 1st mask by the anti-fold of bilateral symmetry.It can also be by the 1st mask with the 2nd mask
It is obtained by the anti-fold of bilateral symmetry, the 3rd mask is to be obtained by the 1st mask by anti-fold symmetrical above and below.4th mask
It is then to carry out the anti-fold of bilateral symmetry by the 1st mask after anti-fold symmetrical above and below and obtain, it can be by the 2nd mask warp
It crosses the anti-fold of bilateral symmetry and obtains.Alternatively, the 4th mask by the 1st mask after the anti-fold of bilateral symmetry, then carry out up and down
It symmetrically instead turns down and obtains, can be obtained by the 3rd mask by anti-fold symmetrical above and below.Namely the 4th is generated by the 1st mask
In mask process, anti-fold symmetrical above and below and the anti-precedence turned down of bilateral symmetry can be overturned.
On the basis of the above embodiments, it is symmetrically turned over symmetrically counter according to first group of mask by reducing in the present embodiment
The step of folding, generation mask sequence, specifically includes: for any non-monodrome mask in first group of mask, by the length and width of the mask
It is reduced into original half respectively, symmetrical above and below and symmetrical new mask is generated based on the mask after diminution；It is right
New mask carries out anti-fold, the anti-fold of bilateral symmetry and anti-fold symmetrical above and below symmetrical above and below respectively and bilateral symmetry is counter turns over
Folding, using three kinds of symmetrical anti-fold results of new mask and new mask as one group of new mask；Any in new mask set is covered
Mould iteration executes the step of reducing symmetrically with symmetrical anti-fold, and all mask sets are constituted mask by the sequencing of generation time
Sequence.It should be noted that since first mask in first group of mask is monodromy matrix, to first in first group of mask
A mask carry out reducing respectively obtain after symmetrical and symmetrical anti-fold be still first group of mask, therefore, first group is not covered
First mask in mould reduce symmetrical.Fig. 3 is 4 masks in second group, wherein first mask in second group
It is to be obtained by second mask in whole mask sequence by reducing symmetrical.As shown in figure 4, first mask is in Fig. 4
The length of first mask in Fig. 4 and width are reduced into original half, as new mask by second mask in first group
The a quarter part in the upper left corner, such as second mask in Fig. 4.Then symmetrical above and below and left and right is generated according to the mask of diminution
Symmetrical new mask, such as the third mask in Fig. 4.Using the third mask in Fig. 4 as first mask in second group.
The 2nd mask in second group is obtained by the anti-fold symmetrical above and below of first mask in second group, and the 3rd in second group covers
Mould is obtained by the anti-fold of first mask bilateral symmetry in second group.Wherein, what is generated after second group and second group is all
Carried out respectively in mask set anti-fold symmetrical above and below and the anti-sequence turned down of bilateral symmetry with first group in carry out respectively it is upper and lower right
Claim anti-fold identical with the anti-sequence turned down of bilateral symmetry.The 4th mask in second group is by first mask in second group
The anti-fold acquisition of bilateral symmetry is carried out again after carrying out anti-fold symmetrical above and below, or left by first mask progress in second group
Anti- fold symmetrical above and below is carried out again after right symmetrical anti-fold to obtain.Similarly, it is covered based on the third in whole mask sequence
Mould carries out reducing symmetrical and symmetrical anti-fold respectively, generates third group mask, is covered based on the 4th in whole mask sequence
Mould generates the 4th group of mask.In next iteration, carries out reducing respectively symmetrical based on acquired each mask and symmetrical counter turn over
Folding, generates new mask set, and so on, until the quantity of the mask set of generation reaches the image total pixel number of target object.
Every one group of 4 masks is set in mask sequence, for s group, s=1,2,3 ..., N/4, wherein N is mask sequence
The total number of mask in column.The number of first mask is n in s group1=4s-3, the number of second mask are n2=4s-2,
The number of third mask is n3The number of=4s-1, the 4th mask are n4=4s.Number is n1Mask be by generating before
Whole mask sequence in s-th of mask by reducing symmetrical generate.Number is n2Mask by numbering be n1Mask pass through
It is symmetrical above and below it is anti-fold and obtain, number n3Mask by numbering be n1Mask obtained by the anti-fold of bilateral symmetry, can also be with
Number is n2Mask by numbering be n1Mask obtained, number n by the anti-fold of bilateral symmetry3Mask by numbering be n1
Mask by it is symmetrical above and below it is anti-fold and obtain.Number is n4Mask by numbering be n1Mask pass through anti-fold symmetrical above and below
And bilateral symmetry is instead turned down and is obtained, and is not limited to anti-fold symmetrical above and below and the anti-sequence turned down of bilateral symmetry, it can also be by numbering
For n2Mask obtain by the anti-fold of bilateral symmetry, or numbering is n3Mask by it is symmetrical above and below it is anti-fold and obtain.
On the basis of the above embodiments, in the present embodiment according to the number of connected region in each mask of mask sequence,
The step of carrying out unusual correction to mask sequence, obtain optimal mask sequence specifically includes: any group of mask sequence is covered
Mould, if the number of connected region is greater than connected region in the third mask of this group of mask in second mask of this group of mask
Number, then by this group of mask second mask and third mask carry out position exchange.
