Identification device, classification method

Abstract

To provide an identification apparatus that can provide a high sensitivity identification apparatus in which an imaging unit may not excessively limit an SNR of an effective output signal corresponding to a projected spectral image.SOLUTION: An identification apparatus comprises: an imaging unit that comprises a plurality of light collecting optical systems, a spectroscopic element which separates a plurality of pieces of light from the plurality of light collecting optical systems, and a plurality of light receiving elements which are arranged in a row direction and a column direction, and in which a plurality of spectroscopic spectra from the spectroscopic element are projected along the row direction; and an acquisition unit that acquires at least any spectroscopic information of a plurality of specimens on the basis of an output signal from the imaging unit. The identification apparatus comprises a sensitization processing unit that performs row direction binning processing which integrates output signals of the plurality of light receiving elements existing at different positions along the row direction.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention , knowledge Separate device , and sorting method Regarding. [Background technology]

[0002] Identification devices that optically identify the properties of specimens using spectroscopic analysis are known. These identification devices are placed along a transport path along which multiple specimens are transported, and are used for product inspection, waste sorting, and the like.

[0003] Spectroscopic analysis does not necessarily require atmospheric control processes such as vacuum reduction, atmospheric control, immersion in liquid, and drying, which limit throughput, and can identify the properties of samples in ambient air. In recent years, attempts have been made to apply this method to the separation of waste resins.

[0004] Known spectroscopic analyses include infrared absorption spectroscopy, which obtains the absorption spectrum of a specimen for incident light including the infrared wavelength range, and Raman scattering spectroscopy, which obtains the scattering spectrum of a specimen for incident light including the ultraviolet wavelength range.Raman scattering spectroscopy is used to identify waste specimens with varying sizes because it is less affected by light attenuation due to sample thickness.

[0005] Non-Patent Document 1 discloses an identification device for waste resin, which includes an identification device having multiple sets of a light collection optical system, a spectroscopic element, and a two-dimensional imaging unit. The identification device described in Non-Patent Document 1 further discloses an electrical sensitization process that performs binning on a group of light receiving elements, on which a spectral image is projected, in a row direction that intersects with the spectrum, to increase the output signal of Raman scattered light.

[0006] Patent Document 1 discloses an identification device equipped with multiple light-collecting optical systems, a spectroscopic element, and a two-dimensional imaging unit. The identification device of Patent Document 1 discloses that the device is made smaller by consolidating the devices subsequent to the multiple light-collecting optical systems into one spectroscopic element and one two-dimensional imaging unit. The identification device of Patent Document 1 also discloses that the two-dimensional imaging unit is a rolling shutter type CMOS image sensor, thereby speeding up the identification process. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Application Publication No. 2019-105628 [Non-patent literature]

[0008] [Non-Patent Document 1] BUNSEKI KAGAKU Vol.61,No.12,pp.1027-1032(2012) Summary of the Invention [Problem to be solved by the invention]

[0009] In the identification device described in Patent Document 1, the effective sensitivity of the detection optical system is limited by the sensitivity characteristics of the imaging element and the amount of light irradiated onto each imaging element. In the identification device described in Patent Document 2, the projection distance from the spectroscopic element to the two-dimensional imaging unit can be adjusted to adjust the resolution in the wavelength direction and the number of scanning lines of the light-receiving element onto which one light-collecting channel is projected. On the other hand, in the identification device described in Patent Document 1, increasing the projection distance can reduce the amount of light per light-receiving element of the projected spectral image, resulting in an insufficient signal-to-noise ratio (SNR).

[0010] Furthermore, although the identification device described in Non-Patent Document 1 performs binning processing in the projection width direction of the spectral image, the sensitivity is insufficient, and there is concern that the SNR for weak scattered light in the high wavenumber range will be insufficient, limiting the identification throughput.

[0011] An object of the present invention is to provide a highly sensitive identification device in which the signal-to-noise ratio of the output signal corresponding to the projected spectral image is not excessively limited by the imaging unit. [Means for solving the problem]

[0012] An identification device according to an embodiment of the present invention comprises: Multiple a color information acquisition unit for acquiring color information of a sample; and Multiple a plurality of light-collecting optical systems that collect scattered light from specimens; a spectroscopic element that disperses the plurality of light beams from the plurality of light-collecting optical systems; an imaging unit that includes a plurality of light-receiving elements arranged in row and column directions, and onto which a plurality of spectral spectra from the spectroscopic elements are projected along the row direction; and an acquisition unit that acquires spectral information of at least any of the plurality of specimens based on an output signal from the imaging unit. 、 a sensitization processing unit that performs row-direction binning processing to integrate output signals from a plurality of the light receiving elements at different positions along the row direction, the sensitization processing unit Based on the color information of the plurality of samples acquired by the color information acquisition unit, The number of binnings of the light receiving elements in the row direction binning process is made different for a first sample having first color information and a second sample having second color information different from the first color information. [Effects of the Invention]

[0013] According to the present invention, it is possible to provide a classification device that can reduce the influence of background noise components in an imaging unit that detects a projected spectral image and acquire an output signal corresponding to the spectral image with high sensitivity. Also, the classification device of the present invention makes it possible to perform classification processing with high throughput. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 1 is a schematic diagram illustrating the configuration of an identification device (a), a spectroscopic imaging system (b), and an imaging unit (c) according to a first embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of a lighting unit. [Figure 3] FIG. 2A is a diagram illustrating the relationship between the light receiving elements of the imaging unit and a monochromatic light projected image according to the first embodiment, and FIG. 2B is a diagram illustrating the operating conditions of binning. [Figure 4] 5A and 5B are diagrams showing the relationship between the light amount distribution of a monochromatic light projected image and the beam diameters Φ(a) and 2Φ(b) according to the first embodiment. [Figure 5] 10A, 10B, and 10C show examples of projected images of an imaging unit in the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0016] Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, within the scope of the present invention, appropriate modifications and improvements to the following embodiments based on the ordinary knowledge of those skilled in the art are also included in the scope of the present invention, provided that they do not deviate from the spirit of the present invention.

