CN113048907A - Single-pixel multispectral imaging method and device based on macro-pixel segmentation - Google Patents
Single-pixel multispectral imaging method and device based on macro-pixel segmentation Download PDFInfo
- Publication number
- CN113048907A CN113048907A CN202110184220.4A CN202110184220A CN113048907A CN 113048907 A CN113048907 A CN 113048907A CN 202110184220 A CN202110184220 A CN 202110184220A CN 113048907 A CN113048907 A CN 113048907A
- Authority
- CN
- China
- Prior art keywords
- pixel
- macro
- dimensional
- light
- sub
- 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
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000000701 chemical imaging Methods 0.000 title claims abstract description 48
- 230000011218 segmentation Effects 0.000 title claims abstract description 46
- 230000003595 spectral effect Effects 0.000 claims abstract description 38
- 238000001228 spectrum Methods 0.000 claims abstract description 30
- 238000005070 sampling Methods 0.000 claims abstract description 26
- 238000005286 illumination Methods 0.000 claims abstract description 11
- 238000001514 detection method Methods 0.000 claims description 16
- 230000003287 optical effect Effects 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 8
- 239000006185 dispersion Substances 0.000 claims description 4
- 230000008878 coupling Effects 0.000 claims description 3
- 238000010168 coupling process Methods 0.000 claims description 3
- 238000005859 coupling reaction Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 239000011248 coating agent Substances 0.000 claims description 2
- 238000000576 coating method Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 16
- 238000003384 imaging method Methods 0.000 description 17
- 238000005516 engineering process Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 9
- 230000010287 polarization Effects 0.000 description 6
- 230000007547 defect Effects 0.000 description 2
- 235000001466 Ribes nigrum Nutrition 0.000 description 1
- 241001312569 Ribes nigrum Species 0.000 description 1
- 238000013398 bayesian method Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000007261 regionalization Effects 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/12—Generating the spectrum; Monochromators
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/12—Generating the spectrum; Monochromators
- G01J2003/1208—Prism and grating
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a single-pixel multispectral imaging method and device based on macro-pixel segmentation. The method carries out macro-pixel segmentation and wavelength-position coding on a spectrum cube to be detected so as to realize sampling of the three-dimensional spectrum cube in a one-time two-dimensional sampling process. The invention further designs a single-pixel multispectral imaging device, which loads a preset two-dimensional code on a spatial light modulator for generating structured light illumination; using a specific array filter to perform macro-pixel segmentation and wavelength-position coding on the illumination light field; detecting the light field intensity value acted by the detected scene by using a single-point detector; and according to the one-dimensional intensity measured value, reconstructing the two-dimensional space information and the spectral information of the scene to be measured by using a reconstruction algorithm, and thus restoring to obtain a multispectral imaging result. The method has simple light path and high sampling speed, and can select proper wave bands and wave band numbers by designing the array filter, thereby realizing rapid and effective single-pixel multispectral imaging.
Description
Technical Field
The invention relates to the field of computational imaging, in particular to a single-pixel multispectral imaging method and device based on macro-pixel segmentation.
Background
The multispectral imaging technology can simultaneously acquire the information of the space dimension and the spectral dimension of an object, and provides an effective means for analyzing and identifying substances, so that the multispectral imaging technology has wide application in the fields of military affairs, biology, medicine, industrial detection and the like. Conventional multi-spectral imaging techniques typically acquire a three-dimensional spectral cube of the target object by scanning over time. The method can acquire complete information of the target object spectrum cube, but has long scanning time and large data redundancy. In order to obtain spectral cubes more efficiently, researchers have proposed a coded aperture spectral imaging technique (CASSI). The technology uses a compressive sensing theory, and can restore the respective spatial information of each spectral band from a single two-dimensional aliasing image containing the information of each spectral band, thereby improving the multispectral imaging efficiency. However, the multispectral imaging technology uses an area array detector such as a CCD for detection, and the detected spectral range is relatively limited.
The single-pixel imaging is a new imaging technology, combines spatial coding and single-point detection, can reconstruct the spatial information of an object from a one-dimensional signal obtained by a single-point detector, and does not need to use an area array detector to detect the object, so that the problem of high cost of a partial waveband array detector can be solved, and the range of a detection spectrum band is expanded. Since single pixel imaging technology focuses light onto a single point detector, detection can also be performed when the light intensity is weak. The existing single-pixel multispectral imaging technologies are mainly divided into the following technologies: (1) the spectrometer is directly used as a detector for detection, so that imaging with high spectral resolution can be realized, but the whole system is expensive and heavy; (2) the spectral dimension is coded by using mechanical structures such as a color wheel, a specially-made coding wheel and the like so as to restore the spectral cube of the target object from the one-dimensional intensity information, and the system has the defects that the spectral coding speed is limited by the mechanical structures and the stability is poor; (3) based on Fourier single-pixel imaging technology, space and color information are simultaneously encoded by using a color coding pattern, but the defects of limited imaging spectrum range, slow projection speed of the color coding pattern and low imaging efficiency exist. Therefore, the multispectral imaging with high imaging efficiency, rich spectrum bands and good system stability is realized by using the single-pixel detector, and the multispectral imaging method has important significance.