Fewer in view of the connected region number of mask, it is bigger which treats imageable target object reconstruction quality contribution,
Above-mentioned symmetrical anti-fold generates n-th2、n3And n4During a mask, taken on the contrary in the pixel value to mask some regional area
When number, the connected domain number in this regional area, but the part-area edge and other region joining parts will not be changed
May because of value variation and the variation of connected domain number occurs, and n-th4The generation of a mask need to be anti-by two sub-symmetries
Fold, it is minimum in s group to the contribution of object reconstruction quality.And n-th is symmetrically generated in above-mentioned diminution1The process of a mask
In, the pixel value of symmetry axis two sides is centainly identical, this makes the part-area edge in the case of this connect generated connected domain
Number variation is minimum, i.e., it is maximum in s group to the contribution of object reconstruction quality.Therefore only n-th2And n3A mask can be sent out
The case where raw connected domain number is successively decreased occurs unusual.Carrying out unusual correct to mask sequence is specially for mask sequence
Any group of mask, if in second mask of this group of mask connected region number be greater than this group of mask third mask in connect
The number in logical region, then by this group of mask second mask and the progress position exchange of third mask.Such as in Fig. 3, the
In two groups of masks, the connected region number of first mask is 3, and connected region number is minimum in this set, and contribution is maximum；4th
The connected region number of a mask is 8, and connected region number is maximum in this set, and contribution is minimum；The connected region of second mask
Domain number is 4, and the connected region number of third mask is 6,4 less than 6, normally.
On the basis of the above embodiments, in the present embodiment according to the number of connected region in each mask of mask sequence,
The step of carrying out unusual correction to mask sequence, obtain optimal mask sequence specifically includes: mask sequence being divided into every four and is covered
One group of mould, all mask set sequences are divided into four sections, are divided into four sub-districts as top, then by each span order
Between, as secondary top layer, and so on, until being sub-divided into four masks is a subinterval；Wherein, the firstth area of top
Between and the 4th section of top in unusual mask set position consistency, by the position of mask set unusual in the first interval of secondary top layer
Determination is set, all mask sets in top 3rd interval are unusual mask set；Respectively by the unusual mask set of mask sequence
In second mask and third mask carry out position exchange.
Specifically, due to the percentage contribution sort descending that obtained exposure mask sequence is according to it to image quality, because
In each mask set, it is n that rule substantially, which should be the relationship between the connected region number of each mask, for this1<n2<n3<n4.Table 1
It is respectively shown with table 2 corresponding according to symmetrical 16 ranks and 64 rank exposure mask sequences obtained with symmetrical anti-fold method of above-mentioned diminution
Connected domain number, wherein every one group of 4 data.16 rank mask sequences are the sequence being made of 16 masks.64 rank mask sequences
For the sequence being made of 64 masks.It has been observed that it is the 3rd group that unusual group, which occur, in 16 rank exposure masks, 64 rank exposure mask sequences occur
Unusual group is the 3rd, 9~12 and 15 group.Unusual mask set distinguished with showing in Tables 1 and 2 by overstriking.The group of so-called abnormality
Refer to that connected domain number is successively decreased in the group.Notice that discussed herein is non-general with group number that 4 exposure masks are one group
Exposure mask serial number.If all mask sets are divided into four sections, it is not difficult to find that 16 ranks and the 2nd of 64 rank mask sequences
The group in section is all normal, and the group in the 3rd section is all unusual.And for the 1st section and the 4th section, 16 ranks are covered with 64 ranks
The result of mode sequence is not consistent.Here can be explained with recurrence: the first interval of 64 ranks is exactly whole groups of 16, therefore
Unusual group number is inevitable identical as 16 ranks；And by comparing discovery, the abnormality in the 4th section of 64 rank is organized anti-with the 1st section of 64 rank
Normal position of the group in section is identical, i.e. the 3rd group of the 1st section and the 3rd group of the 3rd section are unusual group.
The connected domain number of 16 rank mask sequences of the table 1 based on fold mask construction
The connected domain number of 64 rank mask sequences of the table 2 based on fold mask construction
In Fig. 5, above-mentioned rule is verified in the reverse position of the 256 rank mask sequences formed using 256 masks, wherein every
One line segment represents a group, corresponding group of the line segment of digital representation of number below line segment.The line segment of overstriking indicates the group
For unusual mask set, the line segment of not overstriking indicates that the group is normal.256 masks are divided into 4 sections, every a line, i.e., 16
Group is a section, and the 1st section of 256 ranks is consistent with 64 unusual group numbers, and the 2nd section is all normal, and the 3rd section is all anti-
Often, the reverse position in the 4th section is corresponding with the reverse position in the 1st section consistent, i.e. reverse position in the 4th section is the 4th area
Between in the 3rd group, the 9 to 12nd group and the 15th group, the reverse position in the 1st section is the 3rd group, the 9 to 12nd group in the 1st section
With the 15th group.
