A light field regulation detection method and device based on computational spectral reconstruction

Abstract

The application discloses a kind of light field regulation detection method and device based on computing spectral reconstruction, belong to spectral analysis technical field.The method is by introducing medium-metal composite nanostructure, integrates metal nanoisland array and silicon nitride nanocolumn array on quartz substrate, realizes electric field enhancement using the local surface plasmon resonance effect excited by metal nanoisland array, simultaneously realizes wide spectral range phase coding by medium nanocolumn array;Super surface chip is integrated with single point detector, combines pre-calibrated super surface sensing matrix A and enhancement factor matrix E, and reconstructs spectrum using compressed sensing or Tikhonov regularization algorithm.The application keeps the advantages of miniaturization and no moving parts, realizes directional high sensitivity detection to weak signal, and signal-to-noise ratio is improved.

Description

Technical Field

[0001] This invention belongs to the field of spectral analysis technology, and in particular relates to a light field modulation detection method and device based on computational spectral reconstruction. Background Technology

[0002] As a core method for substance detection, spectroscopic analysis technology has long faced the challenge of balancing performance and size in instrument development. Traditional spectrometers rely on dispersive elements such as gratings and prisms to achieve spectral separation through spatial dispersion. However, their inherent mechanism suffers from the classic "luminous flux-resolution" contradiction: to ensure spectral resolution, the entrance slit width must be narrowed, but this leads to a significant attenuation of incident light energy, greatly limiting the improvement of instrument sensitivity. This physical limitation makes traditional spectrometers perform poorly when detecting weak signals.

[0003] With the advancement of micro-nano fabrication technology, the miniaturization of spectrometers has become an important trend. However, existing spectrometers based on microelectromechanical systems (MEMS) or integrated optical waveguides often sacrifice sensitivity in pursuit of compact structures. Due to the reduction in device size leading to shorter optical path lengths or smaller light-transmitting areas, their ability to capture weak signals is more limited, which severely restricts the application of miniature spectrometers in fields such as rapid on-site detection and portable detection.

[0004] Metasurfaces, as artificially designed two-dimensional nanostructure arrays, offer a new path for the miniaturization of spectrometers due to their powerful electromagnetic wavefront modulation capabilities. Through the dispersive or encoding functions of metasurfaces, miniaturized spectrometers without moving parts can be realized. However, existing metasurface spectrometer solutions mostly focus on structural miniaturization, such as using metasurfaces based on wavelength-dependent phase modulation mechanisms to replace traditional dispersive elements, without fundamentally solving the essential problem of insufficient ability to capture weak signals.

[0005] Therefore, developing a novel spectrometer that retains the miniaturization advantages of metasurface spectrometers while significantly improving detection sensitivity through innovative physical mechanisms has become a pressing technical challenge in this field. This invention presents an innovative solution to this problem.

[0006] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a light field modulation detection method and device based on computational spectral reconstruction, which can solve the problem of insufficient capture capability of existing metasurface spectrometers for weak signals.

[0008] To achieve the above objectives, this invention provides a light field modulation detection method based on computational spectral reconstruction, comprising the following steps:

[0009] Step S1: The optical signal to be tested is incident perpendicularly onto the metasurface chip, wherein the metasurface chip comprises a composite nanostructure;

[0010] Step S2: The metasurface chip performs field localization enhancement and spectral encoding on the optical signal to be measured;

[0011] Step S3: Convert the encoded optical signal to be tested into an electrical signal using a detector;

[0012] Step S4: Construct a reconstruction algorithm with an embedded physical enhancement model, call the pre-stored metasurface sensing matrix A and enhancement factor matrix E, calculate and output the spectrum x of the optical signal to be measured based on the electrical signal.

[0013] Optionally, in step S1, the composite nanostructure of the metasurface chip includes a "dielectric-metal" composite nanostructure unit, which includes an integrated array of dielectric nanopillars and an array of metal nanoislands.

[0014] Optionally, in step S2, the interaction of light with the composite nanostructure specifically includes:

[0015] The dielectric nanopillar array is a silicon nitride nanopillar array. Through its geometric layout and periodic structure, it performs broadband phase encoding on the incident light. The phase encoding strategy is based on a pre-designed sensing matrix A to achieve mixing and modulation of optical signals of different wavelengths to support subsequent calculation and reconstruction.

[0016] The shape, size, and position parameters of the metal nanoisland array are determined according to the core optimization objective, which is aimed at the target signal band for a specific application, including the heavy metal characteristic peak in water quality detection. The heavy metal characteristic peak is optimized through electromagnetic simulation or experimental calibration to excite local surface plasmon resonance in this band.