[0017] (First embodiment) An identification device 1 according to a first embodiment of the present invention will be described with reference to Figures 1(a), (b), (c) and Figure 2. Figure 1(a) is a diagram schematically showing the configuration of the identification device 1 according to the first embodiment.

[0018] As shown in FIG. 1(a), the identification device 1 includes a plurality of light-collecting optical units 1103i that collect scattered light from a plurality of specimens 190 (190i, i = a, b, c, etc.) and a spectroscopic element 133 that disperses the plurality of light beams from the plurality of light-collecting optical units 1103i. The identification device 1 further includes an image capture unit 135 that includes a plurality of light-receiving elements 1351 arranged in a row direction 135r and a column direction 135c, and onto which a plurality of spectral spectra 300 from the spectroscopic element 133 are projected along the row direction 135r. The identification device 1 further includes an acquisition unit 140 that acquires spectral information of any of the plurality of specimens 190 based on an output signal from the image capture unit 135, and a sensitization processor 140. The sensitization processor 140 is a characteristic feature of the present invention and performs row-direction binning processing that integrates output signals from a plurality of light-receiving elements 1351 located at different positions along the row direction 135r.

[0019] As shown in FIG. 1( a), the identification device 1 includes an identification unit 100 that optically identifies multiple specimens 190 placed at predetermined positions, and a classification unit 160 that classifies the multiple specimens 190 based on identification information from the identification unit 100. The identification unit 100 includes a placement unit 150 that places the multiple specimens 190 at predetermined positions, and a light-collecting optical unit 1103i that collects scattered light from the placed specimens 190i. The identification unit 100 further includes a spectroscopic imaging unit 130 that includes a spectroscopic element 133 and an imaging unit 135 to acquire a spectral image 300 of the collected light, an acquisition unit 140 that acquires spectral information of the specimens 190 from the spectroscopic imaging unit 130, and a sensitization processing unit 140. In the identification device 1 of this embodiment, the sensitization processing unit 140 is disposed between the imaging unit 135 and the acquisition unit 140 to perform sensitization processing on the output signal from the imaging unit 135 before passing it to the acquisition unit 140. The sensitization processing unit 140 may be implemented in a form other than between the imaging unit 135 and the acquisition unit 140, such that it is included in a signal output unit (not shown) provided in the imaging unit 135, or in a signal input unit (not shown) provided in the acquisition unit 140.

[0020] (Placement section) The specimens 190 move in a scattered state on a belt conveyor 152 of the mounting unit 150 via a supply means such as a vibrating feeder. As shown in FIG. 1(a), the mounting unit 150 includes a belt conveyor 152 that transports a plurality of specimens in a predetermined direction, and a transport control unit 151 that controls the belt conveyor 152. The predetermined direction 220 may be referred to as a transport direction 220. The mounting unit 150 mounts the specimens 190i in an effective light-collecting area of ​​a light-collecting optical unit 1103i, which will be described later. The specimens 190 mounted in the effective light-collecting area are moved outside the effective light-collecting area 190i by the mounting unit 150. In other words, the mounting unit 150, which includes the belt conveyor 152, transports the specimens 190i from the upstream side to the downstream side of the effective light-collecting area 109i in the transport direction 220. The belt conveyor 152 may be replaced with a roller conveyor, a transport path that is vibrated by standing waves to move the specimen in a predetermined direction, or a combination thereof. The mounting unit 150 includes a mechanism that moves the specimen 190 placed in the effective light-collecting area to outside the effective light-collecting area.

[0021] (Separation department) 1(a), the sorting unit 160 performs a sorting operation of selectively storing a plurality of specimens 190 moved by the mounting unit 150 in a collection compartment of a sorting basket 163 based on the identification information from the identification unit 100. The sorting unit 160 includes an air gun 162 that sprays air at a predetermined timing in a sorting zone where specimens move through the air, an air gun control unit 161 that commands the air gun 162 to spray air, and a sorting basket 163. The identification device 1 includes the sorting unit 160 that performs an identification process to identify the properties of the specimens 190 and sorts the specimens 190 according to the identification results, and may be referred to as a specimen sorting system.

[0022] (Spectroscopic Imaging Unit) As shown in FIG. 1( a), the identification device 1 uses Raman scattering spectroscopy as a spectroscopic process for secondary light scattered from the specimen, and therefore the light collection optical system 90 is disposed in an epi-illumination arrangement with respect to the mounting unit 150. That is, in the light collection optical system 90 according to this embodiment, the irradiation system (primary light) and the light collection system (secondary light) are disposed on the same side of the specimen 190 in order to collect Raman scattered light. In other words, the irradiation system (primary light) and the light collection system (secondary light) are disposed on the same side of the mounting unit 150. In a modified embodiment of this embodiment in which the identification device uses infrared spectroscopy, the light collection optical system is disposed in a transmissive arrangement with respect to the mounting unit 150. That is, in a configuration in which infrared spectroscopy is used, the irradiation system (primary light) and the light collection system (secondary light) are disposed on opposite sides of the specimen 190.

[0023] 1(a) and 1(b), the light collection optical system 90 is configured to collect Raman scattered light from the specimen 190 and guide it to the spectroscopic imaging section 130. The light collection optical system 90 has a plurality of light collection optical units 1103i arranged at different positions across the conveyance width of the belt conveyor 152 along which the specimen 190i is transported. Each light collection optical unit 1103i (i = a, b, c, ... e) includes a light collection section 110i and a light guide section 120i (i = a, b, c, ... e) so as to simultaneously collect Raman scattered light from specimens placed discretely in the conveyance width direction of the belt conveyor 152. In other words, the light collecting optical unit 1103i (i = a, b, c, ... e) is equipped with a light collecting section 110i and a light guiding section 120i (i = a, b, c, ... e) so as to collect Raman scattered light in parallel from specimens placed discretely in the conveying width direction of the belt conveyor 152.