For this purpose, the step of section abnormality is corrected in the present embodiment is divided into 4 sections for that will amount to N number of mask sequence, the 1st
The reverse position in a section and the 4th section is determined that the 3rd section is all unusual, by n-th in each group of unusual section by N/42
And n3The position of a mask is exchanged.Due to every one group of 4 masks, being divided compared to previous classification, grouping is finer,
It can avoid occurring the identical mask of a large amount of connected domain numbers in each group to the full extent, to reduce uncertainty, to obtain
Take optimal mask sequence.
On the basis of the various embodiments described above, according to optimal mask sequence in the present embodiment, the step of carrying out image reconstruction
It specifically includes: selecting the mask of predetermined number in optimal mask sequence, the mask of selection is decomposed into two complementary differentials and is covered
Mould carries out complementary differential modulation to light according to two complementary differential masks, and by the corresponding measured value of two complementary differential masks
Subtract each other, obtain the corresponding complementary differential measured value of each mask of selection, by each complementary differential measurement set at measurement vector；It will choosing
The mask selected is converted to row vector by row-major order or column main sequence, and the corresponding row vector of each mask of selection is spliced, structure
At calculation matrix；According to calculation matrix and measurement vector, image reconstruction is carried out based on deterministic algorithm and compressed sensing algorithm；Its
In, in image reconstruction procedure, the multiplying between calculation matrix and measurement vector is converted to the digraph of Weighted Coefficients, power
Value is positive 1 or minus 1, and each directed connection in digraph is directed toward next layer of element from upper one layer of element.
For example, object is stretched as a dimensional vector, i.e., when the pixel of object is p × q=N × 1Most
It is orthogonal matrix that all masks, which stretch matrix composed by splicing, in excellent mask sequence, and inverse of a matrix is itself.From optimal
The mask that M pixel is p × q before choosing in mask sequence, i-th of mask are denoted as patterni, and each mask need to decompose again
For by 0,1 and two complementary differential masks being formed by 1,0, i.e.,WithAnd meetAccording to the mask split
Complementary differential modulation is carried out to light.
Such asThen have And meetIt will be from optimal mask sequence
Each mask of middle selection is by row-major order or column main sequence stretching conversion row vector and is spliced to form calculation matrix A, and size is M × N,
I.e.
Wherein, meet y=Ax+e=A ψ x '+e, each 1 × N of row vector meets 1 × N=p × q, and e is measurement noise.
For example, Choose preceding 4 masks, if each mask by row-major order stretching embark on journey to
Amount is spliced again, by the calculation matrix A of composition 4 × 16, i.e.,
When carrying out complementary differential modulation, the complementary differential measured value of detector is biIt is measured for complementary differential obtained
As a result, setting front and back, measurement result is respectively twiceWithThen haveWherein i value is 1,2,3 ...,
M.Pendulous frequency is M, obtained measurement vector are as follows:
Then, calculation matrix, measurement vector, sparse transformation matrix and basic parameter are combined, is based on deterministic algorithm
Image reconstruction is carried out with compressed sensing algorithm.Wherein, column vector x obtained by object stretches usually can be in some sparse transformation matrix
Rarefaction representation under ψ, expression coefficient are x ', then meet x=ψ x '.Deterministic algorithm includes that ghost imaging, consistency imaging and output are true
Fixed algorithm etc., compressed sensing algorithm include the algorithm using compressive sensing theory thought.In view of by institute in optimal mask sequence
It is orthogonal matrix that mask, which stretches matrix composed by splicing, and inverse of a matrix is itself, and each complementary differential measures real
Matter is to take out row to the matrix to do dot-product operation with the dimensional vector of N × 1 again, and the multiplying of matrix and column vector is now switched to cum rights
It is worth digraph, weight is taken as ± 1, and each directed connection is directed toward next layer of element, according to this, matrix multiplication from upper one layer of element
It can switch to the simple recursion or recurrence plus and minus calculation between column vector element, and operator can be converted by matrix multiplication and inversion operation
Form substitutes matrix manipulation and avoids huge calculation matrix from storing to reduce computation complexity, saves calculating memory source and disappears
Consumption realizes that the super sub-sampling super-resolution of high pixel resolution calculates imaging.
The present embodiment makes full use of contribution of the mask to image reconstruction, introduces image down and symmetrically instead turns down technology with symmetrical
Mask and calculation matrix are constructed, the collating sequence of connected domain number equal uncertain mask is greatly decreased using unusual correction
The probability of appearance optimizes mask construction, shows that important mask preferentially, and the side of operator is introduced in imaging algorithm
Formula so that the calculation matrix of big data quantity in calculating process without storing repeatedly, largely simplification matrix multiplication and ask
Inverse operation reduces computation complexity and calculates memory consumption, and then realizes the overdelicate height of super sub-sampling of high pixel resolution
Speed calculates imaging, clears away the obstacle in CS practicalization, will be widely used calculating imaging field.
A kind of calculating imaging device based on fold mask construction, the device are provided in another embodiment of the present invention
For realizing the method in foregoing embodiments.Therefore, based on the calculating imaging of fold mask construction in foregoing embodiments
Description and definition in method, can be used for the understanding of each execution module in the embodiment of the present invention.Fig. 6 is the embodiment of the present invention
The calculating imaging device overall structure diagram based on fold mask construction provided, the device include generation module 601, optimization
Module 602 and reconstructed module 603；Wherein:
Generation module 601 is used to generate mask sequence by reducing symmetrically with symmetrical anti-fold；Optimization module 602 is used for
According to the number of connected region in each mask of mask sequence, unusual correction is carried out to mask sequence, obtains optimal mask sequence；
Reconstructed module 603 is used to carry out image reconstruction according to optimal mask sequence.
