Metasurface for spectrum detection, spectrum detection element and spectrum camera

Abstract

The utility model discloses a metasurface for spectrum detection, a spectrum detection element and a spectrum camera. The metasurface comprises a substrate and a micro-nano structure arranged on the substrate, an effective area of the metasurface is divided into at least two sub-areas, and a transmission spectrum corresponding to any sub-area is linearly independent of transmission spectrums corresponding to other sub-areas. According to the invention, the metasurface and the photoelectric sensor form a multi-channel spectrum detection element, and then a proper spectrum restoration algorithm is matched, so that high-accuracy spectrum restoration can be realized, and compared with a spectrum detection element formed by a multi-spectrum narrow-band color filter array needing tedious multiple processing steps, the processing steps of the metasurface are simpler, and the processing efficiency is improved. The processing difficulty and cost of the spectrum detection element are reduced; moreover, according to the multi-channel spectrum detection element provided by the invention, the sampling number of the spectrum can be higher than the number of channels, and the sampling efficiency of spectrum recovery is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photoelectric detection, and particularly to a spectrum detection pixel based on a metasurface, a spectrum detection element and a spectrum camera. BACKGROUND

[0002] The spectrum detection element of a conventional spectrum camera is usually composed of RGB three Bayer color filters and a photoelectric sensor. The optical information of only three channels of RGB is limited and cannot accurately reflect the real color of the imaging spot on each pixel. A complex color and color temperature adjustment algorithm is usually required for correction.

[0003] In another prior art, a spectrum camera with a spectrum detection element composed of a multi-spectrum narrow-band color filter array and an image sensor. The color image recovered by the spectrum camera has more accurate color representation and can restore more real ambient light information. However, in actual processing, the color filter of each channel in the pixel on the spectrum detection element needs to be subjected to a separate mask and deposition step. The more channels the pixel contains, the more and more complex the process steps are. The complex manufacturing process results in high manufacturing cost and difficulty in ensuring the product performance of the final product. CONTENT OF THE UTILITY MODEL

[0004] In view of the above technical problems, the present application provides a spectrum detection pixel based on a metasurface, a spectrum detection element and a spectrum camera. The present application forms a multi-channel spectrum detection element by a metasurface and a photoelectric sensor, and then matches a suitable spectrum restoration algorithm to realize high-accuracy spectrum recovery. Compared with the spectrum detection element composed of a multi-spectrum narrow-band color filter array which requires tedious multiple processing steps, the processing steps of the metasurface are simpler, reducing the processing difficulty and cost of the spectrum detection element. Moreover, the multi-channel spectrum detection element provided by the present application can have a higher number of spectrum samples than the number of channels, significantly improving the sampling efficiency of the spectrum recovery process.

[0005] According to an aspect of an embodiment of the present application, a metasurface for spectrum detection is disclosed, comprising: a substrate and a micro-nano structure arranged on the substrate.

[0006] The effective area of the metasurface is divided into at least two sub-areas, wherein the transmission spectrum corresponding to any of the sub-areas is linearly independent of the transmission spectrum corresponding to the remaining sub-areas.

[0007] In an exemplary embodiment of the present application, the correlation coefficient of the transmission spectrum corresponding to all the sub-areas is less than or equal to a preset threshold.

[0008] In an exemplary embodiment of the present application, the preset threshold is 0.2.

[0009] In an example embodiment of the present application, each of the sub-regions is composed of at least one periodic arrangement of micro / nano structures.

[0010] According to an aspect of an embodiment of the present application, a spectral detection element is disclosed, comprising: a metasurface array composed of at least two metasurfaces as described above; a photoelectric sensor array;

[0011] The photoelectric sensor array is disposed on the light-out side of the metasurface array.

[0012] Wherein, each of the sub-regions is matched with a single photoelectric sensor in the photoelectric sensor array.

[0013] According to an aspect of an embodiment of the present application, the spectral detection element further comprises: at least one microlens array.

[0014] The microlens array is used for converging the received light beams.

[0015] According to an aspect of an embodiment of the present application, if the spectral detection element comprises one microlens array, the microlens array is disposed on the light-in side of the metasurface array; or,

[0016] The microlens array is disposed between the metasurface array and the photoelectric sensor array.

[0017] According to an aspect of an embodiment of the present application, if the spectral detection element comprises two microlens arrays, one of the two microlens arrays is disposed on the light-out side of the metasurface array, and the other is disposed between the metasurface array and the photoelectric sensor array.

[0018] According to an aspect of an embodiment of the present application, the microlenses in the microlens array, the single sub-regions contained in the metasurfaces in the metasurface array, and the single photoelectric sensors in the photoelectric sensor array are matched one by one.

[0019] According to an aspect of an embodiment of the present application, a spectral camera is disclosed, comprising: a spectral detection element as described above; an imaging lens group;

[0020] Wherein, the imaging lens group is used for focusing the object to be measured on the spectral detection element in an imaging manner.

[0021] The super surface for spectrum detection provided by the embodiment of the present application comprises: a substrate and a micro-nano structure arranged on the substrate; an effective area of the super surface is divided into at least two sub-areas, wherein the transmission spectrum corresponding to any sub-area is linearly independent of the transmission spectrum corresponding to the remaining sub-areas. The present application forms a multi-channel spectrum detection element by the super surface and a photoelectric sensor, and then matches a suitable spectrum restoration algorithm, so that high-accuracy spectrum recovery can be realized. Compared with a spectrum detection element formed by a multi-spectrum narrow-band color filter array which needs complicated multiple processing steps, the super surface only needs one photoetching to be realized, thereby reducing the processing difficulty and cost of the spectrum detection element. Moreover, the sampling number of the spectrum of the multi-channel spectrum detection element provided by the present application can be higher than the channel number, thereby significantly improving the sampling efficiency of spectrum recovery.

[0022] Other characteristics and advantages of the present application will become apparent from the following detailed description, or will be learned by practice of the present application.

[0023] It should be understood that the foregoing general description and the following detailed description are only exemplary and are not restrictive of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0024] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

[0025] Figure 1 A schematic diagram of a spectrum detection element of a spectrum camera in the prior art is shown.