Disclosure of Invention
In order to overcome the above problems in the prior art, an object of the present invention is to provide a method for multi-spectral imaging with a single pixel based on macro-pixel segmentation. The method can acquire the spectrum cube information in a one-time two-dimensional sampling process, can effectively recover the three-dimensional spectrum cube information of a scene to be detected, and realizes efficient, stable and band-extensible single-pixel multi-spectral imaging.
Another objective of the present invention is to provide a single-pixel multispectral imaging device based on macro-pixel segmentation. The device uses an array filter to perform macro-pixel segmentation and wavelength-position coding on a scene spectrum cube to be detected so as to realize multispectral imaging.
The specific scheme of the invention is introduced as follows:
the invention firstly discloses a single-pixel multispectral imaging method based on macro-pixel segmentation, which comprises the following steps:
1) loading a preset two-dimensional code on an optical modulator, generating two-dimensional structured light by incident light emitted by a light source through the optical modulator, and performing macro-pixel segmentation and wavelength-position coding on the two-dimensional structured light to generate macro-pixel coded structured light capable of simultaneously acquiring spatial position information and spectral information;
2) coupling two-dimensional space information and spectrum information of a scene to be measured into a one-dimensional intensity measurement value by using the macro-pixel coding structure light and the single-point detector;
3) and reconstructing the measured value based on the one-dimensional intensity by using a reconstruction algorithm to restore the two-dimensional spatial information and the spectral information of the scene to be detected, thereby realizing single-pixel multispectral imaging.
The invention also discloses a selectable single-pixel multispectral imaging method based on macro-pixel segmentation, which comprises the following steps:
1) the method comprises the steps of carrying out macro-pixel segmentation and wavelength-position coding on incident light emitted by a light source, and then carrying out two-dimensional coding on the incident light by using a light modulator loaded with preset two-dimensional codes to generate macro-pixel coding structured light capable of simultaneously acquiring spatial position information and spectral information.
2) Coupling two-dimensional space information and spectrum information of a scene to be measured into a one-dimensional intensity measurement value by using the macro-pixel coding structure light and the single-point detector;
3) and reconstructing the measured value based on the one-dimensional intensity by using a reconstruction algorithm to restore the two-dimensional spatial information and the spectral information of the scene to be detected, thereby realizing single-pixel multispectral imaging.
As a preferred aspect of the present invention, the preset two-dimensional code loaded on the spatial light modulator is designed based on macro-pixel division, each macro-pixel region includes a plurality of sub-regions, each sub-region corresponds to a different position in the macro-pixel region, each sub-region is configured to allow light to enter the subsequent optical system or not allow light to enter the subsequent optical system, and the sub-region codes located at the same position in all the macro-pixel regions constitute a two-dimensional sub-code.
As a preferred embodiment of the present invention, the macro-pixel division and the wavelength-position encoding are performed on the two-dimensional structured light, specifically:
sub-region codes positioned at the same position in all macro-pixel regions form a two-dimensional sub-code, and one two-dimensional sub-code corresponds to one sampling waveband;
the sub-areas at different positions in each macro-pixel area allow light with different wavelengths to transmit, and each different sub-area in the macro-pixel area can sample information of a plurality of wave bands of an object to be measured in the macro-pixel area;
so that one macro-pixel of the structured light can acquire information of a plurality of spectral channels corresponding to the pixel;
the macro-pixel segmentation and the wavelength-position coding are realized by adopting one or more of an array filter, a surface coating of the optical modulator and dispersion of a dispersion element.
The reconstruction algorithm is a compressed sensing algorithm, a dictionary learning method or a deep learning algorithm. The macro-pixel coded structured light may be actively illuminated structured light or passively illuminated structured light.
The invention further provides a single-pixel multispectral imaging device based on macro-pixel segmentation, which comprises:
the structure light module is composed of a white light source, a spatial light modulator and an array filter, wherein the white light source is used as an illumination light source, the spatial light modulator loads a preset two-dimensional code, incident light emitted by the white light source passes through the spatial light modulator to generate two-dimensional structure light, the two-dimensional structure light is subjected to macro-pixel segmentation and wavelength-position coding by the array filter to generate structure light capable of simultaneously collecting spatial position information and spectral information, and the spatial position information and the spectral information are collected simultaneously;
the detection module consists of a convergent lens and a single-point detector, the convergent lens is used for collecting diffuse reflection light of a scene to be detected after the scene is illuminated by the structured light, and the single-point detector is used for detecting the intensity value of the diffuse reflection light;
and the reconstruction module recovers the spectrum cube of the scene to be detected from the detected intensity value by using a reconstruction algorithm to realize single-pixel multispectral imaging.
