JP2015075352A - Hyper-spectral camera and program for hyper-spectral camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a hyper-spectral camera which can accelerate a scan speed without requiring complex arithmetic processing.SOLUTION: A hyper-spectral camera 1 for acquiring hyper-spectral data associated with spectral information for each pixel constituting a two-dimensional image of a subject by spectrally dividing light from a subject into a plurality of wavelength regions with a spectroscope 6 and making an image pick-up device 7 receive light for each wavelength region, includes: a scan mechanism 3 which reflects light from the subject and scans the reflected light along a prescribed direction; a correction lens 4 which is provided between the scan mechanism 3 and the spectroscope 6 and adjusts the focal position of the light scanned by the scan mechanism 3; and a slit 5 which is arranged on the focal position of the correction lens 4, makes the light incident on the spectroscope 6, and whose longer direction is formed along a direction substantially vertical to the scan direction of the scan mechanism 3.

Description

  The present invention relates to a hyperspectral camera that acquires hyperspectral data of a subject.

  Conventionally, as shown in FIG. 10, hyperspectral data obtained by associating spectral information with each pixel constituting a two-dimensional image of a subject is acquired, for example, as described in JP2011-89895A. A spectral imaging apparatus is known (Patent Document 1). Further, as a conventional technique of Patent Document 1, a stage scanning method is disclosed in which a stage mounted with a diffraction grating, a CCD, or the like is moved by a scanning mechanism (paragraph [0058] of Patent Document 1 and FIG. 12 of Patent Document 1). ).

JP 2011-89895 A

  However, in the conventional stage scanning method, the entire stage on which the diffraction grating, the CCD, etc. are mounted is moved. For this reason, there is a limit to increasing the scanning speed, and there is a problem that it takes time to acquire hyperspectral data.

  Further, in the invention described in Patent Document 1, since scanning is performed by moving the slit up and down, the imaging position of the spectral spectrum moves with the movement of the slit. For this reason, it is necessary to prepare a large number of wavelength tables indicating the correspondence between the y coordinate on the image sensor and the projected wavelength for each movement position of the slit. In addition, every time the slit is moved, the y coordinate must be converted into a wavelength using a wavelength table, which requires complicated data processing.

  The present invention has been made to solve such problems, and an object of the present invention is to provide a hyperspectral camera that can increase the scanning speed without requiring complicated arithmetic processing. .

  The hyperspectral camera according to the present invention splits light from a subject into a plurality of wavelength regions by a spectrometer, and causes each imaging region to receive light for each wavelength region, thereby forming each two-dimensional image of the subject. A hyperspectral camera that acquires hyperspectral data in which spectral information is associated with each pixel, and a scanning mechanism that reflects light from the subject and scans the reflected light along a predetermined direction; A correction lens that is provided between the scanning mechanism and the spectroscope, adjusts a focal position of light scanned by the scanning mechanism, and is disposed at the focal position of the correction lens. The light is incident and a longitudinal direction is formed along a direction substantially perpendicular to the scanning direction of the scanning mechanism. And a slit are.

  Further, as one aspect of the present invention, the scanning mechanism includes a reflection mirror that reflects the light, and a rotation driving unit that drives the reflection mirror around a rotation axis that is substantially parallel to the longitudinal direction of the slit. May be.

  Furthermore, as one aspect of the present invention, the scanning mechanism may be a MEMS (Micro Electro Mechanical Systems) mirror provided at a focal position of an objective lens on which light of the subject first enters.

  Further, as one aspect of the present invention, each time the scanning mechanism scans a distance corresponding to the width of the slit, a light source control unit that irradiates a disturbance light removing light source for removing the influence of disturbance light, and You may have a spectrum information correction | amendment part which correct | amends the said spectrum information using the spectrum of the light source for disturbance light removal.

  Furthermore, as one aspect of the present invention, in the two-dimensional image obtained by the imaging device, an edge part detection unit that detects an edge part of the subject, and a half-value width of a first derivative function of light intensity in the edge part You may have the half value width calculation part to calculate, and the lens drive part which drives an objective lens so that the said half value width may become the minimum.

  According to the present invention, it is possible to realize a hyperspectral camera capable of increasing the scan speed without requiring complicated arithmetic processing.