[0026] Figure 2 A schematic diagram of a spectrum detection element of another spectrum camera in the prior art is shown.

[0027] Figure 3 A structural schematic diagram of a super surface 1 provided by an embodiment of the present application is shown.

[0028] Figure 4 A structural schematic diagram of a super surface 1 provided by an embodiment of the present application is shown.

[0029] Figure 5 A top view of a super surface 1 provided by an embodiment of the present application is shown.

[0030] Figure 6 A schematic diagram of a single sub-area 110 of a super surface 1 provided by an embodiment of the present application is shown.

[0031] Figure 7 A schematic diagram of a spectrum detection element provided by an embodiment of the present application is shown.

[0032] Figure 8A schematic diagram of a single detection channel of a spectral detection element provided by an embodiment of the present application is shown.

[0033] Figure 9 A schematic diagram of a single detection channel of a spectral detection element provided by an embodiment of the present application is shown.

[0034] Figure 10 A schematic diagram of a single pixel of a spectral detection element 200 provided by an embodiment of the present application is shown.

[0035] Figure 11 A schematic diagram of a single pixel of a spectral detection element 200 provided by an embodiment of the present application is shown.

[0036] Figure 12 A schematic diagram of a single pixel of a spectral detection element 200 provided by an embodiment of the present application is shown.

[0037] Figure 13 A schematic diagram of a single pixel of a spectral detection element 200 provided by an embodiment of the present application is shown.

[0038] Figure 14 A schematic diagram of a spatial multiplexing method provided by an embodiment of the present application is shown.

[0039] Figure 15 A schematic diagram of a spectral camera provided by an embodiment of the present application is shown.

[0040] Figure 16 A structural schematic diagram of a cylindrical micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0041] Figure 17 A structural schematic diagram of a cylindrical micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0042] Figure 18 A structural schematic diagram of a cylindrical micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0043] Figure 19 A structural schematic diagram of a cylindrical micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0044] Figure 20 A structural schematic diagram of a square pillar micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0045] Figure 21 A structural schematic diagram of a cross pillar micro-nano structure and corresponding visible light spectrum provided by an embodiment of the present application are shown.

[0046] Figure 22 A structural schematic diagram of a circular hole-shaped micro-nano structure and a corresponding visible light spectrum are shown.

[0047] Figure 23 A structural schematic diagram of a square hole-shaped micro-nano structure and a corresponding visible light spectrum are shown.

[0048] Figure 24 A structural schematic diagram of a cross hole-shaped micro-nano structure and a corresponding visible light spectrum are shown.

[0049] Figure 25 The transmission spectrum corresponding to each detection channel of a single pixel of the spectral detection element provided in Embodiment 1 is shown.

[0050] Figure 26 (a) of FIG. 1 shows the normalized intensity of the light signal fed back after each detection channel of a single pixel of the spectral detection element provided in Embodiment 1 of the present application receives first color light, Figure 26 (b) of FIG. 1 shows the corresponding spectral recovery.

[0051] Figure 27 (a) of FIG. 2 shows the normalized intensity of the light signal fed back after each detection channel of a single pixel of the spectral detection element provided in Embodiment 1 of the present application receives second color light, Figure 27 (b) of FIG. 2 shows the corresponding spectral recovery.

[0052] Figure 28 The transmission spectrum corresponding to each detection channel of a single pixel of the spectral detection element provided in Embodiment 2 is shown.

[0053] Figure 29 (a) of FIG. 3 shows the normalized intensity of the light signal fed back after each detection channel of a single pixel of the spectral detection element provided in Embodiment 2 of the present application receives third color light, Figure 29 (b) of FIG. 3 shows the corresponding spectral recovery.

[0054] 29(a) of FIG. 4 shows the normalized intensity of the light signal fed back after each detection channel of a single pixel of the spectral detection element provided in Embodiment 1 of the present application receives fourth color light, Figure 30 (b) of FIG. 4 shows the corresponding spectral recovery.

[0055] Reference signs:

[0056] 1 - superlens; 11 - substrate; 12 - micro-nano structure; 110 - sub-region; 200 - spectral detection element; 210 - super surface array; 220 - photoelectric sensor array; 221 - photoelectric sensor; 300 - microlens array; 301 - microlens; 400 - imaging lens group; 510 - Bayer color filter; 520 - photosensitive element; 610 - narrowband filter array; 620 - image sensor. DETAILED DESCRIPTION

[0057] Example implementations are now described with reference to the drawings; however, these descriptions are not intended to limit the scope of the application, but are intended to provide example examples of the application. Examples of the present application can be practiced in a variety of ways, and the description of examples should not be construed as limiting, but merely as illustrative. In the drawings, like reference numerals refer to similar items throughout the various figures. Understanding that these examples are given for illustrative purposes to convey the principles of the present application, various modifications can be made within the scope of the present application.

[0058] Further, various features, structures or characteristics can be combined in any suitable manner in one or more example implementations. In the following description, numerous specific details are provided to give a thorough understanding of example implementations of the present application. One skilled in the relevant art will recognize, however, that the various implementations of the present application can be practiced without one or more of the specific details, or with other modules, components, etc. In other instances, well-known structures, materials, or operations are not shown or described in detail in order to avoid obscuring aspects of the application.

[0059] In the prior art, as shown in Figure 1 , a schematic diagram of a spectral detection element of a spectral camera in the prior art is shown, the spectral detection element of a conventional spectral camera is usually realized by the cooperation of RGB three Bayer color filters 510 and photoelectric sensors 520 to realize color analysis of the imaging light spot, since the optical information of the RGB three channels is limited, the real color of the imaging light spot on each pixel cannot be accurately reflected, and usually complex color and color temperature adjustment algorithms are needed for correction, but the final result is also not very accurate, the method is complex and not accurate enough. Figure 1 If it is necessary to accurately extract the real color composition of the imaging light spot, the most direct way is to improve the sampling number of the color of the imaging light spot on each pixel of the spectral detection element, i.e. the wavelength sampling number, as shown in

[0060] Figure 2 Figure 2 ​​A schematic diagram of a spectral detection element of another prior art spectral camera is shown, which uses an 8-channel narrow-band filter array 610 and an image sensor 620 to form the spectral detection element. This color imaging method in the working waveband is called multi-spectral imaging in the field of visual imaging, and the color image recovered thereby will have more accurate color representation and can restore more realistic environmental light information.