[0024] The light guiding unit 120i guides the light collected from the light collecting unit 110i to the spectroscopic element 133 included in the spectroscopic imaging unit 130. The multiple light guiding units 120i (i = a, b, c, ... e) each have an output optical end 120oei (i = a, b, c, ... e) on the side of the spectroscopic element 133. The output optical ends 120oei (i = a, b, c, ... e) are bundled and aligned in a row in a direction perpendicular to the plane of FIG. 1(b). The multiple Raman scattered light beams guided from the multiple light guiding units 120i (i = a, b, c, ... e) enter the spectroscopic element 133 from the output optical ends 120oei (i = a, b, c, ... e) in a row in a direction perpendicular to the plane of FIG. 1(b). The multiple Raman scattered lights are separated by the spectroscopic element 133 and projected in a direction perpendicular to the arrangement direction of the output optical ends 120oei (i=a, b, c, . . . e) of the light-guiding sections 120i (i=a, b, c, . . . e), i.e., in a direction parallel to the plane of the paper in Figure 1(b).

[0025] The spectroscopic imaging unit 130 includes a spectroscopic element 133 that disperses the Raman scattered light collected by the light collection optical system 90, and an imaging unit 135 that receives the Raman scattered light dispersed by the spectroscopic element, and disperses the Raman scattered light to generate a spectral signal. The spectroscopic imaging unit 130 includes an imaging lens 131, a long-pass filter 132, a spectroscopic element 133 such as a diffraction grating, an imaging lens 134, and the imaging unit 135.

[0026] The imaging lens 131 collimates the light from the optical fiber 121. The long-pass filter 132 is disposed between the imaging lens 131 and the spectroscopic element 133, and removes the remaining excitation light component, allowing only the Raman scattered light to pass through.

[0027] The spectroscopic element 133 disperses the Raman scattered light collected by the light collection optical system 90 and one-dimensionally disperses the Raman scattered light by wavelength. The imaging lens 134 forms an image of the light dispersed by the spectroscopic element 133 on the imaging unit 135. As a result, a line-shaped spectral image is projected on the light receiving surface of the imaging unit 135, as shown in FIG. 1(c). Note that the optical arrangement and spectroscopic method of the components of the spectroscopic imaging unit 130 may be changed to other commonly used forms, such as a Rowland arrangement or a Czerny-Turner system, as appropriate.

[0028] The long-pass filter 132 is disposed between the imaging lens 131 and the spectroscopic element 133, and removes the remaining excitation light component and transmits the Raman scattered light.

[0029] (imaging unit) The imaging unit 135 receives the multiple Raman scattered light beams one-dimensionally dispersed by the spectroscopic element 133 and converts them into electrical signals. The imaging unit 135 is an area image sensor in which pixels, each including a photoelectric conversion element, are two-dimensionally arranged along a direction parallel to the plane of the paper in FIG. 1(c). On the light receiving surface of the imaging unit 135, multiple spectral images, projected by the Raman scattered light beams guided and dispersed by each of the light guides 120i (i = a, b, c, ... e), are arranged along a column direction 135c in FIG. 1(c). Furthermore, on the light receiving surface of the imaging unit 135, the wavelength components of each spectral image are projected along a row direction 135r.

[0030] As shown in FIG. 1( c), the imaging unit 135 has a pixel unit 1351 in which pixels, each including a photoelectric conversion element, are arranged in a matrix. Raman scattered light, which is guided by each of the multiple optical fibers 121 constituting the optical fiber unit 120 and dispersed by the spectroscopic element 133, is focused on the light receiving surface of the imaging unit 135, and multiple spectral images 300 are projected. For ease of explanation, the intensity of light in the spectral images is indicated by dotted lines to make it easier to understand. The spectral images 300 are converted into electrical signals by the imaging unit 135 and output to the acquisition unit 140 as light intensity information for each wavelength, i.e., spectral data. The acquisition unit 140 controls the separation operation of the separation unit 160 based on the spectral information.

[0031] In other words, the pixel section 1351 defines the effective imaging area of ​​the imaging section 135. The pixel section 1351 has pixels arranged two-dimensionally and can capture a two-dimensional image by being scanned using a combination of main scanning and sub-scanning. In this specification, sub-scanning is performed in a direction intersecting the main scanning direction and at a lower frequency than the main scanning. Generally, when the number of pixel rows included in the pixel section 1351 and sub-scanned is N, the main scanning is performed at a frequency that is N times or more the sub-scanning frequency.

[0032] Here, the identification device 1 identifies the type of sample 190 while transporting the sample 190 using the mounting unit 150, and sorts the sample 190 using the identification device 1 (described later) based on the identification results. Therefore, in order to increase the throughput of the sorting process by the identification device 1, it is preferable to increase the transport speed of the sample 190 using the mounting unit 150. The spectral image projected on the imaging unit 135 is formed by Raman scattered light generated from the sample 190 moving on the transport surface of the mounting unit 150. Therefore, the spectral image is formed on the imaging unit 135 while the transported sample 190 is present within the detectable area of ​​the light-collecting optical system 90. For example, if the transport speed of the mounting unit 150 is 2 m / s and the size of the sample 190 is 10 mm, the time required for the imaging unit 135 to detect the spectral image formed by the Raman scattered light generated from the sample 190 is 5 milliseconds or less. Therefore, the imaging unit 135 is required to have a high frame rate. An example of such a high frame rate imaging unit is a CMOS image sensor, and therefore a CMOS image sensor is preferable for the imaging unit 135.

[0033] Furthermore, as described above, the intensity of the Raman scattered light generated from the specimen 190 is extremely weak, and therefore the intensity of the light incident on each pixel of the pixel unit 1351 of the image capturing unit 135 is also extremely weak. Therefore, it is preferable to use an image capturing unit 135 that has high sensitivity in the wavelength region in which the spectral image 300 is acquired. In general, compared to global shutter image sensors, rolling shutter image sensors have a simpler pixel structure, a higher aperture ratio, and larger photoelectric conversion elements, thereby improving sensitivity and dynamic range. Furthermore, due to the simpler pixel structure, rolling shutter image sensors also have the advantage of being less expensive than global shutter image sensors. For these reasons, in this embodiment, a rolling shutter CMOS image sensor is used as the image capturing unit 135.