On the basis of the above embodiments, generation module is specifically used for carrying out initial mask respectively up and down in the present embodiment
Symmetrical anti-fold, the anti-fold of bilateral symmetry and anti-fold symmetrical above and below and the anti-fold of bilateral symmetry, by initial mask and initial
The symmetrical anti-fold result of three kinds of mask is as first group of mask；Wherein, symmetrical above and below counter turn down will be carried out to initial mask
As a result as second mask in first group of mask, the anti-result turned down of bilateral symmetry is carried out as first using to initial mask
Third mask in group mask；Alternatively, carrying out the anti-result turned down of bilateral symmetry as first group of mask for initial mask
In second mask, covered the anti-result turned down symmetrical above and below is carried out as the third in first group of mask to initial mask
Mould；Mask sequence is generated by reducing symmetrically with symmetrical anti-fold according to first group of mask.
On the basis of the above embodiments, generation module is further specifically used in the present embodiment: for first group of mask
In any non-monodrome mask, the length and width of the mask are reduced into original half respectively, based on the mask after diminution
Generate symmetrical above and below and symmetrical new mask；Carry out anti-fold, symmetrically anti-fold symmetrical above and below respectively to new mask,
And anti-fold symmetrical above and below and bilateral symmetry instead turn down, using three kinds of symmetrical anti-fold results of new mask and new mask as one
Group new mask；The step of reducing symmetrically with symmetrical anti-fold is executed to any mask iteration in new mask set, is covered all
Mould group constitutes mask sequence by the sequencing for generating the time.
On the basis of the above embodiments, optimization module is specifically used in the present embodiment: for any group of mask sequence
Mask, if the number of connected region is greater than connected region in the third mask of this group of mask in second mask of this group of mask
Number, then by this group of mask second mask and third mask carry out position exchange.
On the basis of the above embodiments, optimization module is specifically used in the present embodiment: mask sequence is divided into every four
One group of mask, all mask set sequences are divided into four sections, are divided into four sons as top, then by each span order
Section, as secondary top layer, and so on, until being sub-divided into four masks is a subinterval；Wherein, the first of top
The position consistency of unusual mask set in section and the 4th section of top, by mask set unusual in the first interval of secondary top layer
Position determines that all mask sets in top 3rd interval are unusual mask set；Respectively by the unusual mask of mask sequence
Second mask and third mask in group carry out position exchange.
On the basis of the various embodiments described above, reconstructed module is specifically used in the present embodiment: selecting in optimal mask sequence
The mask of selection is decomposed into two complementary differential masks, according to two complementary differential masks to light by the mask for selecting predetermined number
Complementary differential modulation is carried out, and the corresponding measured value of two complementary differential masks is subtracted each other, each mask for obtaining selection is corresponding
Complementary differential measured value, by each complementary differential measurement set at measurement vector；The mask of selection is turned by row-major order or column main sequence
It is changed to row vector, and the corresponding row vector of each mask of selection is spliced, constitutes calculation matrix；According to calculation matrix and survey
Vector is measured, image reconstruction is carried out based on deterministic algorithm and compressed sensing algorithm；Wherein, it in image reconstruction procedure, will measure
Multiplying between matrix and measurement vector is converted to the digraphs of Weighted Coefficients, and weight is positive 1 or minus 1, every in digraph
A directed connection is all directed toward next layer of element from upper one layer of element.
The present embodiment, which utilizes, makes full use of contribution of the mask to image reconstruction, introduces diminution and symmetrically instead turns down technology with symmetrical
To construct mask, is corrected using abnormality and the probability that the identical mask of connected region number occurs is greatly decreased, reduce mask not
Certainty optimizes mask construction, carries out image reconstruction using the mask sequence of optimization, improves the quality and effect of image reconstruction
Rate.
As shown in fig. 7, the present embodiment provides a kind of calculating imaging system based on fold mask construction, including light source 1, expansion
Beam collimating element 3, imaging len and any of the above-described Installation practice, mask generating means 6 include generation module and optimization module,
Reconstructed module includes spatial light modulator 7, collects convergent lens 8, detector 9 and image reconstruction device 10.
Wherein, light source 1 is configured to transmitting light beam.After beam-expanding collimation element 3 is arranged in light source 1, it is configured to expand light
Beam and collimation, so that beam spot size is adapted to 4 size of target object to be imaged, 7 modulation areas size of spatial light modulator, and becomes
At directional light.Imaging len 5 is arranged between target object 4 and spatial light modulator 7 to be imaged, is configured to target to be imaged
Object 4 is imaged in spatial light modulator 7 or the light beam modulated through spatial light modulator 7 is imaged on object 4, such as Fig. 8 institute
Show.Imaging len 5, target object to be imaged 4, collection convergent lens 8 and detector 9 are arranged in same detection optical axis in Fig. 8.