[0061] However, in actual processing, the color filter of each channel in the pixel on the color sensor with the 8-channel narrow-band filter array needs to be subjected to a separate mask and deposition step. The more channels in the pixel, the more and more complex the process steps for manufacturing the spectral detection element, and the complex manufacturing process leads to high manufacturing cost and difficulty in ensuring that the performance of the final product meets the expectations.

[0062] In order to overcome the above-mentioned defects of the related art, the present application provides a metasurface for spectral detection, a spectral detection element, and a spectral camera. The present application forms a multi-channel spectral detection element by a metasurface and a photoelectric sensor, and then matches a suitable spectral restoration algorithm to achieve high-accuracy spectral restoration. Compared with the spectral detection element formed by a multi-spectral narrow-band filter array which needs complicated multiple processing steps, the processing steps of the metasurface are simpler, reducing the processing difficulty and cost of the spectral detection element. Moreover, the multi-channel spectral detection element provided by the present application can have a higher number of spectral samples than the number of channels, significantly improving the sampling efficiency of spectral restoration.

[0063] The present application provides a metasurface 1 for spectral detection, specifically as shown in Figure 3 , Figure 3 A structural schematic diagram of the metasurface 1 provided by an embodiment of the present application is shown, which includes a substrate 11 and a micro-nano structure 12 disposed on the substrate 11, and the micro-nano structure 12 is a sub-wavelength structure. The metasurface 1 mainly performs wavelength response on the light beams received at each place thereon by configuring the material, cross-sectional size, height, arrangement period, and other parameters of the micro-nano structure 12 at each place thereon.

[0064] In the embodiments of the present application, as shown in Figure 3 and Figure 4 , Figure 4 A structural schematic diagram of the metasurface 1 provided by an embodiment of the present application is shown, and the effective area of the metasurface 1 is divided into at least two sub-areas 110, Figure 4The parts separated by the solid lines are the sub-regions 110, and the transmission spectrum corresponding to any sub-region 110 is linearly independent of the transmission spectrum corresponding to the other sub-regions 110, that is, the transmission spectrum corresponding to any sub-region 110 of the metasurface 1 cannot be directly represented by the transmission spectrum corresponding to the other sub-regions 110 or linearly combined by the transmission spectrum corresponding to a limited number of sub-regions 110. The effective region of the metasurface refers to the region provided with micro-nano structures and capable of modulating the received light beam. In this case, the transmission spectrum of each sub-region 110 is unique and can effectively and selectively transmit the light beam with a wavelength in a specific wavelength range, so as to ensure that the spectral information captured by different sub-regions 110 does not interfere with each other, thereby improving the spectral resolution and the accuracy of spectral recovery; and compared with a multi-channel narrow-band color filter array which needs to be obtained through complicated multiple processing steps, the processing steps of the metasurface are simpler, and the processing of each sub-region 110 on the metasurface can be completed by a single photolithography process, greatly reducing the processing difficulty and cost of the spectral detection element.

[0065] And, Figure 4 The metasurface 1 shown in the figure is a metasurface with 9 sub-regions 110, as shown in Figure 5 (a), Figure 5 The figure shows a top view of the metasurface 1 provided by an embodiment of the present application, Figure 5 The parts separated by the solid lines in the figure are the sub-regions 110, as shown in Figure 5 the top view of the metasurface 1 shown in (a), but this is only an example and does not represent that the number of sub-regions 110 contained in the metasurface 1 provided by the present application is limited to this, and the metasurface 1 can also be divided into 25 sub-regions 110 as shown in the top view of another metasurface 1 shown in Figure 5 (b); in addition, Figure 5 (a) and (b) are only used as examples to show the division of the sub-regions 110 on the metasurface 1 provided by the present application, and the actual division is not limited to these two, and can be set according to actual needs, for example, the number of sub-regions 110 can be 12, 16, 20, 36, etc. Further, the shape of the sub-region 110 can be quadrilateral, hexagonal, sector, etc., which can be set according to actual needs.

[0066] It should be noted that in the metasurface, the optical response of micro-nano structures with different shapes and sizes to light beams with different wavelengths is different, and in terms of the energy modulation that can be detected by a common CMOS (complementary metal oxide semiconductor) photosensitive element, it mainly shows different transmission abilities, that is, transmission spectrum. Assuming that the transmission spectrum of the nth sub-region 110 on the metasurface 1 is T n , then for M wavelengths of light beams, the transmission spectrum of this sub-region 110 can be represented as: Tn =[t n,1 ,t n,2 ,…,t n,M ]; where t n,M This represents the transmittance capability of the nth sub-region 110 for the beam of the Mth wavelength.

[0067] In one embodiment, considering the subsequent process of recovering the spectrum using an algorithm, each sub-region 110 on the metasurface 1 should be as unique as possible, and the transmission spectra corresponding to each sub-region 110 should have sufficiently large differences. In this embodiment, the Pearson correlation coefficient is used to quantify the correlation between the transmission spectra of any two sub-regions 110 on the metasurface 1. Then, the mean of the Pearson correlation coefficients between the transmission spectra corresponding to different sub-regions 110 is calculated to obtain the correlation coefficients of the transmission spectra corresponding to all sub-regions 110 on the metasurface 1, and the correlation coefficients of the transmission spectra corresponding to all sub-regions 110 on the metasurface 1 are made less than or equal to a preset threshold. The specific formula is: Where ρ is the correlation coefficient. This indicates selecting two subregions 110 from N subregions 110, where cov(T) i ,T j ) represents the transmission spectrum T corresponding to the i-th sub-region 110 and the j-th sub-region 110. i and T j The covariance, σ(T) i ) and σ(T j The transmission spectra T are respectively i and T j The smaller the correlation coefficient ρ, the greater the difference in physical characteristics of each sub-region 110 in metasurface 1, and the greater the difference in the corresponding transmission spectrum, making it more unique and thus enabling better spectrum recovery and improving the accuracy of the recovered spectrum.