[0017] The local surface plasmon resonance generates a local electric field enhancement, and the enhancement effect is quantified by the enhancement factor matrix E. Furthermore, the phase encoding of the dielectric nanopillar array and the field enhancement of the metal nanoisland array work together to form an integrated "capture-enhancement-encoding" process, which directly amplifies weak light signals, thereby improving the sensitivity of spectral testing.

[0018] Optionally, in step S4, the mathematical model used by the reconstruction algorithm is:

[0019] y = (A · E)x + n;

[0020] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and n is the system noise vector;

[0021] The reconstruction algorithm recovers the spectral vector x from the measured value vector y by solving the above model.

[0022] Optionally, the enhancement factor matrix E is a diagonal matrix;

[0023] Each element on the diagonal of the matrix corresponds to an enhancement factor for a spectral channel, the value of which is derived from electromagnetic simulation calculations or experimental calibration results of the “dielectric-metal” composite nanostructure unit.

[0024] Optionally, in step S4, the spectral reconstruction algorithm is based on the compressed sensing path, and its mathematical model is as follows:

[0025] ;

[0026] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and λ is the regularization parameter. This algorithm is suitable for scenarios where the measured spectrum has sparse characteristics, such as gas detection or emission spectra of fluorescent markers.

[0027] Optionally, in step S4, the spectral reconstruction algorithm is based on the Tikhonov regularization path, and its mathematical model is as follows:

[0028] ;

[0029] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be obtained, and λ is the regularization parameter. This algorithm is suitable for scenarios where the measured spectrum is a wide-spectral continuous variation, such as blackbody radiation spectrum or solution absorption spectrum, to improve the noise resistance and stability of the solution.

[0030] Optionally, the composite nanostructure of the metasurface chip is achieved through the following device fabrication process, specifically including:

[0031] A substrate is provided, wherein the substrate is a quartz substrate;

[0032] Metal nanoisland arrays are fabricated on the substrate by electron beam evaporation and exfoliation processes, wherein the metal nanoisland arrays are used to generate local field enhancement effects;

[0033] An isolation layer, made of silicon nitride, is grown on the metal nanoisland array using atomic layer deposition technology. The isolation layer is used for electrical isolation and optical modulation.

[0034] A dielectric nanopillar array, which is a silicon nitride nanopillar array, is fabricated on the isolation layer using electron beam lithography and dry etching processes to form a hybrid metasurface. The metal nanoisland array and the dielectric nanopillar array work together to simultaneously provide field localization enhancement and spectral encoding functions, thereby achieving high-sensitivity capture and processing of weak light signals.

[0035] A second aspect of the present invention also provides a light field modulation detection device based on computational spectral reconstruction, for implementing the above-described light field modulation detection method based on computational spectral reconstruction, comprising:

[0036] The metasurface chip, consisting of an array of metal nanoislands and an array of silicon nitride nanopillars on a quartz substrate, is used to receive the optical signal to be measured and to achieve field localization enhancement and spectral encoding.

[0037] The detector, integrated with a metasurface chip, is used to receive enhanced and encoded optical signals and convert them into electrical signals;

[0038] The signal processing unit, connected to the detector, is used to run a reconstruction algorithm to calculate the spectral map based on the sensing matrix and the enhancement factor matrix.

[0039] Optionally, the detector is a single-point detector or a small detector array, wherein the single-point detector includes a photodiode, and the small detector array is used to receive optical signals in parallel or sequentially.

[0040] The detector is configured to convert optical signals into electrical signals and generate a measurement vector y output by the detector;

[0041] The metasurface chip and the detector are integrated into a compact structure. The packaging process includes directly bonding or fixing the metasurface chip and the detector, and electrically connecting them to the signal processing unit to ensure that the device has no moving parts, high stability, and miniaturization advantages.

[0042] The light signal to be measured is incident perpendicularly onto the metasurface chip, and after transmission, it is directly coupled into the detector. The direct coupling method includes an optical path alignment design to minimize optical loss and improve signal transmission efficiency, thereby enhancing the device's ability to capture weak light signals.

[0043] Compared with the prior art, the optical field modulation detection method and apparatus based on computational spectral reconstruction according to the present invention have the following advantages or beneficial effects;

[0044] This invention utilizes the localized surface plasmon resonance effect excited by a metal nanoisland array to directly amplify the light field intensity of a specific target wavelength component in the measured optical signal at the physical level. The enhancement effect is quantized by the enhancement factor matrix E. This field-localized enhancement mechanism solves the problem of detecting weak light signals in miniature spectrometers, avoids the contradiction between light flux and resolution in traditional schemes, and improves sensitivity by an order of magnitude.