[0034] The imaging unit 135 is preferably a rolling reset type image sensor that sequentially resets each pixel row, thereby making it possible to extend the exposure time of each pixel row as long as possible and to increase sensitivity.

[0035] The imaging unit 135 preferably has a crop readout function for performing a readout operation on a specific pixel row in the pixel unit 1352. This makes it possible to perform a readout operation on the pixel row corresponding to the light-collecting optical system 90 when, for example, another detection means detects that the specimen 190 has reached the detectable region of the light-collecting optical system 90.

[0036] The imaging unit 135 includes a readout circuit 1353, a horizontal scanning circuit 1354, a vertical scanning circuit 1355, and an output circuit 1356, and sequentially reads out signals from a plurality of pixels arranged in a matrix, row by row. The vertical scanning circuit 1355 selects and drives an arbitrary pixel row in the pixel unit 1351. The readout circuit 207 reads out signals output from pixels in the row selected by the vertical scanning circuit 1355 and transfers them to the output circuit 1356 under the control of the horizontal scanning circuit 1354. This performs readout in the main scanning direction (row direction). Furthermore, the vertical scanning circuit 1355 shifts the row selected, and the readout circuit 1353 performs readout in the main scanning direction under the control of the horizontal scanning circuit 1354. By repeating this process and shifting the selected pixel row in the sub-scanning direction (column direction), signals can be read out from the entire pixel unit 1351. The readout signals are sent to the outside of the imaging unit 135 via the output circuit 1356. At this time, scanning in the main scanning direction is performed at high speed, but scanning in the sub-scanning direction is slower than scanning in the main scanning direction.

[0037] In this embodiment, the optical fiber unit 120, the spectroscopic element 133, and the image capturing unit 135 are arranged so that the spectral images of the multiple Raman scattered lights dispersed by the spectroscopic element 133 are projected along the main scanning direction of the image capturing unit 135. In other words, the optical fiber unit 120, the spectroscopic element 133, and the image capturing unit 135 are arranged so that the spectral direction of the spectroscopic element 133 is aligned with the main scanning direction of the image capturing unit 135. In other words, the exit end and the spectroscopic element 133 (spectroscopic element) are arranged so that the spectral image formed on the image capturing unit 135 corresponding to one of the multiple Raman scattered lights is aligned with the main scanning direction of the image capturing unit 135. In other words, the spectral images of the multiple Raman scattered lights dispersed by the spectroscopic element 133 (spectroscopic element) are projected at different positions on the image capturing unit 135 along the sub-scanning direction. This allows the signals of the spectral images projected on the light receiving surface of the image capturing unit 135 to be read out at high speed, thereby increasing the throughput of the identification process.

[0038] (Lighting section) The light collecting unit 110i of this embodiment will be described in detail below with reference to Fig. 2. The light collecting unit 110i has an illumination optical unit 1107 that irradiates the specimen 190 with light, and a light collecting optical unit 1113 that collects Raman scattered light from the specimen 190.

[0039] The illumination optical unit 1107 is an optical unit having optical elements that illuminate the specimen 190 on the conveyance surface of the belt conveyor 152 of the mounting section 150 with light from the light source 1101. The specimen 109 illuminated by the illumination optical unit 1107 emits Raman scattered light. At this time, the intensity of the Raman scattered light generated from the specimen 190 is weak, about 10-6 times the intensity of Rayleigh scattered light that is elastically scattered on the surface of the specimen 190 and does not exhibit wavelength shift. As shown in FIG. 2, the illumination optical unit 1107 has a semiconductor laser 1101, a laser mount 1102, a laser driver 1103, a collimator lens 1104, a cylindrical lens 1105, and a condenser lens 1106.

[0040] The semiconductor laser 1101 is a light source that illuminates the specimen 190. In this embodiment, the semiconductor laser 1101 is a continuous-wave laser that emits light with an intensity required to generate Raman scattered light from the specimen 190. The efficiency of Raman scattering generation increases as the wavelength of the illumination light becomes shorter and decreases as the wavelength of the illumination light becomes longer. On the other hand, the intensity of fluorescence, which becomes background noise relative to the Raman scattered light, decreases as the wavelength of the illumination light becomes longer and increases as the wavelength of the illumination light becomes shorter. The light from the semiconductor laser 1101 can have a wavelength of, for example, 532 nm, 633 nm, or 780 nm. While the semiconductor laser 1101 has been described as being used as the light source for the illumination optical unit 1107, this is not limiting and other laser light sources, such as semiconductor-pumped solid-state lasers and gas lasers, can also be used. The wavelength of the light source used in the illumination optical unit 11107 is selected taking into consideration the Raman shift amount, signal-to-noise ratio, and other characteristics specific to the material to be identified.

[0041] Laser mount 1102 holds semiconductor laser 1101 and dissipates heat. Laser driver 1103 supplies current to semiconductor laser 1101 via laser mount 1102, causing semiconductor laser 1101 to oscillate and, at the same time, maintaining a constant temperature of semiconductor laser 1101. Note that a laser driver 1103 may be provided for each light-collection optical unit 1103i, or one laser driver 1103 may be provided for multiple light-collection optical units 1103i.

[0042] The collimator lens 1104 and cylindrical lens 1105 limit the spread of the light emitted from the semiconductor laser 1101 and shape it into parallel light. The cylindrical lens 1105 may be replaced with another collimating optical element, such as an anamorphic prism pair. The illumination optical unit 1107i may be provided with a wavelength filter (not shown) including a laser line filter. The wavelength filter may be positioned at the pupil plane of the illumination optical unit 1107i. This improves the wavelength characteristics of the light irradiated onto the specimen 190 by the illumination optical unit 1107. The condenser lens 1106 condenses the light from the semiconductor laser 1101 onto the specimen 190. The collimator lens 1104, cylindrical lens 1105, and condenser lens 1106 are preferably made of synthetic quartz. Because high-intensity light from the semiconductor laser 1101 passes through these lenses, using lenses made of synthetic quartz can reduce background fluorescence and Raman scattered light originating from trace components in the observation system.

