Filter array and light detection system

Abstract

A filter array has a plurality of filters arranged two-dimensionally and each having a different transmission spectrum. The plurality of filters includes a first filter having a transmission spectrum with a first peak group including a first peak and a second peak adjacent to each other, and at least one second filter having a transmission spectrum with a second peak group including a third peak and a fourth peak adjacent to each other, among the plurality of peaks included in the second peak group, the wavelength of the third peak is closest to the wavelength of the first peak. A first interval between the wavelength of the first peak and the wavelength of the second peak is different from a second interval between the wavelength of the third peak and the wavelength of the fourth peak. When an absolute value of a difference between the first interval and the second interval is ΔFSR, and a half-value width of the first peak is σ, ΔFSR / σ ≥ 0.25.

Description

Technical Field

[0001] This disclosure relates to filter arrays and optical detection systems. Background Technology

[0002] By utilizing spectral information from many narrow bands, such as dozens of bands, it is possible to obtain detailed physical properties of objects that cannot be obtained through traditional RGB images. Cameras that acquire such multi-wavelength information are called "hyperspectral cameras." Hyperspectral cameras are used in various fields such as food inspection, biological examination, pharmaceutical development, and mineral composition analysis.

[0003] Patent Document 1 discloses an example of a hyperspectral imaging device utilizing compressed sensing. This imaging device includes: an encoding element, which is an array of multiple optical filters with different wavelength-dependent transmittance; an image sensor that detects light transmitted through the encoding element; and processing circuitry. The encoding element is arranged in the optical path connecting the subject and the image sensor. The image sensor acquires a wavelength-multiplexed image by simultaneously detecting light components superimposed on multiple wavelength bands at each pixel. The processing circuitry applies compressed sensing to the acquired wavelength-multiplexed image using information about the spatial distribution of the spectral transmittance of the encoding element, thereby generating image data for each of the multiple wavelength bands. In the imaging device disclosed in Patent Document 1, an array of optical filters having two or more transmittance peaks (i.e., maxima) in the target wavelength domain is used as the encoding element.

[0004] Patent document 2 discloses an example of a filter array of a Fabry-Perot resonator using a dielectric multilayer film in the reflective layer.

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: US Patent No. 9599511

[0008] Patent Document 2: US Patent No. 9,466,628 Summary of the Invention

[0009] The problem that the invention aims to solve

[0010] The purpose of this disclosure is to provide a light detection system that can improve the wavelength resolution of a hyperspectral camera, and a filter array used in the light detection system.

[0011] Methods used to solve problems

[0012] A filter array according to the present disclosure comprises multiple filters arranged in a two-dimensional manner, each having a distinct transmission spectrum. The multiple filters include: a first filter whose transmission spectrum has a first peak group comprising adjacent first and second peaks; and at least one second filter whose transmission spectrum has a second peak group comprising adjacent third and fourth peaks, wherein, among the peaks in the second peak group, the wavelength of the third peak is closest to the wavelength of the first peak. A first interval between the wavelengths of the first and second peaks is different from a second interval between the wavelengths of the third and fourth peaks. When the absolute value of the difference between the first and second intervals is ΔFSR, and the half-width of the first peak is σ, ΔFSR / σ ≥ 0.25.

[0013] The inclusive or specific form of this disclosure can also be implemented by a system, apparatus, method, integrated circuit, computer program, or recording medium such as a computer-readable recording disk, or by any combination of a system, apparatus, method, integrated circuit, computer program, and recording medium. Computer-readable recording media may include, for example, non-volatile recording media such as CD-ROM (Compact Disc-Read Only Memory). An apparatus may also consist of one or more devices. When an apparatus consists of two or more devices, these two or more devices may be configured in one device or separately in two or more separate devices. In this specification and claims, "apparatus" can refer not only to a single device but also to a system composed of multiple devices.

[0014] Invention Effects

[0015] According to the technology disclosed herein, the wavelength resolution of a hyperspectral camera can be improved. Attached Figure Description

[0016] Figure 1A This is a diagram schematically illustrating an exemplary implementation of a light detection system.

[0017] Figure 1B This is a diagram illustrating a structural example of a light detection system configured with a filter array separated from the image sensor.

[0018] Figure 1C This is a diagram illustrating a structural example of a light detection system configured with a filter array separated from the image sensor.

[0019] Figure 1D This is a diagram illustrating a structural example of a light detection system configured with a filter array separated from the image sensor.

[0020] Figure 2A This is a diagram schematically illustrating an example of a filter array in an exemplary implementation.

[0021] Figure 2B This is a diagram illustrating an example of the spatial distribution of the transmittance of light in multiple wavelength domains contained within the wavelength domain of an object.

[0022] Figure 2C It means Figure 2A A diagram showing an example of the transmission spectrum of a filter included in the filter array.

[0023] Figure 2D It means Figure 2A A diagram showing an example of the transmission spectrum of another filter included in the filter array.

[0024] Figure 3A This is a diagram used to illustrate an example of the relationship between the wavelength domain of an object and the multiple wavelength domains contained within it.

[0025] Figure 3B This is another example of a diagram used to illustrate the relationship between the wavelength domain of an object and the multiple wavelength domains it contains.

[0026] Figure 4A It is a diagram used to illustrate the characteristics of the transmission spectrum of a filter in a filter array.

[0027] Figure 4B It means to Figure 4A The graph shown is a result of averaging the transmission spectrum by each wavelength domain.

[0028] Figure 5 This is a schematic cross-sectional view of a light detection device according to an exemplary embodiment.

[0029] Figure 6 This is a diagram illustrating an example of the transmission spectrum of a Fabry-Perot filter.

[0030] Figure 7 This is a graph showing the transmission spectra of 16 multimode filters used in comparative examples.

[0031] Figure 8 The figure shows an example of a positive resolution image and an example of a separated image reconstructed from a filter array of a comparison example with a single FSR.

[0032] Figure 9 This is a diagram schematically illustrating an example of the transmission spectrum of a Fabry-Perot filter.

[0033] Figure 10 This is a diagram showing an example of a forward-resolution image and an example of a separated image reconstructed from a filter array with two FSRs.

[0034] Figure 11 This is a graph showing the relationship between the mean square error and ΔFSR / σ between the positive solution image and the separated image of band 6, which presents a local solution, in the comparative example.

[0035] Figure 12 This is a graph showing the relationship between the mean square error between the forward and split images for band 6 and the proportion of the second filter included in the filter array.

[0036] Figure 13 This is a graph showing the calculated transmission spectrum of the Fabry-Perot filter according to this embodiment.

[0037] Figure 14A This is a diagram schematically illustrating a first modified example of a light detection device.

[0038] Figure 14B This is a diagram schematically illustrating a second modified example of the optical detection device.

[0039] Figure 14C This is a schematic diagram illustrating a third modified example of the optical detection device.

[0040] Figure 14D This is a schematic diagram illustrating the fourth variation of the light detection device.

[0041] Figure 14E This is a schematic diagram illustrating the fifth modification of the light detection device.

