Spectrometers and methods for detecting electromagnetic (EM) spectra

Abstract

A spectrometer for detecting electromagnetic (EM) spectra having one or more wavelength components within a spectral band of interest, and a method for detecting electromagnetic (EM) spectra having one or more wavelength components within a spectral band of interest. The method utilizes an entrance aperture; a dispersive and imaging optics comprising at least one dispersive element; an exit aperture; a collection optics; and at least one single-pixel detector, each single-pixel detector being sensitive to one or more wavelength components. The method includes the following steps: spatially encoding at least one entrance slit of an entrance aperture in a direction generally transverse to the dispersion direction of the dispersive and imaging optics; creating a dispersive image of the entrance aperture on the plane of the exit aperture using the dispersive and imaging optics, such that the images of different wavelength components are offset by different amounts of displacement along the dispersion direction; spatially encoding a plurality of exit slits of the exit aperture in a direction generally transverse to the dispersion direction, wherein the exit aperture includes a plurality of exit slits arranged along the dispersion direction; collecting the total electromagnetic energy entering and leaving the entrance aperture to at least one of at least one single-pixel detectors using a collection optics; repeatedly changing at least one of the encoding patterns of the at least one entrance slit and the encoding patterns of the plurality of exit slits; and measuring the output of at least one detector for a corresponding number of times the electromagnetic spectrum is reconstructed.

Description

Technical Field

[0001] This invention generally relates to spectrometers, and more specifically, but not limited to, portable high-throughput spectrometers for field use. Background Technology

[0002] Any references to and / or discussion of prior art throughout the specification shall not be construed as an admission that such prior art is well-known or constitutes part of common general knowledge in the art.

[0003] Fourier transform infrared (FTIR) spectrometers are used in a wide range of sensing applications, but these are typically benchtop instruments designed for controlled laboratory environments. They are inherently optical interferometers susceptible to external interference and contain sophisticated optomechanical scanning mechanisms, which makes them difficult to miniaturize for portable field use.

[0004] On the other hand, there is an increasing demand for on-site use of portable spectrometers (especially infrared spectrometers) in various sensing applications such as food and beverage quality assessment, environmental monitoring, precision agriculture, industrial process control, the Internet of Things, and biomedical care points.

[0005] Infrared (IR) spectroscopy is widely used in chemical detection and analysis due to its superior specificity and sensitivity. Molecules can be identified / detected through their unique light absorption characteristics. In recent years, the demand for portable infrared spectrometers for field use has been increasing, with a wide range of applications including environmental monitoring, food and beverage quality assessment, the Internet of Things (IoT), and drug screening.

[0006] (i) The simplest dispersive spectrometer

[0007] As shown in Figure 1, dispersive spectrometers are typically used in the visible (VIS) and near-infrared (NIR) spectral regions. Light passing through a slit is first collimated by a mirror and guided to a diffraction grating, where its spectral components (1 to N) are diffracted to different angles. These diffracted beams are then focused by another mirror onto a linear detector array, where the different spectral components fall onto different pixels for recording. Dispersive spectrometers work well in the VIS / NIR region, but have the following drawbacks in the mid-infrared (MIR) and far-infrared (FIR) regions, where molecules often exhibit absorption characteristics (hence also called MIR / FIR, and mentioned in this paper as the molecular "fingerprint" region):

[0008] (1) Detector arrays are less efficient in photon detection and more expensive in the MIR and FIR regions.

[0009] (2) Flux (or light collection efficiency), meaning the instrument's ability to collect light energy is insufficient. As shown in Figure 1, the narrow slit (i.e., the small width L) sThis must be used for high spectral resolution, thus limiting the light input to the spectrometer. This can be easily understood as follows: if L s Increase (in other words, collect more optical power), its corresponding image L at each wavelength s It will also broaden, thus reducing spectral resolution.

[0010] (3) Distributing the already limited light energy to multiple locations further weakens the signal at each pixel of the detector array. This, combined with the high dark current of the photodetectors in the MIR and FIR regions, significantly reduces the signal-to-noise ratio (SNR) of the spectrometer.

[0011] (ii) Fourier Transform Infrared (FTIR) Spectrometer

[0012] Therefore, FTIR spectrometers are traditionally used in the MIR and FIR regions. [1] As shown in Figure 2, the Michelson interferometer is frequently used to construct FTIR spectrometers. In an FTIR system, a movable mirror scans within a range L, and the change in the interferometer output light intensity I(δ) is recorded as a function of the optical path difference (OPD)δ, commonly referred to as an interferogram. It can be demonstrated that the power spectrum of the light source I... σ The recorded interferogram I(δ) is correlated with the data via a Fourier transform:

[0013]

[0014] Where δ represents the wavenumber. FTIR offers several advantages over dispersive spectroscopy. The most significant are the advantages of multiplexing (using a single-pixel detector, where multiplexed light energy falls onto a single pixel, thus improving the SNR) and high throughput (capable of collecting more light energy compared to dispersive spectroscopy). However, FTIR is typically a benchtop instrument used in controlled laboratory environments. Miniaturizing these FTIRs and making them readily available for field use would be very attractive for a range of applications, including food quality assessment, environmental monitoring, and point-of-care testing. However, the fabrication of portable, high-quality FTIRs has proven challenging over the past few decades. The main reasons are as follows:

[0015] (1) FTIR is essentially an optical interferometer containing two optical paths. As shown in Figure 2, if there is external interference (temperature, airflow, vibration, etc.) in path A or path B or both, even the slightest interference will significantly affect the recorded interferogram, thereby increasing the noise in the reconstructed spectrum.

[0016] (2) FTIR also includes a moving mirror, which requires high-precision alignment / positioning and is also susceptible to external interference.

[0017] (3) Although the flux of FTIR is higher than that of dispersive spectrometers, it is still limited. This can be understood as the working principle of FTIR requiring collimation of the beam in the interferometer. Results show that the maximum permissible half-angle α... max The following conditions must be met:

[0018]

[0019] Where δσ is the wavenumber resolution, σ max This is the maximum wavenumber of the spectrometer. Typically, this maximum value is very small, for example... Modular FTIR spectrometer MIR8035 TM[2] The maximum half-angle of the input is 0.5°. Assuming a collimating lens with a focal length of 50mm, as shown in Figure 2, the permissible aperture size / diameter D can be easily estimated. s The diameter is 0.88 mm, which is very small for collecting light radiation energy.

[0020] These limitations together make it challenging to miniaturize FTIR for field use while maintaining its high spectral resolution at IR wavelengths.

[0021] (iii) Encoded aperture spectrometer

[0022] Encoded aperture spectrometers represent the latest development in the field of spectral sensing technology. [3] As shown in Figure 3, the coded aperture spectrometer is very similar to the conventional grating spectrometer shown in Figure 1, except that the narrow entrance slit is replaced by a wider coded aperture. This replacement results in many “overlapping” images of the aperture on the plane of the detector array (or image sensor), with different wavelengths at different locations. By assuming that the aperture is uniformly illuminated, these “overlapping” images of the coded aperture recorded by the image sensor can then be used to reconstruct the spectrum. The main advantage of this method is that it separates spectral resolution from instrument throughput. In other words, spectral resolution is set by the minimum features of the coded aperture, while throughput is set by the overall size of the aperture. However, as can be seen from the schematic diagram shown in Figure 3, it still uses a detector array for sensing, so one of the main limitations of this method is:

[0023] Encoded aperture spectrometers are primarily used in VIS-NIR systems rather than in MIR and FIR wavelength regions because detector arrays have a lower pixel count (or lower resolution) and are less efficient in photon detection and more expensive in the MIR and FIR wavelength regions.

[0024] (iv) Reusable light-shielding plate spectrometer

[0025] Multi-channel spectral analyzers were proposed decades ago. They are designed to detect spectra using a single-pixel detector, thus making them readily applicable to IR and other wavelengths. These spectrometers can be divided into two configurations. The first configuration (also known as a single-encoded spectrometer), as shown in Figure 4(a), encodes only the light radiation at the exit aperture. [4] Light passing through the narrow slit is scattered and imaged onto the exit aperture. A movable light-blocking plate (typically using a Hadamard coding design) is placed immediately before or after the exit aperture to encode the light. After encoding, the light is collected by collecting optics and received by a single-pixel detector. The spectrum is reconstructed from the data recorded by the single-pixel detector. This system offers the advantage of multiplexing (allowing multiple wavelengths to arrive at the detector simultaneously) but not the advantage of throughput because the narrow slit limits the light energy that the spectrometer can collect.

[0026] The second configuration, schematically shown in Figures 4(b) and 4(c), also known as a dual-coded spectrometer, involves coding of both the incident and exit apertures. This includes Golay's multi-slit spectrometer. [5] and Girard's Grill spectrometer [6] Early designs, including those with larger apertures, offered the advantage of increased throughput by collecting more light energy, but they lost the advantage of multiplexing. Subsequent research and development focused on encoders based on the Adama matrix, which led to multiplexed shading spectrometers possessing both multiplexing and throughput advantages. [7] .

[0027] Figures 4(b) and 4(c) further highlight the two designs for the second configuration. In these designs, both the incident and exit apertures are encoded using a light-shielding plate containing multiple slits, some open to allow light to pass through and some closed to block light. The source radiation passing through the spectrometer is thus encoded twice. The design in Figure 4(b) utilizes the movement of two light-shielding plates driven by two stepper motors to dynamically generate encoded patterns at the incident and exit apertures. [8] A series of coded patterns are created by stepping a light-blocking plate containing multiple slits through the aperture opening. However, the design in Figure 4(c) uses only one movable light-blocking plate at the inlet to encode the light, and the light-blocking plate remains at the outlet. [9] For the two designs shown in Figures 4(b) and 4(c), the spectrum can be reconstructed from the data recorded by the single-pixel detector under various encoding settings.

