Identification device, classification device, method for manufacturing articles, identification method, and program

Abstract

This technology provides advantages for accurately identifying the type of object. [Solution] An identification device for identifying the type of object comprises: an imaging unit that obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; and a control unit that determines the type of object based on a spectrum extracted from a part of the spectral image as a relationship between the wavelength and intensity of the reflected light. The control unit determines the type of object based on the spectrum extracted from a region in the spectral image where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum.

Description

Technical Field

[0001] The present invention relates to an identification device, a classification device, an article manufacturing method, an identification method, and a program.

Background Art

[0002] In recent years, as efforts to achieve a decarbonized society and suppress global warming, the promotion of recycling of plastic resources has been underway in industries such as the automotive industry and the home appliance industry. As a type of recycling, there is material recycling that uses discarded specific types of plastics as new product materials. For example, after recovering iron, aluminum, etc. from waste automobiles and waste home appliances, the residue may be crushed (pulverized) to a size of several tens to several hundreds of millimeters and targeted for recycling. Since such residues contain various types of plastics, in order to achieve material recycling, it is necessary to identify and classify specific types of plastics from the crushed residues. Patent Document 1 discloses an apparatus that measures reflected light from an object irradiated with light to identify the type of the object.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Optical methods such as near-infrared spectroscopy and Raman spectroscopy are known for identifying the type of plastic (object). In these methods, the reflected light (scattered light) from the plastic irradiated with laser light is spectrally analyzed by an imaging unit and imaged. From the resulting spectral image, a spectrum is obtained that shows the relationship between the wavelength and intensity of the reflected light. Different shapes and characteristics appear in the spectrum depending on the type of plastic, so it is possible to identify the type of plastic by analyzing the spectrum.

[0005] However, the intensity of reflected light from plastics can vary due to various factors. For example, the intensity of reflected light may differ depending on the type of plastic, or even within the same type of plastic, it may differ depending on surface contamination and properties. Therefore, if the intensity of reflected light from a plastic exceeds the range of light intensity detectable by the imaging unit, a portion of the spectrum may become saturated, making it difficult to accurately identify the type of plastic from that spectrum. As a countermeasure, it is conceivable to change the laser irradiation intensity or the imaging time of the imaging unit, but in identification processing that requires high throughput, it is difficult to change the irradiation intensity and imaging time for each type of plastic.

[0006] Therefore, the present invention aims to provide a technology that is advantageous for accurately identifying the type of object. [Means for solving the problem]

[0007] To achieve the above objective, an identification device as one aspect of the present invention is an identification device for identifying the type of object, comprising: an imaging unit that obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; and a control unit that determines the type of object based on a spectrum extracted from a part of the spectral image as a relationship between the wavelength and intensity of the reflected light, wherein the control unit determines the type of object based on a spectrum extracted from a region in the spectral image where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum.

[0008] Further objects or other aspects of the present invention will be revealed by preferred embodiments described below with reference to the accompanying drawings. [Effects of the Invention]

[0009] According to the present invention, for example, it is possible to provide a technique that is advantageous for accurately identifying the type of object. [Brief explanation of the drawing]

[0010] [Figure 1] Schematic diagram showing an example of the configuration of an identification device. [Figure 2] A diagram showing an example of a spectral image obtained by the imaging unit. [Figure 3] Figure showing an example of a spectral image and spectrum. [Figure 4] A diagram showing spectral images and spectra obtained from two identified objects. [Figure 5] A diagram illustrating the identification process of the first embodiment. [Figure 6] A diagram illustrating the identification process of the second embodiment. [Figure 7] A diagram illustrating the identification process of the third embodiment. [Figure 8] A diagram illustrating the identification process of the third embodiment (modified example). [Figure 9] Schematic diagram showing an example of a classification device configuration. [Modes for carrying out the invention]

[0011] The embodiments will be described in detail below with reference to the attached drawings. Note that the following embodiments do not limit the invention as defined in the claims. While the embodiments describe multiple features, not all of these features are essential to the invention, and the features may be combined in any way. Furthermore, in the attached drawings, identical or similar configurations are given the same reference numerals, and redundant descriptions are omitted.