[0042] Figure 14F This is a diagram schematically illustrating the sixth variation of the light detection device. Detailed Implementation

[0043] The following describes exemplary embodiments of this disclosure. Furthermore, the embodiments described below are inclusive or specific examples. Therefore, the numerical values, shapes, materials, constituent elements, arrangement and connection methods of constituent elements, steps, and order of steps shown in the following embodiments are examples and are not intended to limit the scope of this disclosure. Moreover, any constituent elements in the following embodiments that are not described in the independent claims representing the highest-level concept are described as arbitrary constituent elements. In addition, the figures are schematic diagrams and are not necessarily strictly illustrated. Furthermore, in the figures, there are instances where substantially identical components are given the same reference numerals, and repeated descriptions are omitted or simplified.

[0044] Before describing the embodiments of this disclosure, the understanding that forms the basis of this disclosure will be explained.

[0045] Patent Document 1 discloses an imaging device capable of generating high-resolution multi-wavelength images, i.e., images for each of multiple wavelength bands. In this imaging device, an optical element called an "encoding element" encodes an image of light from an object and captures it. The encoding element, for example, has multiple regions arranged in a two-dimensional manner. The transmission spectra of at least two of these multiple regions each have maximum transmittance values ​​in multiple wavelength domains within the wavelength domain of the object being imaged. The multiple regions can, for example, be configured to correspond to multiple pixels of an image sensor. In the capture using this encoding element, the data of each pixel contains information for multiple wavelength domains. That is, the image data obtained by the capture is data with wavelength information compressed. Therefore, holding two-dimensional data is sufficient, and the amount of data can be suppressed. For example, even when the capacity of the recording medium is limited, data for long-duration motion images can be obtained. The multi-wavelength image is generated by reconstructing multiple images corresponding to multiple wavelength domains based on the captured image. In the following description, each image generated according to each wavelength domain, i.e., band, is also referred to as a "separate image".

[0046] The encoding element can be implemented, for example, by a filter array comprising multiple filters arranged in a two-dimensional manner. Each of the multiple filters can, for example, have a so-called Fabry-Perot resonator configuration including an interference layer. As a Fabry-Perot resonator, for example, the configuration disclosed in Patent Document 2 can be adopted. The multiple filters can be designed such that the transmission spectrum of each filter has multiple peaks in the wavelength domain of the image object. If the interference layer is thinned, the multiple peaks shift towards shorter wavelengths; if the interference layer is thickened, the multiple peaks shift towards longer wavelengths. Multiple filters with different interference layer thicknesses have mutually different transmission spectra.

[0047] The spacing between adjacent peaks in the transmission spectrum is called the FSR (Free Spectral Range). According to the inventors' research, when generating multi-wavelength images using a filter array with equal FSR for all filters, the separation of images in a certain wavelength domain may be incorrect. In this case, the wavelength resolution of the hyperspectral camera decreases.

[0048] The filter array of the embodiments of this disclosure is designed such that two or more of the multiple filters have different FSRs. Such a filter array reduces the possibility of generating incorrectly separated images. As a result, the wavelength resolution of the hyperspectral camera can be improved. Hereinafter, the filter array of the embodiments of this disclosure and the light detection system of this disclosure equipped with the filter array will be briefly described.

[0049] (Item 1)

[0050] The filter array of the first item of this disclosure includes a plurality of filters arranged in a two-dimensional manner, each having a distinct transmission spectrum. The plurality of filters includes: a first filter whose transmission spectrum has a first peak group comprising adjacent first and second peaks; and at least one second filter whose transmission spectrum has a second peak group comprising adjacent third and fourth peaks, wherein, among the peaks included in the second peak group, the wavelength of the third peak is closest to the wavelength of the first peak. A first interval between the wavelengths of the first and second peaks is different from a second interval between the wavelengths of the third and fourth peaks. When the absolute value of the difference between the first and second intervals is ΔFSR, and the half-width of the first peak is σ, ΔFSR / σ ≥ 0.25.

[0051] In this filter array, the separated image can be reconstructed more accurately in the target wavelength domain W, which is wider than the first interval FSR1 and the second interval FSR2, resulting in improved wavelength resolution of the hyperspectral camera.

[0052] (Item 2)

[0053] In the filter array relating to the first item, it is also possible that the aforementioned at least one second filter includes a plurality of second filters, the number of which is more than 10% of the total number of the plurality of filters.

[0054] In this filter array, the possibility of reconstructing the separation map more accurately in the target wavelength domain W, which is wider than that of the first interval FSR1 and the second interval FSR2, is increased.

[0055] (Item 3)

[0056] In the filter array relating to the first or second item, the first filter may include a first interference layer having a first reflecting surface and a second reflecting surface on the opposite side of the first reflecting surface, wherein the first peak and the second peak are obtained by forming a standing wave within the first interference layer.

[0057] In this filter array, the first filter can be designed such that the transmission spectrum has multiple peaks, including the first peak and the second peak mentioned above.

[0058] (Item 4)

[0059] In the filter array relating to the first or second item, at least one of the second filters may include a second interference layer having a third reflecting surface and a fourth reflecting surface on the opposite side of the third reflecting surface, wherein the third peak and the fourth peak are obtained by forming a standing wave within the second interference layer.

[0060] In this filter array, the second filter can be designed such that the transmission spectrum has multiple peaks, including the aforementioned third and fourth peaks.

[0061] (Item 5)

[0062] The light detection system of the fifth item of this disclosure includes: a filter array according to any one of items 1 to 4; and an image sensor configured at a position to receive light transmitted through the plurality of filters.

[0063] This optical detection system enables the use of a hyperspectral camera with improved wavelength resolution.

[0064] (Item 6)

[0065] The optical detection system of the fifth item of this disclosure may also include a processing circuit that generates spectral image data corresponding to the multiple bands based on data representing the spatial distribution of the transmission spectra of the multiple filters and image data obtained by the image sensor.

[0066] In this disclosure, all or part of a circuit, unit, device, component, or section, or all or part of a functional block in a block diagram, may be implemented by one or more electronic circuits, including semiconductor devices, semiconductor integrated circuits (ICs), or LSIs (large scale integration). An LSI or IC can be integrated onto a single chip or constructed by combining multiple chips. For example, functional blocks other than storage elements can also be integrated onto a single chip. Hereinafter referred to as LSI or IC, but the terminology varies depending on the degree of integration; it may also be called a system LSI, VLSI (very large scale integration), or ULSI (ultra large scale integration). Field Programmable Gate Arrays (FPGAs) that can be programmed after the LSI is manufactured, or reconfigurable logic devices capable of reconfiguring the bonding relationships within the LSI or setting the circuit partitioning within the LSI, can also be used for the same purpose.

[0067] Furthermore, all or part of the functions or operations of a circuit, unit, device, component, or part can be executed through software processing. In this case, the software is recorded on one or more non-transitory recording media such as ROM, optical disk, hard disk, etc. When the software is executed by a processor, the functions determined by the software are executed by the processor and peripheral devices. The system or device may also include one or more non-transitory recording media containing the software, a processor, and necessary hardware devices, such as interfaces.