It is a figure which shows one Embodiment of the hyperspectral camera which concerns on this invention. It is a figure which shows other embodiment of the hyperspectral camera which concerns on this invention. It is a block diagram which shows the memory | storage means and arithmetic processing means of this embodiment. In this embodiment, it is a figure which shows the half value width of the primary differential function of light intensity. It is a flowchart figure which shows operation | movement of the hyperspectral camera of this embodiment. 2 is a photograph showing a captured image in Example 1. FIG. 3 is a graph showing changes in reflectance in Example 1. It is a graph which shows the normalization vegetation index in the present Example 1. It is a figure which shows the usage example of the hyperspectral camera in the present Example 2. FIG. It is a figure which shows an example of hyperspectral data.

  Hereinafter, an embodiment of a hyperspectral camera according to the present invention will be described with reference to the drawings.

  As shown in FIG. 1, the hyperspectral camera 1 of this embodiment mainly includes an objective lens 2 that receives light from a subject, and a scanning mechanism 3 that scans the light from the objective lens 2 along a predetermined direction. A correction lens 4 that adjusts the focal position of the light, a slit 5 that squeezes the scanned light in a straight line, a spectrometer 6 that splits the light transmitted through the slit 5 into a plurality of wavelength regions, and each wavelength region An imaging element 7 that receives each light, a disturbance light removing light source 8 that removes the influence of disturbance light, a storage unit 9 that stores various data, and an arithmetic processing unit 10 that performs various arithmetic processes. It is configured. Hereinafter, each configuration will be described in detail.

  The objective lens 2 receives light from the subject and irradiates the light toward the scanning mechanism 3. The objective lens 2 is driven in the optical axis direction by an actuator (not shown) controlled by a lens driving unit 19 to be described later, so that the subject image is automatically focused on the image sensor 7.

  The scanning mechanism 3 reflects light from the subject and scans the reflected light along a predetermined direction. In the present embodiment, as shown in FIG. 1, the scanning mechanism 3 drives a reflection mirror 31 that reflects light from the objective lens 2, and drives the reflection mirror 31 around a rotation axis that is substantially parallel to the longitudinal direction of the slit 5. Rotating drive means 32. The rotation driving unit 32 is controlled by the scanning control unit 11 described later, and drives the reflection mirror 31 to rotate.

  Note that the reflection mirror 31 and the rotation driving means 32 can be appropriately selected. For example, when a plane mirror having a mirror surface only on one side is selected as the reflection mirror 31, a linear motor, an ultrasonic motor, or the like may be adopted as the rotation drive means 32, and the plane mirror may be rotated around the rotation axis. Further, when a double-sided mirror having mirror surfaces on both sides as a reflecting mirror 31 or a multi-sided mirror having mirror surfaces on each surface of a triangular prism or a hexagonal column is selected, a servo motor with an encoder, a stepping motor, or the like is employed as the rotation driving means 32 The polyhedral mirror may be driven to rotate around the rotation axis.

  In the present embodiment, the scanning mechanism 3 is mechanically driven, but is not limited to this configuration, and may be electrically driven. For example, as shown in FIG. 2, a MEMS (Micro Electro Mechanical Systems) mirror 33 may be provided at the focal position of the objective lens 2 where the light of the subject first enters, and the mirror element may be electrically tilted. Thereby, since the mechanical scanning mechanism 3 becomes unnecessary, it is further reduced in size and weight. When the MEMS mirror 33 and the spectroscope 6 cannot be physically brought close to each other, as shown in FIG. 2, a relay lens 34 is inserted between the MEMS mirror 33 and the spectroscope 6 to ensure a clearance. preferable.

  The correction lens 4 is provided between the scanning mechanism 3 and the spectroscope 6 and adjusts the focal position of the light scanned by the scanning mechanism 3. In the present embodiment, the correction lens 4 matches the focal position of the light scanned by the scanning mechanism 3 with the slit 5.

  As shown in FIG. 1, the slit 5 is an elongated gap having a longitudinal direction along a direction substantially perpendicular to the scanning direction of the scanning mechanism 3. The slit 5 is disposed at the focal position of the correction lens 4 so that the light from the correction lens 4 is focused on a substantially horizontal linear light and is incident on the spectroscope 6. In the present embodiment, the slit 5 is formed in the slit plate 51 and is fixed to the incident portion of the spectrometer 6. However, the present invention is not limited to this configuration, and the slit 5 may be directly formed in front of the spectroscope housing.