[0012] <First Embodiment> The identification device 100 of the first embodiment of the present invention will now be described. The identification device 100 is a device for identifying the type of object. In this embodiment, as an example of an object whose type is to be identified by the identification device 100 (hereinafter sometimes referred to as the object to be identified), crushed pieces of plastic material that are crushed to a length and width of several tens of millimeters will be used for explanation. Examples of types (materials) of plastic material include PE (polyethylene), PP (polypropylene), PC (polycarbonate), and ABS (acrylonitrile, butadiene, styrene). The identification device 100 may also be understood as a material measuring device for measuring the material of the object to be identified.

[0013] The identification device 100 of this embodiment identifies the type of object to be identified using Raman spectroscopy. When laser light (excitation light) is irradiated onto the object to be identified, the wavelength of the reflected light (scattered light) shifts due to Raman spectroscopy. The amount of this wavelength shift varies depending on the material (type) of the object to be identified. Therefore, the shape of the reflected light spectrum can be acquired in advance as reference data (training data) for each material, and the material of the object to be identified can be determined by comparing the shape of the spectrum with the measurement data of the reflected light from the object to be identified and the reference data for each material.

[0014] FIG. 1 is a schematic diagram showing a configuration example of the identification device 100 of the present embodiment. The identification device 100 may include, for example, a light irradiation unit 110, an imaging unit 120, and a control unit 130.

[0015] The light irradiation unit 110 irradiates a part of the identification object 140 with the laser light L. The light irradiation unit 110 of the present embodiment may include a light source unit 111 and an optical scanning mechanism 112.

[0016] The light source unit 111 includes a light source of the laser light L irradiated to a part of the identification object 140. For example, the light source unit 111 may include a laser diode as a light source and may include its power supply device and control device. The identification device 100 of the present embodiment is a device that identifies the type of the identification object 140 from the shape characteristics of the spectrum in a certain wavelength band. Therefore, the laser light L emitted from the light source unit 111 includes a wavelength band in which the shape characteristics peculiar to the material of the identification object 140 can be obtained in the spectrum.

[0017] The optical scanning mechanism 112 is a mechanism that scans the laser light L on the identification object 140 by controlling the emission direction of the laser light L, and may be understood as a mechanism that controls (changes) the irradiation position of the laser light L on the identification object 140. For example, the optical scanning mechanism 112 includes a scan head mechanism including galvanometer mirrors of two orthogonal axes and may include its power supply device and control device. The laser light L emitted from the light source unit 111 is irradiated to a predetermined position on the identification object 140 by controlling the angles of the galvanometer mirrors of two axes in the optical scanning mechanism 112 after passing through an optical system (not shown). In the present embodiment, the galvanometer mirror is described as the optical scanning mechanism 112, but it is not limited thereto. A method of electrically changing the irradiation pattern of the laser light L with a liquid crystal spatial light modulator or a method of scanning the laser light L by driving a lens such as a cylindrical lens may also be used.

[0018] The imaging unit 120 obtains a spectral image in which the light intensity distribution of the reflected light is represented for each wavelength by imaging the reflected light (scattered light) from the identification target 140 partially irradiated with the laser light L by the light irradiation unit 110. The imaging unit 120 of the present embodiment may include a spectroscope 121 and an image sensor 122.

[0019] The spectroscope 121 spectrally decomposes the reflected light from the identification target 140 partially irradiated with the laser light L and emits it at different angles for each wavelength. For example, the spectroscope 121 may include a diffraction element that spectrally decomposes the reflected light from the identification target 140. The diffraction element may be a one-dimensional transmission type diffraction element. The reflected light from the identification target 140 enters the diffraction element of the spectroscope 121 through an optical system not shown. The light transmitted through the diffraction element of the spectroscope 121 is spectrally decomposed (diffracted) at an angle corresponding to the wavelength and enters the imaging surface of the image sensor 122.