[0068] (Implementation Method)

[0069] <Optical Detection System>

[0070] Figure 1A This diagram schematically illustrates an exemplary embodiment of a light detection system 400 according to the present disclosure. The light detection system 400 includes an optical system 40, a filter array 10, an image sensor 60, and a processing circuit 200. The filter array 10 has the same function as the "encoding element" disclosed in Patent Document 1. Therefore, the filter array 10 can also be referred to as an "encoding element." The optical system 40 and the filter array 10 are arranged in the optical path of light incident from the object 70. Figure 1A In the example shown, the filter array 10 is configured between the optical system 40 and the image sensor 60.

[0071] exist Figure 1A In this example, an apple is shown as an object 70. Object 70 is not limited to an apple and can be any object. The processing circuit 200 generates image data for multiple bands within a specific wavelength domain that is the object wavelength domain, based on the image data generated by the image sensor 60. In this specification, this image data is referred to as "spectral image data". Here, the number of bands within the object wavelength domain is set to N (N is an integer greater than or equal to 4). In the following description, the generated spectral image data for multiple bands will be referred to as separated images 220W1, 220W2, ..., 220W... N These are collectively referred to as separate images 220. In this specification, the signal representing an image, that is, the set of signals representing the pixel values ​​of the multiple pixels constituting the image, is sometimes simply referred to as an "image".

[0072] The filter array 10 comprises multiple transparent filters arranged in rows and columns. The filter array 10 is an optical element whose light transmission spectrum, i.e., the wavelength dependence of light transmittance, varies according to the filter. The filter array 10 modulates the intensity of incident light according to each wavelength domain and allows it to pass through.

[0073] exist Figure 1AIn the example shown, the filter array 10 is positioned near or directly above the image sensor 60. Here, "near" means so close that the image of the light from the optical system 40 is formed on the surface of the filter array 10 in a relatively clear state. "Directly above" means so close that there is almost no gap between them. The filter array 10 and the image sensor 60 can also be integrated. In this specification, the device comprising the filter array 10 and the image sensor 60 will be referred to as "light detection device 300".

[0074] The optical system 40 includes at least one lens. Figure 1A The optical system 40 is represented by a single lens, but it can also be composed of a combination of multiple lenses. The optical system 40 forms an image on the imaging surface of the image sensor 60 via the filter array 10.

[0075] The filter array 10 can also be configured separately from the image sensor 60. Figures 1B to 1D This is a diagram illustrating an example structure of a light detection system 400 configured with the filter array 10 separated from the image sensor 60. Figure 1B In this example, the filter array 10 is positioned between the optical system 40 and the image sensor 60, but separate from the image sensor 60. Figure 1C In this example, the filter array 10 is positioned between the object 70 and the optical system 40. Figure 1D In the example, the light detection system 400 has two optical systems 40A and 40B, with a filter array 10 disposed between them. As in these examples, an optical system including one or more lenses may also be disposed between the filter array 10 and the image sensor 60.

[0076] Image sensor 60 comprises multiple photodetector elements arranged in a two-dimensional manner. Image sensor 60 may be, for example, a CCD (Charge-Coupled Device) sensor, a CMOS (Complementary Metal Oxide Semiconductor) sensor, or an infrared array sensor. The photodetector elements may include, for example, photodiodes. Image sensor 60 may be, for example, a monochrome sensor or a color sensor. The target wavelength range can be arbitrarily determined. The target wavelength range is not limited to the visible wavelength range, but may also be the ultraviolet, near-infrared, mid-infrared, far-infrared, or microwave wavelength range.

[0077] exist Figure 1AIn the example shown, multiple photodetectors are configured opposite one of multiple filters. Each photodetector has sensitivity to light in the wavelength range of the object being photographed. Specifically, each photodetector has the substantial sensitivity required to detect light in the wavelength range of the object being photographed. For example, the external quantum efficiency of the photodetector in this wavelength range can be 1% or higher. The external quantum efficiency of the photodetector can also be 10% or higher. The external quantum efficiency of the photodetector can also be 20% or higher. In the following description, the photodetector is also referred to as a "pixel".

[0078] The processing circuit 200 can be, for example, an integrated circuit equipped with a storage medium such as a processor and a memory. Based on the image 120 acquired by the image sensor 60, the processing circuit 200 generates data for multiple separate images 220, each containing information from multiple wavelength bands. Details regarding the processing method of the multiple separate images 220 and the image signals from the processing circuit 200 will be described later. Alternatively, the processing circuit 200 can be integrated into the light detection device 300, or it can be a component of a signal processing device electrically connected to the light detection device 300 via wired or wireless means.

[0079] <Filter Array>

[0080] The filter array 10 of this embodiment will now be described. The filter array 10 is arranged in the optical path of light incident from the object, modulates the intensity of the incident light according to each wavelength, and outputs it. In this specification, this process performed by the filter array, i.e., the encoding element, is referred to as "encoding".

[0081] Figure 2A This is a schematic diagram illustrating an example of a filter array 10. The filter array 10 includes multiple filters arranged in a two-dimensional configuration. Each filter has an individually defined transmission spectrum. Let the wavelength of the incident light be λ, the transmission spectrum is represented by the function T(λ). The transmission spectrum T(λ) can take values ​​greater than 0 and less than 1.

[0082] exist Figure 2A In the example shown, filter array 10 has 48 rectangular filters arranged in 6 rows and 8 columns. This is merely an example; in practical applications, more filters can be designed. The number could, for example, be the same as the number of pixels in image sensor 60. The number of filters included in filter array 10 can range from, for example, tens to tens of millions, depending on the application.

[0083] Figure 2B This is a diagram illustrating an example of the spatial distribution of light transmittance in multiple wavelength domains W1, W2, ..., Wi within the wavelength domain of an object. Figure 2BIn the example shown, the difference in density between the filters represents the difference in transmittance. A lighter filter has higher transmittance, and a denser filter has lower transmittance. For example... Figure 2B As shown, the spatial distribution of light transmittance varies depending on the wavelength domain.

[0084] Figure 2C and Figure 2D They represent Figure 2A The diagram shows an example of the transmission spectra of filters A1 and A2 included in the plurality of filters of filter array 10. The transmission spectra of filter A1 and filter A2 are different from each other. Thus, the transmission spectrum of filter array 10 varies depending on the filter. However, it is not necessary for all filters to have different transmission spectra. In filter array 10, at least two filters among the plurality of filters have different transmission spectra. That is, filter array 10 includes two or more filters with different transmission spectra. In one example, the number of patterns of the transmission spectra of the plurality of filters included in filter array 10 can be the same as or more than the number i of wavelength domains included in the target wavelength domain. Filter array 10 can also be designed such that more than half of the filters have different transmission spectra.

[0085] Figure 3A and Figure 3B This is a diagram used to illustrate the relationship between the target wavelength domain W and the multiple wavelength domains W1, W2, ..., Wi contained within it. The target wavelength domain W can be set to a variety of ranges depending on the application. For example, the target wavelength domain W can be the wavelength domain of visible light from about 400 nm to about 700 nm, the wavelength domain of near-infrared light from about 700 nm to about 2500 nm, or the wavelength domain of near-ultraviolet light from about 10 nm to about 400 nm. Alternatively, the target wavelength domain W can also be the electromagnetic wave domain such as mid-infrared, far-infrared, terahertz waves, or millimeter waves. Thus, the wavelength domain used is not limited to the visible light domain. In this specification, the term "light" is not limited to visible light; for convenience, non-visible light such as near-ultraviolet light, near-infrared light, and electromagnetic waves are also referred to as "light."