The invention also provides another optional single-pixel multispectral imaging device based on macro-pixel segmentation, which comprises:
the structure light module is composed of a white light source, an array filter and a spatial light modulator, wherein the white light source is used as an illumination light source, the array filter is used for carrying out macro-pixel segmentation and wavelength-position coding on incident light emitted by the white light source, then the spatial light modulator loaded with preset two-dimensional codes is used for carrying out two-dimensional coding on the incident light, structure light capable of simultaneously collecting spatial position information and spectral information is generated, and the spatial position information and the spectral information are collected simultaneously;
the detection module consists of a convergent lens and a single-point detector, the convergent lens is used for collecting diffuse reflection light of a scene to be detected after the scene is illuminated by the structured light, and the single-point detector is used for detecting the intensity value of the diffuse reflection light;
and the reconstruction module recovers the spectrum cube of the scene to be detected from the detected intensity value by using a reconstruction algorithm to realize single-pixel multispectral imaging.
Compared with the prior art, the invention provides a single-pixel multispectral imaging system based on macro-pixel segmentation, which performs two-dimensional coding on an illumination light field through a spatial light modulator, performs macro-pixel segmentation and wavelength-position coding by using an array filter, illuminates a measured object by using the generated novel structured light, and detects a one-dimensional intensity value after diffuse reflection of the measured object by using a single-point detector. After multiple measurements, the spectral cube information of the measured object is restored by combining preset two-dimensional coding, wavelength-position coding and reconstruction algorithms (such as a compressive sensing algorithm) so as to realize the multispectral imaging system based on macro-pixel segmentation and a single-pixel detector. The invention provides the ideas of macropixel segmentation and wavelength-position coding, can convert a primary three-dimensional spectrum cube sampling process into a primary two-dimensional sampling process, and completes primary sampling on the three-dimensional spectrum cube of a measured object in a single detection process. Compared with other single-pixel multispectral imaging systems, the multispectral imaging system has the advantages of simpler light path, no limitation of mechanical structures, higher sampling efficiency and expandable spectrum band.
Drawings
FIG. 1 is a schematic diagram of three-dimensional sampling of a data cube.
FIG. 2 is a schematic diagram illustrating the steps of the single-pixel multispectral imaging method based on macro-pixel segmentation according to the present invention.
Fig. 3 is a schematic diagram of a single-pixel multispectral imaging system based on macro-pixel segmentation according to an embodiment of the present invention.
FIG. 4 is a schematic diagram of a process for forming a macro-pixel encoding pattern according to an embodiment of the present invention.
FIG. 5 is a schematic diagram of three-dimensional sampling using a macropixel coding pattern according to an embodiment of the present invention.
Reference numbers in the figures: 1-a broad spectrum light source; 2-LCOS chip; a 3-polarizing beam splitter Prism (PBS); 4-an array filter; 5-a projection lens; 6-the object to be tested; 7-a converging lens; 8-single point detector; 9-a data acquisition card; 10-a computer.
Detailed Description
The invention is explained in more detail below with reference to specific embodiments and the associated drawings.
As a novel imaging technology, the single-pixel imaging technology can reconstruct two-dimensional intensity information of a measured object from one-dimensional intensity measured values detected by a single-point detector. The technology does not need to use an area array detector, combines two-dimensional coding and single-point detection, and can greatly expand the imaging waveband range. In a single-pixel imaging system, a series of two-dimensional coding patterns are preset to sample two-dimensional space information of a measured object for multiple times, the two-dimensional space information is coupled into one-dimensional intensity information to be detected, and then the one-dimensional intensity information is restored by a reconstruction algorithm. In the single-pixel multispectral imaging method and device based on macropixel segmentation, the two-dimensional spatial information and the spectral information of the object to be detected need to be detected simultaneously, namely the three-dimensional data cube information of the object to be detected is detected. In this case, the two-dimensional coding pattern is insufficient to sample the data cube information.
The invention therefore proposes the idea of sampling a data cube using a three-dimensional coding pattern. FIG. 1 is a schematic diagram of three-dimensional sampling of a data cube. The three-dimensional sampling process can be similar to the two-dimensional sampling process in a single-pixel imaging system, and the spectral dimension is increased. In fig. 1, the white squares in the three-dimensional code pattern represent that information of the data cube at the coordinates (x, y, λ) is collected, and the black squares in the three-dimensional code pattern represent that information of the data cube at the coordinates (x, y, λ) is not collected. However, in practical cases, the light modulation device such as LCOS or DMD can only modulate a two-dimensional pattern, and cannot generate the three-dimensional code pattern in fig. 1. Therefore, the present invention is based on the idea of macropixel segmentation, wavelength-position coding, and designs the macropixel coding pattern shown in fig. 1. The macro-pixel coding pattern can approximately replace a three-dimensional coding pattern, and three-dimensional sampling of a measured object data cube is achieved.
