Wide angle spectral imaging method and system

Abstract

The application designs a wide-angle spectral imaging method and system, which comprises the following steps: S1: constructing an optical filter array and a photoelectric sensor array; S2: obtaining spectral data of each optical filter under different incident angles to form a matrix containing multiple different calibration angle spectral information; S3: acquiring the spectrum of the measured target at an incident angle of θ, transmitting to the photoelectric sensor array after differential modulation of the optical filter array, and generating an intensity signal from the angle θ by the photoelectric sensor; S4: selecting an observation matrix corresponding to the calibration angle equal to or closest to the angle θ in the matrix, and reconstructing the spectrum of the measured target by using a spectral reconstruction algorithm. The application realizes high-resolution spectral imaging of the measured target in a wide-angle / full-angle range by constructing a suitable optical filter array to differentially modulate the spectrum of the measured target and combining the reconstruction algorithm to reconstruct the modulated spectral information.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of spectral imaging technology, and particularly relates to a wide-angle spectral imaging method and system. BACKGROUND

[0002] Spectral imaging technology can obtain the transmission spectrum, reflection spectrum or absorption spectrum information of a measured target at different wavelengths or continuous wave bands, and can also provide spatial image information of the measured target. At present, spectral imaging technology has been widely applied to navigation and remote sensing observation, space exploration, environmental monitoring, biological and medical diagnosis and monitoring, and plays an important role in archaeology, forensic science, pharmaceutical industry, astronomy, agriculture, and military and national defense.

[0003] With the rapid development of computer and intelligent algorithm technology, the combination of spectral imaging technology and intelligent algorithm has become one of the important directions of the development of current spectral imaging systems. Compared with traditional spectral imaging technology, the spectral imaging combined with intelligent algorithm improves the efficiency and accuracy of spectral and spatial data acquisition, reduces the cost, and weakens the limitation of instrument volume on spectral resolution; at the same time, the computational spectral imaging technology can obtain rich spectral information of the measured target, break through the limitation of the number of optical filters in the traditional spectral imaging system on resolution and wave band range, and realize extremely high resolution.

[0004] In the prior art, the spectral imaging technology combined with intelligent algorithm usually obtains the spectral information and spatial image information of the measured target under normal incidence condition, and it is difficult to obtain good imaging effect when obtaining the spectral information of the measured target in a wide-angle incidence range. Although the prior art proposes to use an angle-insensitive plasmonic metasurface array to realize spectral imaging within a certain angle range, it still cannot realize high-resolution spectral imaging in a super large angle or even full angle. SUMMARY

[0005] The present application aims to solve the above technical problems, and provides a wide-angle spectral imaging method and system. The spectral data of a measured target in a wide-angle or full-angle range is differentially modulated by constructing an optical filter array, and an intelligent algorithm is combined to realize spectral reconstruction, thereby realizing high-resolution spectral imaging in a super large angle or even full angle.

[0006] To achieve the above-mentioned purposes, the technical solution adopted by the present application is as follows:

[0007] A wide-angle spectral imaging method, comprising the following steps:

[0008] S1: selecting an optical filter to construct an optical filter array, the optical filter array being used for differentially modulating input spectral information;

[0009] The photoelectric sensor array is connected with the optical filter array and is used for photoelectric signal conversion.

[0010] S2: Obtain the spectral data of each optical filter in the optical filter array under different incident angles, form a matrix T containing a plurality of different calibration angle spectral information, and each calibration angle in the matrix T corresponds to an observation matrix;

[0011] S3: Obtain the spectral information of the measured target at an incident angle of θ, make it pass through the optical filter array, use the optical filter array to differentially modulate and transmit it to the photoelectric sensor array, and generate an intensity signal I from the angle θ by the photoelectric sensor array θ ;

[0012] S4: Select the calibration angle θ corresponding to the observation matrix i in the matrix T which is equal to or closest to the angle θ and reconstruct the spectrum of the measured target based on the matrix and the intensity signal I θ .

