CN115755383A - Non-invasive large-field-of-view imaging method and device through scattering medium - Google Patents
Non-invasive large-field-of-view imaging method and device through scattering medium Download PDFInfo
- Publication number
- CN115755383A CN115755383A CN202211446193.4A CN202211446193A CN115755383A CN 115755383 A CN115755383 A CN 115755383A CN 202211446193 A CN202211446193 A CN 202211446193A CN 115755383 A CN115755383 A CN 115755383A
- Authority
- CN
- China
- Prior art keywords
- target
- sub
- imaging
- image
- psf
- 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.)
- Pending
Links
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention belongs to the technical field of optical scattering imaging, and when imaging through a scattering medium is realized based on an optical memory effect, imaging failure is usually caused when an imaging target exceeds a memory effect range, so that the imaging field of view is limited. The invention provides a method and a device for realizing non-invasive large-field-of-view imaging through a scattering medium based on PSF (particle swarm optimization) estimation and related operation, which are used for rebuilding a sub-target O of a target O hidden in the scattering medium by using a phase recovery algorithm when the target is too large to exceed a memory effect range i The image of the sub-targets is obtained based on point spread function estimation and self-correlation and cross-correlation operation, and the sub-targets are subjected to image splicing according to the directions, so that non-invasive large-view-field imaging is realized; the invention can also carry out non-invasive large-field super-resolution imaging beyond the range of memory effect, and can simultaneously give consideration to the imaging field and the imaging resolution. The method is widely applied to the fields of biomedical imaging, optical microscopic imaging and the like.
Description
The technical field is as follows:
the invention belongs to the technical field of optical scattering imaging, and particularly relates to a method and a device for realizing non-invasive large-field imaging through a scattering medium based on PSF (particle swarm optimization) estimation and related operation.
Background art:
in conventional optical imaging, what is seen is what is obtained, but when light is transmitted through a scattering medium, the obtained image information will degrade into a noise-like speckle image, and thus the scattering medium is generally regarded as an obstacle to conventional imaging.
However, the seemingly noisy speckle image contains information that hides the object. Therefore, many efforts have been made by researchers based on the principle of scatter imaging. Currently, scatter imaging techniques involve wavefront shaping, transmission matrix measurement, optical phase conjugation, and speckle-related imaging based on memory effects. Among them, speckle-related imaging based on memory effects is of great interest because it does not require invasive imaging systems.
However, speckle correlation imaging techniques based on memory effects are generally only able to reconstruct a single object with a size smaller than the range of memory effects because the field of view is limited by the memory effects of the scattering medium; once the size of the object exceeds the range of memory effect, this method cannot recover the image information of the hidden object, resulting in reconstruction failure. Therefore, the speckle correlation imaging technology based on the memory effect needs to be developed more widely and applied more widely, and a technical scheme which breaks through the memory effect limitation and has a larger imaging field of view needs to be developed urgently.
The invention content is as follows:
aiming at the defects and problems in the prior art, the invention provides a method and a device for realizing non-invasive large-field imaging through a scattering medium based on PSF (point spread function) estimation and related operation, and the non-invasive large-field imaging is realized through the scattering medium based on Point Spread Function (PSF) estimation and related operation. The invention solves the problem of limited non-invasive imaging field of view in speckle related imaging and realizes super-resolution imaging on the basis. The invention does not need invasive guide star or prior information, the imaging field of view is not limited by memory effect, and the invention has the advantages of large imaging field of view and high imaging resolution and can simultaneously give consideration to both the imaging field of view and the imaging resolution. The method can be widely applied to the fields of biomedical imaging, optical microscopic imaging and the like.
