JP2024028237A - Spectroscopic camera, method for imaging, program, and recording medium - Google Patents

Abstract

PURPOSE: To provide a spectroscopic camera that can obtain a highly precise image.CONSTITUTION: A spectroscopic camera includes: an optical system made of one pair of lenses; a wavelength selection unit having a pair of reflection surfaces facing each other in the positions of pupils of the optical system, the wavelength selection unit changing the interval of the pair of reflection surfaces according to an applied voltage and selectively transmitting the light of the wavelength appropriate to the interval; a lens array having n number (n is a natural number) of lenses each collecting light that has passed through the optical system and the wavelength selection unit; and an imaging element located in a position conjugate to the wavelength selection unit, the imaging element having n number of divided imaging regions for receiving light collected by the n number of lenses.SELECTED DRAWING: Figure 1

Description

The present invention relates to a spectroscopic camera, an imaging method, a program, and a recording medium.

A plenoptic camera is used as a spectroscopic camera that can simultaneously acquire a plurality of pieces of spectral information and generate an image based on the acquired spectral information. In order to measure a spectral spectrum in a plenoptic camera, a measuring device has been proposed in which a plurality of bandpass filters having different spectral transmittances are arranged near the aperture position of a main lens (for example, Patent Document 1).

Japanese Patent Application Publication No. 2015-132594

In the above-mentioned conventional technology, since a multispectral image is acquired using a finite number of fixed bandpass filters, there is a problem in that continuous spectra cannot be acquired.

Furthermore, a technique has long been known in which the wavelength of light transmitted through a pair of reflective surfaces is continuously changed by continuously changing the distance between a pair of reflective surfaces. This technique has a problem in that if the parallelism or surface precision between the reflecting surfaces is poor, the half-width of the transmitted wavelength deteriorates.

An example of the problem to be solved by the present invention is the problem of deterioration of wavelength selectivity due to manufacturing variations in a spectroscopic camera that selects the transmitted wavelength by changing the distance between a pair of reflective surfaces. Can be mentioned.

The invention according to claim 1 has an optical system including a pair of lenses, and a pair of reflective surfaces disposed opposite to each other at the position of a pupil of the optical system, and the pair of reflective surfaces are arranged in accordance with an applied voltage. a wavelength selection section that changes the spacing between surfaces and selectively transmits light of a wavelength corresponding to the spacing; ); and an imaging element provided at a position conjugate with the wavelength selection section and having n divided imaging regions that receive light condensed by the n lenses. It is characterized by

1 is a diagram showing the configuration of a spectroscopic camera 10 of Example 1. FIG. 3 is a diagram schematically showing how light from an object point is received by an image sensor 13. FIG. 3 is a diagram schematically showing how light from one point on the pupil EYE of the optical system 11 is received by the image sensor 13. FIG. FIG. 6 is a diagram showing an example of a case where a physical opening is arranged in the pupil. FIG. 6 is a diagram schematically showing that an image similar to that obtained by arranging a physical opening in the pupil can be obtained by selecting a pixel region. FIG. 4 is a diagram comparing the intensity and half-width of transmitted light between a case where the gap between the reflecting surfaces is substantially uniform and a case where the gap is non-uniform near the end portions. FIG. 6 is a diagram illustrating an example in which an aperture stop is placed immediately in front of a filter. 3 is a flowchart illustrating processing operations for pre-photography. FIG. 6 is a diagram schematically showing an example of irradiating light of a single wavelength in pre-photography. FIG. 6 is a diagram schematically showing an example in which a bandpass filter is attached to a position in front of the optical system 11 and illumination light is irradiated in advance photographing. FIG. 6 is a diagram schematically showing an example in which a bandpass filter is attached to a position in front of a wavelength selection filter FF and illumination light is irradiated in advance photographing. 7 is a flowchart showing the processing operation of actual photographing. 3 is a diagram showing the configuration of a spectroscopic camera 20 in Example 2. FIG. FIG. 3 is a cross-sectional view showing an example of a configuration in which a plurality of bandpass filters having the same transmission wavelength are provided. FIG. 3 is a cross-sectional view showing an example of a configuration in which a plurality of bandpass filters with different transmission wavelengths are provided. FIG. 3 is a cross-sectional view showing an example of a configuration in which a plurality of bandpass filters with different half widths are provided. FIG. 6 is a diagram schematically showing the state of photographing when dust is attached to the wavelength selection filter.