Fig. 2 is a specific flowchart of the single-pixel multispectral imaging method based on macro-pixel segmentation according to the present invention. As shown in fig. 2, the single-pixel multispectral imaging method based on macro-pixel segmentation mainly includes the following steps:
in step S201, a two-dimensional code is preset and loaded on the light modulator, and two-dimensional structured light is generated. The preset two-dimensional code is designed based on the theory of macro-pixel segmentation. In order to finally realize multispectral single-pixel imaging, the preset two-dimensional code needs to be matched with a spectral coding element in the subsequent step. In the two-dimensional coding, the codes at the same position in each macro-pixel region form a two-dimensional sub-code, and each two-dimensional sub-code corresponds to a sampling waveband of the spectrum coding element in the subsequent step. The two-dimensional sub-coding includes but is not limited to random coding, Hadamard coding and stripe coding. In this embodiment, a random code is used as a two-dimensional sub-code, and a corresponding two-dimensional code pattern is loaded on an LCOS chip to generate two-dimensional structured light.
In step S202, macro-pixel segmentation and wavelength-position encoding are performed on the two-dimensional structured light to generate a novel structured light capable of simultaneously collecting spatial information and spectral information. In the present embodiment, the two-dimensional structured light generated in step S201 is subjected to macro-pixel division and wavelength-position encoding using an array filter. Within each macropixel region, different positions transmit different wavelengths of light. The novel structure generated in step S202 can simultaneously sample spatial position information and spectral information.
In step S203, the single-point detector is used to detect the light field intensity value after the measured scene is acted on. In this embodiment, the detection module is composed of a converging lens and a single-point detector, and can couple the information of the sampled portion of the three-dimensional data cube into a one-dimensional intensity measurement.
In step S204, according to the one-dimensional intensity measurement value, the two-dimensional spatial information and the spectral information of the detected scene are reconstructed by using a reconstruction algorithm, so as to obtain a multispectral imaging result by reduction. The reconstruction algorithm comprises algorithms with similar functions, such as a compressed sensing algorithm (convex optimization algorithm, greedy algorithm, combined algorithm and Bayesian method), a dictionary learning method, a deep learning algorithm and the like.
The embodiment of the invention provides a single-pixel multispectral imaging system based on macro-pixel segmentation. The specific system structure is shown in fig. 3, and includes a wide-spectrum light source 1, an LCOS chip 2, a polarization splitting prism 3, an array filter 4, a projection lens 5, a converging lens 7, a single-point detector 8, a data acquisition card 9, and a computer 10. The object to be measured is part 6 in the figure. When the wide-spectrum LCOS chip works, light emitted by the wide-spectrum light source 1 is polarized by the polarization beam splitter prism 3 and then is incident on the LCOS chip 2 to generate two-dimensional structure light. After passing through the polarization beam splitter prism 3, the emergent two-dimensional structured light irradiates the array filter 4 to perform macro-pixel segmentation and wavelength-position coding, so as to form novel structured light (namely a macro-pixel coding pattern shown in fig. 1) capable of simultaneously sampling two-dimensional spatial information and spectral information. The macro-pixel coding pattern is projected on a measured object 6 by a projection lens 5, the diffuse reflection light is converged on the target surface of a single-point detector 8 by a convergent lens 7, and the obtained measured value is converted by a data acquisition card 9 and then input into a computer 10 to finish single detection. By loading a series of preset two-dimensional codes on the LCOS chip 2 and controlling the single-point detector 8 to detect for multiple times, a one-dimensional intensity measurement value for reconstructing data cube information can be obtained.
FIG. 4 specifically illustrates a process for forming a macro-pixel coding pattern according to an embodiment of the present invention. In this embodiment, λ is selected11、λ12、λ13、λ21、λ22、λ23、λ31、λ32、λ33These 9 bands were measured. In fig. 4, a square grid surrounded by a thick gray frame is a macro-pixel region on the preset two-dimensional encoding diagram, the structural diagram of the array filter and the macro-pixel encoding pattern. In the preset two-dimensional coding pattern, the codes at the same position in each macro-pixel region form a two-dimensional sub-code, and the random code is used as the two-dimensional sub-code in the embodiment of the invention. Each macro-pixel area is divided into 3 x 3 small squares, which are used to sample the information of 9 wave bands in the macro-pixel range of the object to be tested. For the purpose of describing the macro-pixel encoding pattern formation process, the content of this section takes a macro-pixel in the black border at the top left corner of fig. 4 as an example. The analysis procedure for the range of macro-pixels is also applicable to other macro-pixels not marked by bold black boxes. The two-dimensional code and the array filter structure loaded on the LCOS chip are respectively shown in fig. 4. Wherein, the preset two-dimensional code displayWithin the range of the black frame, white cells and black cells respectively indicate that the light energy in the region enters the subsequent optical system and cannot enter the subsequent optical system. In the structural diagram of the array filter, within the range of the bold black frame, 9 small grids mark different wave band values, which indicates that only the light of the marked wave band is allowed to pass through in the corresponding grid area. It can be seen that the array filter segment performs macro-pixel area division on the illumination light field, and the cells in each macro-pixel area only allow light of one wavelength band to pass through, i.e., are wavelength-position encoded. Therefore, the array filter is aligned with the two-dimensional code loaded on the LCOS chip, and the illumination light field emitted from the LCOS chip penetrates through the array filter to form the macro-pixel code pattern as shown in fig. 4, i.e., a novel structured light. In FIG. 4, the macropixel coding pattern is within the range of the bold black frame, λ11、λ13、λ21、λ22And λ33The light of the wave band exists in the corresponding small grid area, and the rest black areas are all free of light emission. The described macropixel coding pattern may approximately replace a three-dimensional coding pattern for three-dimensional sampling of a data cube, the process being illustrated in fig. 5. In this process, the intensity of a local region of the corresponding band within a macro-pixel is used as the average intensity of the band within the macro-pixel. For example, FIG. 5 illustrates a macro-pixel encoding pattern with a thick black frame11Corresponding cells corresponding to the macro-pixel region band of λ in the three-dimensional coding pattern11Can approximate λ within the macropixel region in the sample data cube11Information at the wavelength band.