In order to achieve the purpose, the invention is realized by the following scheme:
non-invasion of transmission scattering medium based on PSF estimation and correlation operationAn entrance type large-field-of-view imaging method, which utilizes a phase recovery algorithm to reconstruct sub-targets O of a target O hidden in a scattering medium i The image of (2) obtains the directions of all sub-targets based on point spread function estimation and autocorrelation and cross-correlation operation, and carries out image splicing on the sub-targets according to the directions to enable the target O to realize non-invasive large-field imaging, and the specific steps are as follows:
step 1: scanning a target O hidden in a scattering medium by using active illumination, wherein the target O is scanned and illuminated for n times, and the method comprises the following steps:
wherein, O i Is a sub-target of i-th scanned illumination of an object O to be imaged, and each sub-target O i The target O is beyond the memory effect range of the scattering imaging system after n times of scanning illumination within the respective memory effect range; recording all sub-targets O by area array photoelectric detector i The reflected light of (2) passes through the speckle image I of the scattering medium i ；
Step 2: speckle image I corresponding to each sub-target through phase recovery algorithm i Reconstructing sub-target image O in autocorrelation i '；
And step 3: judging two adjacent sub-targets O i And O i+1 If the distance between two sub-targets O is smaller than the diameter of the memory effect range i And O i+1 With memory effect ranges overlapping each other, two sub-target speckle images I i And I i+1 There is an information correlation between the two sub-target images, and the reconstructed two sub-target images O 'are determined by estimating the point spread function PSF of the speckle images of the two sub-targets and calculating the autocorrelation and cross-correlation of the point spread functions of the two sub-targets' i And O' i+1 Relative position and orientation information therebetween;
and 4, step 4: and (3) shifting and superposing the reconstructed image with correct relative position and direction information obtained in the step (3) to obtain a correctly spliced target O, so that the limitation of the memory effect of the scattering medium is avoided, and the non-invasive large-field imaging through the scattering medium is realized.
Further, in step 3, the specific steps of determining the relative position and the direction information of the two reconstructed sub-target images are as follows:
step 3.1: estimating a point spread function: from speckle image I i And the reconstructed sub-target image O 'obtained by calculation in the step 2' i Estimating speckle image I by wiener deconvolution algorithm i Point spread function PSF' i ：
PSF' i ＝Deconv(I i ,O' i )
Wherein, deconv represents a deconvolution operation;
step 3.2: determining reconstructed sub-target image O' i And O' i+1 Relative position between: separately calculate PSF' i And PSF' i And PSF' i+1 Determining two-dimensional coordinates of respective correlation peaks, and subtracting to obtain a characterization sub-target image O' i And O' i+1 Shift vector of relative position therebetween
In the above formula { } represents a correlation operation, and position { } represents a two-dimensional coordinate of a correlation peak position;
step 3.3: determining reconstructed sub-target image O' i The direction information of (1): the reconstructed sub-target image O' i Is turned left and right, turned up and down, and simultaneously turned left and right and up and down, and sub-target image O' i Obtaining the sub target image O 'per se' i Four different directional states; and then from the speckle image I i And sub target image O' i Is estimated to obtain PSF corresponding to four directions in four different direction states' i Then from four estimated PSFs' i With adjacent speckle image I i+1 Deconvolution reconstruction is carried out to obtain four deconvolution images, and the four deconvolution images are deconvolutedProduct image and sub-target image O 'obtained by phase recovery' i+1 Is compared to determine O' i The correct direction of the light;
step 3.4: two reconstructed sub-target images O 'with correct directions' i And O' i+1 According to the shift vector in step 3.2
Shifting and superposing;
step 3.5: and (5) repeating the step 3.1 to the step 3.4 for all the sub-targets, and determining the relative position relationship and the direction information of all the sub-targets.
A method for improving imaging resolution and realizing non-invasive large-field imaging based on PSF estimation and correlation operation is characterized in that a high-resolution image breaking through the diffraction limit of an imaging system is obtained by utilizing a Gaussian fitting repositioning method based on the method for realizing non-invasive large-field imaging through a scattering medium, the relative position and direction information of sub-target high-resolution images are determined and spliced, so that super-resolution imaging is realized while large-field imaging of a target is realized, and the method comprises the following specific steps:
step 1: repeating the steps 1 to 2 in the method for realizing the non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and obtaining the ith sub-target O of the jth sparse illumination i_j Low resolution picture of O' i_j Then, a Gaussian fitting repositioning method is utilized to process the low-resolution image to obtain a high-resolution image which breaks through the diffraction limit of the imaging system
Step 2: calculating a point spread function PSF 'of the imaging system, and combining the PSF' with the m acquired speckle images I i_j After deconvolution operation one by one, gaussian fitting repositioning processing is carried out to obtain m high-resolution images
And step 3: m high resolution images
Corresponding positions are overlapped to obtain a sub-target high-resolution image breaking through the diffraction limit
And 4, step 4: high-resolution images of each sub-target obtained in the step 3
And (4) operating according to step 3 and step 4 of the method for realizing non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and realizing non-invasive large-field super-resolution imaging through the scattering medium.
