Optical component with waveguide based filter

Abstract

To provide an image sensor.SOLUTION: An image sensor for recording incident radiation may include a first layer for filtering the incident radiation by attenuating the incident radiation having a frequency below a cut-off frequency, and a second light-sensitive layer for absorbing the radiation passing through the first layer. The first layer may precede the second light-sensitive layer in a direction of propagation of the incident radiation and the first layer includes at least one aperture passing through the first layer to the second light-sensitive layer for propagating radiation therethrough. The cross sectional size of the at least one aperture may be configured to provide a cut-off frequency so that incident radiation with a frequency below the cut-off frequency is attenuated inside the at least one aperture and incident radiation with a frequency above the cut-off frequency propagates through the at least one aperture.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates generally to optical components, and more particularly to optical components having a layer that provides a cutoff frequency for incident radiation. The optical components can be active optical components, such as image sensors, or passive optical components, such as lenses. [Background technology]

[0002] Today, important active optical components are image sensors, i.e. electronic devices that record images by measuring the light intensity of different areas by using an array of light sensors. Typical image sensors are semiconductor charge-coupled devices (CCD) and complementary metal-oxide-semiconductor (CMOS) chips. These sensors often use color filter arrays with colored filter elements, for example using the Bayer pattern, to distinguish between different colors of light that hit the image sensor. Another type of image sensor is based on the Foveon method of color separation, which determines the color of light by identifying the depth at which the light is recorded. This method is based on the fact that light with longer wavelengths has a greater penetration depth into the sensor.

[0003] Infrared light is measured by typical sensors, but since this portion of light is undesirable because it contaminates the color of visible light, known image sensors may require an additional infrared filter.

[0004] Furthermore, it is generally desirable to create image sensors with small pixels, high light measurement efficiency, and good color separation. Image sensors with color filter arrays typically lose more than 50% of the light intensity through the color filters. Image sensors with color filter arrays or based on the Foveon approach may have limited color separation. It is also generally desirable to create image sensors with pixels that collect and measure light independent of the angle at which the light strikes the image sensor.

[0005] Efficient filtering is also a concern among passive optical components such as lenses, fiber couplers, etc. Such optical components often have frequency-dependent behavior that is undesirable for imaging applications, such as chromatic aberration in the case of lenses. It would therefore be desirable to have passive optical elements with improved imaging behavior. Summary of the Invention

[0006] One challenge may be to provide an image sensor that does not require a conventional infrared color filter to filter out infrared light. A further challenge may be to provide an image sensor that efficiently separates and measures different colors of light that strike the image sensor. A further challenge may be to provide an image sensor with pixels that have a collection and measurement efficiency that is less dependent on the angle of the light, or that reduces the amount of leakage of light that strikes the image sensor at narrow angles (i.e., highly non-perpendicular to the surface) to nearby pixels. A further challenge may be to provide an image sensor that allows for color separation of small pixels in a structurally simple manner (e.g., without the need for a conventional color filter array) that may simplify the manufacture of such image sensors.

[0007] According to a first aspect of the invention, an image sensor for recording incident radiation may include a first layer for filtering the incident radiation and a second photosensitive layer for absorbing radiation passing through the first layer. The first layer may be formed of one or more materials that are opaque to the incident radiation. The first layer is in front of the second photosensitive layer in a propagation direction of the incident radiation, and the first layer includes at least one opening penetrating through the first layer to the second layer for propagating radiation through the at least one opening. The cross-sectional size of the at least one opening may be configured to provide a cutoff frequency such that incident radiation having a frequency below the cutoff frequency is attenuated inside the at least one opening and incident radiation having a frequency above the cutoff frequency propagates through said at least one opening.

[0008] In an embodiment, the first layer may have many openings with a predetermined number of different cross-sectional sizes (e.g., one size or three different sizes), so that the first layer provides a filter function with one or more different cutoff frequencies at different positions in the plane of the first layer. The second photosensitive layer may record the intensity and position of radiation passing through the different openings of the first layer. Thus, an image sensor according to an embodiment may record light intensity of a specific wavelength range at different positions. Such an image sensor may be used similarly to a conventional image sensor that may be found, for example, in a digital camera or a mobile phone. In other words, the first layer is used to provide color separation into one or more light components, and the second photosensitive layer of the image sensor may record the one or more light components similarly to the photosensitive layer of a conventional image sensor.

[0009] The image sensor addresses one or more of the above-mentioned challenges for the following reasons: The filter layer may have one or more apertures that provide a cutoff frequency that removes infrared light but allows the propagation of light having higher frequencies (e.g., red light, green light, and blue light). The at least one aperture may also provide a cutoff frequency that is in a different range of the light spectrum to measure one or more different colors of the visible light spectrum. For example, apertures of different sizes may be combined to measure and separate the individual colors of light that strike the image sensor by subtracting the measured radiation that propagates through apertures with different cutoff frequencies. Incident radiation that propagates through the at least one aperture may be propagated by coupling to a propagation mode inside the at least one aperture such that the radiation is not refracted, which may reduce light leakage to nearby pixels and / or the dependence of the collection and measurement efficiency on the angle of the light that strikes the sensor. Guiding the propagation of light through the at least one aperture may also reduce light leakage and / or the dependence of the collection and measurement efficiency on the angle of the light. Image sensors with a filter layer with openings can be structurally simple, since they can be manufactured by standard chip manufacturing techniques and do not require the provision of a conventional color filter array with, for example, colored filter elements. This feature can be relevant for small pixels (e.g., pixels with a pixel pitch of less than 1 μm). The degree of color separation, i.e., how strongly a certain range of light (e.g., below a cutoff frequency) is attenuated, can be controlled by the depth of at least one opening. As a result, stronger color separation can be achieved by using deeper openings or by using a thicker first layer that filters the incident light.

[0010] For any one of the embodiments disclosed herein, the at least one opening can be considered a waveguide for light, and the first layer can be considered a waveguide filter. In waveguide technology, it is known that the width of the waveguide for an electromagnetic wave imposes a cutoff frequency, below which the electromagnetic wave will exponentially decay rather than propagate through the waveguide.

[0011] According to a second aspect of the invention, an optical element includes a radiation filtering first layer and a transparent element. The first layer includes at least one aperture penetrating through the first layer to the transparent element for propagating radiation through the at least one aperture. A cross-sectional size of the at least one aperture is configured to provide a cutoff frequency such that radiation having a frequency below the cutoff frequency is attenuated inside the at least one aperture and radiation having a frequency above the cutoff frequency propagates through the at least one aperture.

[0012] A second aspect addresses the problem of providing efficient filtering to passive optical components such as lenses, fiber couplers, or generally transparent elements that are part of a passive optical component. A further problem may be to improve the imaging properties of such optical components by providing optical elements with specific filtering properties that reduce undesirable frequency-dependent behavior of the passive optical components.

[0013] The first and second aspects of the invention may share one or more features, and therefore the embodiment of such shared features may be the same for both aspects. In one example, an embodiment of a first layer that filters radiation may be used in an image sensor and also in an optical element. Thus, any reference made to the first layer of an image sensor also applies to the first layer of an optical element and vice versa. [Brief description of the drawings]

[0014] [Figure 1] 1 is a cross-sectional view of an image sensor according to an embodiment; [Diagram 2] Fig. 2A is a schematic cross-sectional view of a pixel of an image sensor according to an embodiment in which a first layer is provided on the photosensitive layer of the image sensor. Fig. 2B is a schematic cross-sectional view of a pixel of an image sensor according to an embodiment in which a first layer is provided on the photosensitive layer of the image sensor and a transparent layer is interposed between the first layer and the photosensitive layer. Fig. 2C is a schematic cross-sectional view of a pixel of an image sensor according to an embodiment in which a first layer is provided on the photosensitive layer of the image sensor and the openings of the first layer are filled with a transparent material that further provides a transparent protective layer on the first layer. [Figure 3A] 1 is a schematic cross-sectional view of the interaction of an image sensor with incident radiation according to an embodiment having an empty aperture; [Figure 3B] 1 is a schematic cross-sectional view of an image sensor and the interaction of incident radiation according to an embodiment having a filled aperture; [Figure 3C] 1 is a schematic cross-sectional view of an image sensor and the interaction of incident radiation according to an embodiment having a filled opening and a transparent protective layer. [Figure 4] FIG. 2 is a schematic top view of an embodiment in which four adjacent pixels are shown and the openings of two adjacent pixels are perpendicular to each other and have an elongated rectangular cross section. [Diagram 5] FIG. 2 is a schematic top view of an embodiment of an image sensor showing a 3×3 array of pixels with each pixel having an aperture group, the apertures of each pixel having three different cross-sectional sizes. [Figure 6] 1 is a cross-sectional view of an optical element according to an embodiment. [Figure 7] Figure 7A is a schematic diagram of an optical device for measuring radiation within a range according to an embodiment, and Figure 7B is a schematic diagram of a further optical device for measuring radiation within a range according to an embodiment. [Figure 8] 1 is a block diagram of an imaging device having an image sensor according to an embodiment. [Figure 9] FIG. 2 is a flow diagram of a method for computing a representation of an image captured by an image sensor, according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Detailed Description of the Embodiments Incident radiation as used herein, i.e. in any or all of the aspects and embodiments of the present invention, is electromagnetic radiation and may include visible light, or visible light accompanied by infrared light and / or ultraviolet (UV) light.

[0016] Preferably, incident radiation having a frequency below the first cutoff frequency is exponentially attenuated inside the at least one opening. The term "cutoff frequency" is therefore preferably defined as the limit frequency at which the electromagnetic wave exponentially attenuates with increasing depth of the opening. The exponential attenuation is used according to the waveguide theory. Of course, it is possible to adopt a different definition for the cutoff frequency, for example to define further cutoff frequencies, below which the transmitted intensity decreases to a certain value, for example between 90% and 0% of the intensity of the incident radiation.

[0017] Incident radiation with a frequency above the second cutoff frequency propagates through at least one aperture, i.e., inside the aperture there is a mode that allows the radiation to propagate. In a real-world aperture, the intensity of the propagating radiation may still be reduced, but the reduction will be measurably smaller than for radiation below the cutoff frequency. In standard waveguide theory, the cutoff frequency may be equal to one specific value, but in embodiments there may be a transition frequency range between the first cutoff frequency (below which the radiation is attenuated in the at least one aperture) and the second cutoff frequency (above which the radiation propagates through the at least one aperture). The frequency range between the first and second cutoff frequencies may be caused, for example, by imperfections in the aperture or the walls of the aperture or other non-ideal conditions, or even by the particular morphology of the aperture.

[0018] In general, the cutoff frequency as used herein for any or all of the embodiments may be a cutoff frequency range (e.g., limited by a first cutoff frequency and a second cutoff frequency) such that incident radiation having a frequency below the cutoff frequency range is attenuated inside the at least one opening, and incident radiation having a frequency above the cutoff frequency range propagates through the at least one opening. In other words, the term cutoff frequency as used herein may be defined as a cutoff frequency range, and may be replaced with the term cutoff frequency range. According to standard waveguide theory, a waveguide may have a lowest cutoff frequency, below which incident radiation cannot propagate through the at least one opening due to the absence of a propagation mode. In general, the cutoff frequency as used herein for any or all of the embodiments may refer to such a lowest cutoff frequency, and may not include a cutoff frequency greater than the lowest cutoff frequency. There may also be two or more propagation modes with the same lowest cutoff frequency (within a range). In one example, a square opening (i.e., a rectangular opening with equal side lengths) may have two propagation modes with the same lowest cutoff frequency.

[0019] As used herein, expressions such as "equal" and "greater than" should be understood to mean "substantially equal" and "substantially greater than", respectively, rather than in the strict mathematical sense. A possible definition of what values ​​are "substantially equal" can be "equal within the deviations caused by manufacturing tolerances" or "equal within ±10%" or "equal within ±5%" for use in any or all embodiments. Another definition for equal values ​​can be that they are equal within the accuracy of typical measurements.

[0020] As an alternative to being defined by its cross-sectional size, the at least one opening may be sized or sufficiently small or have an extent in two directions perpendicular to its depth such that the at least one opening provides at least a cutoff frequency, such that incident radiation having a frequency below the first cutoff frequency is attenuated within the at least one opening and incident radiation having a frequency above the cutoff frequency propagates through the at least one opening.