[0013] In some embodiments of the present application, step S1 further comprises the following steps:

[0014] Select one or more predetermined correlation coefficients as the evaluation standard of the optical filter array;

[0015] Set the first correlation coefficient threshold and the second correlation coefficient threshold based on the type of the predetermined correlation coefficient, and the first correlation coefficient threshold and the second correlation coefficient threshold can be the same or different;

[0016] Calculate the correlation coefficient between each optical filter in the optical filter array;

[0017] Adjust the composition material of each optical filter in the optical filter array and / or the arrangement mode of the basic unit in the optical filter, so that the correlation coefficient of any two optical filters in the optical filter array is not greater than the first correlation coefficient threshold, and the average correlation coefficient of each optical filter in the optical filter array is not greater than the second correlation coefficient threshold.

[0018] In some embodiments of the present application, the predetermined correlation coefficient is one or a combination of Pearson correlation coefficient, Spearman correlation coefficient and Kendall correlation coefficient.

[0019] In some embodiments of the present application, step S1 further comprises the following steps:

[0020] Set the correlation coefficient difference threshold based on the type of the predetermined correlation coefficient;

[0021] dividing the wavelength range of the optical filter into multiple wavebands;

[0022] respectively calculating the correlation coefficient between each optical filter in each waveband;

[0023] adjusting the composition material of each optical filter in the optical filter array and / or the arrangement mode of the basic unit in the optical filter, so that the difference of the correlation coefficient of any two optical filters in the optical filter array in each waveband is less than the pre-set difference threshold.

[0024] In some embodiments of the present application, the predetermined spectral reconstruction algorithm is one or more of a compressed sensing algorithm, a deep learning algorithm, a machine learning algorithm, a dictionary learning algorithm, a LASSO algorithm, an iterative shrinkage threshold algorithm, a ridge regression algorithm, a semi-quadratic splitting algorithm, a TV algorithm, a neural network algorithm, a network algorithm based on Transformer, a network algorithm based on full connection, a L0 norm reconstruction algorithm, a L1 norm reconstruction algorithm and a L2 norm reconstruction algorithm.

[0025] In some embodiments of the present application, the method for reconstructing the spectrum of the measured target by using the predetermined spectral reconstruction algorithm specifically comprises the following steps:

[0026] S41: selecting a sparse transformation matrix ψ based on the predetermined spectral reconstruction algorithm, so that wherein x is a sparse vector, s is the spectrum to be reconstructed, is a perception matrix;

[0027] S42: calculating the value of the sparse vector x, the sparse vector x satisfying the equation:

[0028]

[0029] wherein ∈ is a positive number;

[0030] S43: calculating the spectrum to be reconstructed s based on the value of the sparse vector x, and the calculation formula of the spectrum to be reconstructed s is:

[0031] s=ψx.

[0032] In some embodiments of the present application, the method further comprises the following steps:

[0033] S5: dividing the wavelength range of the spectrum of the measured target into multiple wavebands, respectively performing spectral imaging on the spectral information of the measured target at an incident angle of θ in each waveband, and merging the spectral imaging results of each waveband.

[0034] Some embodiments of the present application further provide a spectral imaging system for implementing a wide-angle spectral imaging method, which comprises an optical filter array and a photoelectric sensor array;

[0035] The optical filter array is laid on the front side of the photoelectric sensor array, and each optical filter in the optical filter array corresponds to one or more photoelectric sensors in the photoelectric sensor array, or each photoelectric sensor in the photoelectric sensor array corresponds to one or more optical filters in the optical filter array;

[0036] The optical filter array is composed of at least one optical filter arrangement for differentiating modulation of input spectral information;

[0037] When the optical filter array is composed of one optical filter, the optical filter obtains spectral data of multiple different azimuth angles by transforming different azimuth angles;

[0038] When the optical filter array is composed of multiple optical filters, the azimuth angles of the optical filters with the same structure in the array are different, and the azimuth angles of the optical filters with different structures can be the same or different;

[0039] When the optical filter array is composed of multiple optical filters, one or more optical filters in the array obtain spectral data of multiple different azimuth angles by transforming different azimuth angles.

[0040] The photoelectric sensor array is composed of at least one photoelectric sensor arrangement for receiving the spectral information differentiated and modulated by the optical filter array and converting the received spectral information into an electrical signal.

[0041] In some embodiments of the present application, the arrangement mode of the optical filters in the optical filter array includes uniform arrangement along the array surface, ring arrangement, linear arrangement and non-uniform interval arrangement; the array surface includes a plane and a non-plane.

[0042] In some embodiments of the present application, the spectral type of the optical filter includes narrowband spectrum, wideband spectrum, discontinuous spectrum or quasi-continuous spectrum composed of multiple narrowband spectra, and composite spectrum formed by combination of a limited number of narrowband spectra and wideband spectrum.