Fig. 8 is the difference from Fig. 7 is that herein by the way of Structured Illumination, i.e., imaging len 5 is arranged in spatial light modulator 7
Between verifying object 4, and after spatial light modulator 7, it is configured to for the light beam modulated through spatial light modulator 7 being imaged on
On target object 4 to be imaged.
Mask generating means 6 are configured to being sized according to mask, treat imageable target object 4 based on mask and rebuild matter
The principle of the percentage contribution of amount symmetrically constructs corresponding mask sequence with symmetrical anti-fold using the image down of mask, and
The correction of section abnormality is carried out by mask connected domain number incremental order, the mask after correction forms grouping finely and degree of certainty is high
Optimal mask sequence, be sequentially inputted in spatial light modulator 7.Spatial light modulator 7.Configuration generates dress by load mask
The optimal mask sequence exported is set, so that contributing bigger mask preferentially to show image, complementary differential modulation is carried out to light,
Wherein each mask by row-major order be stretched as row vector be stitched together composed by matrix be calculation matrix.Collect convergent lens
8.It is arranged in 9 front end of detector, is configured to converge light on detector 9.Detector 9.It is burnt flat to be arranged in collection convergent lens 8
On face, be configured to using complementary differential measuring technique record correspond to optimal mask sequence in each select mask it is mutual
Mend differential measurement values, composition measurement vector.Image reconstruction device 10.It is raw to be configured to the calculation matrix for exporting detector 9, mask
Calculation matrix composed by the optimal mask sequence exported at device 6, sparse transformation matrix, basic parameter combine, by true
Deterministic algorithm, compressed sensing algorithm reconstruct image, and wherein deterministic algorithm includes ghost imaging, consistency is imaged, output determines
Algorithm etc., compressed sensing algorithm includes the algorithm using compressive sensing theory thought.
Light source 1, beam-expanding collimation element 3 and spatial light modulator 7 are on same primary optical axis；The wherein light after beam-expanding collimation
Beam with vertical incidence or oblique can be mapped to 7 working face of spatial light modulator；It collects convergent lens 8 and detector 9 is arranged in together
In one detection optical axis.Light source 1 is counterfeit hot radiant, hot radiant, single-photon light source, pump light source or entangled light source spatial light tune
Device 7 processed is selected from Digital Micromirror Device, net amplitude LCD space light modulator, pure phase position LCD space light modulator, reflective sky
Between any one of optical modulator, transmissive spatial optical modulator, liquid crystal light valve, mask switch boards.7 He of spatial light modulator
Detector 9 needs to synchronize setting by the way of logical connection or electrical connection, be specifically arranged position and optical path without
It closes.By the way of logical connection or electrical connection, mask generating means 6 are exported for spatial light modulator 7 and mask generating means 6
Optimal mask sequence, as the input of spatial light modulator 7, the specific placement position of mask generating means 6 and optical path are unrelated.It covers
Mould generating means 6, detector 9 are with image reconstruction device 10 using logical connection or by the way of being electrically connected, mask generating means 6
With the output of detector 9, as the input of image reconstruction device 10, the specific placement position of image reconstruction device 10 and optical path without
It closes.Detector 9 includes detector, the photomultiplier transit of point detector, bucket detector, photodiode and analog-digital converter composition
Detector, the superconducting single-photon point detector etc. of pipe, single photon point detector and counter composition, wavelength cover microwave, infrared
Line, visible light, ultraviolet light, X-ray range.When detector 9 selects photomultiplier tube, single photon point detector and counter composition
Detector, superconducting single-photon point detector when, further includes: the optical attenuator component 2 being arranged between light source and detector, configuration
It is horizontal for light decay is reduced to dim light even single photon, and prevent detector 9 supersaturated.
Source emissioning light beam in the present embodiment becomes directional light with collimation through expanding for beam-expanding collimation element, using compression
The mode of imaging or Structured Illumination, using mask generating means, according to being sized for mask, according to mask to object reconstruction
The percentage contribution of quality, using the symmetrical mask sequence corresponding with symmetrical anti-fold construction of image down of mask, and by mask
Connected domain number incremental order carries out the correction of section abnormality, and the mask after correction is sequentially inputted in spatial light modulator, so that
Bigger mask is contributed preferentially to show image, detector records the complementary differential measured value corresponding to the modulation of different masks,
Composition measurement vector reconstructs image by deterministic algorithm, compressed sensing algorithm in conjunction with calculation matrix, calculates in algorithm
In, operator form is converted by matrix multiplication and inversion operation, substitution matrix manipulation avoids huge to reduce computation complexity
Calculation matrix storage, save calculate memory source consumption.The device and method have super sub-sampling, hypersensitive, sampling dimension
The advantages that low, high pixel resolution, real-time high speed.
The present embodiment provides a kind of electronic equipment, Fig. 9 is electronic equipment overall structure provided in an embodiment of the present invention signal
Figure, which includes: at least one processor 901, at least one processor 902 and bus 903；Wherein,
Processor 901 and memory 902 pass through bus 903 and complete mutual communication；
Memory 902 is stored with the program instruction that can be executed by processor 901, and the instruction of processor caller is able to carry out
Method provided by above-mentioned each method embodiment, for example, by reducing symmetrically with symmetrical anti-fold, generate mask sequence；
According to the number of connected region in each mask of mask sequence, unusual correction is carried out to mask sequence, obtains optimal mask sequence；
According to optimal mask sequence, image reconstruction is carried out.