[0068] In one embodiment, the correlation coefficient of the transmission spectrum corresponding to all sub-regions 110 on the metasurface 1 should be less than or equal to a preset threshold of 0.2.

[0069] In one embodiment, the correlation coefficient of the transmission spectrum corresponding to all sub-regions 110 on the metasurface 1 should be less than or equal to a preset threshold of 0.1.

[0070] In an embodiment, in order to improve the richness of the optical response of the metasurface 1 structure, on the metasurface 1, the micro-nano structures 12 in each sub-region 110 are different, which is embodied in the material, shape, cross-sectional size, and height of the micro-nano structures 12. If at least one of the material, shape, cross-sectional size, and height of any two micro-nano structures 12 is different, the two micro-nano structures 12 can be regarded as different types of micro-nano structures. Moreover, the micro-nano structures 12 with different materials, shapes, cross-sectional sizes, and heights correspond to different structure transmission spectra. In this case, the sub-regions 110 with different transmission spectra on the metasurface 1 also each include different types of micro-nano structures 12.

[0071] Moreover, each sub-region 110 is composed of at least one periodic arrangement of micro-nano structures 12, that is, a single sub-region 110 can include only one type of micro-nano structure 12 or multiple types of micro-nano structures 12, as shown in Figure 6 , Figure 6 FIG. 1 shows a schematic diagram of a single sub-region 110 of the metasurface 1 provided by an embodiment of the present application, as shown in Figure 6 (a) of FIG. 1, Figure 6 (a) of FIG. 1 shows that this sub-region 110 is composed of only a periodic arrangement of square-hole-shaped micro-nano structures 12; as shown in Figure 6 (b) of FIG. 1, Figure 6 (b) of FIG. 1 shows that this sub-region 110 is composed of a periodic arrangement of square-hole-shaped micro-nano structures 12 and cross-hole-shaped micro-nano structures 12. It should be noted that the shape of the micro-nano structures 12 is not limited to the square-hole shape and the cross-hole shape described above, and can be selected according to actual needs. For example, the shape of the micro-nano structures 12 can be one or a combination of a cylindrical shape, a hollow cylindrical shape, a circular-hole shape, a hollow circular-hole shape, a square-cylindrical shape, a square-hole shape, a hollow square-cylindrical shape, and a hollow square-hole shape.

[0072] Further, the material of the micro-nano structures 12 needs to have a relatively high transmittance and a relatively low absorption in the working waveband. For example, when the working waveband is the visible light waveband and the near-infrared waveband, the material of the micro-nano structures 12 can be selected from silicon nitride, intrinsic silicon, titanium dioxide, or silicon carbide; when the working waveband is the far-infrared waveband, the material of the micro-nano structures 12 can be selected from intrinsic silicon; and when the working waveband is the ultraviolet waveband, the material of the micro-nano structures 12 can be selected from silicon nitride. Therefore, the selection of the material of the micro-nano structures 12 is determined by the application scenario of the spectral camera, and the corresponding wavebands of the material are visible light, visible light + near-infrared, visible light + near-infrared + far-infrared, ultraviolet + visible light, and ultraviolet + visible light + near-infrared, which respectively correspond to scenarios such as daily color shooting, near-infrared light supplement shooting, industrial thermal imaging shooting, medical detection, and agricultural detection.

[0073] The application also provides a spectral detection element 200 for spectral detection, as shown in Figure 7 , Figure 7 The application provides a spectral detection element, as shown in the schematic diagram, the spectral detection element 200 includes: a metasurface array 210 composed of at least two metasurfaces 1 described in the above embodiments and a photoelectric sensor array 220; and the photoelectric sensor array 220 is arranged on the light-emitting side of the metasurface array 210 and used for receiving a light beam modulated by the metasurface array 210; wherein each single sub-region 110 contained in the metasurface 1 in the metasurface array 210 is matched with a single photoelectric sensor 221 in the photoelectric sensor array 220, that is, for any single sub-region 110 contained in any metasurface 1 in the metasurface array 210, there is a unique corresponding photoelectric sensor 221 in the photoelectric sensor array 220, as shown in Figure 7 , Figure 7 The part separated by the thick solid line in the metasurface 1 in the metasurface array 210, Figure 7 The part separated by the thin solid line in the metasurface 1 is each sub-region 110 of the metasurface 1.

[0074] In the spectral detection element, the metasurface 1 and the multiple photoelectric sensors 221 corresponding to all the sub-regions 110 of the metasurface 1 jointly constitute a single pixel, that is, an image unit, of the spectral detection element, which is the smallest unit for processing input light in the spectral detection element. In the single pixel, multiple detection channels are contained, the metasurface in the single pixel has multiple sub-regions 110 and each sub-region 110 has a corresponding photoelectric sensor 221, the single sub-region 110 and the corresponding photoelectric sensor 221 constitute a single detection channel in the pixel, that is, actually Figure 7 The part separated by the thin solid line in the metasurface 1 and the corresponding photoelectric sensor in the photoelectric sensor array 220 constitute a detection channel in the pixel, Figure 7 The part separated by the thick solid line in the metasurface 1 and the corresponding photoelectric sensor constitute a single pixel of the spectral detection element 200, Figure 7 The part separated by the thin solid line and the corresponding photoelectric sensor in the pixel represent a detection channel, Figure 7 The spectral detection element is shown in which a single pixel contains nine detection channels, but Figure 7 This is only an exemplary description and does not represent the actual number of detection channels in a single pixel in the spectral detection element 200 provided by the application. In this case, the micro-nano structure 12 of the metasurface 1 is compatible with the semiconductor process and can be directly processed on the protective layer on the light-entering side of the photoelectric sensor 221, thereby maintaining a high element integration degree and ensuring the accuracy of spectral recovery in the extremely short transmission distance between the metasurface 1 and the photoelectric sensor 221.