Further, in step 2, a point spread function PSF ' calculation method and a point spread function PSF ' of a method for realizing non-invasive large-field-of-view imaging through scattering media based on PSF estimation and correlation calculation ' i The calculation method is the same.
A device for realizing non-invasive large-field imaging through a scattering medium based on PSF (particle swarm optimization) estimation and related operation is used for realizing the method and comprises a target, an aperture diaphragm, the scattering medium and an area array photodetector, wherein the aperture diaphragm is arranged between the target and the area array photodetector, the scattering medium is tightly attached to one surface of the aperture diaphragm, the planes of the area array photodetector and the scattering medium are parallel to the plane of the target, and the central position of the plane is superposed with the normal of the target; the distance between the scattering medium and the target is larger than the distance between the scattering medium and the area array light detector.
Further, the scattering medium is ground glass, a diffuse reflection wall surface, biological tissues, cloud and mist, turbid water or artificial scattering medium.
The invention realizes the non-invasive large-field super-resolution imaging on the device by utilizing PSF estimation and related operation. According to the invention, firstly, a phase recovery algorithm is utilized to reconstruct images of sub-targets hidden in a scattering medium, then, corresponding PSFs are estimated, autocorrelation and cross-correlation operations are carried out to obtain the directions of the images of the sub-targets, and the images of the sub-targets are spliced according to the obtained directions, so that non-invasive large-field imaging is realized on the whole target hidden in the scattering medium and beyond the range of memory effect, and the imaging effect is obviously improved. The invention is suitable for the occasions where the imaging requirements of turbid scattering media such as biological tissues, cloud and mist, smog, turbid liquid and the like are met, invasive guide star or prior information is not needed, the imaging field of view is not limited by memory effect, and the invention has the advantages of large imaging field of view and high imaging resolution and can simultaneously give consideration to both the imaging field of view and the resolution. The method can be widely applied to the fields of biomedical imaging, optical microscopic imaging and the like.
Description of the drawings:
FIG. 1 is a schematic diagram of the structure of the apparatus of the present invention;
FIG. 2 is a speckle image, sub-target image, and point spread function image of the present invention;
FIG. 3 shows (a 1) - (a 8) reconstructing 9 sub target images O' 1 -O' 9 The relative position of (a);
FIG. 4 shows (a 1) - (a 8) reconstructing 9 sub target images O' 1 -O' 9 Overlapping and splicing results;
FIG. 5 is a result of a method for achieving non-invasive large field-of-view imaging through a scattering medium based on PSF estimation and correlation;
fig. 6 (a) - (i) are sub-target non-invasive super-resolution imaging, and (g) are super-resolution imaging of the method for realizing non-invasive large-field imaging based on PSF estimation and correlation operation for improving imaging resolution;
in fig. 7 (a) is non-invasive large field-of-view imaging; (b) Non-invasive large field of view super-resolution imaging for improved resolution.
In the figure: 1. the system comprises a target 2, an aperture diaphragm 3, a scattering medium 4 and an area array photoelectric detector;
in fig. 1, the solid white circles in the object 1 are memory effect ranges, and the numbers of the corresponding positions represent sub-objects O which are sequentially scanned and illuminated 1 -O 9 In the order of (1), the arrows represent the predicted point spread function PSF 1 -PSF 9 The correlation operation sequence between the two groups;
9 sheets were collected in (a 1) - (a 9) of FIG. 2Speckle image I 1 -I 9 (ii) a (b1) - (b 9) reconstruction of 9 sub-target images O' 1 -O' 9 (ii) a (c1) - (c 9) estimating the corresponding point spread function PSF' 1 -PSF' 9 。
The specific implementation mode is as follows:
the present invention will be described in further detail with reference to the accompanying drawings.
It should be noted that, for convenience of description, the following description of the direction is consistent with the direction of the drawings, but does not limit the structure of the present invention.