Embodiments of the present invention will be described below with reference to the drawings. In the following description of each embodiment and the accompanying drawings, substantially the same or equivalent parts are designated by the same reference numerals.

FIG. 1 is a cross-sectional view showing the configuration of a spectroscopic camera 10 according to the first embodiment. The spectroscopic camera 10 performs imaging by, for example, irradiating light onto an object X (subject) and receiving light reflected from the object X or light transmitted through the object X. The spectroscopic camera 10 includes an optical system 11 , a camera 14 including a microlens array 12 and an image sensor 13 , and an image processing section 15 . The microlens array 12 is arranged at a position conjugate with the object X. Further, the image sensor 13 is arranged at a position conjugate with the pupil of the optical system 11.

The optical system 11 includes a first lens L1 and a second lens L2, which are made of, for example, a convex lens. A wavelength selection filter FF is arranged at the pupil position of the optical system 11 between the first lens L1 and the second lens L2. The optical system 11 has a mechanism (focusing mechanism) that can adjust the distance according to the position of the object X, and is adjusted so as to maintain a conjugate relationship between the object X and the microlens array 12.

The first lens L1 is arranged on the side where the light from the object ). Light from the object X passes through the first lens L1, the wavelength selection filter FF, and the second lens L2, and is focused on the microlens array 12.

The wavelength selection filter FF is composed of, for example, a Fabry-Perot interference filter. The wavelength selection filter FF is made up of reflective films RF1 and RF2 that are arranged to face each other, and transmits light having a wavelength corresponding to the spacing between the reflective films. The wavelength selection filter FF changes the wavelength of transmitted light by changing the distance d between the opposing surfaces of the reflective films RF1 and RF2 (hereinafter referred to as the gap d between the reflective surfaces) upon application of a voltage V. It is configured so that it can be set selectively.

Ideally, the gap d between the reflective surfaces of the wavelength selection filter FF should be formed so that the spacing is uniform, but due to variations during manufacturing and assembly, in reality, there may be uneven portions. (hereinafter referred to as non-uniformity of the gap d between the reflecting surfaces). The spectroscopic camera 10 of this embodiment has a function of correcting the captured image according to the non-uniformity of the gap d between the reflecting surfaces.

The microlens array 12 is composed of n microlenses (n is a natural number) arranged vertically and horizontally. Light from different points on object X enters different microlenses. Each microlens corresponds to a "pixel" in a normal camera, and the number of microlenses corresponds to the "number of pixels" in a normal camera. Each microlens of the microlens array 12 receives the light that has passed through the optical system 11 and focuses the light on the image sensor 13.

The image sensor 13 is an image sensor that receives light collected by the microlens array 12, converts it into electronic information, and obtains a captured image. The image sensor 13 is composed of n divided imaging regions corresponding to each microlens of the microlens array 12. The light incident on each microlens of the microlens array 12 is received by the corresponding divided imaging area of the image sensor 13. The light incident on different microlenses is received by different divided imaging areas, and the divided imaging areas where the light is received do not overlap.

FIG. 2 is a diagram schematically showing how light from one point of the object X (referred to as object point P) is received by the image sensor 13. Light from the object point P passes through different regions on the first lens L1, the pupil EYE, and the second lens L2, enters the microlens ML2 of the microlens array 12, and is received by the divided imaging region R2. . The brightness value of the image of the object point P is represented by the sum of the brightness values of all pixels in the divided imaging region R2. Therefore, by obtaining the sum of the brightness values of all pixels in each divided imaging region, the same imaging result (captured image) as an image captured by a normal camera with an image sensor arranged at the position of the microlens can be obtained.

FIG. 3 is a diagram schematically showing how light from one point on the pupil EYE of the optical system 11 is received by the image sensor 13. Light from a single point on the pupil EYE (that is, light incident on a single point on the pupil EYE) passes through different regions of the second lens L2 and different microlenses of the microlens array 12, and is transmitted to each of the image sensors 13. The light is focused on the same location in the divided imaging area.