[0075] In the embodiments of this application, such as Figure 8 and Figure 9 As shown, Figure 8 A schematic diagram of a single detection channel of a spectral detection element provided in one embodiment of this application is shown. Figure 9 This diagram illustrates a single detection channel of a spectral detection element provided in an embodiment of this application. The single detection channel includes a sub-region 110 and a corresponding photoelectric sensor 221. Figure 9 The arrows indicate the light beam and its propagation direction. The metasurface array 210 on the spectral detection element 200 is used to receive the light beam and modulate it, exhibiting energy modulation with the broad spectral physical properties of micro-nano structures. This means that light beams of different wavelengths are converged onto the photoelectric sensor array 220, enabling the photoelectric sensor array 220 to acquire the spectral information of the light beam. The photoelectric sensor 221 in the photoelectric sensor array 220 does not have wavelength selectivity. The process of the metasurface array 210 modulating the light beam is equivalent to compressing a broad spectral incident signal into a transmitted energy signal. Subsequently, by processing the electrical signals of all detection channels within a single pixel on the spectral detection element, the spectral information of that pixel can be obtained. The spectral information acquired by all pixels on the spectral detection element 220 will form a spectral image.

[0076] It should be noted that the more detection channels a single pixel has, the higher the spectral resolution. However, if the size of a single detection channel remains unchanged and the number of detection channels is increased, the size of a single pixel will be larger and the spatial resolution will be lower. On the other hand, reducing the size of a single detection channel will increase the difficulty of fabricating spectral detection elements.

[0077] In one embodiment, to avoid energy divergence caused by scattering of the incident light beam by the metasurface 1 and large-angle diffraction, the spectral detection element 200 further includes at least one microlens array 300, which is composed of at least one microlens 301.

[0078] Specifically, before the light beam enters the photoelectric sensor array and / or metasurface array, it can pass through a microlens array. The microlens array deflects and focuses the light beam, reducing energy divergence and improving the energy utilization of the optical signal, thereby improving the response of the metasurface array, similar to CRA (Chief Ray Angle) matching in traditional imaging.

[0079] It should be noted that the microlens array 300 can be arranged on the light-incoming side of the metasurface array 210 or on the light-incoming side of the photosensor array 220. Therefore, in an embodiment, when the spectral detection element 200 only includes a single microlens array 300, the microlens array 300 can be arranged on the light-outgoing side of the metasurface array 210 and on the light-incoming side of the photosensor array 220, that is, the microlens 300 is arranged between the metasurface array 210 and the photosensor array 220, as shown in FIG. 2A. Figure 10 and Figure 11 Figure 10 FIG. 2A shows a schematic diagram of a single pixel of the spectral detection element 200 provided by an embodiment of the present application, in which the metasurface array 210 is arranged on the light-incoming side of the photosensor array 220, and the microlens array 300 is arranged on the light-outgoing side of the metasurface array 210. Figure 11 FIG. 2B shows a schematic diagram of a single pixel of the spectral detection element 200 provided by an embodiment of the present application, in which the metasurface array 210 is arranged on the light-incoming side of the photosensor array 220, and the microlens array 300 is arranged on the light-incoming side of the metasurface array 210. Figure 11 The arrows represent the incident light and the propagation direction of the light beams. In fact, that is, the metasurface array 210 is arranged on the surface of the photosensor 220 with the microlens array 300 on the light-incoming side, as shown in FIG. 2A, in which case the light energy with a certain divergence after being modulated by the micro-nano structures on the metasurface sub-regions can be focused on the photosensor by the microlens 301, thereby reducing the loss of light energy and improving the utilization rate of light energy. Figure 11

[0080] In another embodiment, when the spectral detection element 200 only includes a single microlens array 300, the microlens array 300 can be arranged on the light-incoming side of the metasurface array, as shown in FIG. 2B. Figure 12 Figure 12 FIG. 2B shows a schematic diagram of a single pixel of the spectral detection element 200 provided by an embodiment of the present application, in which the metasurface array 210 is arranged on the light-incoming side of the photosensor array 220, and the microlens array 300 is arranged on the light-incoming side of the metasurface array 210.

[0081] In fact, the spectral detection element 200 can also include multiple microlens arrays 300. In an embodiment, when the spectral detection element 200 includes two microlens arrays 300, the two microlens arrays 300 are arranged on the light-incoming side of the metasurface array 210 and on the light-incoming side of the photosensor array 220, that is, one of the two microlens arrays 300 is arranged on the light-incoming side of the metasurface array 210, and the other is arranged between the metasurface array 210 and the photosensor array 220. As shown in FIG. 2C, Figure 13 Figure 13 ​​​​A schematic diagram of a single pixel of the spectral detection element 200 provided by an embodiment of the present application is shown. In this case, the incident light beam is first converged by the microlens array 300 arranged on the light-in side of the super surface array 210, and then projected to the super surface array 210. After the modulation of the super surface array 210, the light energy with a certain divergence is converged by the microlens array 300 arranged on the light-in side of the photoelectric sensor array 220, and then projected to the photoelectric sensor array 220. This further reduces the loss of light energy in the propagation process and improves the accuracy of spectral recovery.

[0082] It should be noted that the spectral detection element 200 can actually further include more microlens arrays 300. However, multiple microlens arrays 300 will inevitably affect the volume of the spectral detection element 200. Therefore, in actual settings, the number of microlens arrays 300 can be fully set by comprehensively considering the volume of the spectral detection element 200 and the degree of loss of light energy. That is, when the volume requirement of the spectral detection element 200 is not high, multiple microlens arrays 300 can be included in the spectral detection element 200.

[0083] Further, the microlens array 300 includes multiple microlens 301. In an embodiment, referring to Figure 10 to Figure 12 , each microlens in the microlens array 300 can be one-to-one matched with a single sub-region 110 on the super surface 1 and a single photoelectric sensor 221 in the photoelectric sensor array 220. That is, the single sub-region 110, the single microlens, and the single photoelectric sensor 221 constitute a single detection channel. In this case, each microlens 301 can individually converge the light beam required to be modulated and / or modulated by the corresponding sub-region 110, thereby minimizing energy loss and improving the performance of the spectral detection element 200. It should be noted that if there are two microlens arrays 300, referring to Figure 13 , there is one microlens 301 in each of the two microlens arrays 300 matched with a single sub-region 110 and a single photoelectric sensor 221, and the two microlens 301 are one-to-one matched.