As shown in FIGS. 1-6, the invention discloses a method for realizing non-invasive large-field imaging through a scattering medium based on PSF estimation and related operations, which reconstructs a sub-target O of a target O hidden in the scattering medium by using a phase recovery algorithm i The image of (2) obtains the directions of all sub-targets based on point spread function estimation and autocorrelation and cross-correlation operation, and carries out image splicing on the sub-targets according to the directions to enable the target O to realize non-invasive large-field imaging, and the specific steps are as follows:
step 1: scanning a target O hidden in a scattering medium by using active illumination, wherein the target O is scanned and illuminated for n times, and the method comprises the following steps:
wherein, O i Is a sub-target of i-th scanned illumination of an object O to be imaged, and each sub-target O i The target O is beyond the memory effect range of the scattering imaging system after n times of scanning illumination within the respective memory effect range; recording all sub-targets O by area array photoelectric detector i The reflected light of (2) passes through the speckle image I of the scattering medium i 。
Step 2: speckle image I corresponding to each sub-target through phase recovery algorithm i Reconstructing sub-target image O 'in autocorrelation' i 。
And step 3: judging two adjacent sub-targets O i And O i+1 If the distance between the two is less than the diameter of the memory effect rangeSub-target O i And O i+1 With memory effect ranges overlapping each other, two sub-target speckle images I i And I i+1 There is an information correlation between the two sub-target images, and the reconstructed two sub-target images O 'are determined by estimating the point spread function PSF of the speckle images of the two sub-targets and calculating the autocorrelation and cross-correlation of the point spread functions of the two sub-targets' i And O' i+1 Relative position and orientation information therebetween; if two sub-targets O i And O i+1 Is greater than the memory effect range, indicating that the two sub-targets do not overlap with each other, and there is no information correlation between the speckle images of the two sub-targets.
The specific steps of determining the relative position and direction information of the two reconstructed sub-target images are as follows:
step 3.1: estimating a point spread function: from speckle image I i And the reconstructed sub-target image O 'obtained by calculation in the step 2' i Estimating speckle image I by wiener deconvolution algorithm i Point spread function PSF' i ：
PSF' i ＝Deconv(I i ,O' i ) (2)
Wherein Deconv denotes a deconvolution operation.
Step 3.2: determining reconstructed sub-target image O' i And O' i+1 Relative position between: separately calculate PSF' i And PSF' i And PSF' i And PSF' i+1 Determining two-dimensional coordinates of respective correlation peaks, and subtracting to obtain a characterization sub-target image O' i And O' i+1 Shift vector of relative position therebetween
In the above formula { } represents the correlation operation, and the position { } represents the two-dimensional coordinates of the correlation peak position.
Step 3.3: determining reconstructed sub-target image O' i The direction information of (1): note that the phase recovery sub-target image O' i Is uncertain, there may be an erroneous direction of left-right flipping, up-down flipping, or both left-right and up-down flipping, but the sub-target image O 'is reconstructed' i Speckle image I of i And sub-target image O' i+1 Speckle image I of i+1 There is a bearing information association between them. In this case, the reconstructed sub-target image O' i Is turned left and right, turned up and down, and simultaneously turned left and right and up and down, and sub-target image O' i Obtaining the sub target image O 'per se' i Four different directional states; and then from the speckle image I i And sub target image O' i The PSF 'corresponding to the four directions is estimated and obtained in the state of the four different directions' i Then from four estimated PSFs' i With adjacent speckle image I i+1 Deconvolution reconstruction is carried out to obtain four deconvolution images, and the four deconvolution images and the sub-target image O 'obtained by phase recovery' i+1 Is compared to determine O' i The correct direction of the light.
Step 3.4: two reconstructed sub-target images O 'with correct directions' i And O' i+1 According to the shift vector in step 3.2
And performing shift superposition.
Step 3.5: and (5) repeating the step 3.1 to the step 3.4 for all the sub-targets, and determining the relative position relationship and the direction information of all the sub-targets.
And 4, step 4: and (3) shifting and superposing the reconstructed image with correct relative position and direction information obtained in the step (3) to obtain a correctly spliced target O, so that the limitation of the memory effect of the scattering medium is avoided, and the non-invasive large-field imaging through the scattering medium is realized.