[0028] For the design shown in Figure 4(a), the main limitations are as follows:

[0029] (1) In order to achieve high spectral resolution, the slit image on the exit aperture plane is from λ1 to λ NThe spatial spacing needs to be large, which means a large exit aperture is required, thus necessitating significant translational movement of the coding shield. This makes it difficult to miniaturize designs implemented with motors for field use.

[0030] (2) It is worth noting that in recent years, there has been considerable interest in using microelectromechanical systems (MEMS) based mirrors and mirror arrays to replace motor-driven coded slit shields. The coded pattern can be dynamically implemented by opening or closing rows / columns of the micromirrors. Typical designs using Texas Instruments' (TI) digital micromirror devices (DMDs) can be found in many recent publications.

[10]

[11] These designs have the potential for miniaturization; however, encoding only the exit aperture does not provide high throughput because the total amount of light energy that can enter the spectrometer is severely limited by the narrow slit or pinhole at the entrance aperture.

[0031] For the designs shown in Figures 4(b) and 4(c), the entrance aperture is encoded and has an extended region. Therefore, this design offers both throughput and multiplexing advantages, potentially enabling high-performance spectrometers. However, their main limitations are as follows:

[0032] (1) For the design in Figure 4(b), the driving mechanism is complex, involving the motion control and synchronization of two coded light shields, so the design is difficult to miniaturize for field use.

[0033] (2) For the design in Figure 4(c), the entrance aperture consists of N slits, and the exit aperture consists of (2N-1) slits. The encoding scheme is based on the following assumption: for any integer t, light entering from the i-th slit and exiting from the j-th slit has the same wavelength as light entering from the (i+t)-th slit and exiting from the (j+t)-th slit. The entire encoding / decoding scheme is based on this assumption. Its meaning is shown in Figure 5(a). The wavelength components λ1, λ2, ..., λ of the light passing through the first entrance slit are... NThe wavelengths should exit from the 1st, 2nd, ..., Nth exit slits respectively. Furthermore, the wavelengths from the 2nd entrance slit to the 2nd exit slit, from the 3rd entrance slit to the 3rd exit slit, ..., from the Nth entrance slit to the Nth exit slit should be the same, and they should be equal to wavelength λ1. In other words, the total width of the entrance aperture is proportional to the total spatial separation of the wavelength components on the exit aperture plane, as shown in Figure 5(a). In a typical high-resolution spectrometer, as shown in Figure 5(b), taking Ocean Optics' USB model VIS-NIR spectrometer as an example, this spatial separation of wavelength components is indeed very large. The direct result is an excessively large entrance aperture, making it difficult to achieve miniaturization of the overall instrument size and balance of optical aberrations in high-performance portable spectrometers. It should be noted that the above assumptions on which the system in Figure 4(c) is based actually combine flux with spectral resolution, making the design of portable and high-resolution systems challenging.

[0034] Embodiments of the present invention provide methods and systems for seeking to solve at least one of the above-described problems. Summary of the Invention

[0035] According to a first aspect of the present invention, a spectrometer is provided for detecting electromagnetic (EM) spectra having one or more wavelength components within a spectral band of interest, comprising:

[0036] Entrance aperture;

[0037] Dispersive and imaging optics that include at least one dispersive element;

[0038] Exit port;

[0039] Collect optical devices; and

[0040] At least one single-pixel detector, each single-pixel detector being sensitive to one or more wavelength components;

[0041] The entrance aperture includes at least one entrance slit, which is spatially encoded along a direction generally transverse to the dispersion direction of the dispersive and imaging optics.

[0042] The dispersive and imaging optics are configured to create a dispersive image of the entrance aperture on the plane of the exit aperture, such that the images at different wavelength components are offset by different amounts of displacement along the dispersive direction.

[0043] The exit aperture includes a plurality of exit slits arranged along the dispersion direction, wherein each exit slit is spatially encoded in a direction generally transverse to the dispersion direction.

[0044] The collecting optics are configured to collect the total electromagnetic wave energy entering the entrance aperture and exiting the exit aperture to at least one of the single-pixel detectors.

[0045] In this embodiment, at least one of the coding patterns of at least one entrance slit and the coding patterns of multiple exit slits is adjustable and configured to be changed multiple times.

[0046] The spectrometer also includes a measurement unit configured to measure the output of at least one detector for a corresponding number of times the EM spectrum is reconstructed.

[0047] According to a second aspect of the invention, a method is provided for detecting an electromagnetic (EM) spectrum having one or more wavelength components within a spectral band of interest, using: an entrance aperture; a dispersive and imaging optics comprising at least one dispersive element; an exit aperture; a collection optics; and at least one single-pixel detector, each single-pixel detector being sensitive to one or more wavelength components. The method includes the following steps:

[0048] Spatially encode at least one entrance slit of the entrance aperture along a direction generally transverse to the dispersion direction of the dispersive and imaging optics.

[0049] Using dispersive and imaging optics, a dispersive image of the entrance aperture is created on the plane of the exit aperture, such that the images at different wavelength components are offset by different amounts of displacement along the dispersive direction.

[0050] The multiple exit slits of the exit aperture are spatially encoded along a direction generally transverse to the dispersion direction, wherein the exit aperture includes multiple exit slits arranged along the dispersion direction.

[0051] The total electromagnetic energy entering the entrance aperture and exiting the exit aperture is collected by collecting optical devices to one of at least one single-pixel detectors.

[0052] The coding pattern of at least one entrance slit and the coding pattern of multiple exit slits are changed multiple times; and

[0053] The output of at least one detector is measured for the corresponding number of times the electromagnetic spectrum is reconstructed.

[0054] Brief description of the attached figures

[0055] From the following written description, which is by way of example only, and in conjunction with the accompanying drawings, those skilled in the art will better understand and more readily comprehend embodiments of the invention, wherein:

[0056] Figure 1 shows a prior art dispersive spectrometer used in the visible and near-infrared spectral regions.

[0057] Figure 2 shows the existing Michelson interferometer used to construct FTIR spectrometers.

[0058] Figure 3 shows a prior art coded aperture spectrometer.

[0059] Figure 4(a) shows a prior art single-code, multiplexed light-shielding spectrometer.

[0060] Figure 4(b) shows a prior art dual-coded, multiplexed light-shielding spectrometer.

[0061] Figure 4(c) shows another prior art dual-coded, multiplexed light-shielding spectrometer.

[0062] Figure 5(a) illustrates a prior art encoding / decoding scheme for the spectrometer in Figure 4(c).

[0063] Figure 5(b) shows the existing high-resolution marine optics USB model VIS-NIR spectrometer.

[0064] Figure 5(c) is a schematic diagram illustrating how, according to an example embodiment, both the entrance slit and the exit slit are encoded in a direction generally transverse to the dispersion direction to decouple flux from spectral resolution.

[0065] Figure 6 A schematic diagram illustrating a spectrometer according to an example embodiment is shown.

[0066] Figure 7(a) shows a schematic diagram illustrating a movable light shield placed immediately after the entrance slit according to an example embodiment.

[0067] Figure 7(b) shows a schematic diagram illustrating a spatially encoded entrance slit of a linear array of micromirrors implemented according to an example embodiment.

[0068] Figure 8(a) shows a schematic diagram illustrating the imaging of a coded entrance slit by means of a dispersive and imaging optics according to an example embodiment.

[0069] Figure 8(b) shows a schematic diagram illustrating a second encoding pattern according to an example embodiment.

[0070] Figure 9(a) shows a schematic diagram illustrating an entrance aperture according to an example embodiment, the entrance aperture having an extended area encoded using an adjustable 2D encoding pattern that enables a movable light shield.

[0071] Figure 9(b) shows a schematic diagram illustrating an adjustable 2D coded pattern implemented as a 2D micromirror array according to an example embodiment.

[0072] Figure 10 A schematic diagram illustrating that the beginning portion of a second encoding pattern according to an example embodiment is aligned with a dispersion image of a first slit to encode light passing through the first slit.

[0073] Figure 11(a) shows a prior art dispersive spectrometer in which the slit image is at a wavelength λ greater than the spot size S, resulting in the spectral resolution being mainly determined by the slit width.

[0074] Figure 11(b) shows a prior art dispersive spectrometer in which the slit image is much smaller than the spot size on its focal plane, resulting in the spectral resolution being mainly determined by the spot size.

[0075] Figure 12 A schematic diagram illustrating an adjustable encoder for an entrance aperture, implemented using a movable light shield in combination with a fixed aperture opening, according to an example embodiment, is shown.

[0076] Figure 13 A schematic diagram illustrating a typical row of a second encoding pattern according to an example embodiment is shown.

[0077] Figure 14 (a) A front view schematic diagram illustrating an implementation of the coding mechanism design according to an example embodiment is shown.

[0078] Figure 14 (b) shows Figure 14 (a) is a cross-sectional schematic diagram of an implementation of the coding mechanism design.

[0079] Figure 14 (c) shows a cross-sectional schematic diagram illustrating an optical position sensor according to an example embodiment, the sensor involving a light-emitting diode, a photodetector and a pair of gratings.