[0084] In another embodiment, considering that equipping each detection channel with a corresponding microlens will result in extremely high processing difficulty, the multiple microlens on the microlens array 300 can be matched with multiple sub-regions 110, that is, in this embodiment, a single microlens in the microlens array 300 is matched with multiple sub-regions 110 and the photoelectric sensor corresponding to the sub-regions 110, that is, a single microlens converges the light beam required to be modulated and / or modulated by multiple sub-regions 110. In this case, the processing difficulty and cost of preparing the microlens array are reduced.

[0085] In another embodiment, only one piece of microlens is included in the microlens array 300, and the light beams modulated by the metasurface array 210 and / or the light beams whose modulation is completed are converged by the single piece of microlens.

[0086] In the embodiments of the present application, as described above, it is assumed that the transmission spectrum of the nth sub-region 110 on the metasurface 1 is T n Then, for the light beams of M wavelengths, the transmission spectrum T n of the sub-region 110 can be expressed as: T n = [t n,1 ,t n,2 ,…,t n,M ], and for the incident spectrum s(λ) of a specific color light, where λ is the wavelength, the sampling wavelengths are λ1, λ2,…, λ M , the light energy I n after the metasurface array 220 is:

[0087]

[0088] where s m is the intensity of the spectrum at the wavelength λ m , and specifically, when the response transmission spectrum is narrowband at λ m , I n = t n,m s m .

[0089] On the spectral detection element 200, the size of a single pixel is very small, and considering that the color light on a pixel is approximately uniform, at this time, the overall response of the micro-nano structure on a pixel with N detection channels can be described by a matrix equation:

[0090]

[0091] that is, I = Ts.

[0092] where I N represents the light intensity distribution corresponding to the Nth detection channel, t N,M represents the transmission capability of the Nth detection channel to the light beam with a wavelength of λ M , and s M represents the intensity of the spectrum at the wavelength λ M .

[0093] When the rank of matrix T is Rank(T) = Rank(T, I) = M, the above matrix equation has a unique solution, and the actual physical meaning is that the number of sub-regions 110 in a single hyper-surface 1 on the spectral detection element 200, i.e., the number of detection channels in a single pixel, is greater than or equal to the number of spectral samples. In fact, only the number of detection channels contained in a single pixel needs to be equal to the number of spectral samples, i.e., N = M, and the matrix T is a reversible matrix, so the recovered spectrum s' can be solved according to the light energy signal received by the photoelectric sensor array 220: s' = T -1 I; wherein s' is the recovered spectrum.

[0094] When the rank of matrix T is Rank(T) = Rank(T, I) < M, the above matrix equation has an infinite number of solutions, and the matrix equation is underdetermined. At this time, the number of detection channels contained in a single pixel on the spectral detection element 200 is less than the number of spectral samples. In theory, a small number of channels cannot accurately solve a large number of spectral samples. However, most natural substances in nature have a simple reflection spectrum, i.e., in the compression sensing theory, it is called a sparse distribution in a specific domain, so only a sparse conversion matrix D that meets a certain condition needs to be found, so that: s = Dv;

[0095] wherein the vector v is sparse and contains only a small number of non-zero elements, and the matrix D can be a fixed matrix, such as a common discrete cosine transform matrix or a wavelet transform matrix, or a dictionary training matrix, which can be adjusted according to actual application. In this case, the problem of solving the matrix equation is converted into a minimum norm optimization problem:

[0096]

[0097] wherein a is a weight coefficient and needs to be adjusted according to actual situation. The reconstruction algorithm of compression sensing can be used to solve the sparse vector v, and then the spectrum s' = Dv can be recovered with high accuracy. In this case, the original high-dimensional spectrum signal can be accurately reconstructed with less low-dimensional information, and the recovery of the original high-dimensional spectrum signal is completed, that is, the spectral sampling number of the spectral detection element provided by the application can be higher than the number of channels, and the sampling efficiency of spectral recovery is significantly improved.

[0098] In an embodiment, common compression sensing algorithms include orthogonal matching pursuit (OMP), fast iterative shrinkage-thresholding algorithm (FISTA), and basis pursuit (BP).

[0099] In an embodiment, during the process of signal collection, due to the existence of background stray light and thermal noise of the spectral detection element, noise signal e is usually accompanied in the light energy signal received by the photoelectric sensor array 220, that is, I = Ts + e; in this case, a suitable noise reduction algorithm is added in the recovery algorithm, and the recovery of the spectrum can still be ensured with high accuracy.

[0100] In an embodiment, the imaging lens, the color filter on the imaging lens, and the photoelectric sensor also have response curves for light beams of different wavelengths, that is, the above matrix equation also needs to consider the influence of the wavelength response matrix H of the imaging lens, the color filter on the imaging lens, and the photoelectric sensor, which is a known quantity, and the matrix equation is expressed as I = THs at this time; in the solving process of spectral recovery, TH can be regarded as a matrix, and the solving process of any one of the above is performed.

[0101] In an embodiment, in order to evaluate the accuracy of spectral recovery, the standard deviation value K can be used:

[0102]

[0103] wherein s i represents the intensity of the spectrum at wavelength λ i , s i ' represents the recovered spectrum at wavelength λ i , and the smaller the standard deviation value K, the higher the accuracy of the spectrum recovery; the preferred value of the standard deviation value K is less than 0.1, and preferably, when the spectrum is completely recovered, s = s', and K = 0.

[0104] Further, the spectral recovery operation described above is performed on all the pixels on the spectral detection element 200, so that the spectral distribution on all the pixels can be obtained, and a high-dimensional spectral data cube (dimension: horizontal coordinate × vertical coordinate × wavelength coordinate), that is, a spectral image, is formed. Subsequently, color conversion operation can be performed on all the spectra, and a real color image containing accurate spectral information can be obtained. In general, a wavelength resolution of 50 nm is a multispectral image, a wavelength resolution of 20 nm is a hyperspectral image, and a wavelength resolution of 1 nm or less is an ultrahyperspectral image. The smaller the spectral resolution wavelength, that is, the more the number of spectral samples that need to be recovered, and the more the number of detection channels required for spectral reconstruction, that is, the more the number of sub-regions 110 required to be contained in the metasurface 1. Therefore, in this case, even if the compressed sensing algorithm shown in the present application is used, the number of detection channels of a single pixel and the number of recovered spectral samples should not be too different, otherwise the standard deviation value K will be large. Preferably, the ratio of the number of detection channels to the number of recovered spectral samples on a single pixel should be greater than 0.5.