Referring again to FIG. 1, the image processing unit 15 is comprised of, for example, a CPU (Central Processing Unit), and performs image correction processing on the image of the object X acquired by the image sensor 13. Specifically, the image processing unit 15 selects a pixel region according to the brightness distribution in each divided imaging region of the image sensor 13 obtained by pre-imaging, and processes the object X using only the information of the selected pixel region. Construct a captured image. In the pre-imaging, the image processing unit 15 records the signals received by the image sensor 13 in a storage unit (not shown) such as a memory while sweeping the voltage applied to the wavelength selection filter FF, and stores the recorded plurality of signals. A brightness distribution indicating the non-uniformity of the transmission wavelength of the wavelength selection filter FF (that is, the non-uniformity of the gap d of the reflecting surface) is obtained based on the signal information. That is, the image processing section 15 has a function as a detection section that detects non-uniformity of the transmission wavelength of the wavelength selection filter FF.

The brightness distribution within the divided imaging area of the image sensor 13 represents the brightness distribution of the light incident on the corresponding microlens of the microlens array 12 when it passes through the pupil. Therefore, if the brightness value of light incident on a certain microlens is calculated as the sum of specific pixel areas in the corresponding divided imaging area, it will pass through the specific area of the pupil (the area corresponding to the pixel area in the divided imaging area). The intensity of the received light can be obtained. Furthermore, if the brightness values of all the microlenses are calculated as the sum of the same pixel areas of the corresponding divided imaging areas, an image based on light that has passed through a specific area of the pupil can be obtained. As a result, an image similar to that obtained when a physical aperture having a predetermined shape is arranged in the pupil of the optical system 11, for example, can be obtained.

FIG. 4 is a diagram showing an example in which a physical opening AP having a heart shape is arranged at the pupil position of the optical system. In each divided imaging area of the image sensor 13, an image obtained by inverting the heart shape (that is, the shape of the opening AP) upside down is obtained.

On the other hand, in the spectroscopic camera 10 of this embodiment, by calculating the brightness value using only the selected pixel area, as shown in FIG. , a heart-shaped opening) can be obtained without arranging a physical opening (in the figure, the heart-shaped image is flipped upside down).

Referring again to FIG. 1, the image processing unit 15 is comprised of, for example, a CPU (Central Processing Unit), and performs image correction processing on the image of the object X acquired by the image sensor 13. Specifically, the image processing unit 15 selects a pixel region according to the brightness distribution in each divided imaging region of the image sensor 13 obtained by pre-imaging, and processes the object X using only the information of the selected pixel region. Construct a captured image. In the pre-imaging, the image processing unit 15 records the signals received by the image sensor 13 in a storage unit (not shown) such as a memory while sweeping the voltage applied to the wavelength selection filter FF, and stores the recorded plurality of signals. A brightness distribution indicating the non-uniformity of the transmission wavelength of the wavelength selection filter FF (that is, the non-uniformity of the gap d of the reflecting surface) is obtained based on the signal information. That is, the image processing section 15 has a function as a detection section that detects non-uniformity of the transmission wavelength of the wavelength selection filter FF.

The brightness distribution within the divided imaging area of the image sensor 13 represents the brightness distribution of the light incident on the corresponding microlens of the microlens array 12 when it passes through the pupil. Therefore, if the brightness value of light incident on a certain microlens is calculated as the sum of specific pixel areas in the corresponding divided imaging area, it will pass through the specific area of the pupil (the area corresponding to the pixel area in the divided imaging area). The intensity of the received light can be obtained. Furthermore, if the brightness values of all the microlenses are calculated as the sum of the same pixel areas of the corresponding divided imaging areas, an image based on light that has passed through a specific area of the pupil can be obtained. As a result, an image similar to that obtained when a physical aperture having a predetermined shape is arranged in the pupil of the optical system 11, for example, can be obtained.