[0043] The present application has the following beneficial effects:

[0044] 1. The present application differentiates and modulates the incident spectrum of the target to be measured by constructing a proper optical filter array, and reconstructs the modulated spectral information by combining a reconstruction algorithm, thereby realizing high-resolution spectral imaging of the target to be measured in a wide-angle / full-angle range;

[0045] 2. The application provides a method for more accurately evaluating and optimizing an optical filter array, by selecting a proper correlation coefficient as an evaluation standard, and adjusting the composition material of each optical filter in the optical filter array or the arrangement mode of the basic unit in the optical filter based on the evaluation standard, so that the light spectrum modulated by the adjusted optical filter array can obtain better imaging effect after calculation;

[0046] 3. The application divides the wavelength range of the measured target spectrum into multiple wave bands, and improves the resolution of the spectral imaging by respectively reconstructing the spectrum and imaging in the multiple wave bands, so as to ensure the uniformity of the spectral imaging resolution in the wideband range; meanwhile, the wideband wave band division method provided by the application can also reduce the number of filters in the optical filter array and the number of detectors in the photoelectric sensor array, thereby effectively reducing the cost of the whole system.

[0047] 4. The optical filter used in the spectral imaging system provided by the application can be sensitive to the light incidence angle or not sensitive to the light incidence angle, and the optical filter which is sensitive to the light incidence angle or not sensitive to the light incidence angle can realize wide-angle or even full-angle spectral imaging. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the specific embodiments of the present application will be described in detail below with reference to the drawings, and other drawings can also be obtained by those skilled in the art without creative labor on the premise of not deviating from the concept of the present application.

[0049] Figure 1 A flow chart of a wide-angle spectral imaging method provided by the application;

[0050] Figure 2 A structural schematic diagram of a wide-angle spectral imaging system provided by the application;

[0051] Figure 3 A comparison diagram of spectral reconstruction results in normal incidence in an embodiment of the application;

[0052] Figure 4 A comparison diagram of spectral reconstruction results in 10° incidence in an embodiment of the application;

[0053] Among them, the reference signs are:

[0054] 1, measured target; 2, light information of the measured target; 3, optical filter array; 4, light signal modulated by the optical filter array; 5, photoelectric sensor array. DETAILED DESCRIPTION

[0055] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is described and explained below in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application. Based on the embodiments provided by the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0056] It should be noted that the terms used herein are only intended to describe specific embodiments, and are not intended to limit the exemplary embodiments according to the present application. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should be understood that the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0057] The embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0058] The technical solutions of the present application are described in detail below in conjunction with specific embodiments and the accompanying drawings.

[0059] As shown in the accompanying drawings, Figure 1 In one illustrative example of a wide-angle spectral imaging method according to the present application, the spectral imaging method covers both spectral reconstruction and spatial imaging, and specifically includes the following steps.

[0060] S1: Select an optical filter to construct an optical filter array 3, which is used to differentially modulate the input spectral information.

[0061] Construct a photoelectric sensor array 5 connected to the optical filter array 3 for photoelectric signal conversion.

[0062] In some embodiments of the present application, step S1 further includes the following steps.

[0063] Select a predetermined correlation coefficient as an evaluation standard for the optical filter array 3, the predetermined correlation coefficient being one or a combination of more than one of Pearson correlation coefficient, Spearman correlation coefficient and Kendall correlation coefficient.

[0064] Set a first correlation coefficient threshold and a second correlation coefficient threshold based on the specific type of the selected predetermined correlation coefficient; wherein the first correlation coefficient threshold and the second correlation coefficient threshold can be the same or different;

[0065] calculating the correlation coefficient between each optical filter in the optical filter array 3;

[0066] adjusting the constituent material of each optical filter in the optical filter array 3 and / or the arrangement of the basic units in the optical filter, so that the correlation coefficient of any two optical filters in the optical filter array 3 is not greater than the first correlation coefficient threshold, and the average correlation coefficient of each optical filter in the optical filter array 3 is not greater than the second correlation coefficient threshold.

[0067] In some embodiments of the present application, step S1 further comprises the following steps.

[0068] setting the predetermined difference threshold based on the type of the predetermined correlation coefficient.