[0086] exist Figure 3A In the example shown, let i be any integer greater than 4, and let the domains obtained by dividing the target wavelength domain W into i equal parts be wavelength domain W1, wavelength domain W2, ..., wavelength domain Wi. However, this is not a limited example. The multiple wavelength domains contained in the target wavelength domain W can also be arbitrarily set. For example, the bandwidth can be non-uniform based on the wavelength domains. There can also be gaps between adjacent wavelength domains. Figure 3B In the example shown, the bandwidth varies depending on the wavelength domain, and there is a gap between two adjacent wavelength domains. Thus, multiple wavelength domains can be determined arbitrarily, as long as they are distinct. The number of wavelength divisions, i, can also be 3 or less.

[0087] Figure 4A This is a diagram used to illustrate the characteristics of the transmission spectrum of a certain filter in the filter array 10. In Figure 4A the example shown, the transmission spectrum has multiple maxima P1 to P5 and multiple minima with respect to wavelengths within the target wavelength range W. In Figure 4A the example shown, normalization is performed so that the maximum value of the light transmittance within the target wavelength range W is 1 and the minimum value is 0. In Figure 4A the example shown, in wavelength ranges such as wavelength range W2 and wavelength range Wi-1, the transmission spectrum has maxima. Thus, in the present embodiment, the transmission spectrum of each filter has maxima in at least two of the multiple wavelength ranges W1 to Wi. According to Figure 4A it can be seen that the maxima P1, P3, P4, and P5 are 0.5 or more.

[0088] As described above, the light transmittance of each filter varies according to the wavelength. Therefore, the filter array 10 allows a larger amount of light in a certain wavelength range of the incident light to be transmitted, and does not transmit much of the light in other wavelength ranges. For example, it can be that for the light in k of the i wavelength ranges, the transmittance is greater than 0.5, and for the light in the remaining i - k wavelength ranges, the transmittance is less than 0.5. k is an integer satisfying 2 ≤ k < i. If the incident light is white light that equally contains all the wavelength components of visible light, the filter array 10 modulates the incident light for each filter into light having discrete multiple intensity peaks with respect to the wavelength, and superimposes and outputs this multi-wavelength light.

[0089] Figure 4B As an example, it shows Figure 4A a diagram of the result of averaging the shown transmission spectrum for each wavelength range W1, wavelength range W2,..., wavelength range Wi. The averaged transmittance is obtained by integrating the transmission spectrum T(λ) for each wavelength range and dividing by the bandwidth of that wavelength range. In this specification, the value of the transmittance averaged for each wavelength range is referred to as the transmittance of that wavelength range. In this example, in the three wavelength ranges where the maxima P1, P3, and P5 are taken, the transmittance becomes prominently high. In particular, in the two wavelength ranges where the maxima P3 and P5 are taken, the transmittance exceeds 0.8.

[0090] The wavelength resolution of the transmission spectrum of each filter can be set to approximately the bandwidth of the desired wavelength domain. In other words, within the wavelength range containing one maximum value in the transmission spectrum curve, the width of the range of values ​​above the minimum value closest to that maximum and the average value of that maximum can be set to approximately the bandwidth of the desired wavelength domain. In this case, if the transmission spectrum is decomposed into frequency components, for example, through a Fourier transform, the values ​​of the frequency components in that wavelength domain become relatively larger.

[0091] Typically, such as Figure 2A As shown, the filter array 10 has multiple filters divided into a grid pattern. Some or all of these filters have different transmission spectra. The wavelength and spatial distribution of the light transmittance of the multiple filters included in the filter array 10 can be, for example, random or quasi-random distributions.

[0092] The considerations for random and quasi-random distributions are as follows. First, each filter in the filter array 10 can be considered as a vector element having a value from 0 to 1 based on the light transmittance. Here, the value of the vector element is 0 when the transmittance is 0, and the value of the vector element is 1 when the transmittance is 1. In other words, the set of filters arranged in a row or column direction can be considered as a multidimensional vector with values ​​from 0 to 1. Therefore, it can be said that the filter array 10 has multiple multidimensional vectors in the column or row direction. In this case, random distribution means that any two multidimensional vectors are independent, i.e., not parallel. Furthermore, quasi-random distribution refers to a structure containing a subset of multidimensional vectors that are not independent. Therefore, in random and quasi-random distributions, a vector whose element is the transmittance of light in the first wavelength domain of each filter belonging to a set of filters arranged in one row or column is independent of a vector whose element is the transmittance of light in the first wavelength domain of each filter belonging to a set of filters arranged in other rows or columns. Similarly, for the second wavelength domain, which is different from the first wavelength domain, the vector of elements consisting of the transmittance values ​​of light in the second wavelength domain of each filter belonging to a set of filters arranged in one row or column is independent of the vector of elements consisting of the transmittance values ​​of light in the second wavelength domain of each filter belonging to a set of filters arranged in other rows or columns.

[0093] When the filter array 10 is positioned near or directly above the image sensor 60, the spacing between the multiple filters included in the filter array 10 can be approximately the same as the pixel spacing of the image sensor 60. In this case, light encoded by a particular filter in the filter array 10 will not mix with light encoded by an adjacent filter before incident on the pixel. By ensuring that light transmitted through each filter is incident only on the corresponding pixel, the calculations described later become easier. When the filter array 10 is positioned separately from the image sensor 60, the spacing between the filters can be narrowed according to the distance.

[0094] exist Figures 2A to 2D In the example shown, filter array 10 has a grayscale transmittance distribution where the transmittance of each filter can take any value between 0 and 1. However, it is not necessarily required to be a grayscale transmittance distribution. For example, a binary-scale transmittance distribution can also be used where the transmittance of each filter can take some value of approximately 0 or approximately 1. In a binary-scale transmittance distribution, each filter allows most of the light from at least two of the multiple wavelength domains included in the target wavelength domain to pass through, while preventing most of the light from the remaining wavelength domains from passing through. Here, "most" means approximately 80% or more.

[0095] Alternatively, a portion, such as half, of the filters can be replaced with transparent filters. Such transparent filters allow light from all wavelengths W1 to Wi within the target wavelength range to be transmitted with the same high transmittance. This high transmittance is, for example, 0.8 or higher. In such a structure, multiple transparent filters can be arranged, for example, in a checkerboard pattern. That is, in the two arrangement directions of the multiple filters in the filter array 10, filters with different transmittances depending on the wavelength and transparent filters can be arranged alternately. Figure 2A In the example shown, the two arrangement directions are horizontal and vertical.

[0096] Data representing the spatial distribution of the spectral transmittance of such filter array 10 is obtained in advance based on design data or experimental calibration and is stored in a storage medium provided by the processing circuit 200. This data is used in the computational processing described later.