[0045] While introducing field enhancement functionality, this invention fully retains all the advantages of computational metasurface spectrometers, such as miniaturization, absence of moving parts, and high stability. After the metasurface chip and detector are integrated and packaged, the device is compact and mechanically stable, suitable for portable or embedded applications, such as devices used in field environmental monitoring or medical diagnostics. By synergistically optimizing the field enhancement characteristics of the metal nanoisland array and the phase encoding function of the dielectric nanopillar array, targeted high-sensitivity detection of specific material characteristic peaks can be achieved.

[0046] This invention can be combined with various types of computational metasurfaces, including metasurface structures based on wavelength multiplexing, polarization multiplexing, or angle multiplexing. This adaptability ensures the broad applicability of the technology, allowing for flexible adjustments to meet different application requirements such as multispectral imaging or polarization-sensitive detection, thereby enhancing the invention's market potential and technological scalability. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating a light field modulation detection method based on computational spectral reconstruction according to an embodiment of the present invention. Figure 1 ;

[0048] Figure 2 This is a flowchart illustrating a light field modulation detection method based on computational spectral reconstruction according to an embodiment of the present invention. Figure 2 ;

[0049] Figure 3 This is a schematic diagram of a "dielectric-metal" composite nanostructure unit according to an embodiment of the present invention;

[0050] Figure 4 This is a simulation diagram of the electric field enhancement effect of a "dielectric-metal" composite nanostructure unit according to an embodiment of the present invention;

[0051] Figure 5 A comparison of the spectral response functions of traditional computational metasurfaces and the sensitivity-optimized metasurface of this invention;

[0052] Figure 6 This is a flowchart of a spectral reconstruction algorithm according to an embodiment of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0055] It should be understood that in the various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0056] It should be understood that in this invention, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0057] It should be understood that in this invention, "multiple" refers to two or more. "And / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, "and / or B" can represent: A existing alone, A and B existing simultaneously, and B existing alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "Contains A, B, and C", "Contains A, B, and C" means that all three A, B, and C are contained; "Contains A, B, or C" means that one of A, B, and C is contained; "Contains A, B, and / or C" means that any one, two, or three of A, B, and C are contained.

[0058] It should be understood that in this invention, "B corresponding to A", "B corresponding to A", "A and B correspond", or "B and A correspond" means that B is associated with A, and B can be determined based on A. Determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information. Matching A and B is defined as a similarity between A and B that is greater than or equal to a preset threshold.

[0059] Depending on the context, "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection."

[0060] The technical solution of the present invention will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and there may be some embodiments for the same or similar concepts or processes, which will not be described again.

[0061] like Figure 1 and Figure 2 As shown, a light field modulation detection method based on computational spectral reconstruction according to a preferred embodiment of the present invention includes the following steps:

[0062] Step S1: The optical signal to be tested is incident perpendicularly onto the metasurface chip, which includes a composite nanostructure. It should be noted that, unlike the traditional design that aims to "uniformly distinguish wavelengths", this embodiment of the invention takes "achieving the highest system signal-to-noise ratio in a preset weak target signal band" as the core optimization goal.

[0063] Step S2: The metasurface chip performs field localization enhancement and spectral encoding on the light signal under test. It should be noted that when the target wavelength in the light under test matches the local surface plasmon resonance wavelength of the metal nanostructure, it excites collective oscillations of electrons on the metal surface, forming an extremely strong local electromagnetic field in the gaps or tips of the nanostructure. This amplifies the originally weak target signal intensity by several orders of magnitude, fundamentally improving the signal's detectability. Different wavelengths of light experience unique phase delays due to optical path differences when passing through the nanopillars. The encoded light field carries a "wavelength characteristic tag," providing "informative measurement data" for subsequent spectral reconstruction.

[0064] Step S3: Convert the encoded optical signal to be tested into an electrical signal using a detector;

[0065] Step S4: Construct a reconstruction algorithm with an embedded physical enhancement model, call the pre-stored metasurface sensing matrix A and enhancement factor matrix E, calculate and output the spectrum x of the optical signal to be measured based on the electrical signal.

[0066] It should be noted that traditional reconstruction algorithms only solve for x based on the measurement matrix, without utilizing the physical prior of field enhancement. This invention constructs a physical enhancement model using the metasurface sensing matrix A and the enhancement factor matrix E, thus incorporating the "enhancing effect of field enhancement on the signal" into the reconstruction process.