[0020] The image sensor 122 images the reflected light spectrally decomposed by the spectroscope 121, that is, the diffracted light emitted from the diffraction element of the spectroscope 121. Thereby, a spectral image in which the light intensity distribution of the reflected light is represented for each wavelength can be obtained. For example, the image sensor 122 is configured as a camera including a photoelectric conversion element such as a CMOS sensor or a CCD sensor, and may include its power supply device and control device. In the image sensor 122, imaging conditions such as the imaging time for imaging the identification target 140 to obtain one spectral image are set in advance. In the present embodiment, it is assumed that no changes and adjustments of imaging conditions that cause complication of processing and reduction of throughput, such as adjustment of the imaging time, are made. Note that the imaging time may be understood as the exposure time or the charge accumulation time of the photoelectric conversion element included in the image sensor 122.

[0021] Figure 2 shows an example of a spectral image 200 obtained by the imaging unit 120. The spectral image 200 is obtained as a two-dimensional image defined by the spectral direction (first direction) in which reflected light from the object to be identified 140 is spectrally separated by the spectrometer 121 of the imaging unit 120, and the radial direction (second direction) perpendicular (intersecting) to the spectral direction in the cross-section of the reflected light. In the spectral image 200 shown in Figure 2, the horizontal direction is the spectral direction and the vertical direction is the radial direction, and the light intensity distribution in the radial direction in the cross-section of the reflected light is represented for each wavelength by the intensity of the color. Note that the spectral image 200 in Figure 2 may be understood as showing only a part of the image obtained by the imaging unit 120.

[0022] The control unit 130 controls the light irradiation unit 110 and the imaging unit 120, and controls the identification process that identifies the type of object to be identified 140 based on the spectral image obtained by the imaging unit 120. The identification process includes extracting a spectrum (spectral waveform) showing the relationship between the wavelength and intensity of reflected light from a part of the spectral image, and determining the type of object to be identified 140 based on the extracted spectrum. The control unit 130 is composed of an information processing device (computer) having a processor such as a CPU (Central Processing Unit) and a storage unit such as memory, and may include a setting unit 131, an extraction unit 132, and a determination unit 133. Here, the control unit 130 may be composed of a PLD (Programmable Logic Device) such as an FPGA (Field Programmable Gate Array), or an ASIC (Application Specific Integrated Circuit), or a general-purpose computer with a program installed, or a combination of all or part of these.

[0023] As shown in Figure 3(a), the setting unit 131 sets an extraction region ER for extracting spectra from the spectral image 200 obtained by the imaging unit 120. Figure 3(a) shows an example of a spectral image 200 obtained by the imaging unit 120. The extraction region ER can be set as a region extended in the spectral direction over a portion of the radial range (local range). The extraction region ER may not be a region with width in the radial direction, but may be set as a straight line extended in the spectral direction.

[0024] As shown in Figure 3(b), the extraction unit 132 extracts (acquires) the spectrum 300 from the extraction region ER set for the spectral image 200 by the setting unit 131. Figure 3(b) shows an example of the spectrum 300 extracted from the extraction region ER of the spectral image 200. For example, the extraction unit 132 can extract the spectrum 300 from the extraction region ER by averaging the intensity of reflected light for each wavelength in the extraction region ER. Here, if the extraction region ER is set as a straight line, the spectrum 300 is extracted based on the intensity of reflected light for each wavelength along the straight line without averaging.

[0025] The determination unit 133 determines (identifies, estimates) the type of object to be identified 140 based on the spectrum 300 extracted by the extraction unit 132. For example, the determination unit 133 can determine the type of object to be identified 140 by comparing the shape of the spectrum 300 extracted by the extraction unit 132 with the shape of the spectrum for each material that has been previously acquired as reference data.

[0026] Incidentally, conventionally, the setting unit 131 set one extraction region ER for one spectral image 200. For example, the setting unit 131 set the extraction region ER in the area (a certain pixel or range of pixels) where the intensity of reflected light is highest in the spectral image 200. The "area where the intensity of reflected light is highest in the spectral image 200" can be identified based on known information such as the positional relationship between the spectrometer 121 and the image sensor 122, and is generally located in the radial center of the spectral image 200, as shown in Figure 3(a).