  The spectroscope 6 separates light from a subject into a plurality of wavelength regions. The spectroscope 6 is provided with a spectroscopic optical element 61. The spectroscopic optical element 61 has, for example, a region from visible light to near-infrared light as an application range, and the region covers a plurality of bands (wavelength regions). Spectroscopy. The spectroscopic optical element 61 is not particularly limited, and may be a diffraction grating having a large number of parallel grooves, a prism that can spectrally decompose incident light, or a photoacoustic element.

  The imaging device 7 forms an image of light for each wavelength region dispersed by the spectroscope 6, photoelectrically converts light and darkness of the image light into an amount of electric charge, and outputs it as an electrical signal. The imaging element 7 is not particularly limited, and for example, a CMOS (Complementary Metal Oxide Semiconductor) image sensor, a CCD (Charge Coupled Device) image sensor, or the like can be used.

  The disturbance light removing light source 8 removes the influence of ambient disturbance light in order to acquire the accurate light intensity of the subject. Usually, when hyperspectral data is acquired under a light source that has a lot of emission line spectra, such as a fluorescent lamp, there arises a problem that only a specific wavelength is visible. For this reason, to obtain accurate hyperspectral data, the light source needs to be spectrally continuous.

Therefore, in this embodiment, in an environment where the light source is not spectrally continuous, a light source such as a strobe or LED that emits light sufficiently stronger than the disturbance light is used as the disturbance light removal light source 8. Hereinafter, a method for removing disturbance light will be described. First, pure light intensity I ₀ (λ) from a subject when there is no disturbance light (noise) is expressed by the following formula (1) using reflectance R (λ) and spectral intensity L (λ) of a continuous spectrum light source: 1).
I ₀ (λ) = R (λ) · L (λ) Formula (1)

However, for example, NDVI (Normalized Difference Vegetation Index), which is an index indicating the distribution status and activity of vegetation represented by the following formula (2), is weighted as L (λ) when the light source changes. Will change and the numbers will be different.
NDVI = (I ₀ (λ ₂ ) −I ₀ (λ ₁ )) / (I ₀ (λ ₂ ) + I ₀ (λ ₁ ))
= (R (λ ₂ ) · L (λ ₂ ) −R (λ ₁ ) · L (λ ₁ ))
/ (R (λ ₂ ) · L (λ ₂ ) + R (λ ₁ ) · L (λ ₁ ))
= (R (λ ₂ ) −R (λ ₁ ) · α) / (R (λ ₂ ) + R (λ ₁ ) · α) (2)
Here, α = L (λ ₁ ) / L (λ ₂ )

On the other hand, when the disturbance light (noise) is N (λ), the light intensity I (λ) received by the image sensor 7 is expressed by the following formula (3) using the above I ₀ (λ).
I (λ) = I ₀ (λ) + N (λ) (3)
However, the value of I ₀ (λ) = I (λ) −N (λ) derived from the above equation (3) is generally inaccurate. Therefore, in the present embodiment, by imposing the condition of I ₀ (λ) >> N (λ) by the disturbance light removing light source 8, it is assumed that I (λ) ≈I ₀ (λ), and the influence of disturbance light is eliminated. It is supposed to be.

  The storage unit 9 stores various data and functions as a working area when the arithmetic processing unit 10 performs arithmetic processing. In the present embodiment, the storage means 9 is composed of a ROM (Read Only Memory), a RAM (Random Access Memory), a flash memory, etc. As shown in FIG. 3, a program storage unit 91, a wavelength table storage unit, and the like. 92 and a spectral spectrum storage unit 93. Hereinafter, each component will be described in more detail.

  The program storage unit 91 is installed with a hyperspectral camera program 1a for controlling the hyperspectral camera 1 of the present embodiment. And the arithmetic processing means 10 functions as each structure part mentioned later by running the said program 1a for hyperspectral cameras.

  The wavelength table storage unit 92 stores a wavelength table indicating wavelengths corresponding to y-coordinates (coordinates in the scanning direction) on the image sensor 7. In the present embodiment, the slit 5 is fixed, and the positional relationship inside the spectroscope 6 does not change, so that the imaging region of the spectral spectrum formed on the image sensor 7 moves in accordance with the scanning operation. There is no. For this reason, it is only necessary to prepare one wavelength table corresponding to the position of the slit 5 in the wavelength table storage unit 92.