[0097] The filter array 10 can be constructed, for example, using multilayer films, organic materials, diffraction grating structures, or microstructures containing metals. When using multilayer films, for example, dielectric multilayer films or multilayer films containing metal layers can be used. In this case, at least one of the thickness, material, and stacking order of each multilayer film can be different for each filter. This allows for different spectral characteristics depending on the filter. By using multilayer films, sharp increases and decreases in spectral transmittance can be achieved. Structures using organic materials can be implemented by using different pigments or dyes, or by stacking different types of materials, depending on the filter. Structures using diffraction grating structures can be implemented by setting different diffraction spacing or depths for each filter. When using microstructures containing metals, spectral analysis based on the plasmon effect can be used.

[0098] <Processing Circuit>

[0099] Next, a method for reconstructing a multi-wavelength separated image 220 using processing circuitry 200 will be described. Here, "multi-wavelength" refers to a wavelength range greater than, for example, the RGB three-color wavelength range obtained by a typical color camera. The number of this wavelength range can be, for example, around 4 to 100. This number of wavelength ranges is also referred to as the "number of optical bands." Depending on the application, the number of optical bands can also exceed 100.

[0100] The desired data is a split image 220, denoted as f. If the number of split frequency bands is w, then f represents the image data f1, f2, ..., f2 of each frequency band. w The integrated data. Here, for example... Figure 1A As shown, let the horizontal direction of the image be the x-direction, and the vertical direction be the y-direction. If we let n be the number of pixels in the x-direction and m be the number of pixels in the y-direction of the image data to be calculated, then the image data f1, f2, ..., f w These are two-dimensional data of n×m pixels. Therefore, data f is three-dimensional data with n×m×w features. On the other hand, the data g of the image 120 obtained by encoding and multiplexing by the filter array 10 has n×m features. Data g can be represented by the following equation (1).

[0101] [Formula 1]

[0102]

[0103] Here, f1, f2, ..., f w It contains data with n×m elements. Therefore, the vector on the right is strictly speaking an n×m×w one-dimensional vector with 1 column. We transform vector g into an n×m one-dimensional vector with 1 column for calculation. Matrix H represents the components f1, f2, ..., f of vector f. wThe transformation involves encoding and intensity modulation using coding information that differs for each wavelength domain, and then summing them. Therefore, H is an n×m row, n×m×w column matrix.

[0104] If vector g and matrix H are given, f should be able to be calculated by solving the inverse problem of equation (1). However, since the number of features n×m×w of the data f to be obtained is greater than the number of features n×m of the data g, this problem is poorly configured and cannot be solved in this state. Therefore, the processing circuit 200 utilizes the redundancy of the image contained in the data f and uses a compressed sensing method to solve it. Specifically, the data f to be obtained is estimated by solving the following equation (2).

[0105] [Formula 2]

[0106]

[0107] Here, f' represents the estimated data of f. The first term within the parentheses in the above equation represents the deviation between the estimated result Hf and the obtained data g, the so-called residual term. Here, the sum of squares is used as the residual term, but the absolute value or the square root of the sum of squares can also be used as the residual term. The second term within the parentheses is the regularization or stabilization term, which will be discussed later. Equation (2) means finding f that minimizes the sum of the first and second terms. The processing circuit 200 can calculate the final solution f' by recursively calculating the solution to converge.

[0108] The first term within the parentheses in equation (2) refers to the operation of calculating the sum of squares of the differences between the obtained data g and Hf obtained by systematically transforming the estimation process f using matrix H. The second term, Φ(f), is the regularization constraint of f, a function reflecting the sparse information of the estimated data. Its function is to smooth or stabilize the estimated data. Regularization terms can be represented, for example, by the discrete cosine transform (DCT), wavelet transform, Fourier transform, or total variation (TV) of f. For example, using the total variation, stable inferred data that suppresses the noise effects of the observed data g can be obtained. The sparsity of the object 70 in the space of each regularization term varies depending on the texture of the object 70. Regularization terms with sparser textures of the object 70 in the space of regularization terms can also be selected. Alternatively, multiple regularization terms can be included in the operation. τ is the weighting coefficient. The larger the weighting coefficient τ, the greater the reduction in lengthy data and the higher the compression ratio. The smaller the weighting coefficient τ, the weaker the convergence to the solution. The weighting coefficient τ is set to an appropriate value that allows f to converge to some extent without becoming over-compressed.

[0109] In addition, Figures 1B to 1DIn the structure, the image encoded by the filter array 10 can be acquired in a blurred state on the imaging surface of the image sensor 60. Therefore, this blur information can be held in advance, and the separated image 220 can be reconstructed by reflecting this blur information onto the system matrix H mentioned above. Here, the blur information is represented by the Point Spread Function (PSF). The PSF is a function that specifies the degree of diffusion of a point image to surrounding pixels. For example, if a point image equivalent to one pixel in the image diffuses to a region of k×k pixels around that pixel through blurring, the PSF can be defined as a group of coefficients, i.e., a matrix, representing the influence of the brightness of each pixel in that region. By reflecting the blurring effect of the PSF-based encoding pattern onto the system matrix H, the separated image 220 can be reconstructed. The position of the filter array 10 can be configured arbitrarily, but a position can be selected where the encoding pattern of the filter array 10 will not diffuse excessively and disappear.

[0110] Furthermore, this example illustrates the operation of compressed sensing using Equation (2), but other methods can also be used to solve it. For example, other statistical methods such as maximum likelihood estimation or Bayesian estimation can be used. Moreover, the number of separated images 220 is arbitrary, and each wavelength domain can also be arbitrarily set. Details of the reconstruction method are disclosed in Patent Document 1. The entire disclosure of Patent Document 1 is incorporated herein by reference.

[0111] <Filter array with Fabry-Perot filter>

[0112] Next, refer to Figure 5 This section describes an example of the specific construction of the filter array 10 in this embodiment. Figure 5 This is a schematic cross-sectional view illustrating an example of a light detection device 300. The cross-sectional view shows... Figure 2A The example shown is a cross-sectional structure of one row of the filter array 10 and the image sensor 60. Figure 5 In the example shown, the filter array 10 is disposed on the image sensor 60. The plurality of photodetectors 60a included in the image sensor 60 are respectively located immediately below a corresponding filter 100 of the plurality of filters 100 included in the filter array 10. The filter array 10 and the image sensor 60 can also be separated. In this case, the plurality of photodetectors 60a can also be disposed at positions that receive light transmitted through a corresponding filter among the plurality of filters. The components can also be configured such that light transmitted through the plurality of filters is incident on the plurality of photodetectors 60a via mirrors. In this case, each of the plurality of photodetectors 60a is not disposed immediately below a corresponding filter among the plurality of filters.