[0069] dividing the wavelength range of the optical filter into multiple wavebands; the wavelength range of each waveband includes both the same and different cases.

[0070] calculating the correlation coefficient between each optical filter in the array in each waveband respectively;

[0071] adjusting the constituent material of each optical filter in the optical filter array 3 and / or the arrangement of the basic units in the optical filter, so that the difference of the correlation coefficient of any two optical filters in the optical filter array 3 in each waveband is less than the predetermined difference threshold.

[0072] It should be noted that in actual application, in order to ensure the uniformity of the resolution of the broadband spectral imaging, the smaller the difference threshold is, the better, and the specific value can be determined based on the spectral imaging index of the measured waveband and application requirements, which is not limited by the present application.

[0073] S2: spectral calibration. The spectral information of each optical filter in the array is calibrated under different incident angles to obtain spectral information data under each angle condition, and there is a certain difference between two adjacent angles.

[0074] The specific implementation method of step S2 is to select multiple different incident angles at a predetermined angle interval within a wide angle / full angle range, for example, selecting an incident angle every 5 degrees or 10 degrees, collecting the spectral information of each optical filter in the array under each incident angle, and forming a matrix T containing multiple different predetermined calibration angle spectral information, each predetermined calibration angle in the matrix T corresponds to an observation matrix where θ i is any selected incident angle, and the maximum value of i is determined by the number of angle intervals within the wide angle / full angle range.

[0075] The spectral information includes a transmission spectrum, a reflection spectrum, and an absorption spectrum. The angular interval mode includes uniform interval and non-uniform interval.

[0076] S3: Obtain the spectral information of the measured target 1 at the incident angle θ, make the spectral information pass through the optical filter array 3, use the optical filter array 3 to differentially modulate the spectral information, obtain the differentially modulated target spectral information, and transmit the target spectral information to the photoelectric sensor array 5, so as to generate the intensity signal I from the angle θ by the photoelectric sensor array 5. θ .

[0077] Specifically, the spectrum of the measured target 1 is set as s M×1 , and each optical filter uses a spectrum calibrated for a corresponding incident angle θ , where M is the number of spectral sampling points, and θ i is the i-th angle calibrated in space.

[0078] The optical filter array 3 is composed of N optical filters, and the spectrum of the optical filter array 3 is written in a matrix form as The photoelectric sensor array 5 receives the light intensity , where a is the receiving efficiency.

[0079] S4: Perform spectral reconstruction and spatial imaging using the angle-calibrated optical filter array 3. In the matrix T, select the calibrated angle θ i corresponding to the angle θ , and based on the observation matrix and the intensity signal I θ , reconstruct the spectrum of the measured target 1 by using a predetermined spectral reconstruction algorithm.

[0080] In some embodiments of the present application, the predetermined spectral reconstruction algorithm is one or more of a compressed sensing algorithm, a deep learning algorithm, a machine learning algorithm, a dictionary learning algorithm, a LASSO algorithm, an iterative shrinkage thresholding algorithm, a ridge regression algorithm, a semi-quadratic splitting algorithm, a TV algorithm, a neural network algorithm, a network algorithm based on Transformer, a network algorithm based on full connection, a L0 norm reconstruction algorithm, a L1 norm reconstruction algorithm, and a L2 norm reconstruction algorithm.

[0081] In some embodiments of the present application, the method for reconstructing the spectrum of the measured target 1 by using the predetermined spectral reconstruction algorithm specifically includes the following steps.

[0082] S41: Select a sparse transformation matrix ψ based on the predetermined spectral reconstruction algorithm, so that , where x is a sparse vector, and s is the spectrum to be reconstructed.

[0083] S42: Calculate the value of the sparse vector x, and the sparse vector x satisfies the equation:

[0084]

[0085] wherein ∈ is a positive number.

[0086] S43: based on the value of the sparse vector x, the spectrum s to be reconstructed is calculated, and the calculation formula of the spectrum s to be reconstructed is:

[0087] s=ψx.

[0088] In some embodiments of the present application, in order to realize higher resolution spectral imaging, the following steps are further included.

[0089] S5: the wavelength range of the measured target spectrum is divided into multiple wavebands, based on the spectral information of the measured target at an incident angle of θ, the optical filter array 3 and the photoelectric sensor array 5 are used to perform spectral imaging in each waveband respectively, and the imaging results of each waveband are merged.