[0027] However, the intensity of reflected light from the object to be identified 140 can vary due to various factors. For example, even for objects of the same type 140, the intensity of reflected light will differ depending on the color, dirt, and surface properties. Therefore, if the intensity of reflected light from the object to be identified 140 exceeds the range of light intensity detectable by the imaging unit 120 (image sensor 122), a portion of the spectrum will become saturated, making it difficult to accurately identify the type of object to be identified 140 from that spectrum.

[0028] Figure 4 shows spectral images and spectra obtained from two different identification objects 141 and 142, respectively. Figure 4(a) shows the spectral image 201 and spectrum 301 obtained for the first identification object 141, and Figure 4(b) shows the spectral image 202 and spectrum 302 obtained for the second identification object 142, which is different from the first identification object 141.

[0029] Here, the extraction region ER in the spectral image 201 in Figure 4(a) and the extraction region ER in the spectral image 202 in Figure 4(b) are set to the same position in the spectral image. The irradiation intensity of the laser light L by the light irradiation unit 110 and the imaging time of the imaging unit 120 are the same for the first identification target 141 and the second identification target 142. The threshold TH is a value used to determine whether the intensity of the reflected light has reached the upper limit of the light intensity detectable by the imaging unit 120 (image sensor 122), that is, whether the intensity of the reflected light has saturated in the imaging unit 120 (image sensor 122).

[0030] In the spectrum 301 of Figure 4(a), the maximum intensity of reflected light is less than the threshold TH, and there is no region where the intensity of reflected light is saturated (hereinafter sometimes referred to as the saturated region). In this case, the determination unit 133 can determine the type of the first object to be identified 141 based on the spectrum 301. On the other hand, in the spectrum 302 of Figure 4(b), the maximum intensity of reflected light reaches the threshold TH and includes the saturated region. This is because the intensity of reflected light from the second object to be identified 142 is greater than the intensity of reflected light from the first object to be identified 141, due to differences in color, surface stains, and properties between the first object to be identified 141 and the second object to be identified 142. In this case, it is difficult for the determination unit 133 to determine the type of the second object to be identified 142 based on the spectrum 302.

[0031] Thus, since the intensity of reflected light may differ for each individual object 140, the maximum intensity of reflected light in the spectrum may reach the threshold TH depending on the color and surface properties of the object 140. In spectra where the maximum intensity of reflected light reaches the threshold TH, it may become difficult to accurately identify the type of object 140.

[0032] Therefore, in this embodiment, the setting unit 131 sets multiple extraction regions ER, which are located at different radial positions from each other, for a single spectral image. Then, the determination unit 133 determines the type of object to be identified 140 based on the spectrum extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH. The spectral extraction in this embodiment will be described below.

[0033] Figure 5 is a diagram illustrating the spectral extraction in this embodiment. Figure 5(a) shows an example of a spectral image 203 obtained by the imaging unit 120, and Figure 5(b) shows an example of spectra 303a to 303c extracted from each extraction region ERa to ERc of the spectral image 203.

[0034] As shown in Figure 5(a), the setting unit 131 sets multiple extraction regions ERa to ERc, which are located at different radial positions from each other, for the spectral image 203 obtained by the imaging unit 120. For example, the setting unit 131 sets the extraction range ERb to the radial center where the intensity of reflected light is highest in the spectral image 203, sets the extraction region ERa above extraction region ERb, and sets the extraction region ERc below extraction region ERb. Each extraction region ERa and ERc should be set near extraction region ERb, because if they are set to a region in the spectral image 203 where the intensity of reflected light is too low, the spectrum extracted from there may be buried in noise. Furthermore, the extraction regions ERa to ERc may be touching each other or separated from each other.

[0035] Here, the number of extraction regions ER set for the spectral image 203 is not limited to three, but may be two or four or more. Also, in this embodiment, each extraction region ER is set by the setting unit 131, but it may be pre-set to be placed at a predetermined position in the spectral image 203 obtained by the imaging unit 120. In this case, the setting unit 131 does not have to be provided in the control unit 130.