[0105] It should be emphasized that, as mentioned above, the accuracy of the recovered spectrum is affected by the observation matrix, regardless of the method of directly solving the matrix equation or the method of compressed sensing recovery, and it is required that the different sub-regions 110 on the metasurface 1 should be as unique as possible, and the transmission spectrum corresponding to each sub-region 110 should be sufficiently different. Otherwise, even if the number of detection channels contained in a single pixel on the spectral detection element 200 is equal to the number of spectral samples, the transmission spectrum of a certain sub-region 110 can be linearly represented by the transmission spectrum of other sub-regions 110, for example, T1 = 0.5T2 + 0.5T3, at this time, there must be Rank(T) = Rank(T, I) < M, which cannot be directly solved by taking the inverse; for the compressed sensing algorithm, the irrelevance of the observation matrix is also a key requirement, so as mentioned above, the Pearson correlation coefficient can be used to quantitatively analyze the correlation of the transmission spectrum of each sub-region 110.

[0106] In an embodiment, a spatial multiplexing method can also be used in the process of spectral recovery, as shown in Figure 14 Figure 14 A schematic diagram of the spatial multiplexing method provided by an embodiment of the present application is shown, when a single pixel on the spectral detection element 200 contains only 9 detection channels, the edge detection channels of adjacent pixels can be extracted, as shown by the dashed boxes in Figure 14 The 9 squares in the center of each dashed box are the 9 detection channels contained in a single pixel, and the squares at the edge of each dashed box are the edge detection channels of adjacent pixels. After extracting the light intensity distribution of the detection channels of a single pixel, the light intensity distribution of the edge detection channels of adjacent pixels is extracted, and then any one of the spectral recovery algorithms mentioned above is performed, and the spectral information corresponding to this pixel can be obtained. However, it should be noted that due to the repeated calculation of the detection channels, the final spectral image may have a sawtooth phenomenon, which needs to be weakened by a digital image processing method.

[0107] The present application also provides a spectral camera for spectral detection, which comprises a spectral detection element 200 and an imaging lens group 400 according to any one of the above embodiments, as shown in Figure 15 Figure 15 A schematic diagram of the spectral camera provided by an embodiment of the present application is shown.

[0108] Specifically, the imaging lens group 400 is used to focus the object to be measured on the spectral detection element 200 of the spectral camera in an imaging manner, that is, the object to be measured is actually photographed by the imaging lens group, and the spectral information of the object to be measured can be detected.

[0109] ​​In an embodiment, in order to avoid the influence of light beams outside the working wavelength band of the spectral detection element, a broadband color filter is arranged between the imaging lens group 400 and the spectral detection element 200 in the spectral camera. The broadband color filter is used to filter light beams of non-working wavelength bands, so that only light beams with wavelengths in the working wavelength band can be transmitted.

[0110] Embodiment 1

[0111] In the visible light application scenario, intrinsic silicon is selected as the material of the micro-nano structure 12 of the metasurface 1, and the micro-nano structure 12 has a height of H = 220 nm, Figure 16-23 The following nine micro-nano structures are listed.

[0112] Figure 16 The structure diagram of the cylindrical micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 16 The arrangement period of the cylindrical micro-nano structure shown in (a) is P = 550 nm, and the diameter is D = 350 nm, Figure 16 (b) shown in (a) is the corresponding visible light spectrum of the cylindrical micro-nano structure.

[0113] Figure 17 The structure diagram of the cylindrical micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 17 The arrangement period of the cylindrical micro-nano structure shown in (a) is P = 350 nm, and the diameter is D = 150 nm, Figure 17 (b) shown in (a) is the corresponding visible light spectrum of the cylindrical micro-nano structure.

[0114] Figure 18 The structure diagram of the cylindrical micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 18 The arrangement period of the cylindrical micro-nano structure shown in (a) is P = 450 nm, and the diameter is D = 200 nm, Figure 18 (b) shown in (a) is the corresponding visible light spectrum of the cylindrical micro-nano structure.

[0115] Figure 19 The structure diagram of the cylindrical micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 19 The arrangement period of the cylindrical micro-nano structure shown in (a) is P = 700 nm, and the diameter is D = 300 nm, Figure 19 (b) shown in (a) is the corresponding visible light spectrum of the cylindrical micro-nano structure.

[0116] Figure 20 The structure diagram of the cylindrical micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 20The arrangement period of the square column micro-nano structure shown in (a) of FIG. 1 is P = 350 nm, and the cross-sectional side length L0= 75 nm, Figure 20 The visible light spectrum corresponding to the square column micro-nano structure is shown in (b) of FIG. 1.

[0117] Figure 21 The structural schematic diagram of the cross column micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 21 The arrangement period of the cross column micro-nano structure shown in (a) of FIG. 2 is P = 450 nm, the arm length L1= 225 nm, and the arm width L2= 75 nm, Figure 21 The visible light spectrum corresponding to the cross column micro-nano structure is shown in (b) of FIG. 2.

[0118] Figure 22 The structural schematic diagram of the circular hole micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 22 The arrangement period of the circular hole micro-nano structure shown in (a) of FIG. 3 is P = 600 nm, and the diameter D = 450 nm, Figure 22 The visible light spectrum corresponding to the circular hole micro-nano structure is shown in (b) of FIG. 3.

[0119] Figure 23 The structural schematic diagram of the square hole micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 23 The arrangement period of the square hole micro-nano structure shown in (a) of FIG. 4 is P = 400 nm, and the cross-sectional side length L0= 300 nm, Figure 23 The visible light spectrum corresponding to the square hole micro-nano structure is shown in (b) of FIG. 4.