[0043] The condenser lens 1106 is not necessarily required as long as Raman scattered light of sufficient intensity for identification can be obtained from the specimen 190. In other words, the illumination optical unit 1107 may be configured to irradiate the specimen 190 with light collimated by the collimator lens 1104 and / or the cylindrical lens 1105 as is.

[0044] The light-collecting optical unit 1103i is an optical unit having optical elements that collect Raman scattered light from the specimen 190 illuminated by the illumination optical unit 1107. The Raman scattered light collected by the light-collecting optical unit 1103i is guided to the spectroscopic element 133 by an optical fiber 121, which is a light-guiding means. The light-collecting optical unit 1103i has an objective lens 1110, an excitation light cut filter 1111, and a fiber condenser lens 1112.

[0045] The objective lens 1110 collects Raman scattered light from the specimen 190 illuminated by the illumination optical unit 1107. Because high-power light may be irradiated depending on the specimen 190, it is preferable to use quartz lenses for each lens constituting the light-collecting optical unit 1103i, such as the objective lens 1110, to reduce background fluorescence and Raman scattered light from the observation system. Similarly, it is preferable not to use cemented lenses to suppress background noise from the balsam and to prevent balsam peeling due to heat generation. In other words, each lens constituting the light-collecting optical unit 1103i, such as the objective lens 1110, is preferably a single lens. Furthermore, it is preferable that the objective lens 1110 be an aspherical lens to improve coupling efficiency to the optical fiber 1114, which serves as the light-guiding means.

[0046] The excitation light cut filter 1111 is a wavelength filter such as a bandpass filter or longpass filter, which blocks light in at least a portion of the wavelength range of the light collected by the objective lens 1110 and transmits the Raman scattered light. This blocks light unnecessary for measuring the Raman scattered light and transmits the Raman scattered light. From the viewpoint of filter characteristics, the excitation light cut filter 1111 is preferably placed in the parallel light beam between the objective lens 1110 and the fiber focusing lens 1112, i.e., on the pupil plane of the light collecting optical unit 1103i.

[0047] The fiber condenser lens 1112 couples the Raman scattered light into the optical fiber 121. When the excitation light cut filter 1111 is inserted, the Raman scattered light from the fiber condenser lens 1112 can be ignored, so it is advisable to prioritize coupling efficiency into the optical fiber 121 and use a cemented lens to suppress aberrations. The cemented lens includes a doublet lens.

[0048] In this embodiment, the illumination optical unit 1107 and the light collecting optical unit 1103i of the light collecting optical system 90i are configured independently, but this is not limiting. That is, the illumination optical unit 1107 and the light collecting optical unit 1103i may share some of the optical elements, such as various lenses, that make up their respective optical systems.

[0049] Furthermore, it is preferable that the multiple light-collection optical systems 90i are also arranged at different positions in the conveyance direction of the mounting section 150. Each light-collection optical system 90i has a certain size because it includes an illumination optical unit 1107i and a light-collection optical unit 1103i. Therefore, by arranging the multiple light-collection optical systems 90i diagonally when viewed from a direction perpendicular to the conveyance surface of the mounting section 150, it is possible to increase the arrangement density of the light-collection optical systems 90i in the width direction of the mounting section 150. This increases the discrimination resolution of the discrimination device 1, making it possible to discriminate smaller-sized specimens 190.

[0050] (binning processing) 1(c) is a diagram showing a projection state of the imaging region 1351 of the imaging unit 135 onto which a plurality of spectral images 300 are projected. The imaging region 1351 projects the spectral images 300i corresponding to the respective fiber positions of the output optical ends 120oei (i=a, b, c, . . . e) of the light guide 120i. The spectral images 300i are converted into electrical signals by the imaging unit 135 and output to the acquisition unit 140 via the sensitization processor 170 as Raman spectra, which are light intensity information for each wavelength.

[0051] In this case, the wavelength resolution of the spectral imaging unit 130 is limited by the fiber core diameter of the light guide 120i and the optical performance of the spectroscopic element 133, so it is not necessarily necessary for the imaging unit 135 to achieve a wavelength resolution equal to or greater than the wavelength resolution of the spectral imaging unit 130. In other words, even if the element size (element pitch) of the light receiving elements along the row direction 135r, which is the projection spectrum direction, is set to a projection condition smaller than the projection beam diameter corresponding to monochromatic light, the effective wavelength resolution will reach its limit and become over-specified. For specimens that are dark or black and have weak Raman scattered light intensity, sensitization is more desirable than increasing the resolution in the wavelength direction. Specimens that are dark or black and have weak Raman scattered light intensity contain polypropylene.

[0052] The sensitization processor 170 of this embodiment reduces noise, increases the signal amount, and improves detection sensitivity by binning detection signals corresponding to the light-receiving elements in the spectral direction of the imaging unit 135. Binning processing in the wavelength direction has the effect of increasing the effective signal-to-noise ratio while increasing the detection signal intensity, because thermal noise and the like contained in detection signals from multiple light-receiving elements has low correlation between the light-receiving elements, while the detection signals increase due to integration.

[0053] Binning is a process of accumulating charges (in the case of a CCD sensor) obtained by photoelectric conversion in a collection of light-receiving elements, such as 5x1 pixels or 4x2 pixels, count values, or data values ​​calculated based on the count values. The accumulating process includes averaging. Binning increases sensitivity without substantially reducing the required resolution, or by sacrificing resolution within an acceptable range, enabling more accurate analysis and sorting of resins. Therefore, at least in the spectral direction of the spectroscopic element 133 (the row direction 135r in which wavelengths are expanded on the imaging unit 135), binning can be performed for the number of light-receiving elements that matches the required resolution.

[0054] (row-wise binning processing) Next, the row-direction binning process performed by the sensitization processor 170 of this embodiment will be described with reference to Figures 3(a), 3(b), 4(a), and 4(b). The sensitization processor 140 performs column-direction binning process to integrate output signals from light receiving elements 1351 located at different positions along the column direction of the imaging section 135.