The present embodiment provides a kind of non-transient computer readable storage medium, non-transient computer readable storage medium storages
Computer instruction, computer instruction make computer execute method provided by above-mentioned each method embodiment, for example, pass through contracting
Small symmetrical and symmetrical anti-fold, generates mask sequence；According to the number of connected region in each mask of mask sequence, to mask sequence
Column carry out unusual correction, obtain optimal mask sequence；According to optimal mask sequence, image reconstruction is carried out.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer readable storage medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed；And storage medium above-mentioned includes: ROM, RAM, magnetic disk or light
The various media that can store program code such as disk.
Electronic equipment embodiment described above is only schematical, wherein unit can as illustrated by the separation member
It is physically separated with being or may not be, component shown as a unit may or may not be physics list
Member, it can it is in one place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of the modules achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case where, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
The method of certain parts of example or embodiment.
Finally, the present processes are only preferable embodiment, it is not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent replacement, improvement and so on should be included in protection of the invention
Within the scope of.
Claims (10)
1. a kind of calculating imaging method based on fold mask construction characterized by comprising
By reducing symmetrically with symmetrical anti-fold, mask sequence is generated；
According to the number of connected region in each mask of the mask sequence, unusual correction is carried out to the mask sequence, is obtained
Optimal mask sequence；
According to the optimal mask sequence, image reconstruction is carried out.
2. the method according to claim 1, wherein generating mask sequence by reducing symmetrically with symmetrical anti-fold
The step of column, specifically includes:
Carry out anti-fold, the anti-fold of bilateral symmetry and anti-fold symmetrical above and below symmetrical above and below and left and right respectively to initial mask
Symmetrical anti-fold, using three kinds of symmetrical anti-fold results of the initial mask and the initial mask as first group of mask；
Wherein, the anti-result turned down symmetrical above and below is carried out as second in first group of mask using to the initial mask
Mask is covered the anti-result turned down of bilateral symmetry is carried out as the third in first group of mask to the initial mask
Mould；Alternatively,
The anti-result turned down of bilateral symmetry is carried out as second mask in first group of mask for the initial mask,
The anti-result turned down symmetrical above and below is carried out as the third mask in first group of mask using to the initial mask；
Mask sequence is generated by reducing symmetrically with symmetrical anti-fold according to first group of mask；
Wherein, the step of carrying out anti-fold symmetrical above and below to the initial mask specifically includes:
It keeps the top half of the initial mask constant, the lower half portion of the initial mask is taken into inverse value；Wherein, described
The lower half portion of the top half of initial mask and the initial mask is at axial symmetry；
The step of bilateral symmetry instead turns down is carried out to the initial mask to specifically include:
It keeps the left-half of the initial mask constant, the right half part of the initial mask is taken into inverse value；Wherein, described
The right half part of the left-half of initial mask and the initial mask is at axial symmetry；
The step of carrying out anti-fold symmetrical above and below and symmetrically anti-fold to the initial mask specifically includes:
The initial mask anti-result turned down symmetrical above and below is subjected to the anti-fold of bilateral symmetry；Alternatively,
The anti-result turned down of initial mask bilateral symmetry is subjected to anti-fold symmetrical above and below.
3. symmetrical and right by reducing according to the method described in claim 2, it is characterized in that, according to first group of mask
The step of claiming anti-fold, generating mask sequence, specifically includes:
For any non-monodrome mask in first group of mask, the length and width of the mask are reduced into original two respectively/
One, symmetrical above and below and symmetrical new mask is generated based on the mask after diminution；
Carry out anti-fold, the anti-fold of bilateral symmetry and anti-fold symmetrical above and below symmetrical above and below and a left side respectively to the new mask
Right symmetrical anti-fold, using three kinds of symmetrical anti-fold results of the new mask and the new mask as one group of new mask；
The step of reducing symmetrically with symmetrical anti-fold is executed to any mask iteration in new mask set, and all mask sets are pressed
The sequencing for generating the time constitutes mask sequence.
4. according to the method described in claim 3, it is characterized in that, according to connected region in each mask of the mask sequence
Number, the step of carrying out unusual correction to the mask sequence, obtain optimal mask sequence, specifically include:
For any group of mask of the mask sequence, if the number of connected region is greater than this in second mask of this group of mask
Group mask third mask in connected region number, then by this group of mask second mask and third mask progress
Position is exchanged.
5. according to the method described in claim 3, it is characterized in that, according to connected region in each mask of the mask sequence
Number, the step of carrying out unusual correction to the mask sequence, obtain optimal mask sequence, specifically include:
The mask sequence is divided into every one group of four masks, all mask sets sequence is divided into four sections, as most pushing up
Layer, then is divided into four subintervals for each span order, as secondary top layer, and so on, until being sub-divided into four masks
Until a subinterval；Wherein, unusual mask set in the 4th section of the first interval of the top and the top
Position consistency, determined by the position of mask set unusual in the first interval of the secondary top layer, in top 3rd interval
All mask sets are unusual mask set；
Respectively by the unusual mask set of the mask sequence second mask and third mask carry out position exchange.