[0021] The filter layer may be formed from a metal layer with an opening or a non-metal layer with an opening with a metal coating. The filter layer may also be formed of a layer of an impermeable material with an opening, as long as there is a sufficient filtering or attenuating effect on light above a certain wavelength or wavelength range. In general, the filter layer may be configured to sufficiently remove or attenuate light above a certain wavelength. The filtering or attenuating effect need not be close to 100%, the filtering or attenuating effect may be lower, for example, within the range of 100% to 10%, such that the transmitted intensity is reduced between 0% and 90%. In general, the filter layer may be configured to allow sufficient propagation of incident radiation above a certain frequency, for example, by coupling the incident radiation to a propagation mode of the opening. Sufficient propagation may be provided if the propagating radiation is measurably less attenuated than the radiation below the cutoff frequency. The difference between the propagating radiation above the cutoff frequency and the attenuated radiation below the cutoff frequency may have a rounded value in the range of 100% to 10%. This may mean that radiation propagating above the cutoff frequency may also be attenuated to a rounded value between 100% and 10%, but still be measurably less attenuated than radiation below the cutoff frequency.

[0022] In general, the filter layer and the at least one aperture may be configured to efficiently couple photons that impinge on the filter layer into one or more modes that propagate through the at least one aperture.

[0023] In the following, exemplary embodiments of the openings are described first.

[0024] The two basic parameters of an aperture are the parameter describing the size of the aperture and the parameter describing the depth of the aperture (or the thickness of the first layer). The calculations below are done for apertures that are not filled with additional transparent material, but such materials can be taken into account along with their optical properties.

[0025] For the propagation of electromagnetic waves through an empty waveguide, the following wave equation (derived from Maxwell's equations in the absence of a vacuum source) is taken as a starting point: TIFF2025039865000002.tif7128 or in a different notation TIFF2025039865000003.tif4128, where f (x,y,z,t) represent the electric and magnetic fields (the underlines represent vectors).

[0026] For a general waveguide in the z direction with constant axial cross section, the following solution for f(x,y,z,t), which describes a wave propagating in the z direction, is appropriate: TIFF2025039865000004.tif4128

[0027] When using this solution, f (x,y,z,t) is a complex function, and a physical quantity such as the electric field E or the magnetic field B is treated as the real part of this function.

[0028] The wave equation is as follows: TIFF2025039865000005.tif11128 or TIFF2025039865000006.tif4128

[0029] This equation, together with the boundary conditions derived from a waveguide with perfect conductors as walls, defines an eigenvalue problem for f(x,y).

[0030] The boundary condition is the normal vector of the wall.n This is in response to. n · B =0 (i.e., the B magnetic field vanishes at the wall in the normal direction) n × E = 0 (i.e., the electric field E vanishes tangentially on the wall).

[0031] For a rectangular shaped waveguide with width a and height b in the x and y directions, respectively, it is common to use a planar wave function that is appropriate for such a geometry. TIFF2025039865000007.tif4128

[0032] From the wave equation we get: TIFF2025039865000008.tif4128

[0033] The boundary conditions give results for the wavenumber. TIFF2025039865000009.tif4128

[0034] From this, the wave number in the z direction is given as follows: TIFF2025039865000010.tif11128

[0035] This means that for a wave propagating in the z direction (i.e. not exponentially decaying), the frequency w is the smallest cutoff frequency TIFF2025039865000011.tif7128 (this is for a>b (and n=1, m=0), TIFF2025039865000012.tif4128).

[0036] Here, the propagation modes can be classified into two main mode types: (i) transverse electric modes (TE_m,n) where E_z=0 (i.e., the electric field is always oriented transverse to the direction of propagation) and (ii) transverse magnetic modes (TM_m,n) where B_z=0 (i.e., the magnetic field is always oriented transverse to the direction of propagation). There are also transverse electromagnetic modes (TEM) where E_z=0 and B_z=0 (i.e., the electric and magnetic fields are always oriented transverse to the direction of propagation), but this requires a waveguide with at least two electrically insulated conducting walls. For a waveguide with a square cross section (i.e. a=b), we have: TIFF2025039865000013.tif4128

[0037] As shown in the textbook "Classical Electrodynamics" by JD Jackson, 6th edition, 1967, the TE mode always has the lowest cutoff frequency TIFF2025039865000014.tif7128, while TM mode has a higher cutoff frequency.

[0038] A waveguide with a square cross section has two separate TE modes with the same cutoff frequency, namely the TE_1,0 and TE_0,1 modes. As a result, such a waveguide is a multimode waveguide, which can have the advantage that photons that strike the waveguide's aperture are more efficiently coupled into the modes of the waveguide, such that the waveguide is more transparent to radiation or light above the cutoff frequency.

[0039] Because the lengths of the waveguides considered here are very short (eg, less than 10 μm or even less than 1 μm), other effects that may be associated with walls made of imperfect conductors may be neglected here.

[0040] For a square waveguide, the two smallest modes (n=1,m=0 and n=0,m=1) give: TIFF2025039865000015.tif4128

[0041] In the case of a non-propagating mode of the waveguide, that is, a mode having an angular frequency smaller than the cut-off frequency \(w < w_c = c(\pi / a)\), the solution of \(k_z\) is a pure imaginary number, leading to an exponential decay of the wave in the z-direction. TIFF2025039865000016.tif18128Here, \(\lambda = 2\pi c / w\) is the wavelength of the electromagnetic (em) wave, and \(\lambda_c = 2a\) is the cut-off wavelength. Equation (*) is a basic equation showing that when the wavelength of the em wave exceeds the cut-off wavelength, the propagation in the waveguide direction decays exponentially as indicated by \(k_z\).

[0042] The above equation for \(k_z\) approximately holds for an empty waveguide. For example, in the case of a waveguide filled with a dielectric material, it is, for example, TIFF2025039865000017.tif4134Here, \(\varepsilon\) is the permittivity of the dielectric transparent material in the waveguide, and \(\mu\) is the permeability of the dielectric transparent material. \(\varepsilon_0\) and \(\mu_0\) are the corresponding constants in vacuum.

[0043] Equation (*) is here used to design an exemplary aperture such that visible light with wavelengths from 400 nm to 700 nm is separated into three components representing blue light, green light, and red light.

[0044] The largest aperture is configured to transmit all visible light while suppressing light with wavelengths above 700 nm (infrared light). As a result, \(a_1 = 350\) nm.

[0045] The medium-sized aperture is configured to allow the propagation of blue light and green light (and yellow light) while suppressing red light with wavelengths above 600 nm. As a result, \(a_2 = 300\) nm.

[0046] The smallest aperture is configured to enable the propagation of blue light while suppressing green light and red light with wavelengths above 500 nm. As a result, \(a_3 = 250\) nm.

[0047] In other embodiments, the cutoff frequency may be different to separate the visible light into different components. As an example, a medium sized aperture may be selected to allow the propagation of blue and green light and suppress yellow and red light above 570 nm. This results in a_2=285 nm.

[0048] The light intensity S is proportional to the square of the electric field of the em wave: S prop(Re f)**2.

[0049] Now, the aperture can be designed such that light with wavelength λ_D, for example 100 nm greater than the cutoff wavelength (λ_D=λ_c+100 nm), is reduced by at least a factor of exp(-2)≈0.135. This means that the electric field is reduced by a factor of exp(-1). This gives the minimum depth l_z of the aperture. TIFF2025039865000018.tif4128

[0050] From this we obtain for the largest aperture (which determines the minimum thickness of the first layer) with a=a_1=350 nm: TIFF2025039865000019.tif4128

[0051] If it is desired to reduce the light intensity for light having wavelength λ_D (λ_D=λ_c+100 nm) by at least a factor of exp(-4)≈0.018, this gives a minimum depth l_z' with a=a_1. TIFF2025039865000020.tif4128

[0052] The calculated l_z' values ​​provide fairly strong color separation, which is even stronger for shorter wavelength light.

[0053] The embodiment with the first layer thickness as specified by l_z or l_z' can be easily manufactured using standard chip manufacturing processes, where a metal layer having a thickness of, for example, 460 nm can be deposited on top of a bare image sensor (i.e., an image sensor without a color filter array) of a design that may be conventional with pixels having a pixel pitch of about 1 to 10 micrometers (or smaller or larger). Openings in the first layer may then be provided, for example by using an etching process. In this way, the pixels are directly coupled to the openings in the metal layer. The openings may have sizes of 350 nm, 300 nm, and 250 nm. All the necessary manufacturing steps are readily available with today's technology.

[0054] A thicker first layer allows for a more strict reduction of light with wavelengths greater than the cutoff wavelength defined by the aperture size. A thicker first layer therefore increases the color separation. A thinner layer therefore reduces the color separation between the different light components. Depending on the application, the color separation may be implemented in a more strict or less strict manner.

[0055] In further embodiments, the thickness of the first layer may also vary from opening to opening.

[0056] In further embodiments, the shape of the openings may be, for example, conical or tapered (i.e., having a larger opening on one side of the filter layer than the other), and the cross-section of the openings may be non-rectangular, such as circular, hexagonal, or other shape.

[0057] In general, it may be desirable to increase the surface of the image chip available for light measurement as much as possible so that the light collection efficiency is increased. As a result, it is possible to couple as many openings as possible to the light-sensitive area of ​​the light-sensitive layer, and to make the wall thickness between the openings as thin as possible.

[0058] For example, assuming that the photosensitive area of ​​an image sensor with a pixel pitch of about 1 micrometer is 800 nm x 800 nm, one embodiment may group four apertures of the same size and directly combine the four apertures into one pixel. In this way, light loss is reduced. It may also be possible for four apertures of the same size (e.g., 350 nm) to be combined into one pixel and nine smaller apertures of the same size (e.g., 250 nm) to be combined into another pixel.

[0059] In general, the number of apertures associated with one pixel may vary according to pixel size, aperture size, and wall thickness. As an example, a square pixel with a size of 4 μm may have the following number of square apertures when the metal wall has a thickness of about 50 nm: for a square aperture with a=350 nm, there may be 9×9 or 10×10 apertures per pixel; for a=300 nm, there may be 11×11 apertures per pixel; and for a=250 nm, there may be 13×13 apertures per pixel. In a further example, the pixel may have a side length of less than 1 μm, and for each pixel, there may be a single aperture associated with the pixel.

[0060] In the embodiment, the openings are a1>a2>a3...>a n n different sizes of side lengths a i The aperture may have a constant square cross section with i=1..n. The aperture may have metal walls and a cutoff frequency with a corresponding cutoff wavelength λ i,cut-off above which the incident radiation exponentially decays inside the at least one aperture, and below which the incident radiation propagates through the aperture of the aperture, where a i is λ i,cut-off / 2. In this example, the thickness of the first layer, l, corresponds to the depth of the opening, d. z is for the vacuum in the opening, TIFF2025039865000021.tif19128 or TIFF2025039865000022.tif4128, where a_1 is the length value of one side a i (i=1..n) is the maximum side length value, λ_D,1=λ D,1 is the cutoff wavelength λ 1,cut-off = 2a1 and with a strength of at least e -2 In other words, if the filter layer has a depth d according to the above formula, the intensity of radiation having wavelength λ_D,1 will be attenuated by at least e -2 The attenuation is λ_D,1 and λ 1,cut-off Using the above formula, it is possible to determine a depth that provides some degree of attenuation, with radiation having wavelengths between λ_D,1 being attenuated less and radiation below λ_D,1 being attenuated more. In one example, λ_D,1 may be set to a depth greater than λ_D,1 for a relatively low level of attenuation or a less steep filtering characteristic. 1,cut-off In a further example, λ_D,1 may be selected to be 100 nm greater than λ for a relatively high level of attenuation or steeper filtering characteristics. 1,cut-off may be selected to be 25 nm greater than

[0061] The above formulas for l_z, vac, or d_vac are valid for a completely empty aperture, an aperture where no vacuum is present, or an aperture filled with a substance such as a dielectric material or air. TIFF2025039865000023.tif19128 or TIFF2025039865000024.tif4128Here, ε is the dielectric constant of the material in the opening, μ is the magnetic permeability of the material in the opening, ε_0 is the dielectric constant of vacuum, and μ_0 is the magnetic permeability of vacuum.