[0113] In this embodiment, all filters 100 included in the filter array 10 have a resonant structure. A resonant structure refers to a structure in which light of a certain wavelength forms a standing wave inside and exists stably. Figure 5 The resonant structure shown includes a first reflective layer 28a, a second reflective layer 28b, and an interference layer 26 between the first reflective layer 28a and the second reflective layer 28b. Reflective surfaces are present at the interfaces between the first reflective layer 28a and the interference layer 26, and at the interfaces between the second reflective layer 28b and the interference layer 26. The reflectivity of each reflective surface can be, for example, 10% or more. The first reflective layer 28a and the second reflective layer 28b can each be formed, for example, from at least one material selected from the group consisting of dielectric multilayer films and metal thin films. The interference layer 26 can be formed from a dielectric or semiconductor that is transparent in a specific wavelength range. The interference layer 26 can be formed, for example, from at least one material selected from the group consisting of Si, Si3N4, TiO2, Nb2O5, and Ta2O5. The refractive index and / or thickness of the interference layer 26 vary depending on the filter 100. The transmission spectra of the plurality of filters 100 each have maximum transmittance values ​​at multiple wavelengths.

[0114] Alternatively, the filter array 10 of this embodiment may also include filters that do not have the resonant structure described above. For example, the filter array 10 may also include wavelength-dependent filters such as transparent filters or ND filters (Neutral Density Filters).

[0115] In this specification, the filter 100 having the resonant structure described above is also referred to as a "Fabry-Perot filter". A Fabry-Perot filter is a type of interference filter. Other types of interference filters, such as color separation filters composed of diffraction gratings, may be used instead of the Fabry-Perot filter of this embodiment.

[0116] Next, an example of the transmission spectrum of filter 100 as a Fabry-Perot filter will be described. When the thickness of the interferometer layer 26 is greater than a specified value, the transmission spectrum of filter 100 has multiple peaks in the target wavelength domain W. These multiple peaks are obtained by forming a standing wave within the interferometer layer 26. In this specification, a filter having multiple peaks in the target wavelength domain W will be referred to as a "multimode filter". Alternatively, instead of the thickness of the interferometer layer 26, the refractive index of the interferometer layer 26 can be appropriately designed to achieve a multimode filter. Or, both the thickness and refractive index of the interferometer layer 26 can be appropriately designed to achieve a multimode filter.

[0117] Figure 6 This is a diagram showing an example of the transmission spectrum of filter 100. In Figure 6In order to facilitate understanding, the multiple peaks in the transmission spectrum are represented by multiple Lorentzian functions. The same applies to subsequent graphs representing the transmission spectrum. Figure 6 The transmission spectrum shown is a simplified representation of the transmission spectrum of a real Fabry-Perot filter. Examples of transmission spectra for real Fabry-Perot filters are described later. Figure 6 In the example shown, using the transmission spectrum represented by the solid line as a reference, if the interference layer 26 is thinned, the transmission spectrum shifts towards the shorter wavelength side, as represented by the dashed line and single-dotted line. If the interference layer 26 is thickened, the transmission spectrum shifts towards the longer wavelength side, as represented by the double-dotted line. Even with different thicknesses of the interference layer 26, the FSR interval between adjacent peaks in the transmission spectrum is equal, which is FSR = 50 nm.

[0118] Before describing the specific characteristics of the filter array 10 in this embodiment, refer to Figure 7 and Figure 8 The comparative example filter array is described below. The comparative example filter array comprises one million filters arranged in a two-dimensional pattern. These one million filters include 16 types of multimode filters arranged in a random or quasi-random distribution.

[0119] Figure 7 This is a graph showing the transmission spectra of 16 comparative multimode filters. The 16 comparative multimode filters each have an interference layer 26 of slightly different thicknesses. Figure 7 In the example shown, the target wavelength domain W (between 450nm and 550nm) is divided into 10 wavelength domains. Hereinafter, these wavelength domains will be referred to as "bands". Each band has a width of 10nm. The 10 bands are named band 1, band 2, ..., band 10, sequentially from the shortest wavelength side. This is achieved by gradually varying the thickness of the interference layer 26 of the multiple filters 100, as... Figure 7 As shown, multiple peaks in the transmission spectrum, represented by the thick line, can be filled with multiple peaks from 15 other transmission spectra. All 16 multimode filters in this comparative example have the same FSR. That is, the filter array in the comparative example has a single FSR. The FSR in this example corresponds to a bandwidth of 5 bands.

[0120] According to the inventors' research, when a photodetector system with a filter array of comparative examples generates a separated image for each band, the separated images for some bands cannot be generated correctly. This problem will be explained below.

[0121] Figure 8These are examples of forward-resolved images of each band obtained when correctly restored, and examples of separated images of each band reconstructed by an optical detection system with a filter array having a single FSR, as shown in the figures. Figure 8 As shown in the upper part, the positive resolution image in this example only has pixel values ​​exceeding the noise level for wavelengths in band 1. In contrast, as... Figure 8 As shown in the lower section, the separation image generated by the optical detection system with the filter array of the comparative example has pixel values ​​exceeding the noise level not only for band 1 but also for band 6. This separation image differs significantly from the forward resolution image. In the case where all multiple filters have a single FSR, such as... Figure 7 As shown, the transmission spectra of all filters exhibit the same dynamics whenever the wavelength shift is equivalent to the FSR. In other words, the transmission spectra of all filters in the comparative example vary with equal periods in the target wavelength domain W. For example, the transmission spectra represented by the thick solid line have peaks in bands 1 and 6, and the transmission spectra represented by the thick dashed line have peaks in bands 3 and 8.

[0122] exist Figure 7 In the example shown, in matrix H of equation (1), the components corresponding to bands 1 to 5 are substantially equal to the components corresponding to bands 6 to 10. As a result, the separated images that should be generated for the original bands can also be generated for different bands. This is because, when the components corresponding to multiple bands in matrix H are equal or similar, in the recursive iterative calculation of equation (2), the image data f may not be the optimal solution but rather reach an incorrect local solution. The smaller the difference between the components corresponding to multiple bands in matrix H, the higher the probability that the image data f reaches a local solution.

[0123] Based on the above, it is not easy to correctly reconstruct the separated image in the object wavelength domain W, which is wider than the FSR, in a filter array of a comparative example with a single FSR. Therefore, the object wavelength domain W of the hyperspectral camera is limited to below the FSR.

[0124] Therefore, the inventors of this disclosure conceived of the structure of the filter array 10 of this embodiment in order to solve the above-mentioned problems. Hereinafter, reference will be made to... Figures 9 to 12 This section describes an example of the filter array 10 in this embodiment. The filter array 10 in this embodiment can reduce the possibility that the image data f reaches a local solution. Therefore, even when, for example, the target wavelength domain W of the hyperspectral camera is expanded to above the FSR, a more accurate separation image can be generated.

[0125] Figure 9This diagram illustrates an example of the transmission spectra of two filters 100 in the filter array 10 of this embodiment. The filter array 10 includes a plurality of first filters and a plurality of second filters. The transmission spectra of the plurality of first filters are different from each other, but they share a common peak interval FSR1. Similarly, the transmission spectra of the plurality of second filters are different from each other, but they share a common peak interval FSR2. Figure 9 The transmission spectrum, represented by a dashed line, illustrates the transmission spectrum of one of a plurality of first filters. This transmission spectrum has multiple peaks, including adjacent first and second peaks. The interval between the first and second peaks is FSR1. Figure 9 The transmission spectrum, represented by solid lines, illustrates the transmission spectrum of one of a plurality of second filters. This transmission spectrum has multiple peaks, including adjacent third and fourth peaks. The wavelength of the third peak is closest to the wavelength of the first peak. The interval between the third and fourth peaks is FSR2. Let the absolute value of the difference between FSR1 and FSR2 be ΔFSR, and let the half-width of the first peak be σ. Figure 9 In the example shown, FSR2 = FSR1 + ΔFSR > FSR1. Specifically, FSR1 = 50 nm, FSR2 = 75 nm, and σ = 12 nm.