The invention can also be realized by adjusting the sequence of the two-dimensional code and the wavelength-position code (macro-pixel segmentation), can also convert the once three-dimensional spectrum cube sampling process into the once two-dimensional sampling process, and completes the once sampling of the three-dimensional spectrum cube of the object to be detected in the single detection process. The method also has the advantages of simpler light path, no limitation of mechanical structure, higher sampling efficiency and expandable spectrum. For example, in an alternative embodiment of the present invention, the single-pixel multispectral imaging device based on macro-pixel segmentation includes a wide-spectrum light source, an array filter, an LCOS chip, a polarization splitting prism, a projection lens, a converging lens, a single-point detector, a data acquisition card and a computer, and the imaging method thereof may include the following steps: the wide-spectrum light source provides illumination for a system, an array filter plate is used for carrying out macro-pixel segmentation and wavelength-position coding on an illumination light field, after polarization is carried out by a polarization beam splitter prism, LCOS (liquid Crystal on silicon) chips loaded with preset two-dimensional codes are used for carrying out two-dimensional coding, macro-pixel coding structural light capable of collecting space position information and spectral information is projected onto a measured object, a convergent lens is used for collecting diffuse reflection light of the measured object after the structured light is illuminated, a single-point detector is used for detecting the intensity value of the diffuse reflection light, the obtained measured value is converted by a data acquisition card and then is input into a computer, and single detection is completed. And changing the two-dimensional coding pattern on the LCOS chip, and detecting for multiple times to obtain a one-dimensional intensity measurement value which is subsequently used for reconstructing data cube information.
In summary, the invention discloses a single-pixel multispectral imaging method and device based on macro-pixel segmentation. The method and the device can sample three-dimensional spectrum cube information in a one-time two-dimensional sampling process by combining a macro-pixel segmentation and wavelength-position coding mode on the basis of a single-pixel imaging system, and can recover a data cube of a scene to be detected from a one-dimensional intensity measured value, thereby realizing multispectral imaging. The invention has the advantages of simpler light path, no limitation of mechanical structure, higher sampling efficiency and expandable spectrum, and can be widely applied to the fields of agriculture, military, medical treatment and the like.
In the embodiments, only a specific experimental device is selected for specifically illustrating the idea and features of the present invention, which is intended to enable those skilled in the art to understand the contents of the present invention and to implement the present invention accordingly, and the protection scope of the present invention is not limited to the above-mentioned embodiments. Therefore, all equivalent changes or modifications made according to the principles and experimental ideas disclosed by the present invention are within the scope of the present invention.
Claims (10)
1. A single-pixel multispectral imaging method based on macro-pixel segmentation is characterized by comprising the following steps:
1) loading a preset two-dimensional code on an optical modulator, generating two-dimensional structured light by incident light emitted by a light source through the optical modulator, and performing macro-pixel segmentation and wavelength-position coding on the two-dimensional structured light to generate macro-pixel coded structured light capable of simultaneously acquiring spatial position information and spectral information;
2) coupling two-dimensional space information and spectrum information of a scene to be measured into a one-dimensional intensity measurement value by using the macro-pixel coding structure light and the single-point detector;
3) and reconstructing the measured value based on the one-dimensional intensity by using a reconstruction algorithm to restore the two-dimensional spatial information and the spectral information of the scene to be detected, thereby realizing single-pixel multispectral imaging.
2. The method according to claim 1, wherein the predetermined two-dimensional code loaded on the spatial light modulator is designed based on macro-pixel division, each macro-pixel region comprises a plurality of sub-regions, each sub-region corresponds to a different position in the macro-pixel region, each sub-region is configured to allow or not allow light to enter the subsequent optical system, and all sub-region codes in the same position in the macro-pixel region constitute one two-dimensional sub-code.
3. The method according to claim 2, wherein the two-dimensional structured light is subjected to macropixel segmentation and wavelength-position coding, in particular:
sub-region codes positioned at the same position in all macro-pixel regions form a two-dimensional sub-code, and one two-dimensional sub-code corresponds to one sampling waveband;
the sub-areas at different positions in each macro-pixel area allow light with different wavelengths to transmit, and each different sub-area in the macro-pixel area can sample information of a plurality of wave bands of an object to be measured in the macro-pixel area; so that one macro-pixel of the structured light can acquire information of a plurality of spectral channels corresponding to the pixel;
the macro-pixel segmentation and the wavelength-position coding are realized by adopting one or more of an array filter, a surface coating of the optical modulator and dispersion of a dispersion element.