[0062] Thus, the first aspect may relate to an image sensor for recording incident radiation. The image sensor may include a first layer for filtering the incident radiation and a second photosensitive layer for absorbing and recording radiation passing through the first layer. The first layer may be formed from one or more materials that are opaque to the incident radiation. In one example, the first layer may be considered to be opaque to the incident radiation because the incident radiation couples to one or more modes provided by the at least one aperture, resulting in attenuation of the incident radiation (or the energy of the incident radiation) or propagation of the incident radiation (or the energy of the incident radiation) depending on the frequency of the incident radiation. The first layer (or filter layer) is in front of the second photosensitive layer in the propagation direction of the incident radiation, and the first layer includes at least one aperture penetrating through the first layer to the second layer to propagate radiation through the at least one aperture. The cross-sectional size of the at least one aperture may be configured to provide a cutoff frequency such that incident radiation having a frequency below the cutoff frequency is attenuated inside the at least one aperture and incident radiation having a frequency above the cutoff frequency propagates through the at least one aperture. In other words, the at least one aperture provides a cutoff frequency, incident radiation having a frequency below the cutoff frequency is attenuated inside the at least one aperture and incident radiation having a frequency above the cutoff frequency propagates through the at least one aperture. Thus, the at least one aperture filters the incident radiation by attenuating incident radiation having a frequency below the cutoff frequency provided by the at least one aperture and by propagating incident radiation having a frequency above the cutoff frequency through the at least one aperture. The terms cutoff frequency, attenuated or attenuating, and propagating may be used as in the field of waveguides. Thus, incident radiation having a frequency below the cutoff frequency may be attenuated exponentially inside the at least one aperture and incident radiation having a frequency above the cutoff frequency may propagate through the at least one aperture by coupling to one or more propagation modes within the at least one aperture.In an embodiment, incident radiation having a frequency above the cut-off frequency can also be exponentially attenuated, but this attenuation is not predicted by waveguide theory and can instead be caused by other reasons (e.g., non-ideal conductors such as walls or other real-world effects). As described above, the term cut-off frequency as used herein can mean a cut-off frequency range (not just one cut-off frequency value predicted by waveguide theory). In this case, incident radiation having a frequency below the cut-off frequency range is attenuated within at least one opening, and incident radiation having a frequency above the cut-off frequency range propagates through at least one opening.

[0063] In an embodiment, at least one opening may be perpendicular to the plane of the first layer, at least one opening has a constant cross-section along the axis of the opening, and the cross-section is symmetric and optionally a square, rectangular, circular, or hexagonal cross-section. The first layer may be a metal or the first layer may include a non-metal layer coated with a metal layer, such that in both cases, at least one opening has one or more metal walls. Optionally, all walls of at least one opening may be metal.

[0064] In an embodiment, the first layer may include two, three, or more openings with different cross-sectional sizes csi (i = 1..n) where cs1 > cs2 > cs3… > csn. Each opening may be configured to attenuate light having a frequency below its respective cut-off frequency w1, w2, w3, …, wn and to propagate light having a frequency above its respective cut-off frequency, where w1 < w2 < w3… < wn. In other words, the cross-sectional size cs1 is configured to provide the cut-off frequency w1, the cross-sectional size cs2 is configured to provide the cut-off frequency w2, and so on.

[0065] In an embodiment, at least one opening may be filled with a transparent material and optionally the transparent material may provide an additional protective layer over the first layer.

[0066] In an embodiment, the at least one aperture may have a constant square cross section with n different sized side length values ​​ai (i=1...n). The at least one aperture may have metal walls and a cutoff frequency may have a corresponding cutoff wavelength λi,cut-off (i=1...n), above which the incident radiation exponentially attenuates inside the at least one aperture and below which the incident radiation propagates through the at least one aperture. The side length value ai may be equal to λi,cut-off / 2 (i=1...n). Optionally, the thickness l_z of the first layer or the depth d of the at least one aperture may be determined by: TIFF2025039865000025.tif4133, where a1 is the maximum side length value of ai (i=1..n), λ_(D,1) is the wavelength of radiation that is greater than the cut-off wavelength λ1,cut-off=2a1 and is attenuated in intensity by at least e**(-2) times inside the at least one aperture when it reaches the second photosensitive layer, ε is the permittivity of the material in the at least one aperture, μ is the magnetic permeability of the material in the aperture, and ε_0 and μ_0 are the corresponding constants of vacuum. The material in the aperture may be, for example, air, a dielectric transparent material or also a vacuum (in which case the factor sqrt((μ_0ε_0) / (με)) equals 1).

[0067] In an embodiment, the photosensitive layer may include an array of pixels, with each opening or each of a plurality of connected openings having the same cross-sectional size overlapping and associated with an individual pixel of the array, In an embodiment, each pixel may include a photosensitive area, and the opening or the plurality of connected openings associated with a pixel may cover an area that is equal to or larger than the photosensitive area of ​​the individual pixel, but less than the entire area of ​​the pixel.

[0068] The image sensor may have different pixel arrays. In one example, the array may have one or a very small number of pixels, in a different example, the array may be 4000x6000 pixels, or much more, such as 100 million pixels (or more), or any number of pixels in between. If the image sensor is used in, for example, a scanning device, the array may also have 1x8000 pixels or 3x8000 pixels. Again, the number of pixels may be more or less.

[0069] In the embodiment, the photosensitive layer of the image sensor may be similar or identical to the photosensitive layer of a conventional image sensor. The image sensor for recording incident radiation according to the embodiment may be similar to a conventional image sensor as far as the circuit is concerned. In other words, the image sensor according to the embodiment may be similar to a conventional image sensor in which a conventional color filter array with colored filter elements is replaced by the first layer according to the embodiment.

[0070] In an embodiment, a transparent layer may be located between the first layer and the photosensitive layer. Optionally, the transparent layer may include walls located above the boundaries of individual pixels, the walls being opaque to incident radiation, preferably metallic.

[0071] In an embodiment, the first layer may include an array of apertures with two or more different cross-sectional sizes. Each aperture in the array may have a constant square cross-section with n differently sized side length values ​​ai (i=1..n), and each aperture may have a metal wall. The apertures may be grouped such that each aperture group has one aperture or two or more apertures that are connected together and have the same cross-sectional size, and each aperture group overlaps and is coupled to only one pixel of the photosensitive layer. The aperture groups may be arranged in a repeating pattern in the first layer.

[0072] In an embodiment, the openings may have three or more different cross-sectional sizes, a first size filtering out infrared light and allowing red, green, and blue light to propagate, a second size filtering out red light and allowing green and blue light to propagate, and a third size filtering out red and green light and allowing blue light to propagate. The repeating pattern in the first layer may be configured to provide measured or interpolated values ​​for red, green, and blue light for each pixel of the pixel array. Optionally, a cutoff wavelength λ_(1) corresponding to a cutoff frequency of the first cross-sectional size is in the range of 600 nm to 750 nm, optionally in the range of 650 nm to 730 nm, a side length value a1 of the first cross-sectional size is in the range of 300 nm to 375 nm, optionally in the range of 325 nm to 365 nm, and a cutoff wavelength λ_(2) corresponding to a cutoff frequency of the second cross-sectional size is in the range of 495 nm to 590 nm, optionally in the range of 515 nm to 570 nm. nm range, a side length value a2 of the second cross-sectional size is in the range of 247 nm to 295 nm, optionally in the range of 257 nm to 285 nm, a cutoff wavelength λ_(3) corresponding to the cutoff frequency of the third cross-sectional size is in the range of 450 nm to 510 nm, optionally in the range of 465 nm to 495 nm, and a side length value a3 of the third cross-sectional size is in the range of 225 nm to 255 nm, optionally in the range of 233 nm to 248 nm.

[0073] In an embodiment, the first layer may include an array of apertures with three or more different cross-sectional sizes, a first size allowing infrared, red, green, and blue light to propagate, a second size removing red light and allowing green and blue light to propagate, and a third size removing red and green light and allowing blue light to propagate. Each aperture of the first size may overlap a pixel area subpixel, and each aperture of the first size may be filled with a filter material that removes infrared light and transmits light having a shorter wavelength than infrared light. Optionally, all apertures are filled with filter material and / or the filter material provides an additional layer on the first layer. In this embodiment, the largest aperture is so large that the largest aperture does not provide a filtering function, for example by removing infrared light. Instead, infrared light is filtered by additional colored filter material, as additional colored filter material may be used in conventional image sensors. Such an embodiment may be more effective to collect all frequencies of light that are recorded in the pixel bounded by the largest aperture.

[0074] In an embodiment, the array of apertures may have three different cross-sectional sizes, and groups of apertures may be arranged in two alternating rows of pixels. The pixels in the first row may alternately measure red, green, and blue light in one pixel and green and blue light in the next pixel, and the pixels in the second row may alternately measure green and blue light in one pixel and blue light in the next pixel. A pixel in the first row measuring red, green, and blue light may have an adjacent pixel in the second row measuring green and blue light.

[0075] In embodiments, the opening may have any one of the following additional sizes, or any combination of sizes: a fourth size that removes a portion of the far-infrared light below the infrared light cutoff frequency and allows the propagation of a further portion of the near-infrared light above the infrared light cutoff frequency and the propagation of red, green, and blue light; a fifth size that removes red and yellow light and allows the propagation of green and blue light; a sixth size that removes a portion of the green light below the red and green light cutoff frequencies and allows the propagation of a further portion of the green light above the green light cutoff frequency and the propagation of blue light; and a seventh size that removes red, green, and blue light and allows the propagation of ultraviolet light.

[0076] In an embodiment, the aperture may have two or more different cross-sectional sizes, one size that removes light below a lower cutoff frequency and allows light above the lower cutoff frequency to propagate, and another size that removes light below a high cutoff frequency and allows light above the high cutoff frequency to propagate, such that the amount of light in the range between the lower and upper cutoff frequencies can be measured in addition to other components of the light. Optionally, the range between the lower and upper cutoff frequencies may include one or more spectral lines, optionally absorption or emission lines of the atoms or molecules, that identify transitions between different states of the atoms or molecules.

[0077] In an embodiment, at least one aperture may be elongated, preferably rectangular in cross section, such that the length of the at least one aperture in the elongated direction defines a cutoff frequency for incident radiation polarized perpendicular to the elongated direction, and the length of the at least one aperture in the direction perpendicular to the elongated direction defines a further cutoff frequency that is greater than the upper limit of the range of incident radiation desired to be analyzed by the image sensor. Optionally, the first layer may include a plurality of first elongated apertures overlapping first pixels of the array of pixels of the photosensitive layer, and a plurality of second elongated apertures overlapping adjacent second pixels and having a cross-sectional shape rotated by 90° around the axis of the second elongated aperture relative to the first elongated aperture.

[0078] In embodiments, at least one opening may taper toward or away from the second layer.

[0079] 1 is a cross-sectional view of an image sensor 10 according to an embodiment. The image sensor 10 has a first layer 11 that filters incident radiation and a second photosensitive layer 12 that absorbs radiation passing through the first layer. The second photosensitive layer 12 is a pixel 14. n,m The array of pixels includes only pixel 14. n,m are combined into 5×5 apertures 13, although other numbers of apertures per pixel are possible depending on the size of the pixel, the apertures and the walls. In an embodiment, the apertures 13 are perpendicular to the plane of the first layer 11 and have a constant square cross section along the axis of the aperture, and are aligned with the pixel 14. n、m 1. The apertures 13 all have the same size to provide one cutoff frequency for . The apertures 13 penetrate through the first layer 11 to the second layer 12, so that radiation above the cutoff frequency provided by the apertures 13 can propagate through the apertures 13 and be measured in the photosensitive second layer 12. In contrast, radiation below the cutoff frequency will be attenuated inside the apertures 13 and will contribute only a small portion to the radiation measured in the second layer 12. By varying the thickness of the first layer 11, the intensity of radiation below the cutoff frequency reaching the second layer can be controlled to be as small as desired (until it actually reaches zero). The apertures 13 have metal walls 20 and act as a waveguide. Incident radiation below the cutoff frequency is attenuated inside the apertures 13 by exponentially attenuating therein, while incident radiation above the cutoff frequency propagates through the apertures 13 by coupling to one or more propagation modes inside the apertures 13. In this embodiment, the cutoff wavelength corresponding to the cutoff frequency is twice the length of one side of the square cross section of the aperture 13 .

[0080] In FIG. 1, a plurality of 5×5 adjacent openings 13 having the same cross-sectional size are arranged to define individual pixels 14 of the array. n,m and are coupled to the respective pixel. Other pixels of the array may be coupled to a further plurality of adjacent apertures having a square cross-section but different cross-sectional sizes.