[0080] Figure 15 A schematic diagram illustrating a spectrometer according to an example embodiment is shown, wherein the dispersive and imaging optics and the collection optics share the same optical setup.

[0081] Figure 16 A schematic diagram illustrating a spectrometer using a single-pixel detector with a small photosensitive area according to an example embodiment is shown.

[0082] Figure 17 A schematic diagram illustrating a spectrometer using a single-pixel detector with a small photosensitive area, according to an example embodiment, is shown.

[0083] Figure 18 A schematic diagram illustrating a second coded pattern at the exit aperture plane, designed to be distorted to match a distorted image of the incident slit, according to an example embodiment, is shown.

[0084] Figure 19 A flowchart illustrating a method for detecting an electromagnetic (EM) spectrum having one or more wavelength components within a spectral band of interest, according to an example embodiment, is shown. Detailed Implementation

[0085] Embodiments of the present invention may provide a miniature spectrometer having one or more of the following significant advantages.

[0086] Unlike FTIR, the embodiments of the present invention are not based on optical interferometers, and are therefore more robust and less sensitive to external interference, making them suitable for field use.

[0087] Compared to the typical slits or pinholes used in traditional grating or prism-based dispersive spectrometers, embodiments of the present invention can have enlarged entrance apertures, thereby significantly enhancing light-gathering capabilities.

[0088] • Embodiments of the present invention may use single-pixel photodetectors or multiple single-pixel photodetectors, thus enabling cost-effective operation at UV and / or IR wavelengths (where image sensor / detector arrays are expensive).

[0089] • Embodiments of the present invention can have the advantage of reuse, thus supporting high SNR detection.

[0090] Typically, embodiments of the present invention relate to a portable spectrometer suitable for field use. According to exemplary embodiments, these spectrometers can have large entrance apertures and therefore high throughput (or light collection efficiency), resulting in a high signal-to-noise ratio (SNR). These spectrometers also utilize single-pixel (or single-element) multiplexing detection technology, thus enabling convenient and low-cost application in UV, VIS, and IR spectroscopy applications.

[0091] In other words, the spectrometer technology according to the example embodiments can be fabricated into compact / portable spectrometers that still offer high resolution and throughput. According to the example embodiments, this technology is similar to FTIR in its use of unit-piece (or single-pixel) detectors instead of detector arrays, thus performing well in IR spectroscopy where detector arrays are expensive and less sensitive. Similar to FTIR, it also offers the advantage of multiplexing, resulting in high SNR during detection. Unlike FTIR, it is not based on optical interferometers, making it less sensitive to positioning errors and external interference, thus facilitating miniaturization for field use while maintaining satisfactory spectral resolution. Finally, and equally importantly, the spectrometers disclosed according to the example embodiments can also have relatively large entrance apertures, thus enabling them to collect more optical radiation energy even compared to FTIR spectrometers.

[0092] Furthermore, as shown in Figure 5(c), in an exemplary embodiment of the invention, to decouple flux from spectral resolution, the entrance and exit slits can be encoded along a direction generally transverse to the dispersion direction. This design allows for the free selection of any number of entrance slits in the entrance aperture. This differs from the spectrometer shown in Figure 5(a), where the number of entrance slits must be approximately half the number of exit slits. Furthermore, according to the exemplary embodiment, the images of the entrance slits do not need to be perfectly aligned with the exit slits, as can be seen from an example highlighted in Figure 5(c) showing a second entrance slit and its image at wavelength λ1. This also differs from the spectrometer shown in Figure 5(a). According to the exemplary embodiment, these unique features decouple flux from spectral resolution. Therefore, the size of the entrance aperture can now be freely chosen, achieving an optimal balance between spectral resolution, flux, and portability.

[0093] Figure 6 A schematic diagram illustrating a spectrometer 600 according to an exemplary embodiment of the present invention is shown. As shown, an entrance aperture 602 consists of at least one slit 604, which is encoded in a direction 606 generally transverse to the dispersion direction 608 of the spectrometer 600. The aperture 602 is uniformly illuminated by a light or EM wave source (denoted as digit 610). Uniform illumination can be achieved in various ways, including using a small lens array, a light equalizer, and an integrating sphere. The light or EM wave passing through the entrance aperture 602 is then encoded for the first time using a first encoding pattern 612 (also referred to as a first encoder) before subsequently entering a dispersive and imaging optics 614. As those skilled in the art will understand, the dispersive and imaging optics 614 is similar to or the same as those used in dispersive spectrometers and may consist of a mirror / lens for imaging and at least one dispersive element (e.g., a grating / prism for dispersion). On the image plane 616 of the dispersive and imaging optics 614, an exit aperture 618 and a second encoding pattern 620 (also referred to as a second encoder) are positioned, which encodes the light a second time. The second encoder 620 includes a plurality of slits, such as 622, which together define an exit aperture 618 in this embodiment, encoded in a direction 606 generally transverse to the dispersion direction 608 of the spectrometer 600. Therefore, the total light attenuation or transmission is determined by superimposing the image of the first encoded pattern 612 onto the second encoded pattern 620. (This is similar to Goli's multi-slit spectrometer.) [5] And Grill spectrometers that allow only a single wavelength component to pass through and reach the detector while blocking other wavelength components. [6] Unlike other examples, according to the example embodiment, the spectrometer 600 allows multiple wavelength components to pass through and reach the photodetector with slight or moderate attenuation, thus having the advantage of multiplexing to achieve enhanced SNR.

[0094] In this embodiment, the first encoding pattern 612 is adjustable or programmable, while the second encoding pattern 620 is fixed. However, in different embodiments, the first encoding pattern may be fixed while the second encoding pattern is adjustable, or both encoding patterns may be adjustable. The adjustable encoding patterns in various embodiments can be implemented in a variety of ways, such as using a programmable micromirror array, for example, from Texas Instruments (TI).

[12] A digital mirror device (DMD) or simply placing a movable light shield directly behind a fixed aperture.

[0095] exist Figure 6 On the image plane 616 of the dispersive and imaging optics 614 shown, due to dispersion, the images of the encoded entrance apertures 612 of different wavelengths are located at different positions on the second encoded pattern along the dispersion direction 606 as shown. Therefore, the light intensity of different wavelengths is encoded differently by the second encoded pattern 620 with a specified weight fixed in this embodiment, thereby allowing it to be successfully decoded to reconstruct the spectrum using an appropriate signal processing algorithm.

[0096] After light passes through the exit aperture 618 / second coded pattern 620, it is collected and focused by the light-collecting optics 624 onto the single-pixel photodetector 626. The light-collecting optics 624, which matches the size of the exit aperture 618 to the size of the photosensitive area on the photodetector 626, can be an imaging optics device such as a condenser or a non-imaging optics device, such as a compound parabolic condenser (CPC).

[13] .

[0097] The general operating principle of the spectrometer 600 according to the example embodiment is described below. Figure 6 As shown, the incident light radiation 610 is in the spectral band of interest [λ]. min , λ max The [contains] N spectral components λ1, λ2, ..., λ3. k 、…λ N (k = 1 to N) and uniformly illuminate the entrance aperture 602. As those skilled in the art will understand, light radiation of other wavelengths outside the spectral band of interest can be easily filtered out by the filter and prevented from entering the entrance aperture 602 of the spectrometer 600 system. The first coded pattern 612 is adjustable, and according to the example embodiment, it is adjusted M times for a complete coded sequence. Figure 6 As shown, the second encoding pattern 620 used in this example embodiment is fixed. Therefore, when the first encoding pattern 612 is set to the j-th pattern (j = 1 to M), the total transmittance of the k-th spectral component through the system is w. jk Similarly, the total transmittance through the l-th spectral component of the system is w jl, which can be determined by the superposition of the first and second coding patterns. It should be noted that w jk usually should be different from w jl because of the dispersion of the system and the overlap of the patterns at different positions. Therefore, the total radiation power that can be transmitted and reach the single-pixel photodetector is a linear combination of all the powers in its spectral components. Thus, the j-th measurement signal from the detector is:

[0098]

[0099] where y j is the measured detector signal of the j-th coding pattern, c is a system constant determined by, for example, the detector sensitivity, and x k is the light intensity of the k-th spectral component at the entrance aperture (i.e., the spectrum to be determined). Next, the first coding pattern is set as the (j + 1)-th pattern to repeat the process until a complete set of M measurements is completed. Alternatively, the above equation, for j = 1 to M, can be represented in matrix form (here, the constant c is removed because only the relative intensities are important for the spectrum):

[0100] Y = WX (4)

[0101] where Y is the measurement vector (M×1) containing the signals from the photodetector 626, X is the vector ((N×1) containing N spectral components, and W is the weighting matrix (M×N) determined by the coding scheme used or the patterns 612, 620 at the entrance and exit apertures. Depending on the number of recorded measurements, there are various methods to solve the above spectral X equation:

[0102] (a) If M = N, then X = W -1 Y, where, W -1 is the inverse matrix of W. It can be seen that the best way to design the coding pattern should follow the Hadamard matrix with weights 1 and -1 and the S matrix with weights 0 and 1 [9] .

[0103] (b) If M < N (i.e., fewer measurements are made than unknowns), then a compressive sensing algorithm

[14] can be used to solve for the spectrum X, provided that the spectrum must be sparse on some basis.

[0104] (c) If M > N (i.e., more measurements are made than unknowns), then various methods can be used to obtain the spectrum X, including the Moore - Penrose generalized inverse

[15] .