[0120] Figure 24 The structural schematic diagram of the cross hole micro-nano structure and the corresponding visible light spectrum provided by an embodiment of the present application are shown. Figure 24 The arrangement period of the cross hole micro-nano structure shown in (a) of FIG. 5 is P = 450 nm, the arm length L1= 325 nm, and the arm width L2= 150 nm, Figure 24 The visible light spectrum corresponding to the cross hole micro-nano structure is shown in (b) of FIG. 5.

[0121] Each detection channel corresponding to the micro-nano structure in each single pixel on the spectral detection element 200 provided by the embodiment 1 is selected from the above-mentioned nine kinds of micro-nano structures, respectively, to form a super surface sub-region corresponding to the nine detection channels of the single pixel on the spectral detection element 200. It should be noted that each super surface sub-region can be composed of one or more micro-nano structures. As shown in Figure 25 Figure 25 ​The transmission spectrum corresponding to each detection channel of the single pixel containing nine detection channels in the spectral detection element provided by Embodiment 1 is shown. If the same number of wavelength samples and detection channels are used, that is, the method of directly solving the matrix equation is used, the transmission spectrum can be down-sampled to 9, and at this time the wavelength resolution is 37.5 nm. The correlation coefficient of the transmission spectrum corresponding to each sub-region 110 contained by each super surface 1 on the spectral detection element 200 is 0.193. Figure 26 (a) of FIG. 1 shows the normalized intensity of the light signal fed back after each detection channel on a single pixel of the spectral detection element provided by Embodiment 1 of the present application receives first color light, Figure 26 (b) of FIG. 1 shows the corresponding spectral recovery, and it can be seen that the deviation of spectral recovery is extremely small, and the standard deviation value K is 7.6x10 -15 . Figure 27 (a) of FIG. 2 shows the normalized intensity of the light signal fed back after each detection channel on a single pixel of the spectral detection element provided by Embodiment 1 of the present application receives second color light, Figure 27 (b) of FIG. 2 shows the corresponding spectral recovery, and it can be seen that the deviation of spectral recovery is also extremely small, and the standard deviation value K is 9.3x10 -16 . It is not difficult to see that the two standard deviation values K are both of the order of 10 -15 , and the spectral recovery accuracy of the spectral detection element 200 provided by Embodiment 1 of the present application is extremely high.

[0122] Embodiment 2

[0123] Each detection channel in a single pixel on the spectral detection element 200 provided by Embodiment 2 is selected from the above-mentioned nine micro-nano structures, and respectively constitutes a super surface sub-region corresponding to twenty-five detection channels of a single pixel on the spectral detection element 200. It should be noted that each super surface sub-region can be composed of one or more micro-nano structures. As shown in Figure 28 , Figure 28 The transmission spectrum corresponding to each detection channel of the single pixel containing twenty-five detection channels in the spectral detection element provided by Embodiment 2 is shown. If the number of wavelength samples is greater than the number of detection channels, that is, the method of compressed sensing recovery is used, the transmission spectrum can be down-sampled to 31 (greater than the number of detection channels 25), and at this time the wavelength resolution is 10 nm. The correlation coefficient of the transmission spectrum corresponding to each sub-region 110 contained by each super surface 1 on the spectral detection element 200 is 0.089. Figure 29 (a) of FIG. 3 shows the normalized intensity of the light signal fed back after each detection channel on a single pixel of the spectral detection element provided by Embodiment 2 of the present application receives third color light, Figure 29Fig. 2(b) shows the corresponding spectral recovery, and it can be seen that the deviation of spectral recovery is very small, and the standard deviation value K is 0.025. Figure 30 Fig. 2(a) shows the normalized intensity of the feedback light signal received by each detection channel on a single pixel of the spectral detection element provided by the embodiment 1 of the present application after the fourth color light is received, Figure 30 Fig. 2(b) shows the corresponding spectral recovery, and it can be seen that the deviation of spectral recovery is very small, and the standard deviation value K is 0.023. It can be seen that the two standard deviation values K are both less than 0.1, and the spectral recovery accuracy of the spectral detection element 200 provided by the embodiment 2 of the present application is very high.

[0124] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the application being indicated by the following claims.

Claims

1. A metasurface for spectral probing, characterized in that, The metasurface comprises a substrate and micro-nano structures arranged on the substrate; An effective area of the metasurface is divided into at least two sub-areas, wherein a transmission spectrum corresponding to any of the sub-areas is linearly independent of transmission spectra corresponding to the rest of the sub-areas.

2. The metasurface of claim 1, wherein, Correlation coefficients of transmission spectra corresponding to all the sub-areas are less than or equal to a preset threshold.

3. The metasurface of claim 2, wherein, The preset threshold is 0.

2.

4. The metasurface of claim 1, wherein, Each of the sub-areas is composed of at least one periodic arrangement of micro-nano structures.

5. A spectral detection element, characterized by The spectral detection element comprises a metasurface array composed of at least two metasurfaces according to any one of claims 1-4; and a photoelectric sensor array. The photoelectric sensor array is arranged on an out-light side of the metasurface array. Wherein, each of the sub-areas is matched with a photoelectric sensor in the photoelectric sensor array one by one.

6. The spectroscopic detection element according to claim 5, characterized in that The spectral detection element further comprises at least one microlens array. The microlens array is used for converging the received light beams.

7. The spectroscopic detection element according to claim 6, characterized in that If the spectral detection element comprises one microlens array, the microlens array is arranged on an in-light side of the metasurface array; or, The microlens array is arranged between the metasurface array and the photoelectric sensor array.

8. The spectroscopic detection element of claim 6, wherein, If the spectral detection element comprises two microlens arrays, one of the two microlens arrays is arranged on the in-light side of the metasurface array, and the other is arranged between the metasurface array and the photoelectric sensor array.

9. The spectroscopic detection element according to any one of claims 6 to 8, characterized in that The microlenses in the microlens array, the sub-areas in the metasurfaces of the metasurface array, and the photoelectric sensors in the photoelectric sensor array are matched one by one.

10. A spectral camera, characterized by The spectral camera comprises the spectral detection element according to any one of claims 5-9; and an imaging lens group. Wherein, the imaging lens group is used for focusing an object to be detected on the spectral detection element in an imaging manner.

Images