[0067] If the target spectrum x is sparsity, a compressed sensing path is adopted, which utilizes the characteristics of "sparse prior + concentrated signal energy after field enhancement" to accurately recover x at a low sampling rate.

[0068] If the spectrum is a continuous and smooth signal, the Tikhonov regularization path is adopted to suppress noise through L2 norm constraints, while field enhancement is used to improve signal dominance, so that the reconstruction result is closer to the real spectrum.

[0069] The final output x is a high signal-to-noise ratio spectrum that has been physically enhanced and denoised by algorithms, which retains the weak signal characteristics of the target band while suppressing background noise.

[0070] This invention, through field enhancement and signal amplification at the physical layer and noise suppression, combined with compressed sensing / Tikhonov regularization denoising at the algorithm layer, achieves a signal-to-noise ratio (SNR) ≥2 times higher than traditional direct detection, meeting the accuracy requirements of trace analysis (such as heavy metal and biomolecule detection in water). The broad-spectrum phase encoding of the nanopillars preserves spectral details, and the enhanced field enhancement improves the contrast of characteristic peaks, making adjacent characteristic peaks easier to distinguish.

[0071] Furthermore, such as Figure 2 As shown, the light under test first enters a "field-localized enhancement metasurface chip," whose core is a dielectric-metal composite nanostructure array. Utilizing the localized field enhancement effect of the nanostructure, the weak incident light is physically enhanced, and the optical field is controlled through coding design. The optical signal, after metasurface coding and physical enhancement, is received by a single-point photodetector, completing the conversion from optical to electrical signal. The electrical signal output from the single-point photodetector is transmitted to an analog front-end, which includes a transimpedance amplifier and / or a filter. The transimpedance amplifier amplifies the weak electrical signal and compensates for signal attenuation from the photodetector, while the filter removes environmental noise, high-frequency interference, and other invalid signals, retaining the electrical signal in the target frequency band and improving the signal-to-noise ratio. The electrical signal processed by the analog front-end enters an analog-to-digital converter (ADC), converting the continuous analog electrical signal into a discrete digital signal. The digital signal is then sent to a digital signal processor (DSP / FPGA), which simultaneously calls three types of key data from a "pre-stored calibration database" to achieve accurate signal analysis and spectral reconstruction. In addition, the power management and control unit provides stable power and timing control for the aforementioned hardware modules (photodetectors, analog front-end, ADC, DSP / FPGA, etc.) to ensure coordinated system operation. The "reconstructed high signal-to-noise ratio spectrum" obtained after DSP / FPGA processing finally enters the "display / storage / output" stage, which can be visualized through a display screen, recorded through a storage device, or transmitted through an external interface.

[0072] In this embodiment of the invention, in step S1, the composite nanostructure of the metasurface chip includes a "dielectric-metal" composite nanostructure unit, which comprises an integrated array of dielectric nanopillars and an array of metal nanoislands. It should be noted that the structural parameters of the "dielectric-metal" composite nanostructure unit, such as the size of the dielectric nanopillar array and the shape and position of the metal nanoisland array, are determined through electromagnetic simulation and reverse design according to the core optimization objective of this invention. Specifically, the dielectric portion achieves broadband phase encoding, while the integrated metal portion excites local surface plasmon resonances in the target wavelength band, generating extremely strong local electric field enhancement. Therefore, this composite unit is the concrete physical embodiment and realization of the core optimization objective.

[0073] Furthermore, such as Figure 3 As shown, the topmost "Incident Light (including target wavelength)" is the unit input. The large rectangular box in the middle, "Three-dimensional side of the composite nanostructure unit," demonstrates the layered design of the structure, consisting of three layers: top, middle, and bottom. The top layer is a dielectric nanopillar (such as silicon nitride Si3N4); the middle layer is a metal nanoisland (such as gold Au), whose surface electrons can be excited by incident light to generate localized surface plasmon resonance (LSPR)—that is, collective oscillation of electrons on the metal surface, which, after coupling with incident light, generates a high-intensity electric field locally at the nanoscale; the bottom layer is a substrate (such as quartz SiO2), which supports the upper nanostructure without interfering with the propagation and interaction of light. The arrows and gray boxes on the left illustrate the excitation of LSPR and the enhancement of the local electric field. The "Outgoing / Transmitted Light" in the rectangular box on the right is the unit output, used for subsequent high-sensitivity detection. The above unit utilizes the nanoscale LSPR effect to enhance the local electric field.