  The spectral spectrum storage unit 93 stores the spectral spectrum of the disturbance light removing light source 8. In this embodiment, before acquiring hyperspectral data, the spectral spectrum of the disturbance light removing light source 8 is measured in advance by the hyperspectral camera 1. A spectrum information correction unit 16 to be described later calculates the spectral reflectance of the subject from the spectral spectrum.

  The arithmetic processing means 10 calculates hyperspectral data based on the electrical signal acquired from the image sensor 7. In the present embodiment, the arithmetic processing means 10 is composed of a CPU (Central Processing Unit), an FPGA (Field Programmable Gate Array), and the like, and executes the hyperspectral camera program 1a installed in the storage means 9. Accordingly, as shown in FIG. 3, the scanning control unit 11, the image data acquisition unit 12, the spectrum information acquisition unit 13, and the HSD (hyperspectral data) acquisition unit 14 function.

  As an optional function, the arithmetic processing means 10 functions as a light source control unit 15 and a spectrum information correction unit 16 to realize a disturbance light removal function, and includes an edge portion detection unit 17 and a half-value width calculation unit 18. By functioning as the lens driving unit 19, an autofocus function is realized. Hereinafter, each component will be described in more detail.

  The scanning control unit 11 controls the rotational driving means 32 of the scanning mechanism 3 to rotationally drive the reflection mirror 31. For example, when a servo motor with an encoder is used as the rotation driving means 32, the scanning control unit 11 outputs a pulse signal while controlling the rotation angle by the encoder and controls the motor to drive the entire subject in a predetermined direction. It is designed to scan.

  The image data acquisition unit 12 acquires image data based on the electrical signal output from the image sensor 7. In this embodiment, when the longitudinal direction of the slit 5 is the x direction and the direction perpendicular to the x direction is the y direction, the light that is linearly narrowed in the x direction by the slit 5 is reflected in the y direction by the spectroscope 6. Spectroscopically forms an image. Therefore, the image data acquisition unit 12 acquires a two-dimensional image composed of a plurality of pixels having (x, y) components.

  The spectrum information acquisition unit 13 acquires spectrum information for each pixel constituting the two-dimensional image. In the present embodiment, the spectrum information acquisition unit 13 calculates the position (x, y) of each pixel in the two-dimensional image acquired by the image data acquisition unit 12 and refers to the wavelength table in the wavelength table storage unit 92. The wavelength λ corresponding to each y coordinate is acquired.

  The HSD acquisition unit 14 acquires hyperspectral data. In the present embodiment, the HSD acquisition unit 14 associates the wavelength λ acquired by the spectrum information acquisition unit 13 with the position (x, y) of each pixel constituting the two-dimensional image acquired by the image data acquisition unit 12. Hyperspectral data that is a data set of (x, y, λ) of each pixel is acquired.

  The light source control unit 15 controls the irradiation of the disturbance light removing light source 8. Specifically, the light source control unit 15 synchronizes with the timing of acquiring image data for one line that has passed through the slit 5, that is, every time the scanning mechanism 3 scans a distance corresponding to the width of the slit 5. Then, control is performed so that the disturbance light source 8 is irradiated. For example, when a servo motor with an encoder is used as the rotation driving means 32, the light source control unit 15 monitors the integrated value of the pulse signal output from the scanning control unit 11 to the servo motor, and the integrated value is equivalent to one line. A trigger pulse is output to the disturbance light removing light source 8 every time the number of pulses is equal to.

  The spectrum information correction unit 16 corrects the spectrum information to a more accurate value. In the present embodiment, the spectrum information correction unit 16 acquires the spectral spectrum of the disturbance light removing light source 8 from the spectral spectrum storage unit 93 and calculates the spectral reflectance of the subject. Then, by using the spectral reflectance, the spectral information is corrected so that the light intensity is constant for each wavelength.