[0081] The filter layer 11 may be fabricated on top of the photosensitive layer 12, which may be of conventional design. The filter layer 11 may be fabricated using standard CMOS or other chip fabrication techniques by providing a metal layer of sufficient thickness, and an opening 13 is fabricated in the filter layer 11. In an embodiment, the opening may be fabricated in a manner similar to that of a zero-mode waveguide used, for example, for single molecule sequencing (see US Patent Application Publication No. US2014 / 0175052A1, which is incorporated herein by reference). The filter layer 11 may also be fabricated by providing a dielectric layer on the image sensor, removing a portion of the dielectric layer to form the opening 13, and then coating the walls 20 on the side and top with a metal layer (see, for example, International Publication No. WO2013 / 109877, which is incorporated herein by reference).

[0082] 2A is a schematic cross-sectional view of a pixel of an image sensor 10 according to an embodiment in which a first layer 11 for filtering incident radiation with a through hole is provided on a second photosensitive layer 12 of the image sensor 10. As shown, the first layer 11 is disposed over and in contact with the second photosensitive layer 12, which absorbs radiation passing through the filter layer 11. The filter layer 11 has an opening 13 with walls 20 that allows radiation to propagate above a cutoff frequency. The thickness of the first layer 11, and therefore the depth of the opening 13, is d.

[0083] The photosensitive layer 12 includes pixels 14 n , 14 n+1, ..., with each aperture 13 overlapping and associated with an individual pixel of the array. Each pixel 14 comprises a light-sensitive area 15, and the apertures 13 may, as shown in a non-limiting manner, only overlie the light-sensitive area 15 of the pixel, thereby blocking unmeasurable light from reaching the image sensor. This can reduce stray light that may reach neighboring pixels, such that no or little light blocking between pixels in the light-sensitive layer is required.

[0084] 2B is a schematic cross-sectional view of a pixel of the image sensor 10 according to an embodiment in which a first layer 11 having a through hole is provided on the second photosensitive layer 12 of the image sensor, and a transparent layer 16 is interposed between the first layer 11 and the second photosensitive layer 12. In this embodiment, the transparent layer 16 is introduced between the photosensitive layer 12 and the filter layer 11 to allow an increase in the area for collecting the light to be measured. n The filter region of the first layer 11 coupled to the pixel 14 n The filter area is larger than the photosensitive area 15 of the pixel 14. n The aperture 13 is approximately as large as the area of ​​the aperture itself. The walls 17 may correspond to the limits of the pixel, for example over the boundary between adjacent pixels. In this embodiment, the aperture 13 covers an area larger than the sensitive area to maximize light collection. In an embodiment, the aperture combined with one pixel may cover as much of the pixel area as possible, taking into account the size of the aperture, the walls, and the pixel.

[0085] The walls 17 of the transparent layer 16 may be opaque (i.e., non-transparent to light) or metallic, and are arranged in a pixel region 14. n Another pixel area 14 n+1 The transparent layer 16 with the walls 17 may be fabricated by first providing an opaque or metallic layer, removing the opaque or metallic layer over the pixel area so that only the walls 17 remain, and then providing the transparent dielectric material of the transparent layer 16.

[0086] FIG. 2C is a schematic cross-sectional view of a pixel of the image sensor 10 according to an embodiment in which a first layer 11 having a through opening is provided on the photosensitive layer 12 of the image sensor 10, and a transparent material 18 is filled in the opening 13 of the filter layer 11, which further provides a transparent protective layer 19 on the filter layer 11. In this embodiment, the transparent material 18 has a thickness greater than that of the filter layer 11, in order to provide an additional layer 19 on the filter layer having an opening. The additional layer 19 may be a protective layer. In a further embodiment, the opening 13 of the filter layer 11 may be filled with a transparent material 18 having the same thickness as the thickness of the filter layer 11. In an embodiment with a transparent material, the calculation regarding the cutoff frequency may have to take into account the dielectric properties of the transparent material 18.

[0087] Figures 3A-3C show diagrammatically the interaction of incident photons with the individual apertures 13 for the cases described above and shown in Figures 2A and 2C. Each of the apertures 13 in these examples has a metal wall 20.

[0088] 3A is a schematic cross-sectional view of the interaction of incident photons with an image sensor according to an embodiment with an empty aperture. In FIG. 3A, the aperture 13 is empty, i.e., not filled with a transparent material corresponding to the example shown in FIG. 2A. If the incident photons impinge on the aperture 13, for example, at an oblique angle of incidence, and the frequency of the incident photons is sufficiently large relative to the size of the aperture 13, a propagating waveguide mode will exist within the aperture. Thus, the photons or the energy of the photons can reach the underlying photosensitive region of the image sensor pixel.

[0089] Fig. 3B is a schematic cross-sectional view of the interaction of the image sensor with incident radiation according to an embodiment with a filled aperture. In Fig. 3B, the aperture is filled with a transparent material corresponding to the example of Fig. 2C, where the thickness of the transparent material 18 is equal to the thickness of the filter layer 11. In other words, the aperture of the filter layer is filled with a transparent material up to the top edge of the wall of the aperture. As in the case of Fig. 3A, incident photons enter the aperture, which may arrive at an oblique, i.e. non-perpendicular, incidence angle, but only if the frequency of the incident photons is large enough relative to the size of the aperture, a waveguide mode is available that propagates the energy of the photons (or a part of the photons) downwards to the photosensitive layer of the image sensor.

[0090] 3C is a schematic cross-sectional view of the interaction of incident radiation with an image sensor according to an embodiment having a filled aperture and a transparent protective layer 19. In FIG. 3C, a transparent material is deposited on the filter layer 11 such that the aperture is filled with the transparent material and there is a transparent protective layer 19 on top of the filter layer 11. In this case, the incident photons are first refracted before they enter the aperture. Again, only if the frequency of the incident photons is large enough will a propagating waveguide mode become available, allowing the photon's energy to reach the underlying photosensitive layer or region of the image sensor substantially unattenuated.

[0091] In a further embodiment, a transparent material may be provided in the form of a microlens (e.g. with a convex upper surface) above each opening of the filter layer 11, so that the incident radiation is focused in the center of each opening. Such a microlens may be used in embodiments where the openings are not filled and in embodiments where the openings are filled with a transparent material. In the second case, the transparent material of the microlens may be the same as the transparent material with which the openings are filled, or the transparent material may be different. The embodiment with the microlens may provide a better coupling of the incident radiation into the propagation mode, so that the recording efficiency of the image sensor is increased.

[0092] In yet another embodiment, a multispectral image may be provided by using a filter array with apertures having more than three different sizes. In one example, an aperture larger than 350 nm may also be used to measure infrared light and distinguish infrared light from other colors being measured. In one example, an aperture smaller than 250 nm may be used to measure UV light and distinguish UV light from other colors being measured. Additionally, apertures having a size between 250 nm and 350 nm may be used to measure light within a narrower frequency band. When attempting to measure narrow band light, it may be advantageous to place pixels that measure light above and below the limits of the narrow band in close proximity, for example as neighboring pixels. This may increase the spatial resolution of the system for narrow band light. In general, such an embodiment may have advantages over standard color filter arrays such as Bayer filters, since different cutoff frequencies may be defined at any value without relying on dyes available for filtering light. Also, the cutoff frequency may provide a sharper limit for light that is measured (e.g., light above the cutoff frequency) as opposed to light that is not measured (e.g., light below the cutoff frequency) by adjusting the depth of the filter layer. The manufacturing method is always similar, only the mask for forming the openings needs to be changed, and this can be done using standard CMOS techniques or other standard chip manufacturing techniques.

[0093] In further embodiments of the invention, the polarization of light may be measured by using a non-square aperture, i.e. an aperture that is longer in one direction than the other. In an embodiment, the long length may be used to provide a filtering function by defining a cutoff frequency, while the short length may be selected small such that the cutoff frequency of the aperture is higher than the radiation desired to be measured (e.g. in the UV frequency range above the blue light range). In this embodiment, the higher cutoff frequency is used not to separate light components or frequency ranges, but to suppress all light desired to be measured and polarized along the long side of the aperture. In this embodiment, the waveguide provided by the aperture may be effectively a single mode TE waveguide, and only modes with electric field vectors perpendicular to the long side of the aperture (and perpendicular to the axis of the aperture, i.e. perpendicular to the normal vector of the filter layer) may propagate through the aperture (if this electric field vector varies in length along the long side of the aperture). The long side of the aperture may be varied to define different cutoff frequencies for light with electric field vectors perpendicular to the long side of the aperture, and therefore different color ranges with such polarization may be measured and distinguished. In such an embodiment, light polarized in the same direction as the single TE mode will couple efficiently into the only propagating mode, while light polarized perpendicular to the single TE mode will couple much less efficiently into the only propagating mode and will therefore be rejected regardless of frequency (assuming the short side of the aperture is small enough so that the measured light is below the cutoff frequency of the short side). In an embodiment, the aperture may be rectangular and have metal walls.

[0094] A pixel with such an elongated aperture may have an adjacent pixel with an aperture of the same size but rotated by 90 degrees. In an exemplary embodiment, two adjacent pixels may be used to separately measure the amount of light above the same cutoff frequency and for each of the two polarization directions. In a further exemplary embodiment, a first pair of adjacent pixels may have an aperture of a first size but rotated by 90 degrees for a different pixel of the first pair, and a second pair of pixels may have an aperture of a second size but also rotated by 90 degrees for a different pixel of the second pair adjacent to the first pair. In such an embodiment, light of two different frequency ranges (and thus also the range between two different cutoff frequencies) and different polarizations may be measured.

[0095] 4 is a schematic top view of an embodiment of a sensor 30 in which four adjacent pixels are shown, and the apertures of two adjacent pixels have an elongated rectangular cross section perpendicular to each other. In this embodiment, the four adjacent pixels can be grouped into two pairs of two adjacent pixels, each pixel of the pair having an elongated rectangular cross section perpendicular to each other. n and 14 n+1 In the first pair, pixel 14 n Pixel 14 n The pixel 14 has a plurality of first elongated openings 13a1 overlapping with each other. n+1 Pixel 14 n+1 The pixel 14 has a plurality of second elongated openings 13b1 overlapping the first elongated openings 13a1 and having a cross-sectional shape equal to but rotated by 90° with respect to the first elongated openings 13a1. n+2 and 14 n+3 The second pair of pixels 14 n+2 and pixel 14 n+3 The elongated opening 13a2 and the elongated opening 13b2 overlap each other and have the same cross-sectional shape except that they are rotated by 90 degrees with respect to each other.

[0096] Pixel 14 n , 14 n+2Correspondingly, the filter layer has openings 13a1, 13a2 which are longer in the vertical direction of the figure, while the pixel 14 n+1 , 14 n+3 The openings 13b1, 13b2 corresponding to pixels 14 are longer in the horizontal direction. The length of the openings along the short side may be equal for all openings of the device and may be small enough to filter out all radiation polarized in the particular direction desired to be measured by the sensor 30. However, the openings 13b1, 13b2 corresponding to pixels 14 may be larger in the horizontal direction. n The opening 13a1 and the pixel 14 n+2 The openings 13a2 of the pixel groups 14 have different lengths along their long sides to provide different cutoff frequencies and to filter out radiation of different wavelength ranges. n+1 The opening 13b1 and the pixel group 14 n+3 The opening 13b2 of the pair of adjacent pixels 14 may have different lengths along the long side. n and 14 n+1 have equal lengths along their long sides and may measure, for example, one color range in two different polarizations. n and 14 n+1 Each of the pixels 13a1, 13b1 utilizes an array of 11×5 apertures 13a1, 13b1. The next two adjacent pixels 14 n+2 , 14 n+3 also have equal lengths along the long sides and measure different color ranges in two different polarizations by utilizing 11×6 apertures 13a2, 13b2, respectively.

[0097] In an embodiment, the apertures may be arranged in a particular pattern: in the following, b apertures represent apertures through which only blue light propagates, gb apertures represent apertures through which only green and blue light propagates, and rgb apertures represent apertures through which red, green, and blue light propagate.

[0098] In an embodiment, the filter array with apertures may be arranged such that each pixel is only coupled to an aperture of the same size. If the apertures are square apertures and are arranged in a square area on the pixel, the number of apertures per pixel may be 1, 4, 9, or other square numbers from 16 to 2500. Depending on the shape of the photosensitive area of ​​the pixel, for example, the number of apertures per pixel may also be a non-square number, such as 20 (4 x 5 apertures) or 130 (10 x 13 apertures).

[0099] For embodiments having openings arranged in a pattern, the following features may be provided.

[0100] It has three types of pixels: (i) rgb for pixels that measure red, green, and blue light, (ii) gb pixels that measure green and blue light, and (iii) b pixels that measure only blue light.