[0055] The number of pixels to be binned in the row direction can be determined by optical simulation or actual measurement. In simulation, the size of the point spread function on the image capture unit 135 when monochromatic light with the same wavelength as the center wavelength of the dispersive element 133 is incident from the light guide unit 120i can be determined as the size of an Airy disk or similar point with a brightness value above a threshold. In actual measurement, the size of the point image on the image capture unit 135 when monochromatic light with the same wavelength as the center wavelength of the dispersive element 133 is incident on the optical fiber of the light guide unit 120i can be used as a reference. Figures 3(a) and 3(b) show the projected image of a point image corresponding to monochromatic light on the image capture unit 135. The number of light receiving elements to be binned can be determined by calculating the size of the point image on the image capture unit 135 through optical simulation or actual measurement. The size of the point image can be measured using the knife-edge method.

[0056] The row binning process involves integrating the output signals of a predetermined number of adjacent light receiving elements along the row direction 135r. The predetermined number of elements can be greater than one if apodization is taken into consideration, but two or more elements are used to achieve a more effective sensitization.

[0057] As shown in Fig. 3(a), the imaging unit 135 has an element pitch Pr in the row direction 135r and an element pitch Pc in the column direction 135c in an imaging region 1351. Meanwhile, without loss of generality, the size Φmt of a projected image of monochromatic light projected onto the imaging unit 135 is assumed to have a length Lmtr in the row direction 135r and a length Lmtc in the column direction 135c. Note that the rectangular lattice region within the imaging unit 135 in Figs. 3(a) and 3(b) corresponds to the imaging region 1351 in Fig. 1(c) where the light receiving elements are arranged.

[0058] As shown in FIG. 3(b), consider the case where the light-receiving element pitch in the row direction 135r of the imaging region 1351 is Pr, and the size Φmt of the projected image of monochromatic light projected onto the imaging unit 135 when the light-collecting optical system 90i collects monochromatic light of a predetermined wavelength is Lmtr in the row direction 135r. In this case, the number of light-receiving elements performing binning in the row direction can be equal to or less than int(Lmtr / Pr) × 2. That is, the upper limit of the number of binning in the row direction can be int(Lmtr / Pr) × 2. Here, F(x) = int(x) denotes an integer of the real number x. The predetermined wavelength can be set to the center wavelength of the spectroscopic element 133 or any wavelength including the center wavelength within a spectroscopic wavelength range determined to suit the target sample to be identified. In other words, the predetermined wavelength is included in the wavelength range projected by the spectroscopic element 133 onto the imaging unit 135.

[0059] Next, the beam diameter Φmt will be explained using Fig. 4(a). Consider the case where the beam diameter Φ is treated as the full width at half maximum (FWHM) in the beam profile of the detected light intensity of a point image corresponding to monochromatic light. In this case, the full width at half maximum FWHM of the wavelength of the monochromatic light is set to be shorter than the wavelength width Δλmin of the spectroscopic limit of the spectroscopic element 133.

[0060] In this case, if the beam diameter Φmt is approximated by a normal distribution with a standard deviation σ=1 and a peak of the detected light intensity of 0.40, the beam diameter corresponding to the full width at half maximum (FWHM) corresponds to -1.175σ to +1.175σ, which is equivalent to 76.0% of the total detected signal corresponding to the point image contained in such a beam diameter Φmt.

[0061] Next, the beam diameter Φmt will be explained using Figure 4(b). Let us consider the case where the beam diameter Φ is treated as twice the full width at half maximum (FWHM) in the beam profile of the detected light intensity of a point image corresponding to monochromatic light. In this case, the beam diameter corresponds to -2.350σ to +2.350σ, and the charge contained in such a beam diameter Φmt is equivalent to 91.1% of the total detected signal corresponding to the point image.

[0062] The upper limit of the row-direction binning number, int(Lmtr / Pr) × 2, mentioned above, means that by treating 91.1% of all detected charges of detection signals corresponding to a point image of monochromatic light as a point image, elements in that wavelength direction can be integrated as spectrally equivalent signals. This upper limit of the binning number is sometimes referred to as the upper limit of the first binning number.

[0063] On the other hand, when it is desired to further increase the resolution in the wavelength direction, the upper limit of the row direction binning number can be set to the upper limit value of the second binning number, which is int(Lmtr / Pr) × 1. With this upper limit value of the second binning number, the upper limit of the sensitization effect is reduced to 0.760 / 0.911 compared to the upper limit value of the first binning number, but the substantial overlap in the wavelength direction after row direction binning processing can be ignored.

[0064] In other words, the acquisition section 140 in this embodiment acquires spectroscopic information relating to one of the plurality of specimens 190 based on the binned signal obtained by the row-direction binning process in the sensitization processor 170 .

[0065] (column-wise binning processing) The column binning process includes integrating the output signals of the photodetectors 1351 corresponding to a predetermined number of adjacent rows along the column direction. The predetermined number of rows is set to a value greater than 1 when apodization is taken into account, but a value of 2 or greater is used to take into account the sensitization effect.

[0066] In the row direction 135r where the light collection optical systems 90i are arranged in multiple rows, the light guiding section 120i can be arranged so that the spectral images 300i do not overlap with each other, thereby preventing crosstalk between the spectral images 300i corresponding to the multiple light collection optical systems 90i.

[0067] When single-mode fibers are used in the light guide 120i, the cladding diameter is sufficiently large relative to the core diameter, reducing the overlap of spectral images of the optical fiber array on the imaging unit, even when the optical fibers are arranged adjacently at a pitch equal to the cladding diameter. When multimode fibers are used in the light guide 120i, the cladding diameter may be slightly larger than the core diameter. In this case, overlap of spectral images may occur depending on the performance of the spectral imaging unit 130. Therefore, to prevent overlap of spectral images, it is advisable to appropriately space the optical fibers in the light guide 120i so that they do not overlap on the imaging unit. The number of binning elements in the multi-row direction of the light guide 120i can be smaller than the spread range of the spectral image corresponding to the wavelength resolution of the spectroscopic element. The multi-row configuration used in the light guide 120i and the light collection optical system 90i is sometimes referred to as concatenation or one-dimensional arraying. When adjacent spectral images overlap, binning may be performed so as not to include the corresponding pixel columns in the row direction. Alternatively, the binning process may be apodization, which weights and accumulates each pixel position in the column direction 135c (each row in the sub-scanning direction). By assigning a low weight to a light receiving element that is close to an adjacent spectral image, crosstalk between the adjacent spectral images can be reduced.