  The edge part detection unit 17 detects the edge part of the subject in the two-dimensional image obtained by the image sensor 7. Specifically, the edge portion detection unit 17 specifies a portion where the brightness of the two-dimensional image changes sharply or discontinuously. For example, the edge portion detection unit 17 calculates a Laplacian (secondary derivative) of the luminance value for each pixel, and detects a position that becomes zero at the positive / negative boundary as an edge. The edge detection method is not particularly limited, and a canny method or the like may be applied.

  In the present embodiment, when detecting the edge portion, the scanning control unit 11 roughly scans the subject by the scanning mechanism 3, and the image data acquisition unit 12 acquires the image data for edge detection. Yes.

  The half-value width calculation unit 18 calculates the half-value width of the first derivative function of the light intensity at the edge portion of the two-dimensional image. In the present embodiment, the full width at half maximum calculation unit 18 differentiates the light intensity of the 0th-order diffracted light that has been dispersed on the image sensor 7 at the edge portion detected by the edge portion detection unit 17, and as shown in FIG. The half-value width (pixel width at half the peak value) in the first-order differential function is calculated.

  The lens driving unit 19 drives the objective lens 2 so that the subject is in focus. In the present embodiment, the lens driving unit 19 acquires the half-value width calculated by the half-value width calculation unit 18 in real time while driving the objective lens 2, and feedback-controls the objective lens 2 so that the half-value width is minimized. To do.

  In the present embodiment, 0th-order diffracted light that is not diffracted by the spectroscopic optical element 61 is used, but the present invention is not limited to this, and 1st-order diffracted light including spectral spectrum information may be used. In general, since the light intensity of the first-order diffracted light continuously changes in the spectral direction, it is difficult to confirm whether or not it is in focus. However, in this embodiment, since the focus is set using the sharpness of the edge as an index, even first-order diffracted light can be used without any problem.

  Next, the operation of the hyperspectral camera 1 of this embodiment will be described with reference to FIG.

  First, when acquiring hyperspectral data of a subject using the hyperspectral camera 1 of the present embodiment, the scanning control unit 11 scans the subject with low accuracy by the scanning mechanism 3 (step S1), and the image data acquiring unit 12 Image data for autofocus is acquired (step S2). This is because the focus processing speed is improved by lowering the scanning accuracy, and real-time autofocus is realized.

  Next, when the edge portion detection unit 17 detects the edge portion of the subject in the image obtained in step S2 (step S3), the half-value width calculation unit 18 uses the half-value width of the first derivative function of the light intensity at the edge portion. Is calculated (step S4). Then, the lens driving unit 19 performs feedback control of the objective lens 2 so that the half width is minimized (step S5). As a result, it is possible to perform autofocus without preparing an autofocus camera or the like separately.

  When the subject is in focus in step S5, the scanning control unit 11 starts the scanning mechanism 3 (step S6). At this time, in the present embodiment, since the reflection mirror 31 in the front stage of the spectroscope 6 changes the direction of the line of sight of the light, the focus surface of the telecentric optical system moves in parallel with the slit 5, and scanning of the subject is performed. It becomes possible.

  Further, in this embodiment, unlike the conventional stage scanning method, the entire spectroscope 6 is not moved, and only the light reflecting mirror 31 is rotated, so that the scanning speed is increased. In particular, if the hyperspectral camera 1 is reduced in size and weight, the weight advantage of driving only the reflecting mirror 31 compared to moving the entire spectrometer 6 as in the conventional stage scanning method. Becomes larger.

  Next, at the same time when the light source control unit 15 irradiates the disturbance light source 8 (step S7), the image data acquisition unit 12 acquires image data (step S8). At this time, the disturbance light removing light source 8 emits light sufficiently stronger than the disturbance light, so that the influence of the disturbance light on the light intensity of the subject is removed.

  Next, the spectrum information acquisition unit 13 calculates the position (x, y) of each pixel in the two-dimensional image obtained in step S8, and acquires the wavelength λ corresponding to each y coordinate from the wavelength table storage unit 92. (Step S9). At this time, in this embodiment, since the slit 5 is not moved, the positional relationship inside the spectroscope 6 is not changed. For this reason, it is not necessary to prepare a wavelength table for each movement position of the slit 5, and conversion processing from the y coordinate to the wavelength is easy.