[0105] In the following description, the method of M=N is used according to a preferred embodiment, which means that the number of measurements is equal to the number of unknown spectral components in the spectrum. Those skilled in the art will appreciate extensions of the method to the case of M≠N according to different embodiments.

[0106] First, consider the case where the entrance aperture defines a single slit 700 as shown in Figure 7. The slit 700 is spatially encoded by first encoding patterns 706a, 706b along its length direction 702 (or in other words, in a direction generally transverse to the dispersion direction 704). These encoding patterns 706a, 706b are adjustable and can be implemented in various ways. Figures 7(a) and 7(b), by way of example and not limitation, describe two preferred methods, respectively.

[0107] As shown in Figure 7(a), a movable light-shielding plate 706a is placed directly behind the slit 700, and the light-shielding plate 706a contains transparent pixels (allowing light to pass through and enter the system, so the pixel weight is "1" in the coding pattern) or opaque pixels (blocking light from entering the system, so the pixel weight in the coding pattern is "0"). Figure 7(a) illustrates the movement of the light-shielding plate 706a perpendicular to the length direction 702 of the slit, and the light-shielding plate 706a contains columns of different coding patterns. As each column moves within the slit 700, a new coding pattern is generated along the slit 700. It should be noted that the movement of the light-shielding plate 706a can also be along the length direction 702 of the slit. In this case, the light-shielding plate design should differ from that shown in Figure 7(a).

[0108] In a second preferred method of spatially encoding the slits as shown in Figure 7(b), a linear array 706b of micromirrors is used. The slits are then formed by the micromirror array 706b itself, with each independently controlled micromirror representing a pixel. When a micromirror is rotated to a specified direction, it reflects light into the spectrometer system, and the pixel represented by that micromirror is in an "on" mode (therefore, the pixel's weight in the encoding pattern is "1"). Alternatively, the micromirror can be rotated to the "opposite" direction, directing light "away" from the spectrometer entrance, thus putting the pixel represented by that micromirror in a "off" mode (therefore, the pixel's weight in the encoding pattern is "0").

[0109] As shown in Figure 7(a), assume that the first encoding pattern is set in the j-th pattern (j = 1 to N), and a jm(m = 1 to N) represents the weight (transparent "1" or opaque "0") of the m-th pixel in the coded pattern. The coded slit should be imaged onto the image plane containing the second coded pattern by dispersive and imaging optics. The corresponding slit image on this image plane is shown in Figure 8(a), where different lighter grayscale colors represent different wavelengths, and the darkest gray represents black spots caused by opaque pixels in the first coded pattern. The light intensity at each spectral component is represented as x. n (n=1~N), by measuring x n To determine the spectrum. It should be noted that, since the dispersive optics are associated with the dispersive direction 800 perpendicular to the slit length direction 801, the slit images of different wavelengths are located at different positions. The second coded pattern 802, as shown in Figure 8(b), contains N×N coded pixels and is aligned with the dispersive slit image. Here, b is used... mn (m = 1 to N and n = 1 to N) represent the weights of pixels with row number m and column number n in the second coded pattern 802 shown in Figure 8(b). Similarly, if a binary mask corresponding to the "transparent" or "opaque" condition of that pixel is used, then b mn The value is either "1" or "0". It should also be noted that the second encoding pattern 802 can also be viewed as a set of N slits arranged one after another with zero gaps along the dispersion direction 800, also referred to here as virtual slits. Each virtual slit is encoded in a direction 801 that is generally transverse to the dispersion direction 800. Those skilled in the art will understand that the second encoding pattern 802, comprising N encoding slits, arranged one after another with narrow, non-zero, opaque gaps in different embodiments, will also operate according to the same principle. Preferably, the gaps are kept as small as possible, but there is no upper limit, although large gaps may degrade performance. In practice, small gaps are unavoidable in implementations, for example, if a micromirror array is used to implement the encoding pattern in the example embodiment, where the gaps between adjacent mirrors are small spaces due to the limitations of microfabrication, this is unavoidable. Clearly, the total transmittance of the light power through the first and second encoders is determined by the superposition of the encoding slit image shown in Figure 8(b) and the second encoding pattern 802. When the first encoding pattern is set to the j-th pattern (j = 1 to N), the signal detected by the single-pixel detector is:

[0110]

[0111] Next, the first encoded pattern is set to the (j+1)th pattern, until a complete set of N measurements is completed. This process again produces a set of N linear equations with N unknowns, which can be conveniently represented in matrix form:

[0112] Y = ABX (6)

[0113] Where Y is a measurement vector (N×1) containing the measurement signal from the single-pixel detector, X is a vector (N×1) containing N spectral components, and A and B are coding matrices (both with N×N dimensions), which depend on the first and second coding patterns, respectively.

[0114] The selection of matrices A and B according to the preferred embodiment will now be described. Let A = S. N and B=S N T S N It is an N-order cyclic S-matrix, where T denotes the matrix transpose. The following characteristics of the cyclic S-matrix can be used to immediately reconstruct the spectrum X, and also to estimate the mean square error in this reconstruction. It has... [9] :

[0115] S N S N T =(I N +J N (N+1) / 4 (7)

[0116] (S N S N T ) -1 =4I N / (N+1)–4J N / (N+1) 2 (8)

[0117] Among them I N J represents an N×N identity matrix. N This represents an N×N matrix where all elements are equal to 1. Spectral reconstruction can be performed using:

[0118] X = (S N S N T ) -1 Y = [4I N / (N+1)–4J N / (N+1) 2 Y (9)

[0119] The mean square error in this reconstruction can be estimated using the following formula. [9] :

[0120] ε=(σ 2 / N)Trace{(S N S N T S N S N T ) -1} (10)

[0121] Where σ 2 This represents the variance of the noise in each measurement, and `trace{}` denotes the trace of the matrix, which is the sum of all diagonal elements. For large numbers N, `Trace{(S...` N S N T S N S N T ) -1} is approximately equal to 16 / N, which indicates that using the proposed method with equation (9), the average mean square error or noise in the reconstructed spectrum is indeed reduced to ε = 16σ. 2 / N 2 Clearly, this reconstruction noise is more significant than the error or noise σ when measuring spectral components one by one. 2 It is much smaller. This is actually the advantage of multiplexing according to this example embodiment, similar to FTIR, which advantageously guarantees enhanced SNR.

[0122] Next, consider the case where the entrance aperture 900 has an extended region as shown in Figure 9(a). Similarly, as shown, the entrance aperture 900 is encoded using first encoding patterns 902a, 902b. Unlike spectrometer systems where the entrance aperture covers a single (encoded) slit (as shown in Figure 7, where the slit contains only a one-dimensional (1D) array of encoded pixels along the slit length), the first encoding patterns 902a, 902b now contain a two-dimensional (2D) array of encoded pixels and the entrance aperture 900 extends across multiple encoded slits, such as 904. To model this system, one can imagine the entrance aperture 900 consisting of multiple encoded slits, such as 904, having the same width Δ, placed one next to another without gaps. Those skilled in the art will understand that these encoded slits, in different embodiments, such as 904, arranged one after another with narrow, non-zero, opaque gaps, will also operate according to the same principle. As mentioned above, the absence of gaps in the encoded slits, such as 904, is also referred to herein as virtual slits. As shown in the figure, each slit, such as 904, contains a column of coded pixels. Or in other words, each slit, such as 904, is encoded along a direction 906 that is generally transverse to the dispersion direction 908.

[0123] Again, as shown in Figure 9(a), assume that the 2D first coded patterns 902a and 902b at this extended entrance aperture 900 are set in the j-th pattern (j = 1 to N) and set a jm (i) (m = 1 to N) represents the coding pattern of the m-th pixel in the i-th slit (or, from the perspective of a two-dimensional array, it is the weight of the pixel with row number m and column number i). Similarly, a jm (i)"1" or "0" corresponds to the transparent or opaque state of a pixel, respectively. According to this example embodiment, the first encoding patterns 902a and 902b are adjustable, and the adjustable 2D encoding patterns can be conveniently implemented using a movable light-shielding plate 902a immediately following the entrance aperture 900 as shown in FIG. 9(a) or a 2D micromirror array 902b (e.g., TI's DMD) as shown in FIG. 9(b).

[0124] The coded entrance aperture is then imaged onto the image plane containing the second coded pattern by dispersive and imaging optics. Because the entrance aperture now contains multiple coded slits, the second coded pattern differs slightly from the coded pattern shown in Figure 8(b) for the case where the entrance aperture has only a single slit. Figure 10 As shown, the second coded pattern 1001b mn The starting portion 1000 of (m = 1 to N and n = 1 to N) is aligned with the dispersive image 1002 of the first slit to encode the light passing through the first slit. Then, as... Figure 10 As shown, the remainder of the second coded pattern 1004 is repeated periodically. In other words, b m(n+N) =b mn Here, the weight b in the encoded pattern mn The value is "1" or "0", which corresponds to the transparent or opaque state of the pixel in the m-th row and n-th column of the second encoding mask, respectively.

[0125] Without loss of generality, consider the i-th coded slit in the entrance aperture as shown in Figure 9(a). It is spatially shifted to the right by a distance (i-1)Δ relative to the first slit, where Δ is the width of the slit. Then, the dispersive image 1006 of the i-th slit on the second coded pattern at each wavelength is also shifted by a distance (i-1)βΔ relative to the image of the first slit at the same wavelength, where β is the dispersive and optical magnification of the imaging optics. Figure 10 The diagram schematically illustrates the dispersive image 1006 of the i-th slit, where different grayscale colors represent its spectral components from 1 to N. Clearly, due to the offset, it cannot be guaranteed that the dispersive image of the i-th slit at each spectral component is aligned with the encoded pixels on the second encoder pattern 1001. This can be achieved through... Figure 10 Mid-wavelength λ n As can be seen from the i-th slit image 1008 highlighted here, the slit image 1008 is actually encoded by two adjacent column encoding patterns 1001 on the second encoder, where each encoding pattern contributes a portion.