[0101] The red light component R may be calculated for a pixel by R=rgb-gb.

[0102] The green light component G can be calculated for a pixel by G=gb-b.

[0103] The blue light component B can be calculated for a pixel by B=b.

[0104] As with the Bayer color filter array, the light range may have to be interpolated for pixels where the light range is not measured. As can be seen above, since the gb pixel values ​​are used to calculate the other two light components, it may be advantageous to have these values ​​with the highest accuracy and with the highest frequency.

[0105] For this reason, an embodiment may have the following pixel array: TIFF2025039865000026.tif57128 points represent a repeating sequence of the indicated series of pixels for each row and column. In one example, rgb gb rgb gb... represents a repeating pattern of rgb and gb pixels.

[0106] An exemplary embodiment of an image sensor 100 having such a pixel array for use at visible wavelengths is shown in Figure 5. Figure 5 is a schematic top view of an embodiment of image sensor 100 showing a 3x3 pixel array where each pixel has an array of apertures, the cross-sectional size of the pixel apertures having three different sizes. As can be seen, the pixel array includes pixels 14 and 16. n、m (n, m=1, 2, 3, ...), with each pixel having a particular array of apertures. For example, the top left, top right, bottom left, and bottom right pixels each have 4x4 square shaped apertures 13rgb. The size of these apertures 13rgb is such that there are cutoff frequencies at frequencies where red, green, and blue light propagate through the apertures while infrared and low frequency light are attenuated. The top center, bottom center, left center, and right center pixels, in turn, each have 5x5 square apertures 13gb sized to pass green and blue light and not pass red light. Finally, the center pixel has 6x6 square apertures 13b sized to pass only blue light.

[0107] To calculate all three light components for every pixel, 5. TIFF2025039865000027.tif18128, where the index i indicates that the value is an interpolation for the pixel rather than a measurement, and P_m,n is the pixel at row m and column n of the 3×3 pattern in FIG.

[0108] The calculation of the interpolated values ​​can use a method similar to the Bayer pattern (where the gb pixels in the embodiment are twice as frequent as the rgb and b pixels and correspond to the green pixels in the Bayer pattern). One simple standard technique used for the Bayer pattern is bilinear interpolation, which interpolates all missing values ​​for each pixel separately for each channel. In an embodiment, this can mean: 1. As has been done here for P_2,2, for all rgb and b pixels from the four neighbouring gb pixels, all gb_i values ​​are calculated in the following way: TIFF2025039865000028.tif41282.All the rgb_i values ​​are calculated for the different pixels as follows: TIFF2025039865000029.tif181283.All the b_i values ​​are calculated for the different pixels as follows: TIFF2025039865000030.tif17128

[0109] Thus, in an embodiment, all relevant frequency ranges are first determined for all pixels by direct measurement or by interpolation, and then the different color components are calculated for each pixel by subtracting the values ​​of the different frequency ranges. Such procedures may also be used for more (or less) than three different frequency ranges as they are described here. In a different embodiment with four different frequency ranges, each second of the gb pixels of the array of FIG. 5 may be used to measure the fourth frequency range. The values ​​of the four different frequency ranges may then be measured or each frequency range may be measured by interpolating from its neighboring pixels. After all four frequency ranges are determined for all pixels, the four different color components may be calculated by subtracting for each pixel.

[0110] Other techniques known in the art for the interpolation of Bayer patterns can be applied in a corresponding manner, including edge-directed interpolation, constant hue-based interpolation, median-based or Freeman interpolation, gradient-based or Laroche-Prescott interpolation, weighted sum interpolation, interpolation using quadratic gradients as corrections, Bayesian interpolation, uniformity-directed interpolation, pattern matching interpolation, alias cancelling interpolation, POCS.

[0111] In different embodiments, the rgb, gb, and b openings may be arranged in a different pattern than that described above, and the rgb (or b) and gb openings may be swapped such that there are twice as many rgb (or b) openings in the pattern as there are in the other two opening types.

[0112] In a further embodiment, the rgb, gb and b openings may be arranged in the following pattern: TIFF2025039865000031.tif54128

[0113] In a further embodiment, the number of rgb openings, gb openings, and b openings may be substantially equal by using the following pattern: TIFF2025039865000032.tif57128

[0114] According to further aspects, embodiments having a filter layer with openings that function as a waveguide may also be used as a filter when provided as an optical element on a transparent optical element different from the image sensor. Such an optical element may be a lens or a glass element. The openings may all have the same size to provide one cutoff frequency for all light passing through the layer, or the openings may have different sizes to pass light with a particular frequency distribution (the distribution of openings of different sizes determines the frequency distribution). Such a layer may provide filter properties not available in other optical filter types. The filter layer may also have elongated openings that remove light with a particular polarization.

[0115] FIG. 6 is a cross-sectional view of an optical element 200 according to an embodiment. The optical element 200 has a first layer 211 that filters radiation and a transparent element 212, which may be a transparent optical element. The transparent element 212 is transparent to the radiation that is desired to be filtered and may be made of glass or transparent plastic or any other suitable material. The optical element 200 may have two air surfaces (one above the first layer 211 and the other below the optical element), or one or two surfaces to other optical elements. The surfaces may be plane-parallel, flat, or curved. The first layer 211 has at least one opening 213 that penetrates through the first layer 211 to the second layer, i.e., the transparent element 212, to propagate radiation through the at least one opening 213. In this embodiment, there are multiple openings 213 as close together as the manufacturing process allows in order to reduce light loss as much as possible. The cross-sectional size of the aperture 213 is configured to provide a cutoff frequency such that radiation having frequencies below the cutoff frequency is attenuated inside the aperture 213 and radiation having frequencies above the cutoff frequency propagates through 213. In this embodiment, radiation having frequencies below the cutoff frequency is exponentially attenuated inside the aperture 213 and radiation having frequencies above the cutoff frequency propagates through the aperture 213 by coupling to one or more propagation modes inside the aperture 213.

[0116] In FIG. 6, the opening 213 is perpendicular to the plane of the first layer 211 and has a constant cross section along the axis of the opening 213. The cross section is symmetrical and of square cross section. However, in different embodiments, the cross section may be rectangular, circular, or hexagonal. In this embodiment, the first layer 211 is metallic and the opening 213 has metal walls 220. In different embodiments, the first layer 211 may include a non-metallic layer coated with a metal layer, such that the opening also has one or more metal walls. In general, the depth of the opening 213 or the thickness of the first layer determines how much radiation or light having a frequency below the cutoff frequency provided by the opening can pass through the opening. The first layer may be made thick enough so that light below the cutoff frequency is attenuated to any desired amount.

[0117] In different embodiments, the first layer 211 may include two, three, or more openings with different cross-sectional sizes csi (i = 1..n) where cs1 > cs2 > cs3…. Each opening or each group of openings of the same size may be configured to attenuate light having a frequency below each respective cut-off frequency w1, w2, w3, … and to propagate light having a frequency above each respective cut-off frequency, where w1 < w2 < w3…. By using openings with different cross-sectional sizes, it is possible to remove light according to the distribution of the cut-off frequencies. In an embodiment, half of the openings may have a cut-off frequency that removes red and green light and passes blue light, and the remaining half of the openings may remove only red light. By using openings with different cross-sectional sizes, it may be possible to obtain filter characteristics that cannot be obtained with conventional filters. In a further embodiment, openings having cross-sectional sizes that change gradually may be used in an optical lens. In this embodiment, an infrared light may be filtered out using a filter layer, but the cross-sectional size becomes smaller as the opening is farther from the center of the lens, and at the periphery of the lens, more infrared light having a higher frequency is removed compared to the center of the lens. This may improve the optical characteristics of the lens.

[0118] In a further embodiment, one or more of the openings may be filled with a transparent material that may further provide an additional layer on the first layer. Such an embodiment is similar to that shown in FIG. 2C, but there is a transparent element 212 instead of the photosensitive layer 12.

[0119] In FIG. 6, the opening 213 has a constant square cross-section having a side length value a1 of a single size. The opening 213 has a metal wall and the cut-off frequency has a corresponding cut-off wavelength λ1,cut-off, above which the radiation exponentially decays inside the opening 213 and below which the radiation propagates through the opening 213. In this embodiment, a1 is equal to λ1,cut-off / 2.

[0120] In further embodiments, the aperture may have more than one side length value, for example, the aperture may have three values ​​a1, a2, and a3. In such an embodiment, there are three cutoff wavelengths λ1,cut-off, λ2,cut-off, and λ3,cut-off, where a1 is equal to λ1,cut-off / 2, a2 ​​is equal to λ2,cut-off / 2, and a3 is equal to λ3,cut-off / 2.

[0121] In an embodiment, the thickness l_z of the first layer 211 or the depth d of the opening 213 is TIFF2025039865000033.tif4132, where a1 is the largest side length value of the side length values ​​ai (i=1...n), λ_D,1 is greater than the cut-off wavelength λ1,cut-off=2a1, and is greater than at least e inside at least one opening when it reaches the second photosensitive layer. -2 where ε is the wavelength of the radiation whose intensity is attenuated by a factor of ε, ε is the permittivity of the material in the aperture, μ is the permeability of the material in the aperture, and ε_0 and μ_0 are the corresponding constants for vacuum. The material in the aperture may be, for example, air, a dielectric transparent material, or a vacuum (in which case the factor sqrt((μ_0ε_0) / (με)) equals 1).

[0122] In a further embodiment, the apertures may be elongated, preferably rectangular in cross section, such that the length of at least one aperture in the elongated direction defines a cutoff frequency for radiation polarized perpendicular to the elongated direction, and the length of at least one aperture perpendicular to the elongated direction defines a further cutoff frequency that is greater than the cutoff frequency for radiation polarized perpendicular to the elongated direction. In an embodiment, the further cutoff frequency may be greater than all relevant radiation, i.e. only radiation with a frequency so high that it is not actually relevant for the application in which the optical element is used can pass through the optical element, even if the radiation has a polarization that is considered to be removed. In one example, the further cutoff frequency may be in the frequency range of ultraviolet light or in a frequency range below ultraviolet light.

[0123] In further embodiments, the opening may be tapered towards or away from the transparent element, which may be done to achieve a particular filtering effect on the incident radiation.

[0124] FIG. 7A is a schematic diagram of an optical device 300 for measuring radiation within a certain range, according to an embodiment. The certain range is defined as a frequency range between a first cutoff frequency and a second cutoff frequency. The optical device 300 includes a first optical element 310, a second optical element 320, a first detector 330 for detecting radiation, a second detector 340 for detecting radiation, a beam splitter 350, and a processing unit 370. The first optical element 310 and the second optical element 320 may correspond to one of the optical elements described above, for example the optical element shown in FIG. 6. In this embodiment, the first optical element 310 has an aperture of a first cross-sectional size that provides a first cutoff frequency, and the second optical element 320 has an aperture of a second cross-sectional size that provides a second cutoff frequency that is greater than the first cutoff frequency. The detectors 330 and 340 may be any type of detector that detects or records radiation or light in the relevant frequency range. The detector may include a CCD or CMOS image sensor with pixels for spatially resolved detection of radiation, or the detector may include a photomultiplier tube that detects the total amount of radiation without spatial resolution of the radiation. The beam splitter is configured to reflect a portion of the impinging radiation and transmit a further portion of the impinging radiation. The beam splitter may preferably be a pellicle mirror, but other types of beam splitters may also be used, such as a beam splitter using two triangular prisms glued together or a beam splitter using a dichroic coating. In one example, the beam splitter may reflect approximately 50% of the incident light within a certain range and transmit approximately 50% of the incident light within a certain range, such that the amount of light between the two cutoff frequencies is equal to the difference in the measurements between the first and second detectors. In this embodiment, the first optical element 310 and the second optical element 320 and the beam splitter 350 are positioned such that an incident beam of radiation 360 first strikes the beam splitter 350 and a transmitted portion 362 and a reflected portion 364 coming from the beam splitter 350 strike the first optical element 310 and the second optical element 320.In FIG. 7A, the reflected beam from the beam splitter impinges on the first optical element 310 and the transmitted beam from the beam splitter impinges on the second optical element 320. In different embodiments, the positions of the first optical element 310 and the second optical element 320 may be swapped, i.e., the reflected beam impinges on the second optical element and the transmitted beam impinges on the first optical element. In either case, the first detector 330 is configured to detect radiation passing through the first optical element 310 and the second detector 340 is configured to detect radiation passing through the second optical element 320. The processing unit 370 is configured to calculate radiation having a frequency between the first cutoff frequency and the second cutoff frequency based on the radiation detected by the first detector 330 and the second detector 340. The processing unit 370 may be a computer communicatively coupled to the first detector and the second detector to receive data identifying the amount of radiation detected by the detectors 330 and 340. Assuming that the beam splitter reflects about 50% of the incident light and transmits about 50% of the incident light, the amount of radiation between the first and second cutoff frequencies can be calculated by subtracting the amount of radiation detected by the second detector from the amount of radiation detected by the first detector. The calculation can be done for the total amount of radiation detected by the first and second detectors (e.g., if the first and second detectors are photomultiplier tubes). However, if the first and second detectors spatially resolve the detected radiation (e.g., if the first and second detectors are CCD or CMOS image sensors having pixels), the calculation can also be done for the spatially resolved radiation, for example, by using values ​​from corresponding pixels of the image sensors used in the first and second detectors.