6. -5 any method according to claim 1, which is characterized in that according to the optimal mask sequence, carry out image
The step of reconstruct, specifically includes:
The mask that predetermined number is selected in the optimal mask sequence, is decomposed into two complementary differentials for the mask of selection
Mask carries out complementary differential modulation to light according to two complementary differential masks, and by two complementary differential masks pair
The measured value answered subtracts each other, and obtains the corresponding complementary differential measured value of each mask of selection, by each complementary differential measurement set
At measurement vector；
The mask of selection is converted into row vector by row-major order or column main sequence, and by the corresponding row of each mask of selection
Vector is spliced, and calculation matrix is constituted；
According to the calculation matrix and the measurement vector, image reconstruction is carried out based on deterministic algorithm and compressed sensing algorithm；
Wherein, in image reconstruction procedure, the multiplying between the calculation matrix and the measurement vector is converted into band
The digraph of weight, weight are positive 1 or minus 1, and each directed connection in the digraph is directed toward next from upper one layer of element
Layer element.
7. a kind of calculating imaging device based on fold mask construction characterized by comprising
Generation module, for generating mask sequence by reducing symmetrically with symmetrical anti-fold；
Optimization module carries out the mask sequence for the number of connected region in each mask according to the mask sequence
Abnormality is corrected, and optimal mask sequence is obtained；
Reconstructed module, for carrying out image reconstruction according to the optimal mask sequence.
8. device according to claim 7, which is characterized in that further include initial module, for initial mask respectively into
Row anti-fold symmetrical above and below, the anti-fold of bilateral symmetry and anti-fold symmetrical above and below and the anti-fold of bilateral symmetry, will be described initial
Three kinds of symmetrical anti-fold results of mask and the initial mask are as first group of mask；
Wherein, the anti-result turned down symmetrical above and below is carried out as second in first group of mask using to the initial mask
Mask is covered the anti-result turned down of bilateral symmetry is carried out as the third in first group of mask to the initial mask
Mould；Alternatively,
The anti-result turned down of bilateral symmetry is carried out as second mask in first group of mask for the initial mask,
The anti-result turned down symmetrical above and below is carried out as the third mask in first group of mask using to the initial mask；
Mask sequence is generated by reducing symmetrically with symmetrical anti-fold according to first group of mask.
9. a kind of electronic equipment characterized by comprising
At least one processor, at least one processor and bus；Wherein,
The processor and memory complete mutual communication by the bus；
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
Enough methods executed as described in claim 1 to 6 is any.
10. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Computer instruction is stored up, the computer instruction makes the computer execute the method as described in claim 1 to 6 is any.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810791144.1A CN109035168B (en) | 2018-07-18 | 2018-07-18 | Computational imaging method and device based on turnover mask structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810791144.1A CN109035168B (en) | 2018-07-18 | 2018-07-18 | Computational imaging method and device based on turnover mask structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109035168A true CN109035168A (en) | 2018-12-18 |
| CN109035168B CN109035168B (en) | 2020-08-18 |
Family
ID=64643322
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810791144.1A Active CN109035168B (en) | 2018-07-18 | 2018-07-18 | Computational imaging method and device based on turnover mask structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109035168B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109946753A (en) * | 2019-03-15 | 2019-06-28 | 西安交通大学 | The coded system and method for ghost imaging are calculated based on low order Hadamard basic vector |
| CN110223238A (en) * | 2019-04-30 | 2019-09-10 | 北京理工大学 | Encoded light shines image reconstruction method and device |
| CN110392193A (en) * | 2019-06-14 | 2019-10-29 | 浙江大学 | A kind of mask plate of mask plate camera |
| CN110487427A (en) * | 2019-07-08 | 2019-11-22 | 清远市天之衡传感科技有限公司 | Single Photon Polarization quantum imaging system based on DMD micro mirror array |
| CN110766625A (en) * | 2019-10-15 | 2020-02-07 | 中国科学院光电技术研究所 | Method for rapidly filling digital image mask |
| CN111968032A (en) * | 2020-07-23 | 2020-11-20 | 太原理工大学 | Self-adaptive sampling single-pixel imaging method |
| CN112950472A (en) * | 2021-03-02 | 2021-06-11 | 北京理工大学 | Single-pixel imaging method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110068268A1 (en) * | 2009-09-18 | 2011-03-24 | T-Ray Science Inc. | Terahertz imaging methods and apparatus using compressed sensing |