The invention also discloses a method for improving imaging resolution and realizing non-invasive large-field imaging based on PSF estimation and correlation operation, which is based on the method for realizing non-invasive large-field imaging through a scattering medium based on PSF estimation and correlation operation, obtains a high-resolution image breaking through the diffraction limit of an imaging system by using a Gaussian fitting repositioning method, determines the relative position and direction information of the high-resolution images of sub-targets and splices the relative position and direction information, so that the target realizes large-field imaging and super-resolution imaging at the same time, and comprises the following specific steps:
step 1: repeating the steps 1 to 2 in the method for realizing the non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and obtaining the ith sub-target O of the jth sparse illumination i_j Low resolution picture of O' i_j Then, a Gaussian fitting repositioning method is utilized to process the low-resolution image to obtain a high-resolution image which breaks through the diffraction limit of the imaging system
If the expected sparse illumination can be further controlled to be generated, the ith sub-target O in the step 1 of the method for realizing the non-invasive large-field-of-view imaging through the scattering medium based on PSF estimation and related operation i Is randomly and sparsely lattice illuminated m times, then there are
Wherein, O i_j Is sub-target O i The result of being illuminated by the jth sparse lattice (i.e., the jth sparse illumination of the ith sub-target O) i_j ) (ii) a Repeating step 2 in the method for realizing non-invasive large-field-of-view imaging through scattering medium based on PSF estimation and related operation, and obtaining ith sub-target O of jth sparse illumination i_j Low resolution image O' i_j And then processing the low-resolution image O 'by utilizing a Gaussian fitting relocation method' i_j Obtaining high resolution image breaking through diffraction limit of imaging system
And 2, step: calculating a point spread function PS for an imaging systemF ', and combining PSF' with m collected speckle images I i_j After deconvolution operation one by one, gaussian fitting repositioning processing is carried out to obtain m high-resolution images
Point diffusion function PSF ' calculation method and point diffusion function PSF ' of method for realizing noninvasive large-field-of-view imaging through scattering medium based on PSF estimation and related calculation ' i The calculation method of (2) is the same.
And step 3: m high resolution images
Corresponding positions are overlapped to obtain a sub-target high-resolution image breaking through the diffraction limit
And 4, step 4: high-resolution images of each sub-target obtained in the step 3
And (4) operating according to the step (3) and the step (4) of the method for realizing the non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and realizing the non-invasive large-field super-resolution imaging through the scattering medium.
The invention also discloses a device for realizing non-invasive large-field imaging through a scattering medium based on PSF estimation and related operation, which is used for realizing the two methods, and comprises a target 1, an aperture diaphragm 2, a scattering medium 3 and an area array photodetector 4, wherein the aperture diaphragm 2 is arranged between the target 1 and the area array photodetector 4, the scattering medium 3 is tightly attached to one surface of the aperture diaphragm 2, in the embodiment shown in figure 1, the scattering medium 3 is positioned on the left side of the aperture diaphragm 2, the same technical effect of the device can be realized by arranging the scattering medium 3 on the other surface of the aperture diaphragm 2, the planes of the area array photodetector 4 and the scattering medium 3 are both parallel to the plane of the target 1, and the central position of the planes of the area array photodetector 4 and the scattering medium 3 is superposed with the normal of the target 1; the distance between the scattering medium 3 and the target 1 is larger than the distance between the scattering medium 3 and the area array light detector 4.
The scattering medium 3 is ground glass, a diffuse reflection wall surface, biological tissues, cloud and mist, turbid water or an artificial scattering medium.
Non-invasive large field of view imaging is achieved:
the apparatus of the invention is constructed in accordance with fig. 1, in the embodiment of fig. 1 the object distance u =94mm from the plane of the object 1 to the plane of the scattering medium 3 and the image distance v =80mm from the plane of the scattering medium 3 to the plane of the area-array photodetector 4. Sequentially scanning illumination sub-targets O 1 -O 9 The scanning illumination sequence is numerically shown in fig. 1. Sub-target O 1 -O 9 The reflected light passes through a 4.5mm aperture diaphragm and penetrates through a scattering medium, and 9 speckle images I are collected by an area array photoelectric detector 4 1 -I 9 In this embodiment, the surface-array photodetector is an sCMOS camera, and the results are shown in fig. 2 (a 1) - (a 9). Then the speckle image I 1 -I 9 Respectively reconstructing through a phase recovery algorithm to obtain 9 sub-target images O' 1 -O' 9 As shown in fig. 2 (b 1) - (b 9). PSF is obtained through deconvolution of corresponding speckle images and sub-target images' 1 -PSF' 9 As shown in fig. 2 (c 1) - (c 9). PSF 'was calculated in the order of the arrows shown in FIG. 1' i And PSF' i And PSF' i+1 Determining two-dimensional coordinates of respective correlation peaks, and subtracting to obtain a characterization sub-target image O' i And O' i+1 8 shift vectors of relative position therebetween
As shown in fig. 3 (a 1) - (a 8). Then determining reconstructed sub-target image O' 1 -O' 9 The direction information of (2). Finally, for the sub-target image O 'with the correct direction' 1 -O' 9 The superposition and splicing are respectively carried out according to 8 shift vectors shown in (a 1) - (a 8) in fig. 3, as shown in (a 1) - (a 8) in fig. 4, a correct overall target O is obtained, as shown in fig. 5, and then the non-invasive large-field imaging can be realized.