[0090] The merging method can be simple splicing, or weighted merging, interpolation merging or other appropriate mathematical or statistical merging techniques. In actual application, the specific merging method can be determined based on the spectral imaging index of the measured waveband and application requirements, and the present application does not make further limitations.

[0091] In some embodiments of the present application, when the number of optical filters in the optical filter array 3 is 1, different spectral information is obtained by changing the spatial azimuth angle of the filter in step S2.

[0092] Some embodiments of the present application further provide a spectral imaging system for realizing the above-mentioned wide-angle spectral imaging method, which at least includes an optical filter array 3 and a photoelectric sensor array 5.

[0093] The optical filter array 3 is laid on the front side of the photoelectric sensor array 5, each optical filter in the optical filter array 3 corresponds to one or more photoelectric sensors in the photoelectric sensor array 5, and each photoelectric sensor in the photoelectric sensor array 5 can also correspond to one or more optical filters in the optical filter array 3.

[0094] The optical filter array 3 is composed of one optical filter, multiple optical filters with the same structure or multiple optical filters with different structures, and is used for differentiating modulation of input spectral information.

[0095] When the optical filter array 3 is composed of one optical filter, the optical filter obtains spectral data at multiple different azimuth angles by transforming different azimuth angles, which is equivalent to multiple different optical filters.

[0096] When the optical filter array 3 is composed of optical filters of the same structure, the optical filters of the same structure in the array are arranged at different azimuth angles.

[0097] When the optical filter array 3 is composed of optical filters of different structures, each type of optical filter in the array contains a plurality of optical filters of the same structure, and the optical filters of the same structure are arranged at different azimuth angles; the optical filters of different structures in the array can be arranged at the same azimuth angle or at different azimuth angles.

[0098] When the optical filter array 3 is composed of a plurality of optical filters, one or more optical filters in the array are arranged at different azimuth angles by transformation to obtain spectral data at multiple azimuth angles.

[0099] The photoelectric sensor array 5 is composed of one or more photoelectric sensors arranged to receive the spectral information modulated by the optical filter array 3 and convert the received spectral information into an electrical signal.

[0100] In some embodiments of the present application, the arrangement of the optical filters in the optical filter array 3 includes uniform arrangement along the surface of the array, ring arrangement, linear arrangement, and non-uniformly spaced arrangement; the surface of the array includes a plane and a non-plane.

[0101] In some embodiments of the present application, the spectral type of the optical filter includes a narrowband spectrum, a broadband spectrum, a discontinuous or quasi-continuous spectrum composed of multiple narrowband spectra, and a composite spectrum formed by a combination of a limited number of narrowband and broadband spectra.

[0102] In some embodiments of the present application, the photoelectric sensor is a charge-coupled device (CCD), a complementary metal-oxide semiconductor (CMOS), a photomultiplier tube, or any other photoelectric device based on the photoelectric effect that can be used for photoelectric signal conversion.

[0103] In some embodiments of the present application, the optical filter constituting the optical filter array 3 includes a single-layer film, a multi-layer film, a one-dimensional photonic crystal, a two-dimensional photonic crystal, a three-dimensional photonic crystal, a grating, a metamaterial, a metasurface, a quantum dot, a non-flat surface, or any other optical device that has a modulation effect on light intensity.

[0104] In some embodiments of the present application, the optical filter constituting the optical filter array 3 can be sensitive to the incident angle of light or not sensitive to the incident angle of light. Both the optical filter sensitive to the incident angle of light and the optical filter not sensitive to the incident angle of light can achieve wide-angle or even full-angle spectral imaging.

[0105] In some embodiments of the present application, the optical filters constituting the optical filter array 3 can be sensitive to light polarization or not. Both the optical filters sensitive to light polarization and the optical filters not sensitive to light polarization can realize wide-angle or even full-angle spectral imaging.

[0106] Among them, the optical filters sensitive to light polarization can realize wide-angle or even full-angle spectral imaging under specific polarized light conditions, including natural light, TE polarized light, TM polarized light, left or right circularly polarized light, elliptically polarized light and vortex light.

[0107] In some embodiments of the present application, the optical filters constituting the optical filter array 3 can be composed of isotropic and anisotropic materials.

[0108] Specifically, in terms of the types of materials, the optical filters can be composed of dielectric, metal and magnetic materials, or a combination of multiple materials.