[0126] In this specification, the filters with peak intervals of FSR1 and FSR2 in the transmission spectrum as described above are referred to as "Filter 1" and "Filter 2," respectively. The interference layers contained in Filter 1 and Filter 2 are referred to as "Interference Layer 1" and "Interference Layer 2," respectively. The two reflecting surfaces on both sides of Interference Layer 1 are referred to as "Reflecting Surface 1" and "Reflecting Surface 2," and the two reflecting surfaces on both sides of Interference Layer 2 are referred to as "Reflecting Surface 3" and "Reflecting Surface 4."

[0127] By gradually varying the thickness of the interference layer 26, and... Figure 7 Similarly, the example shown can yield 16 multimode filters, including 8 first filters with different transmission spectra and 8 second filters with different transmission spectra. The filter array 10 can, for example, include 16 multimode filters with a random or quasi-random distribution having two FSRs. Thus, the filter array 10 of this embodiment includes multiple first filters with different transmission spectra and multiple second filters with different transmission spectra. The transmission spectra of the multiple first filters have the same shape shifted towards the short-wavelength side or the long-wavelength side. The same applies to the transmission spectra of the multiple second filters.

[0128] Figure 10 This diagram illustrates an example of a forward-resolution image and an example of a separated image reconstructed by a light detection system 400 equipped with the filter array 10 of this embodiment. Figure 10In the example shown, the first filter and the second filter included in the filter array 10 are... Figure 9 The example shown also satisfies FSR1 = 50nm, FSR2 = 75nm, and σ = 12nm. (And...) Figure 8 The examples shown in the lower part are different, in Figure 10 In the example shown in the lower part, the separated image for band 6 is roughly black overall, just like the forward resolution image, and only for band 1 can a separated image with pixel values ​​exceeding the noise level be obtained. In the filter array 10 of this embodiment, the transmission spectra of all filters 100 do not change with equal periods. Therefore, in matrix H of equation (1), the possibility of components corresponding to multiple bands being equal or similar can be suppressed. As a result, the filter array 10 of this embodiment, which has multiple FSRs, can avoid local solutions and reconstruct the separated image more accurately.

[0129] Figure 11 This is a graph illustrating the relationship between the mean squared error (MSE) of the separated image relative to the forward solution image in band 6, which presents a local solution in the comparative example, and ΔFSR / σ. The mean squared error is calculated using the following equation (3).

[0130] [Formula 3]

[0131]

[0132] Here, N and M are the number of pixels horizontally and vertically, respectively. i，j It is the pixel value of the forward image at position (i, j). i，j It is the pixel value of the separated image at position (i, j).

[0133] exist Figure 11 In the example shown, the black circle, white quadrilateral, and white triangle represent the relationships for FSR1 / σ = 10, FSR1 / σ = 5, and FSR1 / σ = 2.5, respectively. ΔFSR = 0 corresponds to the case of using a filter array with a single FSR. Figure 11 As shown, the smaller ΔFSR / σ becomes, the greater the MSE in band 6. That is, the smaller ΔFSR / σ becomes, the higher the probability that the image data f in equation (2) will reach a local solution. When ΔFSR / σ < 0.25, the MSE is almost independent of FSR1 / σ, but depends on ΔFSR / σ. In contrast, when ΔFSR / σ ≥ 0.25, the MSE depends on both FSR1 / σ and ΔFSR / σ. When ΔFSR / σ ≥ 0.25 and FSR1 / σ > 5, the MSE effectively decreases with the increase of ΔFSR / σ, becoming below 10.

[0134] Based on the above, when ΔFSR / σ ≥ 0.25, the filter array 10 with multiple FSRs can achieve the effect of avoiding local solutions. As a result, the separated image can be reconstructed more accurately in the target wavelength domain W, which is wider than FSR1 and FSR2. The filter array 10 of this embodiment with multiple FSRs can expand the target wavelength domain W of the hyperspectral camera to above the largest FSR among the multiple FSRs.

[0135] Figure 12 This is a diagram illustrating an example of the relationship between the MSE of the separated image relative to the forward resolution image in band 6 and the proportion of the second filters included in filter array 10. The proportion of the second filters included in filter array 10 is the ratio of the number of second filters to the total number of first filters and second filters included in filter array 10. For example, when the proportion of second filters included in filter array 10 is 10%, the ratio of the number of first filters to the number of second filters in filter array 10 is 9:1. When the proportion of second filters included in filter array 10 is 50%, the ratio of the number of first filters to the number of second filters in filter array 10 is 5:5.

[0136] like Figure 12 As shown, if the proportion of the second filter in the filter array 10 is less than 10%, the MSE increases sharply. Therefore, when the proportion of the second filter in the filter array 10 is greater than 10%, the possibility of reconstructing the separated image more accurately by avoiding local solutions becomes higher.

[0137] Next, refer to Figure 13 This section describes a specific example of the transmission spectrum of the two filters 100 in the filter array 10 of this embodiment. Figure 13 This diagram illustrates an example of the transmission spectrum of the filter 100 according to this embodiment. In this example, the first reflective layer 28a and the second reflective layer 28b in each filter 100 are formed of dielectric multilayer films with alternating layers of TiO2 and SiO2, respectively. The interference layer 26 in the filter 100 is formed of a TiO2 layer. In the calculation of the transmission spectrum, DiffractMOD from RSoft Corporation, an analysis tool based on Rigorous Coupled-Wave Analysis (RCWA), was used.

[0138] Figure 13The solid line and dashed line shown represent the transmission spectra of Fabry-Perot filters with an interference layer thickness of 633 nm and 750 nm, respectively. The filter with an interference layer thickness of 633 nm has a larger FSR than the filter with an interference layer thickness of 750 nm. The filter with an interference layer thickness of 750 nm is the first filter, and the filter with an interference layer thickness of 633 nm is the second filter.

[0139] exist Figure 13 In the example shown, ΔFSR = 5 nm and half-width σ = 14 nm. ΔFSR / σ = 0.36 satisfies the condition ΔFSR / σ ≥ 0.25, which is used to more accurately reconstruct a separated image in a certain wavelength domain.

[0140] In practical Fabry-Perot filters, the half-width σ of the peak may vary depending on the thickness of the interference layer 26 and the wavelength domain. The half-width σ of the peak satisfying ΔFSR / σ ≥ 0.25 is the half-width of the first peak present in the wavelength domain of interest among the peaks of the plurality of first and second filters included in the filter array 10. Figure 11 In the example shown, the wavelength domain of interest is band 6. Furthermore, in this embodiment, the half-width of the first peak of the transmission spectrum of the first filter in the filter array 10 is set to σ, but it is also possible to set the largest half-width among the peak values ​​of the transmission spectra of all filters to σ.