[0125] 7B is a schematic diagram of a further optical device 305 for measuring radiation within a certain range according to an embodiment. In this embodiment, the optical device 305 includes the same elements as the optical device 300, namely a first optical element 310, a second optical element 320, a first detector 330 for detecting radiation, a second detector 340 for detecting radiation, a beam splitter 350, and a processing unit 370. However, an incident beam of radiation 360 first impinges on the first optical element 310, and filtered radiation 366 passing through the first optical element 310 then impinges on the beam splitter 350. A transmitted portion 362 and a reflected portion 364 coming from the beam splitter 350 impinge on the first detector 330 and the second optical element 320. In FIG. 7B, the reflected beam 364 from the beam splitter 350 impinges on the first detector 330, and the transmitted beam 362 from the beam splitter 350 impinges on the second optical element 320. In a different embodiment, the positions of the first detector 330 and the second optical element 320 may be swapped, i.e. the reflected beam impinges on the second optical element and the transmitted beam impinges on the first detector. In either case, the second detector 340 is positioned to detect radiation passing through the second optical element 320. Again, the processing unit 370 is configured to calculate radiation having a frequency between the first cutoff frequency and the second cutoff frequency based on the radiation detected by the first detector 330 and the second detector 340. Again assuming that the beam splitter reflects about 50% of the incident light and transmits about 50% of the incident light, the amount of radiation between the first and second cutoff frequencies can be calculated by subtracting the amount of radiation detected by the second detector 340 from the amount of radiation detected by the first detector 330. Depending on the first and second detectors, as in the embodiment of Figure 7A, a calculation can be made for the total amount of radiation detected by the first and second detectors, or for the spatially resolved radiation detected by the first and second detectors, for example by using values ​​from corresponding pixels of the image sensors used in the first and second detectors.

[0126] Incident radiation as disclosed herein includes or is light in the range starting at, and including, ultraviolet light and ending at, and including infrared light.

[0127] In embodiments of the image sensor as disclosed herein, the first layer may be opaque to incident radiation that does not propagate through the at least one aperture and is not attenuated inside the at least one aperture, and / or the second photosensitive layer 12 may absorb only radiation that has passed through the at least one aperture 13. In other words, only radiation that propagates through the at least one aperture and is attenuated inside the at least one aperture passes through the first layer. The first layer may be considered to be opaque to or block incident radiation that does not pass through the at least one aperture, such that only radiation that passes through the at least one aperture 13 passes through the first layer 11. The first layer may also be considered to be opaque to incident radiation except for incident radiation that passes through the at least one aperture. In such embodiments, the first layer provides for the passage of incident radiation through the at least one aperture as the only mode in which incident radiation passes through the first layer, and the first layer does not provide for any additional modes in which incident radiation passes through the first layer. In particular, the first layer does not provide a plasmon mode or any kind of plasma excitation mode that allows incident radiation to pass through the first layer. For embodiments having a first layer 11 with multiple openings 13, the first layer 11 may be opaque to incident radiation that does not pass through one of the openings 13 (i.e., incident radiation that does not pass through one of the openings 13 and is not attenuated inside one of the openings 13). As a result, the second photosensitive layer 12 absorbs only radiation that has passed through the openings 13. Similarly, for the optical element 200 as disclosed herein, the first layer 211 may be opaque to incident radiation that does not propagate through at least one opening 213 and is not attenuated inside at least one opening 213, and / or only radiation that passes through at least one opening 213 passes through the first layer 211. For embodiments having a first layer 211 having multiple openings 213, the first layer 211 may be opaque to incident radiation that does not pass through the openings 213 (i.e., incident radiation that does not propagate through the openings 213 and is not attenuated inside the openings 213).

[0128] In embodiments of an image sensor having an array of pixels as disclosed herein, each pixel of the array of pixels may record an amount of intensity of radiation propagating through one or more apertures of an aperture group associated with the pixel, along with an amount of intensity of the radiation attenuated inside the one or more apertures of the aperture group. In such embodiments, each pixel of the array of pixels may not record any more radiation.

[0129] In an embodiment of an image sensor having an array of pixels as disclosed herein, the first layer may have walls between apertures associated with the same pixel that are thinner than a cross-sectional size of the apertures associated with the same pixel, which may be equal to the length of one side of a square aperture, equal to the diameter of a circular aperture, or equal to the largest diagonal of a rectangular or polygonal aperture.

[0130] 8 is a block diagram of an imaging device 400 having an image sensor 100 with an array of pixels as disclosed herein, a data storage unit 410, and a processing unit 420. The data storage unit 410 may be communicatively coupled to the image sensor 100 as represented by the line connecting the data storage unit 410 and the image sensor 100. The imaging device 400 may be configured to store radiation intensity values ​​measured by pixels of the array of pixels of the image sensor in the data storage unit 410. The processing unit 420 may be communicatively coupled to the image sensor 100 and the storage unit 410. The communicative coupling between the components of the imaging device 400 may be configured to transfer data from the image sensor 100 to the data storage unit 410 for storage therein. Data from the image sensor 100 may be transferred directly to the data storage unit 410 or may be transferred to the processing unit 420, processed therein, and then stored in the data storage unit 410. In an embodiment, the processing unit 420 may access or retrieve data from the data storage unit 410, process them, and store the converted data in the data storage unit 410. In a further embodiment, the imaging device 400 may have an image sensor according to image sensor 10 or 30 as disclosed herein.

[0131] The imaging device 400 may be a digital camera or a digital camera component of a larger device such as a smartphone, tablet, or laptop. The data storage unit 410 may include a memory device such as a random access memory (RAM) or a non-volatile memory device such as a flash memory. The data may be stored on a memory card that can be retrieved from the imaging device 400 and read by a different device that can read and process the data. The data storage unit 410 may be located in a different housing than the image sensor 100. The processing unit 420 may be an integrated circuit or computer processing unit for use in a digital camera or other device having a digital camera component.

[0132] The processing unit 420 may be configured to calculate a representation of the image captured by the image sensor 100. The representation may include one or more tuples of values ​​for each pixel of the array of pixels, with each tuple representing a color value for a pixel of the representation. Color value is used here in a general sense. For image sensors with apertures that all have the same cross-sectional size, the color value may be a value that represents the luminance value of a pixel. For image sensors with apertures that have two, three, four, or more different cross-sectional sizes, the color value may represent the light by characterizing the components of the light for a range between one or two cutoff frequencies provided by the apertures of the image sensor.

[0133] In an embodiment, the first layer of image sensor 100 may have an array of apertures having two or more different cross-sectional sizes, and processing unit 420 may be configured to calculate the tuples for pixels of the array of pixels based on radiation values ​​measured by the pixel for the wavelength range and / or based on radiation values ​​interpolated for the pixel for one or more further wavelength ranges. Processing unit 420 may also be configured to calculate the tuples using a calibration matrix that associates one or more radiation values ​​(which may be measured or interpolated) for the pixel to each value of the tuple for the pixel.

[0134] In an embodiment, the aperture of image sensor 100 may have three or more different cross-sectional sizes, and the tuple for each pixel includes a value R representing a red component, a further value G representing a green component, and a still further value B representing a blue component. In such an embodiment, the representation may be or include a color space according to which the RGB values ​​are interpreted. Possible color spaces include sRGB, Adobe RGB, and ProPhoto RGB.

[0135] As disclosed in FIG. 5 , the three light components R, G, and B (rgb for red light, green light, and blue light, gb for green light and blue light, and b for blue light) for a pixel measuring light within three different wavelength ranges can be written as follows: TIFF2025039865000034.tif18128

[0136] By removing the index i, which indicates that the range values ​​are interpolated and not measured, the equation can be expressed as a matrix. TIFF2025039865000035.tif5128, where c i represents the three color components R, G, and B for i=1, 2, and 3, respectively, and v j represents the three measured or interpolated radiance values ​​for the light ranges rgb, gb, and b for j = 1, 2, and 3, and a ij is an element of a calibration matrix that relates a measured or interpolated radiation intensity value of a pixel to one or more values ​​of a tuple for that pixel. A tuple may be a representation of a color or light value at the pixel's location.

[0137] The above matrix equation can be generalized to any number of different wavelength ranges, and even different polarizations, measured by the pixels of an image sensor as disclosed herein, where v is the measured or interpolated value for one or more wavelength ranges associated with the cutoff frequency provided by the embodiment. j can have p different values ​​(i.e., j=1,...,p) in the above matrix equation. The above matrix equation can be further generalized to any number of values ​​of the tuple representing the generalized color value at a pixel location. In this case, the value of the tuple of pixels, c ican have q different values ​​(i.e., i=1, ..., q) in the matrix equation above. In one example, an image sensor may have apertures with four different cross sections arranged in a repeating pattern, providing four cutoff frequencies ranging above ultraviolet light and below infrared light. Such an image sensor may record colors more accurately and may record more colors (i.e., provide a larger color gamut) than an image sensor that measures light in only three different wavelength ranges (either by conventional design or by embodiments as disclosed herein). The index j is then summed for values ​​from 1 to 4. Further increasing the number of cutoff frequencies in the visible light range may further increase the precision with which colors in the visible light range are measured, and may further increase the number of colors that can be measured and distinguished. In one embodiment, even if the number of different cutoff frequencies, and therefore wavelength ranges, is greater than three, the tuple may still have only three components: red, green, and blue. However, in another embodiment, the tuple may have four components corresponding to the four ranges limited by the different cutoff frequencies. In a further embodiment, the image sensor may have six different cross-sectional size apertures, four of which provide cutoff frequencies in the visible light range, one of which provides cutoff frequencies above the infrared wavelength range, and one of which provides cutoff frequencies below the ultraviolet wavelength range. In this case, the tuple may have six values, one value identifying the light component in the ultraviolet range, four for the visible light range, and one for the infrared range. In a different embodiment, the tuple may have five values, one for the ultraviolet range, three for the visible light range, and one for the infrared range. In a further embodiment, two or more cutoff frequencies may be placed in the infrared light range to distinguish different spectral distributions within the infrared light range. In one example, the image sensor may have apertures providing three different cutoff frequencies in the infrared range, and one for distinguishing infrared light from visible light. In this case, it is possible to characterize infrared light having three components for three ranges limited by the four cutoff frequencies.It may further be possible to use a fourth component to characterize the infrared content in the wavelength range above the maximum cutoff wavelength corresponding to the lowest cutoff frequency.Similarly, two or more cutoff wavelengths in the ultraviolet range can be used to distinguish light having different content within subranges of the ultraviolet range.

[0138] Element a of the calibration matrix ij may be determined by comparing a representation of a captured image of an object of known color as obtained by the embodiment with known values ​​of the object. The known values ​​of the object may be obtained from representations of these objects obtained through other calibrated cameras. The elements of the calibration matrix may then be optimized to match the known colors as closely as possible. The result of the optimization may depend on the value of the cutoff frequency provided by the aperture of the image sensor and also on the thickness of the first layer, which affects how much light below the cutoff frequency is attenuated inside the aperture. In an embodiment, the object may be a sheet of paper with many different color patches, the representations of which are known or have been measured. The elements a of the calibration matrix ij Once determined, the elements are stored and retrieved to determine the color component c according to the matrix equation above. i , i.e., the value of the tuple for each pixel.