[0109] In some embodiments of the present application, the optical filters constituting the optical filter array 3 are composed of non-single materials, and the arrangement mode of each basic unit in the optical filter is one or a combination of periodic arrangement, quasi-periodic arrangement and disordered arrangement. It should be noted that air or high vacuum is also a special material among the non-single materials.

[0110] In some embodiments of the present application, the optical filters constituting the optical filter array 3 can be composed of materials modulated by external fields. Among them, the materials modulated by external fields include nonlinear optical materials modulated by light field, electro-optic materials modulated by electric field, magneto-optic materials modulated by magnetic field, liquid crystal materials, thermo-optic materials modulated by heat, piezoelectric materials modulated by external force and acousto-optic materials modulated by sound, etc. The filters composed of the materials modulated by external fields can obtain rich spectral imaging information by changing the external field conditions.

[0111] The beneficial effects of the present application will be described below in combination with a specific embodiment:

[0112] In the present embodiment, a complementary metal-oxide-semiconductor (CMOS) device is selected as the photoelectric sensor, and the optical filter array 3 composed of different two-dimensional planar photonic crystals is completely laid in front of the photoelectric sensor array 5, which contains 25 planar photonic crystals, each of which is composed of two layers, the lower layer is a quartz substrate with a thickness of 1.5 mm and a refractive index of 1.45, and the upper layer is a two-dimensional planar photonic crystal composed of air holes in a silicon plate arranged periodically along the direction perpendicular to the hole, with a thickness of 175 nm. Among them, the photonic crystal structure involves two lattice structures of square lattice (including square lattice and rectangular lattice) and regular hexagonal lattice, and five hole structures of circular hole, square hole (including square hole and rectangular hole), square hole rotated by 45°, cross-shaped hole and cross-shaped hole rotated by 45°.

[0113] The results of reconstructing the spectrum under normal incidence condition and the spectrum under 10° incidence condition of the measured object using the optical filter array 3 and the photoelectric sensor array 5 respectively are shown in Figs. 8 and 9. Figure 3 and Fig. 9. Figure 4 As shown in Figs. 8 and 9, the dashed line in Figs. 8 and 9 is the actual spectral image of the measured object under incidence, i.e. the original spectrum; the solid line is the reconstructed spectrum obtained by the spectral imaging method provided in the embodiment, which is shown in Figs. 10 and 11. Figure 3 and Fig. 11. Figure 4 As can be clearly seen, in the embodiment, the reconstructed spectrum under normal incidence and the reconstructed spectrum under 10° incidence are both very consistent with the original spectrum, and have very good reconstruction effect. Figure 3 Figure 4 Finally, it should be noted that: the embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same and similar parts of each embodiment can be referred to.

[0114] Finally, it should be noted that: the embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same and similar parts of each embodiment can be referred to.

[0115] The above embodiments are only used to illustrate the technical solutions of the present application but not to limit it; although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the specific implementation of the present application can be modified or some technical features can be replaced by equivalent; without departing from the spirit of the technical solutions of the present application, they should be covered in the technical solution range of the present application.​

Claims

1. A wide-angle spectral imaging method, characterized in that, Includes the following steps: S1: Select optical filters to construct an optical filter array, which is used to differentially modulate the input spectral information; A photoelectric sensor array is constructed, which is connected to the optical filter array for photoelectric signal conversion; S2: Obtain the spectral data of each optical filter in the optical filter array at different incident angles to form a matrix T containing spectral information at multiple different incident angles. Each incident angle in the matrix T corresponds to an observation matrix. S3: Acquire the spectral information of the target at the incident angle θ, perform differential modulation on it in the optical filter array, and then transmit it to the photoelectric sensor array. The photoelectric sensor array generates an intensity signal I from the incident angle θ. θ ; S4: Select a calibration angle θ that is equal to or closest to the incident angle θ in the matrix T. i The corresponding observation matrix And based on the matrix and the intensity signal I θ The spectrum of the target under test is reconstructed using a predetermined spectral reconstruction algorithm.