[0036] Next, as shown in Figure 5(b), the extraction unit 132 extracts spectra 303 from each of the multiple extraction regions ERa to ERc set for the spectral image 203 by the setting unit 131. For example, the extraction unit 132 extracts spectrum 303a by averaging the intensity of reflected light for each wavelength in extraction region ERa. Similarly, the extraction unit 132 extracts spectrum 303b by averaging the intensity of reflected light for each wavelength in extraction region ERb, and extracts spectrum 303c by averaging the intensity of reflected light for each wavelength in extraction region ERc.

[0037] As shown in Figure 5(b), in the extraction region ERb set in the radial center, the maximum intensity of reflected light reaches the threshold TH, and spectrum 303b containing a saturated portion is obtained. However, the light intensity distribution for each wavelength, represented by the shades of color in the spectral image 203, is a Gaussian distribution where the light intensity is high in the radial center and decreases towards the periphery. Therefore, in the extraction regions ERa and ERc set outside the radial center, spectra 303a and 303c can be obtained, respectively, where the maximum intensity of reflected light is less than the threshold TH and does not contain a saturated portion.

[0038] Next, the determination unit 133 selects a spectrum 303 from among the spectra 303a to 303c extracted by the extraction unit 132 in which the maximum intensity of reflected light is less than the threshold TH, and determines the type of object to be identified 140 based on the selected spectrum 303. In the example in Figure 5(b), the maximum intensity of reflected light in spectra 303a and 303c is less than the threshold TH. Therefore, the determination unit 133 can determine the type of object to be identified 140 based on at least one of spectra 303a and 303c.

[0039] Here, the determination unit 133 may select a spectrum 303 by further using a selection condition that the difference or ratio between the maximum intensity and minimum intensity of reflected light is greater than or equal to a specified value, in addition to the selection condition that the maximum intensity of reflected light is less than the threshold TH. This is because even if a spectrum 303 has a maximum intensity of reflected light less than the threshold TH, if the difference or ratio between the maximum and minimum intensity of reflected light is too small, it becomes difficult to compare it with the spectral shape of the reference data. Therefore, the specified value may be set to the value of the difference or ratio that makes it possible to compare it with the spectral shape of the reference data. For example, the determination unit 133 may determine the type of object to be identified 140 based on spectrum 303a, 303c, in which the maximum intensity of reflected light is less than the threshold TH, and the spectrum 303a has the largest difference or ratio between the maximum and minimum intensity of reflected light.

[0040] Furthermore, in this embodiment, a spectrum 303 was extracted for each of the multiple extraction regions ER, but the spectrum 303 may be extracted only for the selected extraction region ER from among the multiple extraction regions ER. That is, the spectrum 303 does not need to be extracted for the extraction regions ER that were not selected from among the multiple extraction regions ER. For example, if it is predicted that the maximum intensity of reflected light in the extraction region ERc will be less than the threshold TH even if the color of the object to be identified 140 changes, the spectrum 303c may be extracted only for the extraction region ERc. That is, the spectra 303a to 303b do not need to be extracted for the extraction regions ERa to ERb.

[0041] As described above, the identification device 100 of this embodiment sets multiple extraction regions ER in a spectral image and determines the type of object to be identified 140 based on the spectrum extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH. This makes it possible to accurately identify the type of object to be identified 140 even when the intensity of reflected light differs depending on the color of the object to be identified 140. In other words, the type of object to be identified 140 can be accurately identified with simple processing and configuration, while being robust to the color of the object to be identified 140.

[0042] <Second Embodiment> A second embodiment of the present invention will now be described. In this embodiment, an example will be described in which the imaging unit 120 acquires multiple spectral images by spectrally analyzing reflected light and imaging under each of multiple conditions. In this embodiment, the multiple conditions differ in that the portion of the object to be identified 140 irradiated by the laser light L by the light irradiation unit 110 is different from one another. Note that this embodiment basically follows the first embodiment, and except for the matters mentioned below, it may follow the first embodiment. For example, the configuration of the identification device 100 is as described in the first embodiment, and crushed plastic material is exemplified as the object to be identified 140. In addition, in this embodiment, as in the first embodiment described above, a method of setting multiple extraction regions ER for a single spectral image may be applied.