[0074] In this embodiment of the invention, in step S2, the interaction between light and the composite nanostructure specifically includes:

[0075] The dielectric nanopillar array is a silicon nitride nanopillar array. Through its geometric layout and periodic structure, it performs broadband phase encoding of the incident light. The phase encoding strategy is based on a pre-designed metasurface sensing matrix A to achieve mixing and modulation of optical signals of different wavelengths to support subsequent calculation and reconstruction.

[0076] The shape, size, and position parameters of the metal nanoisland array are determined according to the core optimization objective, which targets the target signal band for a specific application, including the heavy metal characteristic peak in water quality detection. The heavy metal characteristic peak is optimized through electromagnetic simulation or experimental calibration to excite local surface plasmon resonance in this band.

[0077] The local surface plasmon resonance generates a local electric field enhancement, and the enhancement effect is quantified by the enhancement factor matrix E. Furthermore, the phase encoding of the dielectric nanopillar array and the field enhancement of the metal nanoisland array work together to form an integrated "capture-enhancement-encoding" process, which directly amplifies weak light signals, thereby improving the sensitivity of spectral testing.

[0078] It should be noted that, according to Figure 4 The simulation diagram shows the electric field enhancement effect of the "dielectric-metal" composite nanostructure unit. The horizontal axis represents the lateral position along the nanostructure, and the vertical axis represents the modulus of the electric field intensity E (|E|). 2 The normalized values ​​indicate that the larger the value, the more significantly the electric field at that location is enhanced relative to the reference electric field without the nanostructure. The red line describes the continuous trend of the normalized magnitude of the electric field at different lateral positions. At approximately 360 nm, the peak value is 10.2, representing the most significant enhancement of the electric field across the entire range. This corresponds to surface plasmon resonance between the incident light and the nanostructure, resulting in a substantial amplification of the local electric field. |E| 2 The "double peak + trough" distribution pattern intuitively demonstrates that the local enhancement effect of the electric field by the nanostructure is not uniform globally, but rather induces a strong electric field enhancement only at specific lateral locations.

[0079] It should be noted that, Figure 5 The graph compares the spectral response functions of a traditional computational metasurface and the sensitivity-optimized metasurface of this invention. The traditional metasurface (blue) has a peak signal-to-noise ratio (SNR) of approximately 4 across the entire wavelength range, exhibiting a flat overall gain with a low upper limit. The sensitivity-optimized metasurface (red) shows a significant peak (8.5) at 600 nm, more than twice the peak value of the traditional metasurface; it also exhibits a secondary peak (approximately 7.9) at 750 nm, demonstrating "high sensitivity across multiple wavelengths." This invention, through structural optimization, overcomes the performance bottleneck of the traditional metasurface's SNR, achieving a leap forward in signal detection capability at the target wavelength.

[0080] In this embodiment of the invention, the mathematical model used in step S4 of the reconstruction algorithm is:

[0081] y = (A · E)x + n;

[0082] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and n is the system noise vector;

[0083] Specifically, the parameters are explained as follows: The measurement vector y output by the detector is a set of light intensity readings measured by a single-point detector (physical device). Assuming the metasurface is divided into M "superpixels," the detector will receive light transmitted / reflected from these superpixels sequentially (or in parallel) at M different times (or through M different channels), resulting in an M×1 column vector y. Each element y... i y represents the total light intensity obtained from the i-th measurement, and y is the measurement value vector output by the detector, which is the raw data processed by the algorithm.

[0084] A is the metasurface sensing matrix, an M×N matrix that describes the metasurface's spectral encoding capability for light without considering field enhancement effects. Here, N is the number of spectral channels to be restored. The physical meaning of the matrix element A{i,j} is the transmittance (or reflectance) response coefficient of the metasurface for light from the j-th wavelength channel in the i-th measurement state. It encodes how different wavelengths are "mixed" into a single measurement value. The metasurface sensing matrix A is entirely determined by the metasurface's nanostructure layout, shape, periodicity, and other geometric parameters. It can be pre-calculated through rigorous electromagnetic field simulation of the design drawings or obtained through experimental calibration of the fabricated chip using a standard light source with a known spectrum.

[0085] E is the enhancement factor matrix, an N×N diagonal matrix. Its diagonal element E{j,j} is a dimensionless amplification factor, representing the square of the factor by which the local electric field intensity at the j-th wavelength channel is enhanced due to the plasmon resonance effect of the metal nanoislands in the composite nanostructure (because the detector responds to the light intensity, i.e., the modulus square of the electric field). The enhancement factor matrix E directly corresponds to the characteristics of the metallic portion in the composite nanostructure, such as material, size, and shape. It is also based on the physical design of the device, obtained through electromagnetic simulation (such as calculating the field distribution around the nanostructure) or experimental calibration of the response at a specific wavelength.