FIG. 4 is a diagram showing an example in which a physical opening AP having a heart shape is arranged at the pupil position of the optical system. In each divided imaging area of the image sensor 13, an image obtained by inverting the heart shape (that is, the shape of the opening AP) upside down is obtained.

On the other hand, in the spectroscopic camera 10 of this embodiment, by calculating the brightness value using only the selected pixel area, as shown in FIG. , a heart-shaped opening) can be obtained without arranging a physical opening (in the figure, the heart-shaped image is flipped upside down).

By the way, the Fabry-Perot interference filter that constitutes the wavelength selection filter FF is an interference filter that is composed of a set of parallel reflective films, and by changing the gap between the reflective surfaces, the wavelength of the transmitted light can be changed. Change. When the reflectance of the reflective surface is R, the gap between the reflective surfaces is h ₀ , and the wavelength of the incident light is λ, the transmittance T(λ) of the Fabry-Perot filter is expressed by the following equation (1).

A reflective surface with R=1 has a value only when λ=2h ₀ , and becomes a completely monochromatic filter that does not allow any other wavelengths to pass. However, in reality, R<1, so a certain range of wavelengths centered around λ=2h ₀ will be passed. The full width at half maximum (FWHM) is expressed by the following equation (2). In the limit of R→1, FWHM→0.

In the limit of R→1, FWHM→0. However, even if the reflectance R is 1, the transmitted light will not be completely monochromatic because the gap between the reflective surfaces is non-uniform in Fabry-Perot filters that are actually produced.

Figures 6 (a) and (b) compare the intensity and half-width of transmitted light between an ideal case where the gap between the reflective surfaces is almost uniform and a case where the gap is non-uniform near the edges. FIG.

As shown in FIG. 6(a), when the gap has a uniform value (d=h ₀ ) in all areas within the reflective surface of the filter, the wavelength of the transmitted light at all locations within the reflective surface is 2h ₀ , and the half width is approximately 0.

On the other hand, as shown in Figure 6(b), if the gap is non-uniform, in the region where d≠h ₀ , light with a wavelength other than λ = 2h ₀ will be transmitted, and the half-value width becomes wider.

In order to improve such a broadening of the half-width, it is generally considered to place an aperture stop immediately before or after the filter. For example, as shown in FIG. 7A, by arranging an aperture diaphragm just before the filter, light can be transmitted only through a region with a uniform gap around the optical axis, thereby improving the half-width. However, although this method can improve the half-width, the amount of transmitted light becomes small.

Additionally, if the gap between the reflective surfaces has non-uniformity that is asymmetrical with respect to the optical axis, the amount of transmitted light will further decrease if an aperture stop is arranged symmetrically with respect to the optical axis as shown in Figure 7(b). Therefore, as shown in FIG. 7(c), it is necessary to arrange an aperture stop according to the non-uniformity of the gap. However, it is difficult to arrange an appropriate aperture diaphragm because the portions that are nonuniform actually vary depending on the filter.

In contrast, in the spectroscopic camera 10 of this embodiment, it is possible to construct an image by selectively using only specific pixel areas in each divided imaging area of the image sensor 13. Therefore, without providing a physical aperture stop, it is possible to obtain an image using only the light that has passed through the area where the spacing between the reflective films RF1 and RF2 of the wavelength selection filter FF is uniform.

Next, a method of imaging and selecting a pixel region using the spectroscopic camera 10 of this embodiment will be described. First, preliminary photographing before photographing (actual photographing) of a subject will be described with reference to the flowchart of FIG. 8 and FIG. 9.

In the preliminary photographing, as shown in FIG. 9, light with a single wavelength λ ₁ emitted from the laser LR is irradiated onto the screen SC via the lens L3 (step S101).

The spectroscopic camera 10 photographs the screen SC while sweeping the voltage V applied to the wavelength selection filter FF. The image processing unit 15 records the signal received by the image sensor 13, which is an imaging device (step S102).

By changing the voltage V applied to the wavelength selection filter FF, the interval between the reflective films RF1 and RF2 is changed. The image of each divided imaging area of the image sensor 13 becomes the brightest when the interval is such that the most light of wavelength λ ₁ passes through. At this time, a region with high brightness in the image VD obtained in the divided imaging region is a region in which the corresponding region on the pupil passes light of wavelength λ ₁ , and the gap d between the reflective surfaces (i.e., the reflective The distance between the films RF1 and RF2) reflects a uniform area.