2. The wide-angle spectral imaging method according to claim 1, characterized in that, Step S1 further includes the following steps: One or more predetermined correlation coefficients are selected as evaluation criteria for the optical filter array; A first correlation coefficient threshold and a second correlation coefficient threshold are set based on the predetermined type of correlation coefficient; the first correlation coefficient threshold and the second correlation coefficient threshold can be the same or different; Calculate the correlation coefficient between each optical filter in the optical filter array; Adjust the constituent materials of each optical filter in the optical filter array and / or the arrangement of the basic units in the optical filter so that the correlation coefficient of any two optical filters in the optical filter array is not greater than the first correlation coefficient threshold, and the average correlation coefficient of each optical filter in the optical filter array is not greater than the second correlation coefficient threshold.

3. The wide-angle spectral imaging method according to claim 2, characterized in that, The predetermined correlation coefficient is one or more combinations of Pearson correlation coefficient, Spearman correlation coefficient and Kendall correlation coefficient.

4. The wide-angle spectral imaging method according to claim 2, characterized in that, Step S1 further includes the following steps: A predetermined difference threshold is set based on the type of the predetermined correlation coefficient; The wavelength range of the optical filter is divided into multiple bands; Calculate the correlation coefficients between each optical filter in each band; Adjust the constituent materials of each optical filter in the optical filter array and / or the arrangement of the basic units in the optical filter so that the difference in the correlation coefficient of any two optical filters in the optical filter array within each band is less than the predetermined difference threshold.

5. The wide-angle spectral imaging method according to claim 1, characterized in that, The predetermined spectral reconstruction algorithm is one or more of the following: compressed sensing algorithm, deep learning algorithm, machine learning algorithm, dictionary learning algorithm, LASSO algorithm, iterative shrinking threshold algorithm, ridge regression algorithm, semi-quadratic splitting algorithm, TV algorithm, neural network algorithm, Transformer-based network algorithm, fully connected network algorithm, L0 norm reconstruction algorithm, L1 norm reconstruction algorithm, and L2 norm reconstruction algorithm.

6. The wide-angle spectral imaging method according to claim 1 or 5, characterized in that, The method for reconstructing the spectrum of the target object using a predetermined spectral reconstruction algorithm specifically includes the following steps: S41: Based on the predetermined spectral reconstruction algorithm, select a sparse transformation matrix ψ such that... Where x is a sparse vector and s is the spectrum to be reconstructed. For the perception matrix; S42: Calculate the values ​​of the sparse vector x, which satisfies the equation: Where ∈ is a positive number; S43: Based on the value of the sparse vector x, calculate the spectrum s to be reconstructed. The formula for calculating the spectrum s to be reconstructed is: s = ψs.

7. The wide-angle spectral imaging method according to claim 1, characterized in that, Further steps include: S5: Divide the wavelength range of the spectrum of the target under test into multiple bands, perform spectral imaging in each band based on the spectral information of the target under test with an incident angle of θ, and merge the spectral imaging results of each band.

8. A wide-angle spectral imaging system for implementing the wide-angle spectral imaging method according to any one of claims 1-7, characterized in that, Including optical filter arrays and photoelectric sensor arrays; The optical filter array is laid on the front side of the photoelectric sensor array, and each optical filter in the optical filter array is connected to one or more photoelectric sensors in the photoelectric sensor array. The optical filter array consists of at least one optical filter and is used to differentially modulate the input spectral information. When the optical filter array is composed of a single optical filter, the optical filter acquires spectral data at multiple different azimuth angles by changing different azimuth angles; When the optical filter array is composed of multiple optical filters, the azimuth angles of optical filters with the same structure in the array are different, and the azimuth angles of optical filters with different structures can be the same or different. When the optical filter array is composed of multiple optical filters, one or more optical filters in the array can acquire spectral data at multiple different azimuth angles by changing different azimuth angles; The optical detector array is composed of at least one photoelectric sensor arranged together to receive spectral information after differential modulation by the optical filter array, and to convert the received spectral information into an electrical signal.

9. The wide-angle spectral imaging system according to claim 8, characterized in that, The optical filters in the optical filter array can be arranged in various ways, including uniform arrangement, circular arrangement, linear arrangement, and non-uniform spacing along the array surface; the array surface can be planar or non-planar.

10. The wide-angle spectral imaging system according to claim 8 or 9, characterized in that, The optical filter's spectral types include narrowband spectrum, broadband spectrum, discontinuous or quasi-continuous spectrum composed of multiple narrowband spectra, and composite spectrum formed by a combination of a finite number of narrowband and broadband spectra.

Images