[0126] As previously stated, according to the example embodiment, the entrance aperture is uniformly illuminated, and for wavelength n, the intensity of the spectral component is x. nTo simulate the percentage of light power incident on the i-th slit, this light power can be transmitted through the first and second encoders and reach the single-pixel photodetector. The integer l is defined as follows:

[0127] l=floor{(i-1)βΔ / D} (11)

[0128] The function floor{x} rounds object x to the nearest integer in the direction of negative infinity, and D is the pixel width in the second coded pattern. The total transmittable power incident on the i-th slit is:

[0129]

[0130] Where i represents the i-th slit, j represents the case where the first encoder is set in the j-th pattern, c is a constant related to the encoding pixel area and the loss in the optical system, and α1 i and α2 i These are two coefficients representing the percentage contribution from two adjacent coded columns. α1 i and α2 i The following can be estimated:

[0131]

[0132]

[0133] The optical power P reaching the single-pixel detector j i The resulting signal

[0134]

[0135] Where η is the sensitivity of the photodetector. In equation (15) above, it is assumed that η is independent of wavelength. However, this assumption does not lose any generality, because the spectral correlation sensitivity of the photodetector can always be calibrated and subsequently removed using a light source with a known spectrum. Since the application we are considering here is spectroscopy, therefore x n The relative values ​​of are important, while their absolute values ​​are not. Therefore, the constant ηc in equation (15) can be removed, resulting in a simpler form:

[0136]

[0137] After all N measurements are completed (i.e., the first coded pattern is adjusted from j=1 to j=N), the above linear equation can be expressed in matrix form as follows:

[0138]

[0139] Where Y iX is a vector containing the measurement signal from the i-th slit, and A is a vector containing the spectrum. (i) B and R are encoding matrices that depend on the first and second encoding patterns, respectively, and R is a shift matrix defined as follows:

[0140]

[0141] R l =RRR…R, that is, matrix multiplication by l. Equation (17) is obtained based on the following facts: (i) the periodicity of the second encoder, i.e. b m(n+N) =b mn and (ii) Properties of the shift matrix:

[0142] The first row and

[0143] Similarly, according to the example embodiment, the first and second encoding patterns can be designed such that matrix A (i) Since B and A are cyclic, according to the characteristics of cyclic matrices, AB is also a cyclic matrix. Using the properties of cyclic matrices and shift matrices, it is easy to obtain (A... (i) B)R l =R l (A (i) B), therefore, equation (17) can be rewritten as:

[0144]

[0145] Considering that the entrance aperture contains multiple coded slits, the overall signal output of the single-pixel detector Z is:

[0146] Z = ∑ i Y i (twenty one)

[0147] The spectrum can then be reconstructed from equation (21). Before discussing the details of this spectral reconstruction, the flux and spectral resolution trade-offs in existing grating-based dispersive spectrometer systems will be considered, which will help to understand how the method according to the example embodiment can successfully overcome this trade-off.

[0148] Figure 11 schematically illustrates a typical conventional dispersive spectrometer 1100. As shown, three points 1, 2, and 3 (point 2 at the center, points 1 and 3 at the edges) in the entrance slit 1102 are scattered and imaged onto its focal plane by the dispersive and imaging optics 1104. At wavelength λ, the images of these three points are 1', 2', and 3', respectively. From wavelength λ1 to λ2... NThe total separation distance of the dispersive slit image within the spectral band of interest is L. This separation distance L is measured along the dispersion direction. As shown in Figure 11, due to diffraction and aberrations in the optical device 1104, the focused light has a finite spot size S. As shown in Figure 11(a), if the slit image at wavelength λ (i.e., βΔ, where β is the optical magnification and Δ is the slit width) is larger than the spot size S, the spectral resolution is mainly determined by the slit width, and the total number of resolvable spectral components is approximately equal to L / (βΔ). As shown in Figure 11(b), if the slit image is much smaller than the spot size on its focal plane, the spectral resolution is mainly determined by the spot size, and the total number of resolvable spectral components is approximately equal to L / S. It is clear from Figure 11 that in order to achieve more resolvable spectral components in the obtained spectrum (or in other words, better spectral resolution of the spectrometer), a smaller slit width is required. However, a smaller slit width also limits the amount of light that can be collected (i.e., the flux), resulting in a lower SNR. This represents a typical trade-off between throughput and spectral resolution in traditional dispersive spectrometers.

[0149] However, advantageously, the aforementioned trade-off does not exist in embodiments of the invention because the entrance aperture can comprise multiple coded slits. In the spectrometer system according to the exemplary embodiment, the image width of a single slit (βΔ) can be selected to be smaller than the spot size, such that it does not affect the spectral resolution and the number of resolvable spectral components is therefore equal to L / S, thus enabling the spectrometer to achieve high spectral resolution. However, the throughput of the spectrometer system depends on the total area of ​​the entrance aperture, not the area of ​​a single slit. And because embodiments of the invention can have multiple slits in the entrance aperture, the throughput of the spectrometer according to the exemplary embodiment can still be high even if it is designed to have high spectral resolution.

[0150] The following describes one of the possible designs of the coding pattern according to the example embodiments to verify that the spectrum can be easily reconstructed using equation (21). It should be noted that the invention is not limited to these examples, and as those skilled in the art will understand, many other possible coding pattern designs can achieve the same goal of reconstructing the spectrum using equation (21) in various other example embodiments.

[0151] In an example design, matrix A in equation (20) is set. (i) The same for all i, or A (i) =A, which means that all the slits in the entrance aperture are coded with the same pattern. Then equation (20) becomes,

[0152]

[0153] In addition, let A = S N S NIt is an N-order cyclic S-matrix. In other words, all entrance slits are adjusted in the same way according to the same cyclic S-matrix. Such an adjustable encoder for the entrance aperture according to the example embodiment can be easily implemented using, for example, a MEMS micromirror array or a movable light-shielding plate combined with a fixed aperture opening, the latter in… Figure 12 The diagram is schematically shown. As shown, the light-shielding plate 1200 consists of a set of opening and closing slots, arranged parallel to the dispersion direction of the instrument along its length. As shown, the movement of the light-shielding plate 1200 is perpendicular to the dispersion direction. In this way, when the light-shielding plate 1200 moves by a predetermined amount, for example, a step equal to the width of the slot, a new encoding pattern is changed. Following the above modeling method, this encoder can be considered as consisting of many tiny virtual slits, such as... Figure 12 As shown, these slits are placed side-by-side without any gaps. Clearly, this design according to the example embodiment satisfies the encoding matrix A. (i) The same applies to all slits (i = 1, 2, ...). In the exit aperture plane, the fixed coding pattern of the second encoder is designed such that the coding matrix B = S. N T , where T represents the matrix transpose.

[0154] Figure 13 A typical row of the second encoding pattern is further illustrated schematically, where D represents the width of the encoded pixel. For example... Figure 13 As shown, the dispersive image of the first slit in the entrance aperture spans a first portion of the second encoder, comprising N pixels from the 1st to the Nth pixel. Subsequent portions of the second encoder, starting from the (N+1)th pixel, are simply periodic repetitions of the first portion. It can be seen that a preferred choice for the pixel width D is to make it equal to the spot size. This is because if D is larger than the spot size, the spectrometer's resolution can still be improved by decreasing D. If D is equal to or smaller than the spot size, decreasing the value of D will not further improve the spectral resolution. According to the example embodiment, the spot size affected by the diffraction limit (Airy spot size) and the aberrations presented in the optical system will ultimately determine the spectrometer's spectral resolution. Figure 13 In the case shown, the number of resolvable spectral components is equal to N.

[0155] As previously described, the entrance aperture according to the example embodiment comprises multiple narrow virtual slits, each with a small width Δ. Therefore, at the exit aperture plane, the dispersive image of one slit at each wavelength is moved a distance βΔ along the dispersive direction to the dispersive image of the adjacent slit of the same wavelength. As shown, βΔ is much smaller than the pixel size D in the example embodiment. In fact, the design can be adapted according to the exemplary embodiment such that D / (βΔ) is an integer p, i.e., p = D / (βΔ), and the integer p is greater than 1. Then, the relative positions of the images of the first, second, and third virtual slits on the second encoder plane are... Figure 13 The following is explained. Using equations (13), (14), and (22), the signal contribution of each slit can be determined as follows:

[0156] For the first slit, i = 1, l = 0, α1 1 =0,α2 1 =1, and,

[0157] Y 1 =R 0 ABX (23)

[0158] Where R 0 It equals the identity matrix I. For the second slit, i = 2, l = 0, α1 2 =1 / p,α2 2 = (p-1) / p, and,

[0159]

[0160] Similarly, for the third slit, i = 3, l = 0, α1 3 =2 / p,α2 3 = (p-2) / p, and,

[0161]

[0162] Continuing in this manner, for the p-th slit, i = p, l = 0, α1 p =(p-1) / p,α2 p =1 / p, and,

[0163]

[0164] Then, for the next (p+1)th slit, i = p+1, l = 1, α1 p+1 =0,α2 p+1 =1, and,

[0165] Y p+1 =R 1 ABX (27)

[0166] For the (p+2)th slit, i = p+2, l = 1, α1 p+2 =1 / p,α2 p+2 = (p-1) / p, and,

[0167]

[0168] Following this pattern, it's easy to see that if there are a total of (pq+1) slits, where q is also an integer and q>1, then the total signal Z from the single-pixel detector is the sum of all signals from all slits. Furthermore, Z can be written as follows:

[0169]

[0170] After simplifying the above formula, we can obtain:

[0171]

[0172] Since p >> 1, a further approximation is as follows:

[0173]

[0174] Considering that the spectral vector X is related to the relative intensity, the wavelength-independent number p can be removed, thus,

[0175]

[0176] Define the following cyclic matrix C,

[0177]

[0178] The first row of C is determined by the following formula.