The image processing unit 15 acquires the brightness distribution of the divided imaging areas that reflects the non-uniformity of the gap d between the reflecting surfaces of the wavelength selection filter (non-uniformity of the transmitted wavelength) based on the plurality of recorded signal information. (Step S103).

Note that when performing preliminary imaging for pixel area selection, instead of using a screen SC illuminated with a single wavelength laser beam, a screen SC illuminated with sunlight or halogen light is used with a wavelength λ ₁ . The image may be photographed through a bandpass filter that passes light.

For example, as shown in FIG. 10, during pre-photography, a bandpass filter that transmits light of wavelength λ ₁ is placed on the optical path in front of the optical system 11 (that is, in front of the first lens L1) when viewed from the incident side. Photographing is performed by disposing a BPF and irradiating illumination light WL, and during actual photographing, the band pass filter BPF is removed and photographing is performed. As a result, it is possible to obtain information on pixel regions showing the distribution of brightness according to the gap between the reflective surfaces of the wavelength selection filter FF, and to construct an image using only pixel regions with high brightness. Furthermore, as shown in FIG. 11, a bandpass filter BPF may be placed on the optical path in front of the wavelength selection filter to perform preliminary imaging. That is, the bandpass filter BPF only needs to be placed on the optical path.

Next, actual photographing of a subject will be described with reference to the flowchart of FIG. 12.

First, the spectroscopic camera 10 is set on the subject (object X) (step S201). The spectroscopic camera 10 photographs the object X while sweeping the voltage V applied to the wavelength selection filter FF (step S202). At this time, unlike pre-photography, the subject is photographed using illumination light such as sunlight or white light rather than single wavelength light. The image processing unit 15 then records the signal received by the image sensor 13, which is an imaging device.

The image processing unit 15 determines the reflection during actual imaging based on the information indicating the non-uniformity of the gap d between the reflective surfaces of the wavelength selection filter obtained by the pre-imaging (that is, the information about the brightness distribution within the divided imaging area). Calculate first image information obtained by receiving light that has passed through an area where the gap d between surfaces is uniform, and second image information obtained by receiving light that has passed through an area where the gap d between surfaces is uneven. (Step S203). The image processing unit 15 obtains a final captured image based on the first image information and the second image information (step S204).

As described above, according to the spectroscopic camera 10 of this embodiment, image information indicating the brightness distribution according to the gap d between the reflective surfaces of the wavelength selection filter FF is acquired in advance photography, and the brightness value is determined in advance photography. By selectively using only the pixel regions with high values, it is possible to obtain a captured image constructed using only light that has passed through a region where the gap d between the reflective surfaces of the wavelength selection filter FF is uniform. Therefore, even if there is a region where the transmitted wavelength is non-uniform in the filter, it is possible to obtain a highly accurate image.

FIG. 13 is a cross-sectional view showing the configuration of a spectroscopic camera 20 according to the second embodiment. The spectroscopic camera 20 of this embodiment differs from the spectroscopic camera 10 of the first embodiment in that a bandpass filter BPF is fixedly provided on the exit side surface of one of the microlenses constituting the microlens array 12. different.

A bandpass filter BPF with a transmission wavelength λ ₁ (the center wavelength of the passband is λ ₁ ) is provided on the exit side surface (that is, the surface on the image sensor 13 side) of the microlens ML1 located at the end of the microlens array 12. is provided. The microlens ML1 functions as a pixel selection microlens, and the other microlenses function as imaging microlenses.

The image sensor 13 has divided imaging regions R1 corresponding to the microlenses ML1 of the microlens array 12. The divided imaging region R1 is provided at a position corresponding to the microlens ML1 (for example, at the end of the image sensor). The divided imaging region R1 functions as a divided imaging region for pixel selection, and the other divided imaging regions function as divided imaging regions for imaging.