[0068] Similar to the row-direction binning process, the column-direction binning process will be described below using a general formula. The column-direction light receiving element pitch of the imaging region 1351 is Pc, and the column-direction length Lmtc is the size Φmt of the projected image of monochromatic light projected onto the imaging section 135 when the light-collecting optical unit 1103i collects monochromatic light of a predetermined wavelength. In order to reduce crosstalk between adjacent projected spectra, the predetermined number of rows corresponding to the first upper limit of the number of binning processes in the column direction is int(Lmtc / Pc)×2.

[0069] In order to further reduce crosstalk between adjacent projected spectra, it is preferable that the predetermined number of rows corresponding to the upper limit of the number of binning processes in the column direction is set to int(Lmtc / Pc).

[0070] The row-direction binning process includes an apodization process that assigns predetermined weights to output signals corresponding to a plurality of light-receiving elements, and a process that accumulates the apodized output signals.

[0071] The sensitization processor 140 performs a saturation determination for the light receiving elements 1351 and performs row-direction binning processing based on the result of the saturation determination. In this embodiment, the number of binning elements in the wavelength axis direction in FIGS. 3(a) and 3(b) is set uniformly regardless of the projection position on the projection spectrum. However, other embodiments may be employed in which the projected spectrum includes regions with different numbers of binning elements in the wavelength axis direction depending on the projection position on the projection spectrum. That is, other embodiments may be employed in which the projected spectrum includes regions with different numbers of binning elements in the wavelength axis direction depending on the projection position on the projection spectrum. The other embodiments described above can be rephrased as forms in which the projected spectrum includes regions with different numbers of binning elements in the wavelength axis direction depending on the spectral wavelength on the projection spectrum. That is, other embodiments may be employed in which the number of binning elements in the wavelength axis direction varies depending on the spectral wavelength on the projection spectrum.

[0072] Furthermore, depending on the performance of the spectral imaging unit 130, the influence of aberration may be greater in the peripheral parts of the imaging unit 135, so the number of binning elements does not need to be uniform across the imaging unit. Figure 5(c) shows an image of a point image when aberration occurs. Since the image may become unclear in the peripheral parts of the imaging unit (areas close to the projection edge in the row direction and wavelength direction), the number of binning elements may be increased compared to the center of the imaging unit.

[0073] Although Fig. 1 shows an example of a separation unit that uses Raman scattering spectroscopy, the configuration of the light collection optical system, the spectroscopic element, etc. may be changed as appropriate to perform analysis by infrared spectroscopy. The configuration of the spectroscopic element in Fig. 1(b) and Fig. 2 may be changed as appropriate, for example, to a system that uses a concave reflecting mirror or a system that uses a prism.

[0074] Furthermore, a configuration can be adopted in which the number of light-receiving elements for binning processing is set based on at least one of information regarding the properties of the specimen and information regarding the candidate materials identified by the acquisition unit. At least one of the information regarding the properties of the specimen and the information regarding the candidate materials can be used as advance information before the time when the light-collecting optical system 120 collects light. In other words, at least one of the information regarding the properties of the specimen and the information regarding the candidate materials can be used as advance information before the time when the imaging unit 130 captures an image.

[0075] For example, when this identification device is used to identify waste resin, there are cases where the properties of the input resin, such as its type, composition ratio, shape, and particle size, are generally known, or where the input resin contains unspecified resin types, but the types of resin that are valuable are limited and the resin to be recovered is specified.

[0076] Furthermore, based on such prior information, it is possible to adaptively select whether to prioritize the spectral sensitivity corresponding to the spectral intensity or the spectral resolution. For example, if the spectral sensitivity is prioritized, the binning number in the wavelength direction can be adaptively set to be higher than int(Lmtr / Pr). Similarly, if the spectral resolution is prioritized, the binning number in the wavelength direction can be adaptively set to be equal to or lower than int(Lmtr / Pr). Such information regarding the properties of the specimens includes prior information linked to the specimen group in advance, color information acquired by an imaging unit such as a hyperspectral camera, etc.

[0077] The information about the properties of the specimen as such prior information includes at least one of information about the spectral reflectance of the specimen in the visible light range, information about the height of the specimen, and information about the surface roughness of the specimen.

[0078] According to the identification device of this embodiment, information from multiple light-collecting optical systems is measured simultaneously using a single spectroscopic element, and binning processing is performed at the required resolution, making it possible to provide a highly sensitive analytical device or sorting unit while constructing a small, low-cost device.

[0079] (Second embodiment) Next, row-direction binning processing according to the second embodiment will be described. In the second embodiment, processing is performed after determining whether or not each pixel is saturated in the sensitization processor 170. Since the device configuration is the same as in Figures 1(a) to 1(c), a description thereof will be omitted in this embodiment, and only the processing in the sensitization processor 170 will be described.

[0080] In particular, in Raman scattering spectroscopy, some specimens produce strong fluorescence, which can cause the signal to exceed the dynamic range of the sensor and become saturated. Depending on the signal intensity, a situation may occur in which some parts of each spectral image 301 of the light-collecting optical system 103 are saturated while others are not. In such a case, performing binning processing as in the first embodiment would result in a loss of quantitation of the pixel data after binning. Therefore, in this embodiment, a saturation determination is performed for each line in the spectral direction before binning processing, and lines with saturated pixels are excluded from the binning processing target, thereby avoiding the effects of saturation and acquiring Raman spectral data. Furthermore, since the number of pixels actually binned in each binning pixel area varies, the binning processing is performed by averaging rather than accumulating.