4. The method of claim 1, wherein the reconstruction algorithm is a compressed sensing algorithm, a dictionary learning method, or a deep learning algorithm.
5. The method of claim 1, wherein the macropixel encoded structured light is either actively illuminated structured light or passively illuminated structured light.
6. The method of claim 1, wherein step 1) is replaced with:
the method comprises the steps of carrying out macro-pixel segmentation and wavelength-position coding on incident light emitted by a light source, and then carrying out two-dimensional coding on the incident light by using a light modulator loaded with preset two-dimensional codes to generate macro-pixel coding structured light capable of simultaneously acquiring spatial position information and spectral information.
7. A single-pixel multispectral imaging device based on macropixel segmentation, comprising:
the structure light module is composed of a white light source, a spatial light modulator and an array filter, wherein the white light source is used as an illumination light source, the spatial light modulator loads a preset two-dimensional code, incident light emitted by the white light source passes through the spatial light modulator to generate two-dimensional structure light, the two-dimensional structure light is subjected to macro-pixel segmentation and wavelength-position coding by the array filter to generate structure light capable of simultaneously collecting spatial position information and spectral information, and the spatial position information and the spectral information are collected simultaneously;
the detection module consists of a convergent lens and a single-point detector, the convergent lens is used for collecting diffuse reflection light of a scene to be detected after the scene is illuminated by the structured light, and the single-point detector is used for detecting the intensity value of the diffuse reflection light;
and the reconstruction module recovers the spectrum cube of the scene to be detected from the detected intensity value by using a reconstruction algorithm to realize single-pixel multispectral imaging.
8. A single-pixel multispectral imaging device based on macropixel segmentation, comprising:
the structure light module is composed of a white light source, an array filter and a spatial light modulator, wherein the white light source is used as an illumination light source, the array filter is used for carrying out macro-pixel segmentation and wavelength-position coding on incident light emitted by the white light source, then the spatial light modulator loaded with preset two-dimensional codes is used for carrying out two-dimensional coding on the incident light, structure light capable of simultaneously collecting spatial position information and spectral information is generated, and the spatial position information and the spectral information are collected simultaneously;
the detection module consists of a convergent lens and a single-point detector, the convergent lens is used for collecting diffuse reflection light of a scene to be detected after the scene is illuminated by the structured light, and the single-point detector is used for detecting the intensity value of the diffuse reflection light;
and the reconstruction module recovers the spectrum cube of the scene to be detected from the detected intensity value by using a reconstruction algorithm to realize single-pixel multispectral imaging.
9. The apparatus according to claim 7 or 8, wherein the predetermined two-dimensional code loaded on the spatial light modulator is designed based on macro-pixel division, each macro-pixel region comprises a plurality of sub-regions, each sub-region corresponds to a different position in the macro-pixel region, each sub-region is configured to allow or not allow light to enter the subsequent optical system, and all sub-region codes in the same position in the macro-pixel region constitute a two-dimensional sub-code.
10. The device of claim 7 or 8, wherein the array filter is formed by an arrangement of macro-pixel cells, each macro-pixel cell comprising a plurality of sub-regions, different sub-regions allowing light of different wavelengths to pass through; therefore, one macro-pixel of the array filter can be used for collecting information of a plurality of spectral channels corresponding to the pixel.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110184220.4A CN113048907B (en) | 2021-02-08 | 2021-02-08 | Single-pixel multispectral imaging method and device based on macro-pixel segmentation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110184220.4A CN113048907B (en) | 2021-02-08 | 2021-02-08 | Single-pixel multispectral imaging method and device based on macro-pixel segmentation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113048907A true CN113048907A (en) | 2021-06-29 |
CN113048907B CN113048907B (en) | 2022-04-22 |
Family
ID=76509034