[0179]

[0180] Then, equation (32) can be rewritten in the following form, taking into account the matrix A = S. N and B=S N T ,

[0181] Z = CABX = CS N S N T X (35)

[0182] Since C is a cyclic matrix, it is usually not singular if N is odd, which is typically the case for cyclic S matrices. Therefore, C is invertible and can be easily computed using the method provided in reference

[16] . -1 Clearly, the spectrum X can be conveniently calculated using the above formula.

[0183] X = (S N S N T ) -1 C -1 Z = [4I N / (N+1)–4J N / (N+1) 2 C -1 Z (36)

[0184] Figure 14 It shows according to Figure 12 The illustrative example embodiment demonstrates one of the many possible ways to design a coding mechanism. Figure 14 (a) and (b) show the front view and cross-sectional view, respectively. As shown, the encoding mechanism mainly consists of two layers 1400 and 1402, where layer 1400 has an entrance aperture 1404 and layer 1402 has a movable light shield 1406. The entrance aperture 1404 is fixed, while the movable light shield 1406 is suspended by a series of bending springs 1408, which restrict the movement of the light shield 1406 to only one degree of freedom (1DOF) in the desired direction. This two-layer structure can be manufactured in a variety of ways, including but not limited to additive manufacturing (i.e., 3D printing), precision machining, and MEMS microfabrication processes. To achieve high encoding speed and a large encoding area (i.e., a large range of motion for the light shield 1406), the suspended light shield 1406 can be driven by at least one actuator 1410 to oscillate at its natural frequency. Since the light-shielding plate 1406 performs continuous movement, a triggering mechanism 1411, including a sensor structure, is preferably used to accurately detect the position of the light-shielding plate 1406 to trigger a data sampling process to generate the desired coded pattern when the light-shielding plate moves to a designated position. As an example, the sensor structure can be a piezoresistive (i.e., strain gauge), electromagnetic, piezoelectric, or optical position sensor, and they can be directly integrated onto the encoder. Figure 14 In the structural cross-sectional schematic diagram shown in (c), the optical position sensor according to the preferred embodiment includes a light-emitting diode (LED) 1414, a photodetector (PD) 1416, and a pair of gratings 1418 and 1420 as shown as an example. The relative motion of the gratings 1418 and 1420 generates an oscillating electrical signal that can be used to determine the position of the light-shielding plate 1406, and thus helps to trigger the data acquisition process (e.g., comparison) at the correct position of the light-shielding plate 1406 in the spectrometer system according to the example embodiment. Figure 6 (to record signals from a single-pixel detector.)

[0185] For reference again Figure 6 The schematic diagram of the spectrometer 600 of the example embodiment shown will now be described, with advantages in further miniaturization according to another example embodiment. Figure 15As shown, in the spectrometer 1500 according to such an embodiment, the dispersive and imaging optics and the collecting optics share the same optical setup 1502. Light 1504 entering the coded entrance aperture 1506 is scattered and imaged onto the coded exit aperture 1508 by the dispersive and imaging optics having two mirrors 1510, 1512 and a dispersive element (i.e., a diffraction grating 1514). The encoding mechanism is slightly modified at the exit aperture 1508 where the second encoding shield is located. Here, if the coded pixel is transmissive, allowing light to pass through and be absorbed by the absorber 1516, the pixel has a weight of "0", or in other words, the pixel is in a "closed" state. Conversely, if the coded pixel is reflective, reflecting light back, the pixel has a weight of "1", or in other words, the pixel is in a "closed" state. The reflected light returns in the opposite direction along almost the same path as its incident direction. Therefore, the same mirrors 1510, 1512 and grating 1514 now act as collecting optics. (Refer to...) Figure 6 In the described embodiment, this reverse process dedisperses the light beams and focuses them to form an image 1518 of the entrance aperture near their physical location. Figure 15 As shown, to separate the entrance aperture and its image 1518, a mirror 1520 can be used to reflect the image 1518 of the entrance aperture in different directions. The mirror 1520 can be positioned slightly above or below the initial forward optical path to avoid blocking the incident light, and to facilitate this, preferably, the reflected light rays extend slightly beyond the plane, which can be easily achieved by tilting the second encoder at the coded exit aperture 1508. It should be noted that the size of the final focused image 1518 is approximately the same as the size of the entrance aperture 1506. If the final image 1518 is still larger than the photosensitive area of ​​the detector, a condenser and focusing optics can be inserted between them to match their sizes.

[0186] Figure 15 The design shown is advantageous because it is compatible with... Figure 6 Compared to the design shown, which uses a photodetector and its light-collecting optics directly after the second encoder at the exit aperture, this design allows for the use of a much smaller photodetector and its light-collecting optics. This is because the exit aperture of a high-resolution spectrometer is typically much larger than the entrance aperture due to dispersion. Smaller detectors are generally preferred, not only because of their size but also because of their high-speed response and low dark current. Figure 15 The example embodiment in which the same optical setup is shared between the dispersive and imaging optics and the collection optics, but uses a design that uses a angular reflector instead of an coded exit aperture in the reflection mode, can be referenced in reference

[17] , which relates to the spectrometer type described above with reference to FIG4(a).

[0187] Other example embodiments are now described to illustrate detection preferably using a single-pixel detector with a small photosensitive area. For example... Figure 16 As shown, light 1600 enters system 1602 through coded entrance aperture 1604 and propagates through a dispersion and imaging system with two mirrors A and B and a grating 1606, further generating a dispersion image of the entrance aperture at the plane of exit aperture 1608 where the second encoder is located. Here, the second encoder 1608 is in transmission mode. In other words, the transmission pixels allow light to pass through and have a weight of "1", while the opaque pixels block light and therefore have a weight of "0". Light passing through coded exit aperture 1609 is collected by collection optics 1610 and focused onto single-pixel photodetector 1612. In this embodiment, imaging collection optics 1610 are used, where light propagates through collection optics 1610 and forms an image of the grating on which the photosensitive area of ​​single-pixel detector 1612 is placed. It should be noted that although the dispersive image of the entrance aperture 1604 can span a large area on the second encoder 1608 to achieve high spectral resolution, as light propagates through the collecting optics 1610, the light after the collecting optics 1610 converges and overlaps to form a grating image. This process also effectively dedisperses the light beams and focuses them to form the grating image. Furthermore, the collecting optics 1610 can be designed to produce a reduced image of the diffraction grating. Therefore, this design of the collecting optics 1610 in this example embodiment is advantageous in reducing the photosensitive area of ​​the usable detector 1612.

[0188] In addition, such as Figure 16 As shown, field lens B can be placed close to exit aperture 1608 to further reduce the image of the grating, thereby reducing the size of detector 1610 or the size of collecting optics 1610 to make it more compact. Furthermore, field lens A can be placed close to entrance aperture to image the exit pupil of the preceding optics 1614 (i.e., illumination optics, etc.) onto diffraction grating 1606, which improves the light utilization efficiency of the entire spectrometer system 1602 in this example embodiment. It should be noted that such field lenses are located at or near entrance aperture 1604 and exit aperture 1608 where light is focused, therefore they have little or no effect on the resolution and F-number of the spectrometer.

[0189] Figure 17An example embodiment is also shown, which can also achieve detection using single-pixel detectors 1700, 1702 with small photosensitive areas. In this embodiment, non-imaging collection optics are used. As shown, light 1704 enters the system 1706 through the coded entrance aperture 1708 and propagates through the dispersion and imaging system, generating a dispersive image of the entrance aperture 1708 at the exit aperture plane where the second encoder 1710 is located. Here, the second encoder 1710 also operates in transmission mode. The coded light passing through the encoder 1710 is collected by non-imaging collection optics in the form of condensers 1712, 1714. Typically, to achieve high spectral resolution, the dispersive image of the entrance aperture can span a large area on the second encoder. If only a single detector 1700 is used, focusing from a large area on the second encoder to a small photosensitive area on the single-pixel detector 1700 may require a very large light collection ratio, making condenser design challenging. However, in this exemplary embodiment, this problem can be conveniently solved by using multiple single-pixel detectors 1700, 1702. like Figure 17 As shown, the area on the second encoder 1710 that collects light is further divided into multiple detection areas, with smaller regions such as region 1, region 2, etc. Each region is then aligned with a condenser 1712, 1714, which collects the light passing through it and converges it to the single-pixel detectors 1700, 1702. This reduces the focusing ratio of each condenser 1712, 1714, making its design easier. In this embodiment, the total encoded light energy transmitted through the spectrometer system 1706 can be equal to the sum of the signals from all the single-pixel photodetectors 1700, 1702, and the spectrum can then be reconstructed using the methods described in this invention.