[0043] In the identification device 100 of this embodiment, laser light L is irradiated onto each of the multiple parts of the object to be identified 140 by the light irradiation unit 110, and a spectral image is obtained by the imaging unit 120. Irradiation of each part of the object to be identified 140 with laser light L can be performed using the optical scanning mechanism 112 of the light irradiation unit 110. Furthermore, if the identification device 100 is provided with a detection unit for detecting the position and size of the object to be identified 140, the control unit 130 can determine the multiple parts of the object to be identified 140 to which laser light L will be irradiated based on the detection results of the detection unit.

[0044] Figure 6(a) shows an example of multiple parts of the object to be identified 140 that are irradiated with laser light L. In Figure 6(a), the multiple parts are exemplified as a first part 140a and a second part 140b. For example, the first part 140a may be set to include the center of gravity of the object to be identified 140, and the second part 140b may be set at a position separated from the first part 140a.

[0045] Figure 6(b) shows the spectral image 204 and spectrum 304 obtained for the first part 140a of the object to be identified 140. The spectral image 204 can be obtained by irradiating the first part 140a with laser light L by the light irradiation unit 110, and then spectrally analyzing and imaging the reflected light from the first part 140a with the imaging unit 120. The spectrum 304 can be extracted by the extraction unit 132 from specific extraction regions ERs set in the spectral image 204. For example, specific extraction regions ERs can be pre-set in the spectral image 204 to be the central part in the radial direction where the intensity of the reflected light from the first part 140a is highest.

[0046] Figure 6(c) shows the spectral image 205 and spectrum 305 obtained for the second part 140b of the object to be identified 140. The spectral image 205 can be obtained by irradiating the second part 140b with laser light L by the light irradiation unit 110, and then spectrally analyzing and imaging the reflected light from the second part 140b with the imaging unit 120. The spectrum 305 can be extracted by the extraction unit 132 from specific extraction regions ERs set in the spectral image 205. For example, specific extraction regions ERs can be pre-set in the spectral image 205 to be the central part in the radial direction where the intensity of the reflected light from the second part 140b is highest. Here, the position in the spectral image where specific extraction regions ERs are set can be the same for the spectral image 204 of the first part 140a and the spectral image 205 of the second part 140b.

[0047] Generally, the reflected light from the object to be identified 140 is a weak signal, and the intensity of the reflected light can change sensitively due to surface contamination and properties. In other words, in the object to be identified 140, the intensity of the reflected light from the first part 140a and the intensity of the reflected light from the second part 140b may differ due to differences in surface contamination and properties between the first part 140a and the second part 140b. As a result, the spectrum 304 of the first part 140a and the spectrum 305 of the second part 140b may differ from each other. In the spectrum 304 of the first part 140a shown in Figure 6(b), the maximum intensity of the reflected light reaches the threshold TH and includes a saturation region. On the other hand, in the spectrum 305 of the second part 140b shown in Figure 6(c), the maximum intensity of the reflected light is less than the threshold TH and does not include a saturation region. Note that Figure 6 shows an example where the reflected light intensity of the first part 140a is greater than that of the second part 140b. However, depending on the surface contamination and properties, the reflected light intensity of the second part 140b may be greater than that of the first part 140a.

[0048] The determination unit 133 selects a spectrum from spectra 304 to 305 in which the maximum intensity of reflected light is less than the threshold TH, and determines the type of object to be identified 140 based on the selected spectrum. In the example in Figure 6, the maximum intensity of reflected light in spectrum 305 is less than the threshold TH. Therefore, the determination unit 133 can determine the type of object to be identified 140 based on spectrum 305.