[0086] The reconstruction algorithm recovers the desired spectral vector x from the measured value vector y by solving the above model. Further, the enhancement factor matrix E is a diagonal matrix; each element on the diagonal corresponds to the enhancement factor of a spectral channel, and the value of the enhancement factor is derived from the electromagnetic simulation calculation results or experimental calibration results of the "dielectric-metal" composite nanostructure unit. This step ensures that the mathematical processing model accurately matches the actual response of the physical device, thereby enabling high-fidelity reconstruction of the true spectrum, especially for weak target signals, from the original measured values.

[0087] In this embodiment of the invention, in step S4, when the following conditions ①②③ are met simultaneously, the spectral reconstruction algorithm is based on the compressed sensing path:

[0088] ① The spectrum of the analyte is known or reasonably believed to be sparse, such as gas detection or the emission spectrum of fluorescent labels;

[0089] ②The number of measurements M is much smaller than the number of spectral channels N, i.e., there is severe undersampling, but the conditions for a compressed sensing path are met;

[0090] ③The core optimization objective is to extract weak, sharp feature peaks from strong background or noise.

[0091] Based on the compressed sensing path, its mathematical model is as follows:

[0092] ;

[0093] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and λ is the regularization parameter. ‖·‖2 is the L2 norm (ensuring data fitting), and ‖·‖1 is the L1 norm (promoting the sparsity of the solution). This algorithm is based on the premise that the measured spectrum x is sparse under a certain basis (such as the wavelength itself or a certain transform domain), that is, the spectrum consists of a few obvious peaks, and the intensity of most channels is close to zero. This algorithm is suitable for scenarios where the measured spectrum has sparse characteristics, such as gas detection or emission spectra of fluorescent markers.

[0094] In this embodiment of the invention, in step S4, when the following conditions ①②③ are met simultaneously, the spectral reconstruction algorithm is based on the Tikhonov regularization path:

[0095] ① The spectrum to be measured is a wide-band, continuously changing spectrum that does not have obvious sparsity characteristics (such as blackbody radiation spectrum or some solution absorption spectrum).

[0096] ②The measurement noise is relatively large. The primary requirement is the stability and smoothness of the solution to avoid non-physical turbulence in the reconstruction results.

[0097] ③ If the number of measurements M is equal to or greater than N, the problem is not seriously underdetermined.

[0098] Based on Tikhonov regularization, its mathematical model is as follows:

[0099] ;

[0100] Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be obtained, λ is the regularization parameter, and ||x||² is the L² norm, used to constrain large amplitude values ​​of the solution, making the solution tend to be smooth and stable. This algorithm does not assume sparse spectra, but rather assumes that the energy of the solution should not be too large, i.e., the spectrum should be smooth or gradually changing. It is mainly used to handle ill-posed problems, obtaining numerically more stable and noise-resistant solutions by relaxing the fitting precision slightly. This algorithm is suitable for scenarios where the measured spectrum is a wide-spectral continuous variation, such as blackbody radiation spectrum or solution absorption spectrum, to improve the noise resistance and stability of the solution.

[0101] It should be noted that, as Figure 6 The flowchart of the spectral reconstruction algorithm is shown. The input layer includes the detector measurement vector y, the basic sensing matrix A, and the enhancement factor matrix E. Then, the model construction layer builds an enhanced reconstruction model y = A·E·x + n. By introducing E, the model combines the "sensing matrix A" with the "field enhancement mechanism," allowing the reconstruction process to utilize the enhancement effect of E on x, laying the foundation for subsequent signal-to-noise ratio (SNR) improvement. Next, the algorithm selection layer provides two reconstruction algorithm paths adapted to different spectral characteristics (sparseness / smoothness): Compressed sensing / convex optimization utilizes the "sparseness" of the spectrum, achieving high-precision reconstruction under undersampling through L1 norm constraints, suitable for weak signals and sparse features; Compressed sensing / Tikhonov regularization suppresses noise through L2 norm constraints, making the solution smoother, suitable for continuous spectra and spectra with low noise sensitivity but requiring interference resistance. Finally, the spectrum after model enhancement and algorithm denoising is obtained, with a significantly higher SNR than the result directly analyzed from y, improving both the SNR and accuracy of the spectral reconstruction.