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110184220.4A Active CN113048907B (en) | 2021-02-08 | 2021-02-08 | Single-pixel multispectral imaging method and device based on macro-pixel segmentation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113048907B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114441035A (en) * | 2021-12-31 | 2022-05-06 | 复旦大学 | Multispectral imaging method and device based on high-speed adjustable multicolor LED light source |
CN115144075A (en) * | 2022-06-30 | 2022-10-04 | 北京理工大学 | High-speed spectral imaging method and device |
WO2023182237A1 (en) * | 2022-03-24 | 2023-09-28 | パナソニックIpマネジメント株式会社 | Distance measuring device |
WO2023193712A1 (en) * | 2022-04-06 | 2023-10-12 | 华为技术有限公司 | Multispectral processing method and multispectral detection apparatus |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6480273B1 (en) * | 2000-05-10 | 2002-11-12 | Trw Inc. | Multispectral imaging system and method |
CN1900741A (en) * | 2005-11-18 | 2007-01-24 | 北京航空航天大学 | High spectrum full polarization three dimension imaging integrate detecting system |
CN101387552A (en) * | 2008-10-28 | 2009-03-18 | 重庆大学 | Optical spectrum detecting method and spectrometer based on space light modulator |
CN102175323A (en) * | 2011-01-31 | 2011-09-07 | 重庆大学 | Hadamard transform imaging spectrometer based on bridge type grating light modulator |
CN102914367A (en) * | 2012-10-25 | 2013-02-06 | 浙江大学 | Multispectral imaging device and multispectral imaging method based on compressed sensing |
CN103822577A (en) * | 2014-03-13 | 2014-05-28 | 中国电子科技集团公司第三十八研究所 | Single-pixel terahertz holographic imaging device and method |
US20150229851A1 (en) * | 2014-02-12 | 2015-08-13 | Xerox Corporation | Hyperspectral single pixel imager with fabry perot filter |
CN106404171A (en) * | 2016-10-21 | 2017-02-15 | 北京理工大学 | Liquid crystal hyperspectral calculation imaging measurement device and method of three-dimensional encoding |
CN106679808A (en) * | 2016-12-26 | 2017-05-17 | 中国科学院合肥物质科学研究院 | Correlated imaging system and method based on compressed spectrum |
CN108801457A (en) * | 2018-03-27 | 2018-11-13 | 浙江大学 | Three-dimensional collection of illustrative plates based on the design of coded sample plate and second energy about beam alignment obtains and method for reconstructing |
CN108895982A (en) * | 2018-07-03 | 2018-11-27 | 清华大学 | High-speed structures light generating device and three-dimensional single pixel imaging system |
CN109791073A (en) * | 2016-04-29 | 2019-05-21 | 西利奥斯技术公司 | Multispectral imaging equipment |
CN109984723A (en) * | 2019-05-15 | 2019-07-09 | 中国科学院宁波工业技术研究院慈溪生物医学工程研究所 | The multispectral fundus imaging equipment of hand-held and system |
CN110175971A (en) * | 2019-05-27 | 2019-08-27 | 大连海事大学 | A kind of deep learning image reconstructing method of multispectral single pixel imaging |
CN110186566A (en) * | 2019-05-30 | 2019-08-30 | 上海交通大学 | Two-dimentional true temperature field imaging method and system based on the multispectral thermometric of light-field camera |
CN110425986A (en) * | 2019-07-17 | 2019-11-08 | 北京理工大学 | Three-dimensional computations imaging method and device based on single pixel sensor |
CN111047681A (en) * | 2019-11-07 | 2020-04-21 | 北京理工大学 | Single-pixel three-dimensional end-to-end reconstruction method and device based on deep learning |
CN111198037A (en) * | 2020-01-12 | 2020-05-26 | 杭州电子科技大学 | Rapid multispectral imaging device |
US20200370959A1 (en) * | 2019-05-20 | 2020-11-26 | Centre National De La Recherche Scientifique | Time-resolved hyper-spectral single-pixel imaging |
CN111999616A (en) * | 2020-08-31 | 2020-11-27 | 南京工程学院 | Single-pixel solar blind ultraviolet imaging system based on space coding mask plate |
-
2021
- 2021-02-08 CN CN202110184220.4A patent/CN113048907B/en active Active
Patent Citations (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6480273B1 (en) * | 2000-05-10 | 2002-11-12 | Trw Inc. | Multispectral imaging system and method |
CN1900741A (en) * | 2005-11-18 | 2007-01-24 | 北京航空航天大学 | High spectrum full polarization three dimension imaging integrate detecting system |
CN101387552A (en) * | 2008-10-28 | 2009-03-18 | 重庆大学 | Optical spectrum detecting method and spectrometer based on space light modulator |
CN102175323A (en) * | 2011-01-31 | 2011-09-07 | 重庆大学 | Hadamard transform imaging spectrometer based on bridge type grating light modulator |
CN102914367A (en) * | 2012-10-25 | 2013-02-06 | 浙江大学 | Multispectral imaging device and multispectral imaging method based on compressed sensing |
US20150229851A1 (en) * | 2014-02-12 | 2015-08-13 | Xerox Corporation | Hyperspectral single pixel imager with fabry perot filter |
CN103822577A (en) * | 2014-03-13 | 2014-05-28 | 中国电子科技集团公司第三十八研究所 | Single-pixel terahertz holographic imaging device and method |
CN109791073A (en) * | 2016-04-29 | 2019-05-21 | 西利奥斯技术公司 | Multispectral imaging equipment |
CN106404171A (en) * | 2016-10-21 | 2017-02-15 | 北京理工大学 | Liquid crystal hyperspectral calculation imaging measurement device and method of three-dimensional encoding |
CN106679808A (en) * | 2016-12-26 | 2017-05-17 | 中国科学院合肥物质科学研究院 | Correlated imaging system and method based on compressed spectrum |