[0190] It should be noted that although squares and rectangles are used schematically in the illustrations of the various embodiments described to represent the coded pixels of the first and second encoders, this is not necessarily the case in practice. In fact, in some embodiments, using distorted coded patterns may be advantageous, especially for the second encoder. This can be understood as follows: Suppose a first encoder is formed at the entrance aperture using straight slits and square or rectangular coded pixels in these slits. Due to dispersion and residual aberrations of the imaging system, the image of these slits may be distorted in the exit aperture plane. In this case, as... Figure 18 As shown, the second coded pattern 1800 at the exit aperture plane can be designed to be distorted accordingly to match the distorted image of the slits, such as the dispersive image of the first slit 1802. In this way, the residual optical aberrations have a much smaller impact on the spectrometer performance.

[0191] In summary, embodiments of the present invention provide a method and system for constructing a novel spectrometer that uses a single-pixel detector instead of a detector array and encodes the incident and exit apertures to achieve high throughput and high spectral resolution, while facilitating miniaturization for portable field use. Both the incident and exit apertures are encoded in a direction generally transverse to the dispersion direction to decouple throughput from spectral resolution, thereby allowing the spectrometer to be easily designed with high performance in a compact package. Miniaturization of the spectrometer can be further achieved using MEMS technology. Such spectrometers have a wide range of applications.

[0192] In one embodiment, a spectrometer is provided for detecting electromagnetic (EM) spectra having one or more wavelength components within a spectral band of interest, comprising an entrance aperture; a dispersive and imaging optics including at least one dispersive element; an exit aperture; a collection optics; and at least one single-pixel detector, each single-pixel detector being sensitive to one or more wavelength components. The entrance aperture includes at least one entrance slit spatially encoded along a direction generally transverse to the dispersion direction of the dispersive and imaging optics; the dispersive and imaging optics are configured to generate a dispersive image of the entrance aperture on the plane of the exit aperture, such that the images at different wavelength components are offset by different amounts along the dispersion direction; the exit aperture includes a plurality of exit slits arranged along the dispersion direction, each exit slit being spatially encoded along a direction generally transverse to the dispersion direction; a collection optics are configured to collect the total electromagnetic energy entering and leaving the entrance aperture to at least one of the single-pixel detectors; at least one of the encoding patterns of the at least one entrance slit and the encoding patterns of the plurality of exit slits is adjustable and configured to be changed multiple times; the spectrometer further includes a measurement unit configured to measure the output of the at least one detector for a corresponding number of times the EM spectrum is reconstructed.

[0193] Adjustable coded patterns for at least one entrance slit and / or multiple exit slits can be implemented using microelectromechanical systems (MEMS) technology. Adjustable coded patterns for at least one entrance slit and / or multiple exit slits can also be implemented using MEMS micromirror arrays.

[0194] An adjustable coding pattern for at least one entrance slit and / or multiple exit slits can be achieved using a movable light-shielding plate placed near a fixed aperture opening. The movable light-shielding plate can move in a preferred direction, preferably along or perpendicular to the dispersion direction, or any other direction. The movable light-shielding plate can be suspended by springs, and the plate is configured to be driven into an oscillating motion at its natural frequency to achieve the advantages of high-speed and large-area coding.

[0195] Dispersive and imaging optics, as well as collecting optics, can be configured to share some or all of the same optical elements. The coded pattern of multiple exit slits can be configured to operate in reflection mode.

[0196] The collecting optics may include an imaging collecting optics that forms an image of at least one dispersive element on the photosensitive area of ​​at least one single-pixel detector by an imaging device. The spectrometer may include a first field lens B placed adjacent to the exit aperture to further reduce the image on the photosensitive area of ​​at least one single-pixel detector or reduce the size of the collecting optics.

[0197] The spectrometer may also include a second field lens positioned adjacent to the entrance aperture to image the exit pupil of the front optics onto at least one dispersive element.

[0198] The spectrometer may include non-imaging collection optics, such as a condenser that focuses light from an exit aperture onto a photosensitive region of at least one single-pixel detector. The non-imaging collection optics may include multiple condensers, each focusing a portion of the light from the exit aperture onto a photosensitive region of a corresponding one of the multiple detectors.

[0199] At least the coded pixels of the exit aperture can be configured to match the distorted image of the slit of the entrance aperture.

[0200] An entrance aperture may include multiple entrance slits that are either not spaced apart from each other or have narrow, non-zero gaps.

[0201] An exit orifice may include multiple exit slits that are either not gapped together or have a narrow, non-zero gap.

[0202] Figure 19A flowchart 1900 illustrates a method for detecting electromagnetic (EM) wave spectra having one or more wavelength components within a spectral band of interest using an entrance aperture according to an example embodiment; a dispersive and imaging optics comprising at least one dispersive element; an exit aperture; a collection optics; and at least one single-pixel detector, each single-pixel detector being sensitive to one or more wavelength components. In step 1902, at least one entrance slit of the entrance aperture is spatially encoded along a direction generally transverse to the dispersive direction of the dispersive and imaging optics. In step 1904, a dispersive image of the entrance aperture is created on the plane of the exit aperture using the dispersive and imaging optics, such that the respective images at different wavelength components are offset by different amounts of displacement along the dispersive direction. In step 1906, a plurality of exit slits of the exit aperture are spatially encoded along a direction generally transverse to the dispersive direction, wherein the exit aperture comprises a plurality of exit slits arranged along the dispersive direction. In step 1908, the total EM wave energy entering and exiting the entrance aperture to reach one of the at least one single-pixel detectors is collected using the collection optics. In step 1910, at least one of the coding patterns of at least one entrance slit and the coding patterns of multiple exit slits is changed multiple times. In step 1912, the output of at least one detector is measured for a corresponding number of times the EM spectrum is reconstructed.

[0203] Embodiments of the present invention may have one or more of the following features and related benefits / advantages.

[0204]

[0205]

[0206] In other words, the spectrometer system according to the example embodiment may have one or more of the following significant advantages.

[0207] Compared to dispersive spectrometers, it has the advantages of FTIR spectrometers, namely multiplexing, thus supporting high SNR detection and the use of single-element photodetectors, enabling cost-effective operation at IR wavelengths.

[0208] Compared to FTIR spectrometers, it has the advantages of dispersive spectrometers. More specifically, this technology is not based on optical interferometers, making it more robust and less sensitive to external interference, thus facilitating miniaturization for field use.

[0209] Most importantly, compared to dispersive grating spectrometers or FTIR spectrometers, the spectrometer system proposed in this invention has a significantly larger entrance aperture, enabling it to collect more optical power and subsequently provide enhanced SNR.

[0210] Therefore, embodiments of the present invention can potentially realize high-spectral-resolution portable IR spectrometers, which can facilitate real-time on-site detection and analysis of materials. Due to the aforementioned advantages, spectrometers constructed based on the present invention should be robust, less sensitive to external interference, possess high spectral resolution within the spectral band of interest, and have high throughput (i.e., the ability to collect more light energy). Embodiments of the present invention can be constructed at a relatively low cost. Application opportunities include spectrometer modules combined with stand-alone handheld modules or portable smartphones, providing a range of new possibilities for sensing in fields such as agriculture, environmental monitoring, healthcare, and industrial process monitoring.

[0211] For example, embodiments of the present invention can be used to develop handheld infrared or Raman spectrometers with high spectral resolution for field use in a range of applications, including but not limited to industrial process control, material (especially polymers and plastics) identification and verification, environmental monitoring, Internet of Things (IoT) sensors, biomedical point-of-care testing, food and beverage quality assessment, coating and film analysis, forensics and law enforcement, and drug research and development.

[0212] Aspects of the systems and methods described herein, such as the movement of components of a spectrometer and the analysis of detected signals, can be implemented as functions programmed into any of a variety of circuits, including programmable logic devices (PLDs), such as field-programmable gate arrays (FPGAs), programmable array logic (PAL) devices, electrically programmable logic and memory devices, and standard cell-based devices, as well as application-specific integrated circuits (ASICs). Some other possibilities for implementing system aspects include: microcontrollers with memory (e.g., electrically erasable programmable read-only memory (EEPROM)), embedded microprocessors, firmware, software, etc. Furthermore, system aspects can be embodied in microprocessors with software based on circuit simulation, discrete logic (sequential and combinational), custom devices, fuzzy (neural) logic, quantum devices, and hybrids of any of the above device types. Of course, the underlying device technologies can be provided in various component types, such as metal-oxide-semiconductor field-effect transistor (MOSFET) technologies, such as complementary metal-oxide-semiconductor (CMOS), bipolar technologies such as emitter-coupled logic (ECL), polymer technologies (such as silicon conjugated polymers and metal conjugated polymer-metal structures), analog and digital hybrids, etc.

[0213] The illustrative embodiments of the systems and methods described above are not intended to be exhaustive or to limit the systems and methods to the precise forms disclosed. While specific embodiments and examples of system components and methods have been described herein for illustrative purposes, various equivalent modifications are possible within the scope of the systems, components, and methods, as will be recognized by those skilled in the art. The teachings of the systems and methods provided herein can be applied to other processing systems and methods, and are not limited to those described above.

[0214] Elements and actions of the various embodiments described above can be combined to provide further embodiments. Based on the detailed description above, these and other changes can be made to the system and method.

[0215] Generally, the terminology used in the following claims should not be construed as limiting the systems and methods to the specific embodiments disclosed in the specification and claims, but should be interpreted as encompassing all processing systems operating under the claims. Therefore, the systems and methods are not limited by this disclosure, and their scope is fully defined by the claims.