Realizing non-invasive large-field super-resolution imaging:
first sub-target O 1 Is randomly and sparsely illuminated by the lattice for j times to obtain j speckle images I 1_j In this embodiment, j is 500, and one speckle image I is selected 1_56 And phase recovery is carried out to obtain a low-resolution sub-image O' 1_56 Combining with Gaussian fitting repositioning processing to obtain high-resolution sub-image breaking through diffraction limit of imaging system
Then collecting speckle images I 1_56 And high resolution sub-image
Performing wiener deconvolution to calculate a point spread function PSF 'of the speckle image' i . The point diffusion function PSF 'obtained by calculation' i And 500 speckle images I 1_j Performing deconvolution operation one by one, and performing Gaussian fitting repositioning processing to obtain 500 high-resolution sub-images
500 high-resolution subimages
Corresponding positions are superposed, and a sub-target image O breaking through the diffraction limit can be obtained 1 As shown in fig. 6 (a). The above steps are repeated one by one for the remaining 8 sub-targets, and super-resolution imaging is performed as shown in fig. 6 (b) - (i). Finally, combining the obtained super-resolution images (6 (a) - (i)) with the non-invasive large-field super-resolution imaging, thereby realizing the non-invasive large-field super-resolution imaging, as shown in fig. 6 (g).
By taking common scattering media such as ground glass, diffuse reflection wall surfaces, biological tissues, cloud and mist, turbid water bodies and artificial scattering media as examples, non-invasive large-field-of-view imaging can be realized based on the scattering media.
The area array photoelectric detector 4 can also select a CCD, and the non-invasive large-field imaging can also be realized according to the method of the invention.
When the target is too large to exceed the memory effect range, reconstructing the image of the sub-targets of the target hidden in the scattering medium by using a phase recovery algorithm, obtaining the orientations of all the sub-targets based on point spread function estimation and autocorrelation and cross-correlation operation, and performing image splicing on the sub-targets according to the orientations to realize non-invasive large-field imaging, as shown in fig. 7 (a); by using the PSF estimation and correlation operation-based method for realizing non-invasive large-field imaging for improving imaging resolution, which is provided by the invention, a high-resolution image breaking through the diffraction limit of an imaging system is obtained by using a Gaussian fitting repositioning method, the relative position and direction information of the high-resolution images of sub-targets is determined and spliced, so that super-resolution imaging is realized while large-field imaging of the target is realized, and as shown in FIG. 7 (b), the imaging field and the imaging resolution can be simultaneously considered.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may occur to those skilled in the art without departing from the principle of the invention, and are considered to be within the scope of the invention.
Claims (6)
1. A method for realizing non-invasive large-field-of-view imaging through a scattering medium based on PSF estimation and related operation is characterized in that a phase recovery algorithm is used for reconstructing a sub-target O of a target O hidden in the scattering medium i The image of (2) obtains the directions of all sub-targets based on point spread function estimation and autocorrelation and cross-correlation operation, and carries out image splicing on the sub-targets according to the directions to enable the target O to realize non-invasive large-field imaging, and the specific steps are as follows:
step 1: scanning a target O hidden in a scattering medium by using active illumination, wherein the target O is scanned and illuminated for n times, and the method comprises the following steps:
wherein, O i Is a sub-target of the i-th scanned illumination of the target O to be imaged, each timeSub target O i The target O is beyond the memory effect range of the scattering imaging system after n times of scanning illumination within the respective memory effect range; recording all sub-targets O by area array photoelectric detector i The reflected light of (2) passes through the speckle image I of the scattering medium i ；
Step 2: speckle image I corresponding to each sub-target through phase recovery algorithm i Reconstructing sub-target image O 'in autocorrelation' i ；
And step 3: judging two adjacent sub-targets O i And O i+1 If the distance between two sub-targets O is smaller than the diameter of the memory effect range i And O i+1 With memory effect ranges overlapping each other, two sub-target speckle images I i And I i+1 There is an information correlation between the two sub-target images, and the reconstructed two sub-target images O 'are determined by estimating the point spread function PSF of the speckle images of the two sub-targets and calculating the autocorrelation and cross-correlation of the point spread functions of the two sub-targets' i And O' i+1 Relative position and orientation information therebetween;
and 4, step 4: and (4) shifting and superposing the reconstructed image with correct relative position and direction information obtained in the step (3) to obtain a correctly spliced target O, so that the limitation of the memory effect of the scattering medium is avoided, and the noninvasive large-field imaging through the scattering medium is realized.