When photographing the object X, of the light incident on the microlens ML1, only the light with wavelength _λ1 passes through the bandpass filter BPF and is received by the divided imaging region R1 of the image sensor 13. Therefore, in the divided imaging region R1, an image VD for pixel selection having a brightness distribution reflecting the gap d between the reflective surfaces of the wavelength selection filter FF is acquired.

On the other hand, light that has passed through areas other than the microlens ML1 of the microlens array 12 is received by other divided imaging areas of the image sensor 13 without passing through the bandpass filter BPF. Therefore, in the divided imaging regions other than the divided imaging region R1 of the image sensor 13, a normal captured image of the object X is obtained.

The image processing unit 15 performs image correction processing on captured images obtained outside the divided imaging region R1, based on the image information of the pixel selection image VD obtained in the divided imaging region R1. Specifically, the image processing unit 15 selects a pixel area according to the brightness distribution shown in the image VD for pixel selection, and configures a captured image of the object X using only the selected pixel area.

Therefore, according to the spectroscopic camera 20 of this embodiment, unlike in the first embodiment, in which it is necessary to perform imaging in two steps: preliminary imaging and actual imaging, only the pixel areas with high brightness are used in one imaging. It becomes possible to obtain a captured image.

Note that, unlike the above configuration, a configuration may be adopted in which a plurality of band pass filters BPF are provided. For example, as shown in FIG. 14, two bandpass filters BPF, both of which have a transmission wavelength (center wavelength of the passband) of λ ₁ , are provided on the exit side surfaces of the microlenses at both ends of the microlens array 12, Two corresponding divided imaging regions (divided imaging regions R1 and Rx) are set as divided imaging regions for pixel selection. As a result, a plurality of images VD for pixel selection are obtained, so that pixel selection can be performed with high precision.

Furthermore, as shown in FIG. 15, a configuration may be adopted in which a plurality of bandpass filters BPF1 (transmission wavelength λ ₁ ) and BPF2 (transmission wavelength λ ₂ ) having different transmission wavelengths are provided. The image for pixel selection in the divided imaging region R1 and the image for pixel selection in the divided imaging region Rx are acquired when different voltages V are applied to the wavelength selection filter FF. For example, when voltage V= _V1 , a pixel selection image VD1 is obtained in the divided imaging region R1, and when voltage V= _V2 , a pixel selection image VD2 is obtained in the divided imaging region Rx. Even in this case, since pixels can be selected based on a plurality of images for pixel selection, pixel selection can be performed with high accuracy.

Further, a configuration may be adopted in which a plurality of bandpass filters having different half widths are provided. For example, as shown in FIG. 16(a), there is a band pass filter BPFA whose transmission wavelength is centered at λ ₁ and whose half-width is W ₁ , and a band-pass filter BPFA whose transmission wavelength is centered at λ ₁ and whose half-width is W ₂ . A bandpass filter BPFB is provided on the exit side surface of the microlenses at both ends of the microlens array 12. When half width W ₂ > half width W ₁ , as shown in FIG. 16(b), the amount of transmitted light is small in the band pass filter BPFA with a small width at half value, and the amount of transmitted light is small in the band pass filter BPFB with a large width at half value. The amount of light increases. Therefore, based on the image VD1 for pixel selection obtained in the divided imaging area A and the image VD3 for pixel selection obtained in the divided imaging area B, pixel selection is performed while switching between pixel selection giving priority to half width and pixel selection giving priority to transmitted light amount. be able to.

Note that the embodiments of the present invention are not limited to those shown in Examples 1 and 2 above. For example, in the above embodiment, a configuration has been described in which a pixel region with a high luminance value is selected within the divided imaging region of the image sensor 13 and image processing is performed in a state where a certain voltage V is applied to the wavelength selection filter FF. For example, if an image like that shown in Figure 9 is obtained in a divided imaging area when a certain monochromatic light is incident, only the pixels in the pixel area with high brightness (the area represented by white in the figure) are considered to be effective pixels. , the decision was made to discard information on pixels in other areas (areas shown in gray and black in the figure). However, since the pixel region with high brightness changes by changing the voltage applied to the wavelength selection filter FF, a configuration may be adopted in which image information is obtained by adding information at a plurality of voltages.