CN108801457A (en) * | 2018-03-27 | 2018-11-13 | 浙江大学 | Three-dimensional collection of illustrative plates based on the design of coded sample plate and second energy about beam alignment obtains and method for reconstructing |
CN108895982A (en) * | 2018-07-03 | 2018-11-27 | 清华大学 | High-speed structures light generating device and three-dimensional single pixel imaging system |
CN109984723A (en) * | 2019-05-15 | 2019-07-09 | 中国科学院宁波工业技术研究院慈溪生物医学工程研究所 | The multispectral fundus imaging equipment of hand-held and system |
US20200370959A1 (en) * | 2019-05-20 | 2020-11-26 | Centre National De La Recherche Scientifique | Time-resolved hyper-spectral single-pixel imaging |
CN110175971A (en) * | 2019-05-27 | 2019-08-27 | 大连海事大学 | A kind of deep learning image reconstructing method of multispectral single pixel imaging |
CN110186566A (en) * | 2019-05-30 | 2019-08-30 | 上海交通大学 | Two-dimentional true temperature field imaging method and system based on the multispectral thermometric of light-field camera |
CN110425986A (en) * | 2019-07-17 | 2019-11-08 | 北京理工大学 | Three-dimensional computations imaging method and device based on single pixel sensor |
US20210021799A1 (en) * | 2019-07-17 | 2021-01-21 | Beijing Institute Of Technology | Three-dimensional computational imaging method and apparatus based on single-pixel sensor, and non-transitory computer-readable storage medium |
CN111047681A (en) * | 2019-11-07 | 2020-04-21 | 北京理工大学 | Single-pixel three-dimensional end-to-end reconstruction method and device based on deep learning |
CN111198037A (en) * | 2020-01-12 | 2020-05-26 | 杭州电子科技大学 | Rapid multispectral imaging device |
CN111999616A (en) * | 2020-08-31 | 2020-11-27 | 南京工程学院 | Single-pixel solar blind ultraviolet imaging system based on space coding mask plate |
Non-Patent Citations (2)
Title |
---|
YE, ZHIYUAN: "Secured regions of interest (SROIs) in single-pixel imaging", 《SCIENTIFIC REPORTS》 * |
王金成: "基于压缩感知的光谱成像研究", 《中国优秀硕士学位论文全文数据库(电子期刊)基础科学辑》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114441035A (en) * | 2021-12-31 | 2022-05-06 | 复旦大学 | Multispectral imaging method and device based on high-speed adjustable multicolor LED light source |
WO2023182237A1 (en) * | 2022-03-24 | 2023-09-28 | パナソニックIpマネジメント株式会社 | Distance measuring device |
WO2023193712A1 (en) * | 2022-04-06 | 2023-10-12 | 华为技术有限公司 | Multispectral processing method and multispectral detection apparatus |
CN115144075A (en) * | 2022-06-30 | 2022-10-04 | 北京理工大学 | High-speed spectral imaging method and device |
CN115144075B (en) * | 2022-06-30 | 2023-08-29 | 北京理工大学 | High-speed spectrum imaging method and device |
Also Published As
Publication number | Publication date |
---|---|
CN113048907B (en) | 2022-04-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113048907B (en) | Single-pixel multispectral imaging method and device based on macro-pixel segmentation | |
Liang | Punching holes in light: recent progress in single-shot coded-aperture optical imaging | |
RU2653772C1 (en) | System for forming broadband hyperspectral image based on compressible probing with a random diffraction grating | |
Cao et al. | A prism-mask system for multispectral video acquisition | |
CN103363924B (en) | A kind of three-dimensional computations ghost imaging system of compression and method | |
US20170026633A1 (en) | Coded access optical sensor | |
US11175489B2 (en) | Smart coded access optical sensor | |
CN107817641B (en) | Single-pixel camera system based on multimode optical fiber | |
EP3129813B1 (en) | Low-power image change detector | |
WO2005010799A2 (en) | Optical encoding and reconstruction | |
CN105116542B (en) | A kind of double vision field computation relevance imaging system and method | |
CN102914367A (en) | Multispectral imaging device and multispectral imaging method based on compressed sensing | |
CN114279564B (en) | Parallel compressed sensing computed tomography spectrometer and imaging spectrum reconstruction method thereof | |
CN114721006B (en) | High-precision single-pixel imaging method and system | |
US20240019714A1 (en) | Spectral polarization integrated imaging system based on spatial dimension coding and design method thereof | |
CN108896183A (en) | Aperture encoded-polarization optical spectrum imaging device | |
CN114659634A (en) | Miniature snapshot type compressed spectrum imaging detection device and detection method | |
CN114264370B (en) | Compressed sensing computed tomography spectrometer system and imaging method | |
KR101986998B1 (en) | Hyperspectral Imaging Device | |
CN114710615B (en) | Efficient single-pixel imaging method and system | |
Ji et al. | Separating reflective and fluorescent components for dynamic scenes | |
JP2022546720A (en) | Optical device and method | |
CN117091700B (en) | System for realizing hyperspectral imaging by using non-array detector | |
CN112866532B (en) | Compact double-optical-path single-pixel imaging system and uneven light source correction method | |
CN116156144B (en) | Integrated system and method for hyperspectral information acquisition and transmission |
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 |