| CN103020579A (en) * | 2011-09-22 | 2013-04-03 | 上海银晨智能识别科技有限公司 | Face recognition method and system, and removing method and device for glasses frame in face image |
| CN105737992A (en) * | 2016-02-04 | 2016-07-06 | 北京理工大学 | Compressed sensing-based dual wavelength temperature field imaging device, system and method |
| CN205483257U (en) * | 2016-01-18 | 2016-08-17 | 天津大学 | Terahertz radiation intensity distribution now detecting device based on compressed sensing |
| CN106706638A (en) * | 2016-12-14 | 2017-05-24 | 北京理工大学 | Phase imaging device and method based on dark stripe logic judgment |
-
2018
- 2018-07-18 CN CN201810791144.1A patent/CN109035168B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110068268A1 (en) * | 2009-09-18 | 2011-03-24 | T-Ray Science Inc. | Terahertz imaging methods and apparatus using compressed sensing |
| CN103020579A (en) * | 2011-09-22 | 2013-04-03 | 上海银晨智能识别科技有限公司 | Face recognition method and system, and removing method and device for glasses frame in face image |
| CN205483257U (en) * | 2016-01-18 | 2016-08-17 | 天津大学 | Terahertz radiation intensity distribution now detecting device based on compressed sensing |
| CN105737992A (en) * | 2016-02-04 | 2016-07-06 | 北京理工大学 | Compressed sensing-based dual wavelength temperature field imaging device, system and method |
| CN106706638A (en) * | 2016-12-14 | 2017-05-24 | 北京理工大学 | Phase imaging device and method based on dark stripe logic judgment |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109946753A (en) * | 2019-03-15 | 2019-06-28 | 西安交通大学 | The coded system and method for ghost imaging are calculated based on low order Hadamard basic vector |
| CN109946753B (en) * | 2019-03-15 | 2020-04-28 | 西安交通大学 | Encoding method for calculating ghost imaging based on low-order Hadamard basis vector |
| CN110223238A (en) * | 2019-04-30 | 2019-09-10 | 北京理工大学 | Encoded light shines image reconstruction method and device |
| CN110392193A (en) * | 2019-06-14 | 2019-10-29 | 浙江大学 | A kind of mask plate of mask plate camera |
| CN110392193B (en) * | 2019-06-14 | 2020-09-15 | 浙江大学 | Mask plate of mask plate camera |
| CN110487427A (en) * | 2019-07-08 | 2019-11-22 | 清远市天之衡传感科技有限公司 | Single Photon Polarization quantum imaging system based on DMD micro mirror array |
| CN110766625A (en) * | 2019-10-15 | 2020-02-07 | 中国科学院光电技术研究所 | Method for rapidly filling digital image mask |
| CN110766625B (en) * | 2019-10-15 | 2022-04-19 | 中国科学院光电技术研究所 | Method for rapidly filling digital image mask |
| CN111968032A (en) * | 2020-07-23 | 2020-11-20 | 太原理工大学 | Self-adaptive sampling single-pixel imaging method |
| CN111968032B (en) * | 2020-07-23 | 2022-09-09 | 太原理工大学 | Self-adaptive sampling single-pixel imaging method |
| CN112950472A (en) * | 2021-03-02 | 2021-06-11 | 北京理工大学 | Single-pixel imaging method and device |
| CN112950472B (en) * | 2021-03-02 | 2022-08-26 | 北京理工大学 | Single-pixel imaging method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109035168B (en) | 2020-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109035168A (en) | A kind of calculating imaging method and device based on fold mask construction | |
| CN105492878B (en) | Apparatus and method for snapshot light spectrum image-forming | |
| Brady | Optical imaging and spectroscopy | |
| JP7246093B2 (en) | Wavefront sensor and method of use | |
| Horstmeyer et al. | Solving ptychography with a convex relaxation | |
| CN102818631B (en) | Compression spectral imaging system based on random wavefront phase modulation | |
| Wu et al. | Hybrid neural network-based adaptive computational ghost imaging | |
| CN113762460B (en) | Multimode optical fiber transmission image migration reconstruction algorithm based on numerical value speckle | |
| CN104796609A (en) | Large-visual-field high-resolution microscopic imaging method based on optimal Hadamard codes | |
| CN103913229B (en) | Coding template multi-target active imaging spectrum system and method | |
| CN107817641A (en) | Single pixel camera system based on multimode fibre | |
| Gureyev et al. | Complementary aspects of spatial resolution and signal-to-noise ratio in computational imaging | |
| Han et al. | Deep learning simulations of the microwave sky | |
| Yu et al. | Ghost imaging based on Pearson correlation coefficients | |
| Schretter et al. | Regularized non-convex image reconstruction in digital holographic microscopy | |
| Xu et al. | Single-shot grating-based X-ray phase contrast imaging via generative adversarial network | |
| Zhou et al. | Multi-resolution single-pixel imaging via Hadamard ‘pipeline’coding | |
| US20220187138A1 (en) | Totagraphy: Coherent Diffractive/Digital Information Reconstruction by Iterative Phase Recovery Using Special Masks | |
| Li et al. | Illumination schemes for coded coherent diffraction imaging: A comprehensive comparison | |
| Zhang et al. | Bi-channel compressive hyperspectral imager based on polarization conversion metasurface | |
| Armitage-Caplan et al. | PReBeaM for Planck: A Polarized Regularized Beam Deconvolution Map-Making Method | |
| Hu et al. | Multispectral video acquisition using spectral sweep camera | |
| Hengeveld et al. | An optical processor for matrix-by-vector multiplication: an application to the distance geometry problem in 1D | |
| Li et al. | An automatic and optimal MPA design method | |
| CN112229514B (en) | Three-dimensional data reconstruction method of liquid crystal hyperspectral calculation imaging system |
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 |