[0081] Furthermore, when saturation of a specific pixel is determined, the presence or absence of saturated pixels may be displayed as output information from the sensitization processor 106. In this way, the user can explicitly recognize at which position in the spectrum or at which pixel on the imaging unit 135 saturation occurs. In this case, the sensitization processor 106 may be configured so that the user can set a binning exclusion region and change the setting so that this is reflected in subsequent analyses or sorting. Furthermore, when the user can explicitly recognize information about saturated pixels, binning may be performed including saturated pixels.

[0082] According to this embodiment, by performing a saturation determination and then performing a binning process based on the results of the saturation determination, analysis and sorting can be performed without significantly reducing measurement accuracy even if saturation occurs. [Explanation of symbols]

[0083] 1 Separation section 90i daylighting optical system 133 Spectroscopic element 135 Imaging unit 170 Sensitizing Processing Unit 140 Acquisition Department 135r row direction 135c Column direction

Claims

1. A color information acquisition unit that acquires color information of a plurality of specimens; a plurality of light collecting optical systems that collect scattered light from the plurality of specimens after the color information is acquired by the color information acquisition unit; an imaging unit including a spectroscopic element that separates the plurality of light beams from the plurality of light-collecting optical systems, and a plurality of light-receiving elements arranged in a row direction and a column direction, and on which the plurality of spectral beams from the spectroscopic element are projected along the row direction; an acquisition unit that acquires spectroscopic information of at least one of the plurality of specimens based on an output signal from the imaging unit; a sensitization processing unit that performs row-direction binning processing to integrate output signals from a plurality of the light receiving elements located at different positions along the row direction; and an identification device characterized in that the sensitization processing unit, based on the color information of the plurality of samples acquired by the color information acquisition unit, changes the number of binning of the light receiving elements in the row direction binning process between a first sample having first color information and a second sample having second color information different from the first color information.

2. 2. The identification device according to claim 1, wherein the row-direction binning process includes a process of integrating output signals from a predetermined number of the light-receiving elements adjacent to each other along the row direction.

3. 3. The identification device according to claim 2, wherein the predetermined number of elements is two or more.

4. The identification device described in claim 2 or 3, characterized in that when the element pitch in the row direction of the light receiving elements is Pr and the row length in the row direction of the size Φmt of the projected image of the monochromatic light projected onto the imaging unit when the light collection optical system collects monochromatic light of a predetermined wavelength is Lmtr, the predetermined number of elements is int(Lmtr / Pr) x 2 or less. Here, F(x)=int(x) means an integer of real number x.

5. 5. The identification device according to claim 4, wherein the predetermined number of elements is equal to or less than int(Lmtr / Pr).

6. 6. The identification device according to claim 2, wherein the predetermined number of elements varies depending on the wavelength of the spectrum.

7. The identification device according to claim 1 , wherein the acquisition unit includes the sensitization processing unit.

8. 7. The identification device according to claim 1, wherein the imaging unit includes the sensitization processing unit.

9. 9. The identification device according to claim 1, wherein the acquisition unit acquires spectroscopic information relating to the plurality of specimens based on binned signals that have been subjected to the row-direction binning process.

10. 10. The identification device according to claim 1, wherein the sensitization processing unit performs column-direction binning processing to integrate output signals of the light receiving elements at different positions along the column direction of the imaging unit.

11. 11. The identification device according to claim 10, wherein the column-direction binning process includes a process of integrating output signals from the light-receiving elements corresponding to a predetermined number of rows adjacent along the column direction.

12. The identification device according to claim 11, wherein the predetermined number of rows is two or more.

13. An identification device as described in claim 11 or 12, characterized in that when the element pitch in the column direction of the light receiving elements is Pc and the column length in the column direction of the size Φmt of the projected image of the monochromatic light projected onto the imaging unit when the light collection optical system collects monochromatic light of a predetermined wavelength is Lmtc, the predetermined number of rows is int(Lmtc / Pc) x 2 or less. Here, F(x)=int(x) means an integer of real number x.

14. The identification device according to claim 13, wherein the predetermined number of rows is equal to or less than int(Lmtc / Pc).

15. 15. The identification device according to claim 4, wherein the predetermined wavelength is included in a wavelength range projected by the spectroscopic element onto the imaging unit.

16. 16. The identification device according to claim 4, wherein the full width at half maximum of the wavelength of the monochromatic light is shorter than the wavelength width of the spectroscopic limit of the spectroscopic element.

17. 17. The identification device according to claim 1, wherein the light collecting optical system collects Raman scattered light from the plurality of specimens.

18. 18. The identification device according to claim 1, wherein the light-collecting optical system has an effective light-collecting area.

19. 19. The identification device according to claim 18, further comprising a placement unit for placing the plurality of specimens in the effective light-collecting area.

20. 20. The identification device according to claim 19, wherein the placement unit includes a mechanism for moving the plurality of specimens placed in the effective light-collecting area to the outside of the effective light-collecting area.

21. 21. The identification device according to claim 19, wherein the mounting unit is configured to move the plurality of specimens in a predetermined direction.

22. 22. The identification device according to claim 21, further comprising a sorting unit that is located downstream of the light-collecting optical system in the predetermined direction and that sorts the plurality of specimens.

23. 23. The identification device according to claim 22, wherein the acquisition unit controls the separation operation of the separation unit based on the spectral information.

24. The identification device described in any one of claims 1 to 23, characterized in that the row-direction binning process includes an apodization process that assigns a predetermined weight to the output signal depending on the plurality of light-receiving elements, and a process that accumulates the apodized output signal.

25. 25. The discrimination device according to claim 1, wherein the sensitization processing unit performs saturation determination of the light receiving elements, and performs the row-direction binning process based on the result of the saturation determination.

26. a supply step of supplying a plurality of specimens to the identification device according to any one of claims 1 to 25; an identification step of identifying the properties of the plurality of specimens using the identification device; a separation step of separating the plurality of specimens based on the results obtained in the identification step; Separation methods including:

Images