2. The method for realizing non-invasive large-field-of-view imaging through a scattering medium based on PSF estimation and correlation operation as claimed in claim 1, wherein in the step 3, the specific steps of determining the relative position and direction information of the two reconstructed sub-target images are as follows:
step 3.1: estimating a point spread function: from speckle image I i And the reconstructed sub-target image O 'obtained by calculation in the step 2' i Estimating speckle image I by wiener deconvolution algorithm i Point spread function PSF of i '：
PSF i '＝Deconv(I i ,O′ i )
Wherein, deconv represents a deconvolution operation;
step 3.2: determining reconstructed sub-target image O' i And O' i+1 Relative position between: separately calculating PSF i ' autocorrelation and PSF i 'and PSF' i+1 Determining two-dimensional coordinates of respective correlation peaks, and subtracting to obtain a characterization sub-target image O' i And O' i+1 Shift vector of relative position therebetween
In the above formula { } represents a correlation operation, and position { } represents a two-dimensional coordinate of a correlation peak position;
step 3.3: determining reconstructed sub-target image O' i The direction information of (1): reconstructing sub-target image O' i Is turned left and right, turned up and down, and simultaneously turned left and right and up and down, and sub-target image O' i Obtaining the sub target image O 'per se' i Four different directional states; and then from the speckle image I i And sub target image O' i The PSF 'corresponding to the four directions is estimated and obtained in the state of the four different directions' i Then from four estimated PSFs' i With adjacent speckle image I i+1 Deconvolution reconstruction is carried out to obtain four deconvolution images, and the four deconvolution images and the sub-target image O 'obtained by phase recovery' i+1 Is compared to determine O' i The correct direction of the light;
step 3.4: two reconstructed sub-target images O 'with correct directions' i And O' i+1 According to the shift vector in step 3.2
Shifting and superposing;
step 3.5: and (5) repeating the step 3.1 to the step 3.4 for all the sub-targets, and determining the relative position relationship and the direction information of all the sub-targets.
3. A PSF estimation and correlation operation based method for realizing non-invasive large-field imaging based on improving imaging resolution, which is based on the PSF estimation and correlation operation based method for realizing non-invasive large-field imaging through a scattering medium in claim 1 or 2, is characterized in that a Gauss fitting relocation method is used for obtaining a high-resolution image breaking through the diffraction limit of an imaging system, determining the relative position and direction information of sub-target high-resolution images and splicing the images, so that super-resolution imaging is realized while the target realizes large-field imaging, and the method comprises the following specific steps:
step 1: repeating the steps 1 to 2 in the method for realizing the non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and obtaining the ith sub-target O of the jth sparse illumination i_j Low resolution image O' i_j Then, a Gaussian fitting repositioning method is utilized to process the low-resolution image to obtain a high-resolution image which breaks through the diffraction limit of the imaging system
Step 2: calculating the point spread function PSF 'of the imaging system, and comparing the PSF' with the m acquired speckle images I i_j After deconvolution operation one by one, gaussian fitting repositioning processing is carried out to obtain m high-resolution images
And step 3: m high resolution images
Corresponding positions are overlapped to obtain a sub-target high-resolution image breaking through the diffraction limit
And 4, step 4: high-resolution images of each sub-target obtained in the step 3
And (4) operating according to the step (3) and the step (4) of the method for realizing the non-invasive large-field imaging through the scattering medium based on PSF estimation and related operation, and realizing the non-invasive large-field super-resolution imaging through the scattering medium.
4. The PSF estimation and correlation based method for achieving non-invasive large field of view imaging with improved imaging resolution as claimed in claim 3, wherein in step 2, the point spread function PSF' is calculated by the PSF estimation and correlation based method and the point spread function PSF is calculated by the PSF estimation and correlation based method for achieving non-invasive large field of view imaging through scattering media i ' the calculation method is the same.