  Moreover, in this embodiment, the spectrum information correction | amendment part 16 acquires the spectral spectrum of the light source 8 for disturbance light removal from the spectral spectrum memory | storage part 93, calculates a spectral reflectance, and uses the said spectral reflectance and spectral information is calculated | required. Correction is performed (step S10). Thereby, more accurate spectral information can be obtained under any light source environment.

  Subsequently, the HSD acquisition unit 14 associates the wavelength acquired by the spectrum information acquisition unit 13 with each pixel constituting the two-dimensional image acquired by the image data acquisition unit 12, and sets (x, y) for each pixel. , Λ) is acquired (step S11). Thereby, hyperspectral data for one line in the subject image that has passed through the slit 5 is accumulated in the storage means 9.

  While scanning the subject, the light source control unit 15 monitors whether or not the integrated value of the pulse signal from the scanning control unit 11 has reached the number of pulses corresponding to one line (step S12). Then, only when the integrated value becomes the number of pulses corresponding to one line (step S12: YES), the scanning control unit 11 determines whether or not the scanning of the subject is completed (step S13). Then, as long as the scan is not completed (step S13: NO), the processing of step S7 to step S12 described above is looped. Thereby, complete hyperspectral data regarding the subject is acquired.

According to the present embodiment as described above, the following effects can be obtained.
1. Scanning can be performed only by rotating the reflection mirror 31 without moving the entire spectroscope 6, and the scanning speed can be increased.
2. Since the positional relationship inside the spectroscope 6 does not change, the arithmetic processing for acquiring the spectrum information can be simplified.
3. The influence of disturbance light can be removed and accurate hyperspectral data can be acquired.
The autofocus function can be realized using not only the 4.0th-order diffracted light but also the first-order diffracted light.
5. By making the scanning mechanism 3 smaller and lighter, it is possible to realize an ultra-small hyperspectral camera 1 module that can be incorporated into a portable terminal, an industrial robot, or the like.
6). By fusing hyperspectral data with digital camera images, phenomena that are invisible to ordinary people, such as freshness of vegetables and the eyes of craftsmen, can be visualized and digitized.
can do.

  Next, specific examples of the hyperspectral camera 1 according to the present invention will be described. Note that the technical scope of the present invention is not limited to the features shown by the following embodiments.

  In Example 1, an experiment was performed in which a leafy vegetable was photographed using the hyperspectral camera 1 and its freshness was measured. Specifically, a 250 W halogen lamp was applied to spinach for 10 hours and the freshness was deteriorated, and the reflection spectrum was measured with the hyperspectral camera 1 every hour. The captured images of the hyperspectral camera 1 and the digital camera at this time are shown in FIG. 6, and the change in reflectance at each wavelength is shown in FIG.

  As shown in FIGS. 6 and 7, the ratio of red (620 to 750 nm) to green (495 to 570 nm) increases with the passage of time, so it can be seen that the apparent color gradually fades to yellow. . In the absorption band (680 nm) of chlorophyll (chlorophyll), the reflectance increases with time. That is, since it can be said that the amount of light absorption necessary for photosynthesis is decreasing, it can be said that the photosynthesis capability is reduced.

  Leafy vegetables have the property of absorbing the wavelength (R) in the red region and reflecting the near-infrared wavelength region (IR). This is related to chlorophyll contained in the leaves of plants. By applying these two wavelength ranges to the above-described NDVI (normalized vegetation index), an index NDVI [= (IR−R) / (IR + R)] indicating the vegetation distribution state and activity is obtained. In the present Example 1, the graph of the NDVI value with time passage is shown in FIG.

As shown in FIG. 8, the NDVI of leafy vegetables shows a tendency to averagely decrease over time while fluctuating up and down, and to rapidly decrease from a certain critical point (= NDVI _min ). For this reason, below NDVI _min , the life activity as a plant will be stopped. When the maximum value of NDVI that can be taken as a plant is NDVI _max , the freshness value of leafy vegetables can be expressed by the following formula (4).

Freshness value = [(NDVI−NDVI _min ) / (NDVI _max −NDVI _min )] × 100 Equation (4)
However, NDVI _max and NDVI _min depend on the type of leafy vegetable and must be determined individually. By measuring various leafy vegetables, a dedicated freshness value can be created.

  From the above Example 1, it was shown that the hyperspectral camera 1 according to the present invention can visualize and digitize the freshness of leafy vegetables.