For example, by changing the voltage applied to the wavelength selection filter FF (hereinafter referred to as the applied voltage) before and after V= _V1 , the brightness value of the pixel area represented in gray when the applied voltage is _V1 is changed to the applied voltage. Suppose that the brightness value of the pixel area reaches a maximum at _V2 , and the brightness value of the pixel area, which was represented in black at the applied voltage _V1 , reaches the maximum at the applied voltage _V3 . In this case, by adding the information of the pixel area at each voltage value and calculating the brightness value of the corresponding microlens, an image of wavelength λ ₁ can be obtained from the information of the entire surface of the wavelength selection filter FF, The loss of light amount can be further reduced. In order to obtain image information of one wavelength in this way, by using multiple applied voltages and pixels (pixel areas) within the divided imaging area corresponding to the applied voltages, we can maximize light utilization efficiency while ensuring maximum light utilization efficiency. , it becomes possible to improve the half width.

Furthermore, when the applied voltage is V ₁ , the gray area is information on another wavelength λ ₂ , and the black area is information on another wavelength λ ₃ , so information on multiple wavelengths is stored for each applied voltage. It is possible to obtain. Therefore, by holding this information in a memory (not shown) or the like while sweeping the applied voltage V, and finally calculating the brightness value of each microlens by adding the information of the same wavelength, Similar results can be obtained.

Note that in the present invention, the wavelength selection filter FF is arranged in the pupil portion of the optical system 11, so that light that has passed through a certain coordinate in the pupil uniformly enters all the microlenses of the microlens array 12. Therefore, if there is a non-uniform part (unevenness) in the gap d between the reflecting surfaces of the wavelength selection filter FF, where on the filter the light passes and enters the microlens array 12 changes depending on the time. Since the timing at which the light enters each microlens is the same, there is no unevenness in the image obtained by the image sensor 13. Further, if there is dirt or a defective portion on the filter, the brightness will be reduced by the area of the dirt or defect, but no pixel defect will occur in the image obtained by the image sensor 13.

Further, according to the spectroscopic camera of the present invention, even if dust or the like is attached to the wavelength selection filter, it is possible to avoid this and photograph the subject. For example, as shown in FIG. 17, when dust GB is attached to the wavelength selection filter FF, when a screen SC illuminated with illumination light WL such as white light is photographed, all divided imaging areas of the image sensor 13 are Garbage GB will be reflected. Therefore, by selecting a pixel area while avoiding the area where the dust GB is reflected, it becomes possible to take an image while avoiding the dust GB. In addition to dust, it is also possible to deal with foreign objects whose locations cannot be predicted, such as dirt and scratches.

Furthermore, in the above embodiments, an example has been described in which the spectroscopic cameras 10 and 20 have a microlens array 12 in which n microlenses are arranged vertically and horizontally. However, the type and arrangement of lenses is not limited to this, and the spectroscopic camera only needs to have a lens array having n lenses.

In addition, in the above-mentioned Example 2, an example was explained in which the band-pass filter is provided on the surface of the exit surface of the microlens, but the position where the band-pass filter is arranged is not limited to this, and the light is focused on the microlens. It is sufficient if it is placed on the optical path of light. Moreover, it is not limited to the microlenses at both ends, but may be arranged at a position corresponding to any one of the n microlenses.

Furthermore, the series of processes described in each of the above embodiments can be performed by computer processing according to a program stored in a recording medium such as a ROM.

10, 20 Spectroscopic camera 11 Optical system 12 Microlens array 13 Image sensor 14 Camera 15 Image processing section

Claims (1)

An optical system consisting of a pair of lenses,
The optical system has a pair of reflective surfaces arranged opposite to each other at the position of the pupil, and the distance between the pair of reflective surfaces is changed according to an applied voltage to selectively emit light with a wavelength corresponding to the distance. a wavelength selection section that transmits the
a lens array having n lenses (n is a natural number) each condensing light that has passed through the optical system and the wavelength selection section;
an imaging element provided at a position conjugate with the wavelength selection section and having n divided imaging regions that receive light condensed by the n lenses;
A spectroscopic camera characterized by having.

Images