[0102] In this embodiment of the invention, the composite nanostructure of the metasurface chip is achieved through the following device fabrication process, specifically including:

[0103] A substrate is provided, wherein the substrate is a quartz substrate;

[0104] Metal nanoisland arrays are fabricated on the substrate by electron beam evaporation and exfoliation processes, wherein the metal nanoisland arrays are used to generate local field enhancement effects;

[0105] An isolation layer, made of silicon nitride, is grown on the metal nanoisland array using atomic layer deposition technology. The isolation layer is used for electrical isolation and optical modulation.

[0106] A dielectric nanopillar array, which is a silicon nitride nanopillar array, is fabricated on the isolation layer using electron beam lithography and dry etching processes to form a hybrid metasurface. The metal nanoisland array and the dielectric nanopillar array work together to simultaneously provide field localization enhancement and spectral encoding functions, thereby achieving high-sensitivity capture and processing of weak light signals.

[0107] The present invention also provides an embodiment of a light field modulation detection device based on computational spectral reconstruction, used to implement the above-described light field modulation detection method based on computational spectral reconstruction, the device comprising:

[0108] The metasurface chip, consisting of an array of metal nanoislands and an array of silicon nitride nanopillars on a quartz substrate, is used to receive the optical signal to be measured and to achieve field localization enhancement and spectral encoding.

[0109] The detector, integrated with a metasurface chip, is used to receive enhanced and encoded optical signals and convert them into electrical signals;

[0110] The signal processing unit, connected to the detector, is used to run the reconstruction algorithm described above and calculate the spectral map based on the sensing matrix and the enhancement factor matrix.

[0111] In this embodiment of the invention, the detector is a single-point detector or a small detector array, wherein the single-point detector includes a photodiode, and the small detector array is used to receive optical signals in parallel or sequentially; the detector is configured to convert the optical signal into an electrical signal and generate a measurement value vector y output by the detector; the metasurface chip and the detector are integrated into a compact structure through packaging, wherein the packaging process includes directly bonding or fixing the metasurface chip to the detector and electrically connecting it to the signal processing unit, to ensure that the device has no moving parts, high stability and miniaturization advantages, and the metasurface chip is integrated on top of the detector (no mechanical scanning, no complex optical path), and its volume is only 1 / 10-1 / 5 of that of a traditional spectrometer; the design without moving parts avoids the impact of vibration and wear on accuracy, and can work stably in complex environments such as in the field and in vehicles; the optical signal to be measured is incident perpendicularly on the metasurface chip, and after transmission, it is directly coupled into the detector, wherein the direct coupling method includes an optical path alignment design to minimize optical loss and improve signal transmission efficiency, thereby enhancing the device's ability to capture weak optical signals.

[0112] This invention patent utilizes the localized surface plasmon resonance effect excited by a metal nanoisland array to achieve field localization enhancement at the physical level, directly amplifying weak light signals. This solves the problems of the "light flux-resolution" contradiction in traditional spectrometers and the insufficient sensitivity of miniaturized devices. Simultaneously, it innovatively integrates the enhancement effect with the broadband phase encoding function of the dielectric nanopillar array. Through deep fusion of the metasurface sensing matrix A and the enhancement factor matrix E with an adaptive reconstruction algorithm (compressed sensing / Tikhonov regularization), it achieves directional high-sensitivity detection of specific material characteristic peaks while maintaining all the advantages of computational metasurface spectrometers, such as miniaturization, no moving parts, and high stability. For example, in water quality detection, nanostructure parameters can be specifically designed to amplify heavy metal characteristic peaks (such as the absorption bands of lead or mercury), thereby improving the detection signal-to-noise ratio and application specificity, outperforming traditional broadband average enhancement schemes. Furthermore, it improves the signal-to-noise ratio, and this technical framework is adaptable to various metasurface types, such as wavelength multiplexing and polarization multiplexing, demonstrating significant universality advantages.

[0113] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for detecting light field modulation based on computational spectral reconstruction, characterized in that, Includes the following steps: Step S1: The optical signal to be tested is incident perpendicularly onto the metasurface chip, wherein the metasurface chip comprises a composite nanostructure; Step S2: The metasurface chip performs field localization enhancement and spectral encoding on the optical signal to be measured; Step S3: Convert the encoded optical signal to be tested into an electrical signal using a detector; Step S4: Construct a reconstruction algorithm with an embedded physical enhancement model, call the pre-stored metasurface sensing matrix A and enhancement factor matrix E, calculate and output the spectrum x of the optical signal to be measured based on the electrical signal.