5. A device for realizing non-invasive large-field imaging through a scattering medium based on PSF estimation and correlation operation is used for realizing the method of any one of claims 1 to 4, and is characterized by comprising a target (1), an aperture diaphragm (2), a scattering medium (3) and an area array photodetector (4), wherein the aperture diaphragm (2) is arranged between the target (1) and the area array photodetector (4), the scattering medium (3) is tightly attached to one surface of the aperture diaphragm (2), the planes of the area array photodetector (4) and the scattering medium (3) are both parallel to the plane of the target (1), and the center position of the plane coincides with the normal of the target (1); the distance between the scattering medium (3) and the target (1) is larger than the distance between the scattering medium (3) and the area array optical detector (4).
6. The device for realizing non-invasive large-field-of-view imaging through scattering media based on PSF estimation and correlation operations as claimed in claim 5, wherein the scattering media (3) are ground glass, diffuse reflection wall surfaces, biological tissues, cloud mist, turbid water bodies or artificial scattering media.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211446193.4A CN115755383A (en) | 2022-11-18 | 2022-11-18 | Non-invasive large-field-of-view imaging method and device through scattering medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211446193.4A CN115755383A (en) | 2022-11-18 | 2022-11-18 | Non-invasive large-field-of-view imaging method and device through scattering medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115755383A true CN115755383A (en) | 2023-03-07 |
Family
ID=85373247
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211446193.4A Pending CN115755383A (en) | 2022-11-18 | 2022-11-18 | Non-invasive large-field-of-view imaging method and device through scattering medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115755383A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116242252A (en) * | 2023-05-11 | 2023-06-09 | 之江实验室 | Scattering imaging method with positioning and size measuring functions |
-
2022
- 2022-11-18 CN CN202211446193.4A patent/CN115755383A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116242252A (en) * | 2023-05-11 | 2023-06-09 | 之江实验室 | Scattering imaging method with positioning and size measuring functions |
| CN116242252B (en) * | 2023-05-11 | 2023-08-15 | 之江实验室 | Scattering imaging method with positioning and size measuring functions |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2239706B1 (en) | Method and camera for the real-time acquisition of visual information from three-dimensional scenes | |
| US9679360B2 (en) | High-resolution light-field imaging | |
| US20110043603A1 (en) | System And Method For Dehazing | |
| JP2013542468A (en) | Scanning projection lensless microscope system | |
| CN113962866B (en) | Scattering imaging target reconstruction method based on physical perception learning | |
| CN115755383A (en) | Non-invasive large-field-of-view imaging method and device through scattering medium | |
| CN101865673A (en) | Microcosmic optical field acquisition and three-dimensional reconstruction method and device | |
| US7336372B2 (en) | Noninvasive optical imaging by speckle ensemble | |
| Ali et al. | Anisotropic motion estimation on edge preserving Riesz wavelets for robust video mosaicing | |
| CN115235999A (en) | Polarization imaging system with super-resolution fusion | |
| CN115220211A (en) | Microscopic imaging system and method based on deep learning and light field imaging | |
| Ichimaru et al. | Unified underwater structure-from-motion | |
| Li et al. | Visibility enhancement of underwater images based on active polarized illumination and average filtering technology | |
| CN112070675A (en) | Regularization light field super-resolution method based on graph and light field microscopic device | |
| CN110827375A (en) | Infrared image true color coloring method and system based on low-light-level image | |
| Je et al. | Disparity-based space-variant image deblurring | |
| US11480917B2 (en) | Method and apparatus for reconstructing three-dimensional image by using diffraction grating | |
| EP4068207A1 (en) | Method of pixel-by-pixel registration of an event camera to a frame camera | |
| Muniraj et al. | Subpixel based defocused points removal in photon-limited volumetric dataset | |
| CN112950481A (en) | Water bloom shielding image data collection method based on image mosaic network | |
| Kim et al. | Environment modelling using spherical stereo imaging | |
| CN111272271A (en) | Vibration measurement method, system, computer device and storage medium | |
| Zhu et al. | Ground target classification from sar image with the pixel complement for target shape | |
| Liu et al. | Optical imaging study of underwater acousto-optical fusion imaging systems | |
| CN114972653B (en) | Monocular polarization three-dimensional reconstruction imaging method based on zenith angle positioning |
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 |