[0141] In the case of multiple first filters with different transmission spectra achieved by gradually varying the thickness of the interferometer layer 26, the actual FSRs of the multiple first filters may differ to some extent. In this case, the maximum value of the FSRs among the multiple first filters is also set as the FSR. 1max Set the minimum value to FSR 1min At that time, one can think of as long as (FSR) 1max -FSR 1min If σ < 0.25, all first filters will have the same FSR1. The same applies to the FSR2 of multiple second filters.

[0142] Alternatively, based on the above results, it can be considered that all filters with ΔFSR / σ < 0.25 belong to filter type 1, and filters with ΔFSR / σ ≥ 0.25 belong to filter type 2. In this case, if the proportion of filter type 2 is greater than 10%, the separated image can be reconstructed more accurately in the target wavelength domain W exceeding the FSR.

[0143] Next, refer to Figures 14A to 14F ,illustrate Figure 5 A modified example of the optical detection device 300 shown.

[0144] Figures 14A to 14F It is a schematic representation Figure 5 A diagram of a modified example of the optical detection device 300.

[0145] like Figure 14A As shown, multiple filters 100 can be divided in the filter array 10. It is not necessary to divide all filters 100. Alternatively, a portion of the filters 100 can be divided.

[0146] like Figure 14B As shown, filters 100 may not be configured on some of the photodetector elements 60a. In other words, at least one of the multiple filters 100 in the filter array 10 may also be transparent.

[0147] like Figure 14C As shown, a space can also be provided between the filter array 10 and the image sensor 60. In other words, the filter array 10 and the image sensor 60 can also be separated by a space.

[0148] like Figure 14D As shown, one filter 100 can also be configured across multiple photodetector elements 60a. In other words, the interference layer 26 can also be continuously arranged across two or more filters 100. The first reflective layer 28a and / or the second reflective layer 28b can also be continuously arranged across two or more filters 100.

[0149] like Figure 14E and Figure 14F As shown, a transparent layer 27 can also be configured to flatten the step difference of the filter array 10. In other words, the filter array 10 may also include a transparent layer 27 to flatten the step difference of two or more filters 100 having the above-described resonant structure. Figure 14E In the example shown, a step difference exists on the upper surface of the second reflective layer 28b of the filter array 10. Figure 14F In the example shown, a step difference exists on the lower surface of the first reflective layer 28a of the filter array 10. By flattening the step difference between two or more filters 100 using the transparent layer 27, it is easy to configure other components on the transparent layer 27.

[0150] like Figure 14E and Figure 14F As shown, multiple microlenses 40a can also be configured on the filter array 10. Each microlens 40a is configured on one of the multiple filters 100. In other words, the filter array 10 also includes two or more microlenses 40a. Each of the two or more microlenses 40a is configured on one of the two or more filters 100 having the resonant structure described above. By focusing the incident light using two or more microlenses 40a, light can be detected efficiently.

[0151] Industrial availability

[0152] The optical detection system and filter array disclosed herein are useful, for example, for cameras and measuring devices that acquire two-dimensional images of multiple wavelengths. The optical detection system and filter array disclosed herein can also be applied to sensing for biological organisms, medical and cosmetic applications, foreign object / pesticide residue detection systems for food, remote sensing systems, and vehicle-mounted sensing systems, etc.

[0153] Label Explanation

[0154] 10 filter arrays

[0155] 26 Interference Layers

[0156] 27 Transparent Layer

[0157] 28a First Reflective Layer

[0158] 28b Second reflective layer

[0159] 40 Optical System

[0160] 60 Image Sensor

[0161] 60a Optical Detection Element

[0162] 70 Objects

[0163] 100 filter

[0164] 120 images

[0165] 200 processing circuit

[0166] 220 Separated Images

[0167] 300 optical detection device

[0168] 400 optical detection system

Claims

1. A filter array, wherein, It has multiple filters arranged in a two-dimensional manner, each with a different transmission spectrum. The aforementioned filters include: The first filter has a transmission spectrum containing a group of first main peaks, including mutually adjacent first main peaks and second main peaks; At least one second filter, the transmission spectrum having a group of second main peaks including mutually adjacent third and fourth main peaks; and The third filter has a transmission spectrum with two adjacent fifth and sixth main peaks. Among the multiple peaks included in the second main peak group, the wavelength of the third main peak is closest to the wavelength of the first main peak. The first interval between the wavelength of the first dominant peak and the wavelength of the second dominant peak is different from the second interval between the wavelength of the third dominant peak and the wavelength of the fourth dominant peak. Let the absolute value of the difference between the first interval and the second interval be ΔFSR. Let the half-value width of the first dominant peak be σ. ΔFSR / σ≥0.25, The half-width of the first dominant peak represents the wavelength difference between the wavelength corresponding to half the transmittance of the first dominant peak. The transmission spectrum of the third filter is different from that of the first filter, and the third interval between the wavelengths of the fifth main peak and the sixth main peak is the same as the first interval.

2. The filter array as described in claim 1, wherein, The aforementioned at least one second filter includes multiple second filters. The number of the aforementioned second filters is more than 10% of the total number of the aforementioned filters.

3. The filter array as described in claim 1 or 2, wherein, The first filter described above includes a first interference layer having a first reflecting surface and a second reflecting surface located on the opposite side of the first reflecting surface. The first and second main peak values ​​mentioned above are obtained by forming standing waves within the first interference layer.

4. The filter array as described in claim 1 or 2, wherein, The aforementioned at least one second filter includes a second interference layer having a third reflecting surface and a fourth reflecting surface located on the opposite side of the third reflecting surface. The aforementioned third and fourth main peak values ​​are obtained by forming standing waves within the aforementioned second interference layer.

5. The filter array as described in claim 1, wherein, When σ M When referring to the half-value width of the main peak with the largest half-value width among the multiple main peaks contained in the first main peak group and the multiple main peaks contained in the second main peak group, ΔFSR / s M ≥0.25, The half-value width of the main peak with the maximum half-value width represents the wavelength difference between the wavelength corresponding to half the transmittance of the main peak with the maximum half-value width.

6. A light detection system, wherein, have: The filter array according to any one of claims 1 to 5; and An image sensor is positioned to receive light that has been transmitted through the aforementioned multiple filters.

7. The optical detection system as described in claim 6, wherein, It also includes a processing circuit that generates a spectroscopic image dataset that corresponds one-to-one with each of the multiple bands, based on data representing the spatial distribution of the transmission spectra of the multiple filters and image data obtained from the image sensor.

8. The optical detection system as described in claim 6, wherein, The aforementioned image sensor includes a first detection element and a second detection element. The first detection element detects light having a spectrum including the first main peak and the second main peak. The second detection element detects light having a spectrum that includes the third main peak and the fourth main peak.

9. The optical detection system as described in claim 7, wherein, Each of the above-mentioned spectroscopic image datasets contains n pixel values ​​horizontally and m pixel values ​​vertically. The number of the above multiple bands is w. The data representing the spatial distribution of the transmission spectra of the aforementioned filters corresponds to a matrix H with n×m rows and n×m×w columns.

Images