2. The optical field modulation detection method based on computational spectral reconstruction according to claim 1, characterized in that, In step S1, the composite nanostructure of the metasurface chip includes a "dielectric-metal" composite nanostructure unit, which includes an integrated array of dielectric nanopillars and an array of metal nanoislands.

3. The optical field modulation detection method based on computational spectral reconstruction according to claim 1, characterized in that, In step S2, light interacts with the composite nanostructure, specifically including: The dielectric nanopillar array is a silicon nitride nanopillar array, which performs broadband phase encoding of incident light through its geometric layout and periodic structure. The shape, size, and position parameters of the metal nanoisland array are determined according to the core optimization objective, which targets the target signal band for a specific application, including the heavy metal characteristic peak in water quality detection. The heavy metal characteristic peak is optimized through electromagnetic simulation or experimental calibration to excite local surface plasmon resonance in this band. The localized surface plasmon resonance generates a localized electric field enhancement, and the enhancement effect is quantified by the enhancement factor matrix E; moreover, the phase encoding of the dielectric nanopillar array works in synergy with the field enhancement of the metal nanoisland array.

4. The optical field modulation detection method based on computational spectral reconstruction according to claim 1, characterized in that, In step S4, the mathematical model used in the reconstruction algorithm is: y = (A · E)x + n; Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and n is the system noise vector; The reconstruction algorithm recovers the spectral vector x from the measured value vector y by solving the above model.

5. The optical field modulation detection method based on computational spectral reconstruction according to claim 4, characterized in that, The enhancement factor matrix E is a diagonal matrix; Each element on the diagonal of the matrix corresponds to an enhancement factor for a spectral channel, the value of which is derived from electromagnetic simulation calculations or experimental calibration results of the "dielectric-metal" composite nanostructure unit.

6. The optical field modulation detection method based on computational spectral reconstruction according to claim 5, characterized in that, In step S4, the spectral reconstruction algorithm is based on the compressed sensing path, and its mathematical model is as follows: ； Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and λ is the regularization parameter.

7. The optical field modulation detection method based on computational spectral reconstruction according to claim 5, characterized in that, In step S4, the spectral reconstruction algorithm is based on the Tikhonov regularization path, and its mathematical model is as follows: ； Where y is the measured value vector output by the detector, A is the metasurface sensing matrix, E is the enhancement factor matrix, x is the reconstructed spectral vector to be determined, and λ is the regularization parameter.

8. The optical field modulation detection method based on computational spectral reconstruction according to claim 1, characterized in that, The composite nanostructure of the metasurface chip is achieved through the following device fabrication process, specifically including: A substrate is provided, wherein the substrate is a quartz substrate; Metal nanoisland arrays are fabricated on the substrate by electron beam evaporation and exfoliation processes, wherein the metal nanoisland arrays are used to generate local field enhancement effects; An isolation layer, made of silicon nitride, is grown on the metal nanoisland array using atomic layer deposition technology. The isolation layer is used for electrical isolation and optical modulation. A dielectric nanopillar array, which is a silicon nitride nanopillar array, is fabricated on the isolation layer using electron beam lithography and dry etching processes to form a hybrid metasurface.

9. A light field modulation and detection device based on computational spectral reconstruction, used to implement the light field modulation and detection method based on computational spectral reconstruction as described in any one of claims 1-8, characterized in that, include: The metasurface chip, consisting of an array of metal nanoislands and an array of silicon nitride nanopillars on a quartz substrate, is used to receive the optical signal to be measured and to achieve field localization enhancement and spectral encoding. The detector, integrated with a metasurface chip, is used to receive enhanced and encoded optical signals and convert them into electrical signals; The signal processing unit, connected to the detector, is used to run the reconstruction algorithm and calculate the spectrum based on the sensing matrix and the enhancement factor matrix.

10. The optical field modulation and detection device based on computational spectral reconstruction according to claim 9, characterized in that, The detector is a single-point detector or a small detector array, wherein the single-point detector includes a photodiode, and the small detector array is used to receive optical signals in parallel or sequentially. The detector is configured to convert optical signals into electrical signals and generate a measurement vector y output by the detector; The metasurface chip is integrated and packaged with the detector, wherein the packaging process includes directly bonding or fixing the metasurface chip to the detector and electrically connecting it to the signal processing unit. The optical signal to be tested is incident perpendicularly onto the metasurface chip, and after transmission, it is directly coupled into the detector. The direct coupling method includes optical path alignment design.

Images