  In the second embodiment, the hyperspectral camera 1 is mounted on a portable terminal, while the analysis processing is separately performed at the analysis center. Specifically, the wavelength table storage unit 92, the spectrum information acquisition unit 13, and the HSD acquisition unit 14 described above are provided on the analysis center side. As a specific usage method, as shown in FIG. 9, the analysis application is downloaded from the analysis center to the mobile terminal, and the subject is photographed by the hyperspectral camera 1.

  As a result, the analysis application selects a band, transmits data corresponding to the band to the analysis center for analysis, receives a spectrum image from the analysis center as an analysis result, and fuses the spectrum image with the digital camera image. A series of processes for displaying the fusion result is automatically executed.

  According to the second embodiment described above, it is possible to incorporate the hyperspectral camera 1 according to the present invention by causing the analysis center to perform an analysis part that requires arithmetic processing even in a portable terminal having weak arithmetic processing capability. Indicated.

Note that the hyperspectral camera 1 according to the present invention is not limited to the above-described embodiments, and is expected to be applied to various fields as described below.
(1) Application to the medical field (visualization of malignant tumors, etc.)
(2) Application in the field of remote sensing (observation of vegetation from helicopters, etc.)
(3) Application in the bio field (observation through connection with a microscope, etc.)
(4) Application to the industrial field (electronic circuit wiring check, etc.)
(5) Application to food field (visualization of freshness of meat and fish)
(6) Application in the cosmetics field (facial stains, visualization of skin moisture, etc.)
(7) Application in the security field (detection of intruders, etc.)
(8) Application to agriculture (growth monitoring, flowering / harvest time prediction, etc.)
(9) Application to the industrial field (inspection in conjunction with POS system, foreign matter / fatigue inspection, etc.)

DESCRIPTION OF SYMBOLS 1 Hyperspectral camera 1a Hyperspectral camera program 2 Objective lens 3 Scan mechanism 4 Correction lens 5 Slit 6 Spectroscope 7 Imaging device 8 Light source for disturbance light removal 9 Storage means 10 Arithmetic processing means 11 Scan control part 12 Image data acquisition part 13 Spectral information acquisition unit 14 HSD acquisition unit 15 Light source control unit 16 Spectral information correction unit 17 Edge portion detection unit 18 Half width calculation unit 19 Lens drive unit 31 Reflection mirror 32 Rotation drive unit 33 MEMS mirror 34 Relay lens 61 Spectral optical element 91 Program Storage unit 92 Wavelength table storage unit 93 Spectral spectrum storage unit

Claims (5)

Spectral information is associated with each pixel constituting the two-dimensional image of the subject by dispersing the light from the subject into a plurality of wavelength regions by a spectroscope and causing the image sensor to receive light for each wavelength region. A hyperspectral camera that acquires hyperspectral data
A scanning mechanism that reflects light from the subject and scans the reflected light along a predetermined direction;
A correction lens that is provided between the scanning mechanism and the spectroscope and adjusts the focal position of the light scanned by the scanning mechanism;
And a hyper slit that is disposed at the focal position of the correction lens and causes the light to enter the spectroscope and has a longitudinal direction formed in a direction substantially perpendicular to the scanning direction of the scanning mechanism. Spectrum camera.   2. The hyperspectrum according to claim 1, wherein the scanning mechanism includes a reflection mirror that reflects the light, and a rotation driving unit that drives the reflection mirror around a rotation axis substantially parallel to a longitudinal direction of the slit. camera.   2. The hyperspectral camera according to claim 1, wherein the scanning mechanism is a micro electro mechanical systems (MEMS) mirror provided at a focal position of an objective lens on which light of the subject first enters. 3. A light source controller that irradiates a disturbance light removing light source for removing the influence of disturbance light every time the scanning mechanism scans a distance corresponding to the width of the slit;
The hyperspectral camera according to any one of claims 1 to 3, further comprising: a spectrum information correction unit that corrects the spectrum information using a spectrum of the disturbance light removing light source. An edge part detection unit for detecting an edge part of the subject in the two-dimensional image obtained by the imaging device;
A full width at half maximum for calculating a full width at half maximum of a first derivative function of light intensity at the edge portion;
The hyperspectral camera according to claim 1, further comprising: a lens driving unit that drives the objective lens so that the half-width is minimized.