[0139] FIG. 9 is a flow diagram of a computer-implemented method 500 for calculating a representation of an image captured by an image sensor 10, 30, 100. The image sensor 10, 30, 100 may comprise a first layer 11 that filters incident radiation by attenuating incident radiation having a frequency below a cutoff frequency, and a second photosensitive layer 12 that absorbs radiation passing through the first layer 11. The first layer 11 may be in front of the second photosensitive layer 12 in the propagation direction of the incident radiation. The photosensitive layer 12 may comprise an array of pixels, and the image sensor 10, 30, 100 may have pixels of the same type or two or more different types of pixels. A type of pixel may be defined as follows: a first pixel is of a different type from a second pixel if and only if the first layer 11 provides a cutoff frequency for the first pixel that is different from the cutoff frequency for the second pixel provided by the first layer 11. Thus, if the cutoff frequencies provided by the first layer for the two different pixels are the same, the two pixels are of the same type.

[0140] In the case of an image sensor having pixels that are all of the same type, the method 500 may include a simple version of the calculation operation that relies only on directly measured radiation values ​​and not on interpolated radiation values. In such a case, the method 500 may include identifying 560 a pixel of the array of pixels for the tuple calculation operation, accessing 570 a radiation value measured by the pixel, and calculating 590 one or more values ​​of a tuple of one or more values ​​for the pixel using the radiation value measured by the pixel. In an embodiment, the tuple may have a single value that represents an amount of radiation that has passed through an aperture in the first layer 11 that all has the same cross-sectional size. The tuple may represent a generalized color value with respect to a particular representation, such as a luminance value for a black and white representation. The method 500 may further include repeating 600 the tuple calculation operation for all pixels of the array of pixels. This may mean repeating 600 accessing 570 a radiation value measured by the pixel and calculating 590 one or more values ​​of a tuple for the pixel for all pixels of the array of pixels. The result of the method 500 is a representation of the image captured by the image sensor 10 , 30 , 100 .

[0141] The computer-implemented method 500 may be performed by any kind of computing device. Such a computing device may be provided by or housed within the imaging device 400 as disclosed herein, or by a general computing device configured to load instructions and execute those instructions. The instructions may control the computing device such that the operations of the method 500 are performed by the computing device. The general computing device may be a laptop, a personal computer, a server, or a larger computing device, which may also be a tablet or a smartphone that includes the imaging device 400 as a component.

[0142] If the image sensor has two or more different types of pixels, the method 500 may further include an interpolation operation. The interpolation operation may include, prior to identifying 560 the pixel, identifying 510 a further pixel of the array of pixels, accessing 520 one or more radiation values ​​measured by the further pixel for a further cutoff frequency different from the cutoff frequency at which the radiation value was measured by the further pixel, and interpolating 530 the radiation value for the further pixel for the further cutoff frequency. If the image sensor has three or more different types of pixels, the method 500 may further include repeating 540 the accessing 520 one or more radiation values ​​and interpolating 530 the radiation values ​​for all types of pixels of the array different from the type of the further pixel for the further pixel. In other words, the method includes interpolating radiation values ​​for each different cutoff frequency provided by the image sensor for the pixel being processed. The method 500 may further include repeating 550 the previous interpolation operation for all pixels of the array. The interpolation operation may further include accessing 520 one or more radiation values ​​measured by further pixels, interpolating 530 radiation values ​​for further pixels, and repeating 540 the accessing 520 one or more radiation values ​​and the interpolating 530 radiation values ​​for further pixels. This may complete the interpolation operation, but the interpolated values ​​may be used to extend the tuple calculation operation to take the interpolated values ​​into account. Thus, after identifying 560 a pixel of the array for the tuple calculation operation, the method 500 may further include accessing 580 one or more interpolated radiation values ​​for the pixel, calculating 590 a tuple for the pixel using the radiation values ​​measured by the pixel and the one or more interpolated radiation values ​​for the pixel, and repeating 600 accessing 580 one or more interpolated radiation values ​​for the pixel and calculating 590 a tuple for the pixel of all pixels of the array of pixels.

[0143] In an embodiment of the method 500, computing 590 the tuple for the pixel may include using a calibration matrix that associates each measured or interpolated radiation value for the pixel with each value of the tuple for the pixel. In further embodiments, the image sensor 10, 30, 100 may have more than two types of pixels, and the tuple for the pixel may have a value representing a red component, a further value representing a green component, and yet a further value representing a blue component.

[0144] In an embodiment of method 500, the first layer 11 of the image sensor 10, 30, 100 may be opaque to incident radiation that does not propagate through the at least one opening 13 and is not attenuated inside the at least one opening 13, and / or the second photosensitive layer 12 absorbs only radiation that has passed through the at least one opening 13.

Claims

1. An image sensor (10, 30, 100) for recording incident radiation, a first layer (11) for filtering incident radiation by attenuating said incident radiation having a frequency below a cutoff frequency; a second photosensitive layer (12) for absorbing radiation passing through said first layer (11); Equipped with the first layer (11) is in front of the second photosensitive layer (12) in a propagation direction of the incident radiation, and the first layer (11) comprises at least one opening (13) penetrating through the first layer (11) to the second photosensitive layer (12) for allowing radiation to propagate through the at least one opening (13); a cross-sectional size of the at least one opening (13) configured to provide a cutoff frequency such that incident radiation having a frequency below the cutoff frequency is attenuated inside the at least one opening (13) and incident radiation having a frequency above the cutoff frequency propagates through the at least one opening (13). Image sensor (10, 30, 100).

2. 2. The image sensor (10, 30, 100) of claim 1, wherein incident radiation having a frequency below said cutoff frequency is exponentially attenuated inside said at least one opening (13).

3. An image sensor (10, 30, 100) as described in any one of claims 1 to 2, wherein incident radiation having a frequency above the cutoff frequency propagates through the at least one opening (13) by coupling into one or more propagation modes inside the at least one opening (13).

4. The image sensor (10, 30, 100) of any one of claims 1 to 3, wherein the at least one opening (13) is perpendicular to the plane of the first layer (11), and the at least one opening (13) has a constant cross-section along an axis of the at least one opening (13), the cross-section being symmetrical and optionally a square, rectangular, circular or hexagonal cross-section.

5. An image sensor (10, 30, 100) according to any one of claims 1 to 4, wherein the first layer (11) is metallic or the first layer (11) comprises a non-metallic layer coated with a metallic layer, whereby in both cases the at least one opening has one or more metallic walls (20).

6. said first layer (11) comprises two, three or more openings (13) of different cross-sectional sizes csi (i=1..n), cs1>cs2>cs3...>csn; Each cross-sectional size is configured to attenuate light having a frequency below a respective cutoff frequency w1, w2, w3, ..., wn, and to propagate light having a frequency above a respective cutoff frequency, where w1 < w2 < w3 ... < wn. An image sensor (100) according to any one of the preceding claims.

7. The image sensor (10, 30, 100) according to any one of the preceding claims, wherein the at least one opening (13) is filled with a transparent material (18).

8. The image sensor (10, 30, 100) of claim 7, wherein the transparent material (18) provides an additional layer (19) on the first layer (11).

9. The at least one opening (13) has n different sizes of side length values ​​ai (i=1 . . . n), said at least one opening (13) has a metal wall (20), and said cutoff frequency corresponds to a cutoff wavelength λ i,cut-off (i=1 . . . n), above which the incident radiation exponentially decays inside the at least one aperture (13) and below which the incident radiation propagates through the at least one aperture (13), where ai is λ i,cut-off / 2(i=1 9. The image sensor (10, 30, 100) according to any one of claims 1 to 8, wherein the number of pixels is equal to n.

10. The thickness of the first layer (11) is l z Or the depth d of the at least one opening is: and Here, a 1 is the length value ai (i=1 . . . n) is the maximum side length value, λ D,1 is the cutoff wavelength λ 1,cut-off = 2a 1 and at least e inside said at least one opening (13). -2 is the wavelength of the radiation whose intensity is attenuated by a factor of 1, ε is the permittivity of the material in the at least one opening, μ is the permeability of the material in the at least one opening, ε_0 is the permittivity of a vacuum, and μ_0 is the permeability of a vacuum; An image sensor (10, 30, 100) according to claim 9.

11. An image sensor (10, 30, 100) according to any one of claims 1 to 10, wherein the photosensitive layer (12) comprises an array of pixels, and each opening (13) or each of a number of adjacent openings (13) having the same cross-sectional size overlaps and is coupled to an individual pixel of the array.

12. 12. The image sensor (10, 30, 100) of claim 11, wherein each pixel comprises a photosensitive area, and the opening or the plurality of adjacent openings associated with the pixel covers an area that is equal to or larger than the photosensitive area of ​​the individual pixel and smaller than the area of ​​the pixel.

13. The image sensor (10, 30, 100) according to any one of claims 11 to 12, wherein a transparent layer (16) is located between the first layer (11) and the second photosensitive layer (12).

14. 14. The image sensor (10, 30, 100) of claim 13, wherein the transparent layer (16) comprises walls (17) located above the boundaries of individual pixels, the walls (17) being opaque to the incident radiation and optionally metallic.

15. said first layer (11) comprising an array of openings (13rgb, 13gb, 13b) having one or more different cross-sectional sizes; the openings are grouped such that each opening group has one opening or multiple openings that are adjacent and have the same cross-sectional size, each opening group overlaps and is coupled to one and only one pixel of the photosensitive layer, and the opening groups are arranged in a repeating pattern in the first layer; An image sensor (100) according to any one of claims 11 to 14.

16. 16. The image sensor (100) of claim 15, wherein the openings (13rgb, 13gb, 13b) have three or more different cross-sectional sizes, a first size filtering out infrared light and allowing the propagation of red, green, and blue light, a second size filtering out red light and allowing the propagation of green and blue light, and a third size filtering out red and green light and allowing the propagation of blue light, and wherein the repeating pattern in the first layer is configured to provide measurements or interpolations of red, green, and blue light for each pixel of the array of pixels.

17. Each aperture of the array has a constant square cross-section with n different sizes of side length values ​​ai (i=1 ..n), each aperture has a metal wall (20), and the cutoff wavelength λ corresponding to the cutoff frequency of the first cross-sectional size 1 is in the range of 600 nm to 750 nm, optionally in the range of 650 nm to 730 nm, the length value a1 of one side of the first cross-sectional size is in the range of 300 nm to 375 nm, optionally in the range of 325 nm to 365 nm, and the cutoff wavelength λ corresponding to the cutoff frequency of the second cross-sectional size is 2 is in the range of 495 nm to 590 nm, optionally in the range of 515 nm to 570 nm, the length value a2 of one side of the second cross-sectional size is in the range of 247 nm to 295 nm, optionally in the range of 257 nm to 285 nm, and the cutoff wavelength λ corresponding to the cutoff frequency of the third cross-sectional size is 3 17. The image sensor (100) of claim 16, wherein a side length value a3 of the third cross-sectional size is in the range of 450 nm to 510 nm, optionally in the range of 465 nm to 495 nm, and a side length value a4 of the third cross-sectional size is in the range of 225 nm to 255 nm, optionally in the range of 232 nm to 247 nm.

18. the first layer (11) comprises an array of openings having three or more different cross-sectional sizes, a first size allowing the propagation of infrared light, red light, green light, and blue light, a second size filtering out red light and allowing the propagation of green and blue light, and a third size filtering out red and green light and allowing the propagation of blue light; each opening of the first size overlaps a pixel area that is equal to or smaller than the pixel; each opening of said first size is filled with a filter material that filters out infrared light and transmits light of shorter wavelengths than infrared light, optionally with all openings filled with said filter material and / or with said filter material providing an additional layer on said first layer; The image sensor (100) of claim 15.

19. 19. The image sensor (100) of any one of claims 16 to 18, wherein the array of apertures (13rgb, 13gb, 13b) has three different cross-sectional sizes, and the groups of apertures are arranged in two alternating rows of pixels, a first row of pixels measuring red, green, and blue light alternately in one pixel and green and blue light in the next pixel, and a second row of pixels measuring green and blue light alternately in one pixel and blue light in the next pixel, and a pixel in the first row measuring red, green, and blue light has an adjacent pixel in the second row measuring green and blue light.

20. The openings have the following additional sizes: a fourth size that removes a portion of infrared light below the infrared light cutoff frequency and allows the propagation of a further portion of infrared light above the infrared light cutoff frequency as well as the propagation of red, green, and blue light; a fifth size that removes red and yellow light and allows the propagation of green and blue light; a sixth size that removes a portion of green light below the red and green light cutoff frequency and allows the propagation of a further portion of green light above the green light cutoff frequency and the propagation of blue light; a seventh size that removes red, green, and blue light and allows the propagation of ultraviolet light. The image sensor (100) of any one of claims 15 to 18, having any one of the sizes or any combination of said sizes.