[0216] Unless the context explicitly requires otherwise, throughout the specification and claims, the words “comprising,” “including,” “including,” etc., shall be interpreted in an inclusive sense, not an exclusive or exhaustive sense; that is, in the sense of “including but not limited to.” The use of singular or plural terms shall also include the plural or singular, respectively. Furthermore, the words “here,” “below,” “above,” “and above,” and similar terms refer to the entire application, not any particular part of it. When the word “or” is used to refer to a list of two or more items, the word covers all of the following interpretations: any item in the list, all items in the list, and any combination of items in the list.

[0217] References:

[0218] [1] RJ Bell, Introductory Fourier Transform Spectroscopy, Academic Press, New York and London, 1972.

[0219] [2]http: / / sa-photonics.com / products / spectroscopy-systems / modular-ir-fourier-spectrometers-2 /

[0220] [3] MEGehm, STMcCain, NPPitsianis, DJBrady, P.Potuluri, andM.E.Sullivan, "Static two-dimensional aperture coding for multimodal, multiplexspectroscopy," Applied Optics, Vol.45, No.13, p.2965 (2006).

[0221] [4]R.N.Ibbett,D.Aspinall,and J.F.Grainger,“Real-Time Multiplexing ofDispersed Spectra in Any Wavelength Region,”Applied Optics,Vol.7,No.6,p.1089(1968).

[0222] [5]M.J.E.Golay,“Static multislit spectrometry,and its application tothe panoramic display of infrared spectra,”J.O.S.A,Vol.41,p 468(1951).

[0223] [6]A.Girard,“Spectrometre a Grilles,”Applied Optics,Vol.2,No.1,p.79(1963).

[0224] [7]M.Harwit,P.G.Phillips,T.Fine,and N.J.A.Sloane,“Doubly MultiplexedDispersive Spectrometers,”Applied Optics,Vol.9,No.5,p.1149(1970).

[0225] [8]M.Harwit,P.G.Phillips,L.W.King,and D.A.Briotta Jr.,“Two AsymmetricHadamard Transform Spectrometers,”Applied Optics,Vol.13,No.11,p.2669(1974).

[0226] [9]M.Harwit and N.J.A.Sloane,“Hadamard Transform Optics,”AcademicPress,New York,1979.

[0227]

[10] Z.Lu,J.Zhang,H.Liu,J.Xu,and J.Li,“The Improvement on thePerformance of DMD Hadamard Transform Near-Infrared Spectrometer by DoubleFilter Strategy and a New Hadamard Mask,”Micromachines 2019,10,149;doi:10.3390 / mi10020149.

[0228]

[11] P.Nelson,“ Technology for Spectroscopy,”white paper from TI,online:http: / / www.ti.com / lit / wp / dlpa048a / dlpa048a.pdf

[0229]

[12] B.Lee,“Introduction to±12 Degree Orthogonal Digital MicromirrorDevices(DMDs),”online document from TI,http: / / www.ti.com / lit / an / dlpa008b / dlpa008b.pdf

[0230]

[13] https: / / en.wikipedia.org / wiki / Nonimaging_optics

[0231]

[14] E.J.Candès,J.K.Romberg,and T.Tao,″Stable signal recovery fromincomplete and inaccurate measurements,″Communications on Pure and AppliedMathematics,Vol.59(8),p.1207–1223(2006).

[0232]

[15] https: / / en.wikipedia.org / wiki / Moore%E2%80%93Penrose_inverse

[0233]

[16] https: / / en.wikipedia.org / wiki / Circulant_matrix

[0234]

[17] J.A.Decker Jr.,“Experimental Realization ofthe MultiplexAdvantage with a Hadamard-Transform Spectrometer,”Applied Optics,Vol.10,No.3,p.510(1971).

Claims

1. A spectrometer for detecting electromagnetic spectra having one or more wavelength components within a spectral band of interest, comprising: Entrance aperture; Dispersive and imaging optics that include at least one dispersive element; Exit port; Collect optical components; and At least one single-pixel detector, each single-pixel detector being sensitive to one or more of the wavelength components; The entrance aperture includes at least one entrance slit, which is spatially encoded along a direction generally transverse to the dispersion direction of the dispersion and imaging optics. The dispersive and imaging optics are configured to create a dispersive image of the entrance aperture on the plane of the exit aperture, such that each image at different wavelength components is offset by a different amount of displacement along the dispersive direction. The exit aperture includes a plurality of exit slits arranged along the dispersion direction, wherein each exit slit is spatially encoded along the direction generally transverse to the dispersion direction; The collecting optics are configured to collect the total electromagnetic energy entering the entrance aperture and exiting the exit aperture to one of the at least one single-pixel detectors; Wherein, at least one of the coding patterns of the at least one entrance slit and the coding patterns of the plurality of exit slits is adjustable and configured to be changed multiple times; The spectrometer further includes a measurement unit configured to measure the output of the at least one single-pixel detector for a corresponding number of times the electromagnetic spectrum is reconstructed; and The spectrometer is configured such that the image width of a single entrance slit on a single wavelength component is smaller than the spot size of the spectrometer at the exit aperture, thereby making the spectral resolution of the spectrometer independent of the number of single entrance slits contained in the entrance aperture.

2. The spectrometer according to claim 1, wherein, The adjustable coding pattern of the at least one entrance slit and / or the plurality of exit slits is implemented using microelectromechanical systems (MEMS) technology.

3. The spectrometer according to claim 2, wherein, The adjustable coding pattern of the at least one entrance slit and / or the plurality of exit slits is implemented using a microelectromechanical system micromirror array.

4. The spectrometer according to claim 1, wherein, The adjustable coding pattern of the at least one entrance slit and / or the plurality of exit slits is achieved using a movable light shield placed near the fixed aperture opening.

5. The spectrometer according to claim 4, wherein, The movable light-shielding plate can move in a preferred direction, which is along the dispersion direction, perpendicular to the dispersion direction, or any other direction.

6. The spectrometer according to claim 4, wherein, The movable light shield is suspended by a spring and is configured to be driven into an oscillating motion at its natural frequency to achieve the advantages of high-speed and large-area encoding.

7. The spectrometer according to claim 5, wherein, The movable light shield is suspended by a spring and is configured to be driven into an oscillating motion at its natural frequency to achieve the advantages of high-speed and large-area encoding.

8. The spectrometer according to any one of the preceding claims, wherein, The dispersive and imaging optics and the collecting optics are configured to share some or all of the same optical elements.

9. The spectrometer according to claim 8, wherein, The coded patterns of the plurality of exit slits are configured to operate in reflection mode.

10. The spectrometer according to any one of claims 1-7, wherein, The collecting optics includes an imaging collecting optics that forms an image of the at least one dispersive element on the photosensitive region of the at least one single-pixel detector.

11. The spectrometer according to claim 10, wherein, The spectrometer includes a first field lens placed adjacent to the exit aperture to further reduce the image on the photosensitive region of the at least one single-pixel detector or reduce the size of the collecting optics.

12. The spectrometer according to any one of claims 1-7, wherein, The spectrometer further includes a second field lens placed adjacent to the entrance aperture to image the exit pupil of the preceding optics onto the at least one dispersive element.

13. The spectrometer according to any one of claims 1-7, wherein, The spectrometer includes non-imaging collection optics, such as a concentrator that focuses light from the exit aperture onto the photosensitive region of the at least one single-pixel detector.

14. The spectrometer according to claim 13, wherein, The non-imaging collection optics includes multiple concentrators, each concentrator focusing a portion of the light from the exit aperture onto the photosensitive region of a corresponding one of the multiple single-pixel detectors.

15. The spectrometer according to any one of claims 1-7, wherein, At least the coded pixels for the exit aperture are configured to match the distorted image of the slit of the entrance aperture.

16. The spectrometer according to any one of claims 1-7, wherein, The entrance aperture includes a plurality of entrance slits, wherein there is no gap between the plurality of entrance slits or there is a narrow non-zero gap.

17. The spectrometer according to any one of claims 1-7, wherein, The exit aperture includes multiple exit slits, and the multiple exit slits have no gap or have a narrow non-zero gap between them.

18. A method for detecting an electromagnetic spectrum having one or more wavelength components within a spectral band of interest, using: Entrance aperture; Dispersive and imaging optics that include at least one dispersive element; Exit port; Collect optical components; and At least one single-pixel detector, each single-pixel detector being sensitive to one or more of the wavelength components; The method includes the following steps: At least one entrance slit of the entrance aperture is spatially encoded along a direction generally transverse to the dispersion direction of the dispersive and imaging optics. Using the dispersive and imaging optics, a dispersive image of the entrance aperture is created on the plane of the exit aperture, such that the images of different wavelength components are offset by different displacements along the dispersive direction. The plurality of exit slits of the exit aperture are spatially encoded in a direction generally transverse to the dispersion direction, wherein the exit aperture includes a plurality of exit slits arranged along the dispersion direction. The collecting optics are used to collect the total electromagnetic energy entering the entrance aperture and exiting the exit aperture to one of the at least one single-pixel detectors, wherein the image width of a single entrance slit in a single wavelength component is smaller than the spot size of the spectrometer at the exit aperture, thereby making the spectral resolution of the spectrometer independent of the number of single entrance slits contained in the entrance aperture. The coding pattern of the at least one entrance slit and the coding pattern of the plurality of exit slits are changed multiple times; and The output of the at least one single-pixel detector is measured for the corresponding number of times the electromagnetic spectrum is reconstructed.

Images