21. 16. The image sensor (100) of claim 15, wherein the opening has two or more different cross-sectional sizes, one size that removes light below a low cutoff frequency and allows light above the low cutoff frequency to propagate, and another size that removes light below a high cutoff frequency and allows light above the high cutoff frequency to propagate, such that an amount of light within a range between the low cutoff frequency and the high cutoff frequency can be measured.

22. The image sensor (100) of claim 19, wherein the range between the lower cutoff frequency and the upper cutoff frequency includes one or more spectral lines.

23. 9. The image sensor (30) according to any one of claims 1 to 8, wherein the at least one aperture (13a1, 13a2; 13b1, 13b2) is elongated and optionally rectangular in cross section, whereby a length of the at least one aperture in the direction of the elongation defines a cutoff frequency for incident radiation polarized perpendicular to the direction of the elongation, and a length of the at least one aperture in a direction perpendicular to the direction of the elongation defines a further cutoff frequency that is greater than an upper limit of a range of incident radiation analyzed by the image sensor.

24. The first layer comprises: a plurality of first elongated openings (13a1) overlapping first pixels of the array of pixels of the second photosensitive layer; and a plurality of second elongated openings (13b1) overlapping adjacent second pixels of the array and having a cross-sectional shape rotated 90° relative to the first elongated openings (13a1).

25. An image sensor (10, 100) according to any one of claims 1 to 3, 5 to 8 or 11 to 14, without reference to claims 4, 9 or 10, wherein the at least one opening tapers towards or away from the second layer.

26. the first layer (11) is opaque to incident radiation that does not propagate through the at least one opening (13) and is not attenuated inside the at least one opening (13); and / or the second photosensitive layer (12) absorbs only the radiation that has passed through the at least one opening (13); An image sensor (10, 30, 100) according to any one of the preceding claims.

27. An image sensor (10, 30, 100) according to any one of the preceding claims, wherein the incident radiation comprises light in a range starting from and including ultraviolet light and ending with and including infrared light.

28. An image sensor (10, 30, 100) according to any one of claims 15 to 27, wherein each pixel of the array of pixels records an amount of intensity of radiation propagating through the one or more apertures of the aperture group associated with the pixel together with an amount of intensity of radiation attenuated inside the one or more apertures of the aperture group associated with the pixel.

29. An image sensor (100) according to any one of claims 15 to 28, a data storage unit (410) communicatively coupled to the image sensor (100); Equipped with configured to store in said data storage unit (410) radiation values ​​measured by pixels of said array of pixels; Imaging device (400).

30. a processing unit (420) communicatively coupled to the image sensor (100) and / or the storage unit (410) and configured to compute a representation of an image captured by the image sensor (100); Further equipped with the representation comprises one or more value tuples for each pixel of the array of pixels, the tuples representing color values ​​of the representation; 30. The imaging device (400) of claim 29.

31. 31. The imaging device (400) of claim 30, wherein the first layer (11) of the image sensor (100) comprises an array of apertures (13rgb, 13gb, 13b) having two or more different cross-sectional sizes, and the processing unit (420) is configured to calculate a tuple for a pixel of the array of pixels based on radiation values ​​measured by the pixel for a wavelength range and based on one or more radiation values ​​interpolated for the pixel for one or more further wavelength ranges.

32. 32. The imaging device (400) of claim 31, wherein the processing unit (420) is configured to calculate the tuple for the pixel using a calibration matrix that associates each measured or interpolated radiation value for the pixel with each value of the tuple for the pixel.

33. An imaging device (400) according to any one of claims 29 to 32, wherein the apertures (13rgb, 13gb, 13b) have three or more different cross-sectional sizes and the tuple for the pixel includes a value representing a red component, a further value representing a green component, and a yet further value representing a blue component.

34. A computer-implemented method (500) for calculating a representation of an image captured by an image sensor (10, 30, 100) comprising a first layer (11) for filtering incident radiation by attenuating incident radiation having a frequency below a cutoff frequency, and a second photosensitive layer (12) for absorbing radiation passing through said first layer (11), comprising: said first layer (11) being in front of said second photosensitive layer (12) in the propagation direction of said incident radiation, said second photosensitive layer (12) comprising an array of pixels, The method (500) comprises: Identifying a pixel of the array of pixels (560); accessing (570) a radiation value measured by the pixel; calculating (590) one or more values ​​of one or more value tuples for the pixel using the radiation values ​​measured by the pixel, the tuples representing color values; repeating (600) accessing (570) radiation values ​​measured by said pixels and calculating (590) one or more values ​​of a tuple for all pixels of said array of pixels to obtain a representation of an image captured by said image sensor (10, 30, 100); A computer-implemented method (500) comprising:

35. the image sensor (10, 30, 100) has two or more different types of pixels, said first pixel being of a different type from said second pixel if and only if said first layer (11) provides a cutoff frequency for the first pixel which is different from the cutoff frequency for the second pixel provided by said first layer (11); The method (500) comprises: identifying (510) a further pixel of the array of pixels prior to identifying (560) the pixel; accessing (520) one or more radiation values ​​measured by further pixels for further cutoff frequencies different from the cutoff frequencies at which radiation values ​​were measured by the further pixels; Interpolating (530) radiation values ​​for the further pixels for the further cutoff frequencies; repeating (540) accessing one or more radiation values ​​(520) and interpolating (530), for the further pixel, radiation values ​​for all types of pixels in the array different from the type of the further pixel; Repeating (550) the previous interpolation operation for all pixels of the array; Following identifying the pixel of the array of pixels (560), accessing one or more interpolated radiation values ​​for the pixel (580); calculating (590) the tuple for the pixel using the radiation value measured by the pixel and one or more interpolated radiation values ​​for the pixel; repeating (600) accessing (580) one or more interpolated radiation values ​​for the pixel and calculating (590) the tuples for all pixels of the array of pixels; 35. The computer-implemented method (500) of claim 34, further comprising:

36. 36. The computer-implemented method (500) of claim 35, wherein calculating (590) the tuple for the pixel comprises using a calibration matrix that associates each measured or interpolated radiation value for the pixel with each value of the tuple for the pixel.

37. 37. The computer-implemented method (500) of any one of claims 35 to 36, wherein the image sensor (10, 30, 100) has three or more types of pixels, and the tuples for the pixels include a value representing a red component, a further value representing a green component, and a yet further value representing a blue component.

38. the first layer (11) comprises at least one opening (13) penetrating through the first layer (11) to the second photosensitive layer (12) for transmitting radiation through the at least one opening (13); the cross-sectional size of the at least one opening (13) is configured to provide a cutoff frequency such that incident radiation having a frequency below the cutoff frequency is attenuated inside the at least one opening (13) and incident radiation having a frequency above the cutoff frequency propagates through the at least one opening (13); the first layer (11) is opaque to incident radiation that does not propagate through and is not attenuated inside the at least one opening (13); and / or the second photosensitive layer (12) absorbs only the radiation that has passed through the at least one opening (13); A computer implemented method (500) according to any one of claims 34 to 37.

39. a first layer (211) for filtering radiation by attenuating incident radiation having a frequency below a cutoff frequency; A transparent element (212) Equipped with the first layer (211) comprises at least one opening (213) penetrating through the first layer (211) to the transparent element (212) for transmitting radiation through the at least one opening (213); the cross-sectional size of the at least one opening (213) is configured to provide a cutoff frequency such that radiation having a frequency below the cutoff frequency is attenuated inside the at least one opening (231) and radiation having a frequency above the cutoff frequency propagates through the at least one opening (213). An optical element (200).

40. 40. The optical element (200) of claim 39, wherein the radiation having a frequency below the cutoff frequency is exponentially attenuated inside the at least one opening (213).

41. An optical element (200) according to any one of claims 39 to 40, wherein radiation having a frequency above the cut-off frequency propagates through the at least one opening (213) by coupling into one or more propagation modes inside the at least one opening (213).

42. 42. The optical element (200) of any one of claims 39 to 41, wherein the at least one opening (213) is perpendicular to the plane of the first layer (211), and the at least one opening (213) has a constant cross-section along an axis of the at least one opening (213), the cross-section being symmetrical and optionally a square, rectangular, circular or hexagonal cross-section.

43. An optical element (200) according to any one of claims 39 to 42, wherein the first layer (211) is metallic or the first layer (211) comprises a non-metallic layer coated with a metallic layer, whereby in both cases the at least one opening has one or more metallic walls (220).

44. the first layer (211) comprises two, three or more openings of different cross-sectional sizes csi (i=1..n), cs1>cs2>cs3...; Each aperture is configured to attenuate light having a frequency below a respective cutoff frequency w1, w2, w3, ... and to propagate light having a frequency above a respective cutoff frequency, where w1 < w2 < w3 ... An optical element (200) according to any one of claims 39 to 43.

45. The optical element (200) according to any one of claims 39 to 44, wherein the at least one opening is filled with a transparent material.

46. 46. ​​The optical element (200) of claim 45, wherein the transparent material provides an additional layer on the first layer.

47. The at least one opening (213) has a side length value ai (i=1 . . . n), said at least one opening has a metal wall (220), and said cutoff frequency has a corresponding cutoff wavelength λ i,cut-off (i=1 . . . n), above which the radiation attenuates exponentially inside the at least one aperture (213) and below which the radiation propagates through the at least one aperture, where ai is λ i,cut-off / 2(i=1 . . . n)。 The optical element (200) according to any one of claims 39 to 46, wherein the optical element (200) is equal to (n).

48. The thickness of the first layer (211) is l z Or the depth d of the at least one opening (213) is and Here, a 1 is the length value a of one side i (i=1 . . . n) and λ D,1 is the cutoff wavelength λ 1,cut-off = 2a 1 and at least e inside at least one opening (213). -2 is the wavelength of the radiation whose intensity is attenuated by a factor of 1, ε is the permittivity of the material in the at least one opening (213), μ is the magnetic permeability of the material in the at least one opening (213), ε_0 is the permittivity of a vacuum, and μ_0 is the magnetic permeability of a vacuum; 48. The optical element (200) of claim 47.

49. 47. An optical element (200) according to any one of claims 39 to 46, wherein the at least one aperture is elongated, preferably rectangular in cross section, whereby a length of the at least one aperture in the direction of the elongation defines a cutoff frequency for radiation polarised perpendicular to the direction of the elongation, and a length of the at least one aperture in a direction perpendicular to the direction of the elongation defines a further cutoff frequency that is greater than the cutoff frequency for radiation polarised perpendicular to the direction of the elongation.

50. The optical element (200) of any one of claims 39-41, 43-46 or 49, when not referring to claims 42 or 47-48, wherein the at least one opening tapers towards or away from the transparent element.

51. the first layer (211) is opaque to incident radiation that does not propagate through the at least one opening (213) and is not attenuated inside the at least one opening (213); and / or Only radiation passing through said at least one opening (213) passes through said first layer (211); An optical element (200) according to any one of claims 39 to 50.

52. An optical device (300, 305) for measuring radiation having a frequency between a first cutoff frequency and a second cutoff frequency, comprising: A first optical element (310) according to any one of claims 39 to 43 and 45 to 51, A second optical element (320) according to any one of claims 39 to 43 and 45 to 51; a first detector (330) for detecting radiation; a second detector (340) for detecting radiation; a beam splitter (350) that reflects a portion of the radiation and transmits a further portion of the radiation; a processing unit (370) configured to calculate the radiation having a frequency between a first cutoff frequency and a second cutoff frequency based on the radiation detected by the first detector (330) and the second detector (340); Equipped with the first optical element has an aperture of a first cross-sectional size providing the first cutoff frequency, the second optical element has an aperture of a second cross-sectional size providing the second cutoff frequency, the second cutoff frequency being greater than the first cutoff frequency; the first optical element, the second optical element and the beam splitter are (i) an incident beam of radiation (360) is positioned such that it first impinges on the beam splitter (350) and a transmitted portion (362) and a reflected portion (364) coming from the beam splitter (350) impinge on the first optical element (310) and the second optical element (320); the first detector (330) is positioned to detect radiation passing through the first optical element (310); and the second detector (340) is positioned to detect radiation passing through the second optical element (320); or (ii) an incident beam of radiation (360) is positioned such that it first impinges on the first optical element (310), filtered radiation (366) passing through the first optical element (310) impinges on the beam splitter (350), and transmitted and reflected portions (362 and 364) coming from the beam splitter (350) impinge on the first detector (330) and the second optical element (320); and the second detector (340) is positioned to detect radiation passing through the second optical element (320); Optical device (300, 305).

Images