[0049] Here, the determination unit 133 may select a spectrum by further using a selection condition that the difference between the maximum and minimum intensity of reflected light is greater than or equal to a specified value, in addition to the selection condition that the maximum intensity of reflected light is less than the threshold TH. For example, the determination unit 133 may determine the type of object to be identified 140 based on the spectrum with the largest difference between the maximum and minimum intensity of reflected light.

[0050] Furthermore, in this embodiment, spectra were extracted from each of the multiple spectral images 204 to 205, but spectra may also be extracted only from the spectral images 204 to 205 in which the maximum intensity of reflected light is less than the threshold TH. In other words, the type of object to be identified 140 may be determined using the spectral images 204 to 205 in which the maximum intensity of reflected light is less than the threshold TH.

[0051] Furthermore, in this embodiment, the method described in the first embodiment may be applied to each of the multiple spectral images 204 to 205. Specifically, multiple extraction regions ER may be set for the spectral image 204 of the first part 140a, and the spectrum 304 may be extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH. Similarly, multiple extraction regions ER may be set for the spectral image 205 of the second part 140b, and the spectrum 305 may be extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH.

[0052] As described above, the identification device 100 of this embodiment acquires multiple spectral images by spectrally analyzing the reflected light with the imaging unit 120 for each of the multiple parts of the object to be identified 140 and capturing the image. Then, the type of object to be identified 140 is determined using the spectral image in which the maximum intensity of the reflected light is less than the threshold TH. In this embodiment as well, similar to the first embodiment, the type of object to be identified 140 can be accurately identified even when the intensity of the reflected light differs depending on the color of the object to be identified 140. In other words, the type of object to be identified 140 can be accurately identified with simple processing and configuration, while being robust to the color of the object to be identified 140.

[0053] <Third Embodiment> A third embodiment of the present invention will now be described. In this embodiment, an example will be described in which the imaging unit 120 acquires multiple spectral images by spectrally analyzing and imaging the reflected light under each of several conditions. In this embodiment, the multiple conditions are characterized by different timings in which the imaging unit 120 images the reflected light during the period in which the light irradiation unit 110 irradiates a part of the object to be identified 140 with laser light L. This embodiment basically follows the first embodiment, and can be followed in all respects except for those mentioned below. For example, the configuration of the identification device 100 is as described in the first embodiment, and crushed plastic material is exemplified as the object to be identified 140. In this embodiment, as in the first embodiment described above, a method of setting multiple extraction regions ER for a single spectral image may also be applied.

[0054] For example, in Raman spectroscopy, it is known that when a portion (the same portion) of an object to be identified 140 is continuously irradiated with laser light L, the reflected light from that portion decreases in proportion to the irradiation time of the laser light L. In other words, the intensity of the reflected light may be lower in the spectral image obtained by the imaging unit 120 after the laser light L has been irradiated for a while compared to the spectral image obtained by the imaging unit 120 in the initial stage when the irradiation of a portion of the object to be identified 140 with laser light L has started. In this case, the spectral image (spectrum) in the initial stage may contain a saturation region where the maximum intensity of the reflected light reaches the threshold TH, whereas the spectral image (spectrum) after the laser light L has been irradiated for a while may contain a saturation region where the maximum intensity of the reflected light is lower than the threshold TH.

[0055] Therefore, in the identification device 100 of this embodiment, a spectral image is obtained by the imaging unit 120 at each of several different timings during the period in which a portion of the object to be identified 140 is irradiated with laser light (hereinafter sometimes referred to as the irradiation period). Irradiation of the object to be identified 140 with laser light L can be performed using the optical scanning mechanism 112 of the optical irradiation unit 110. Furthermore, the portion of the object to be identified 140 irradiated with laser light L (hereinafter sometimes referred to as the irradiation target portion) can be determined, for example, to include the center of gravity of the object to be identified 140. If the identification device 100 is provided with a detection unit that detects the position and size of the object to be identified 140, the control unit 130 can determine the portion of the object to be identified 140 irradiated with laser light L based on the detection result of the detection unit.

[0056] Figure 7(a) shows the irradiated portion 140c of the object 140, the spectral image 206 obtained by the imaging unit 120, and the spectrum 306 extracted from the spectral image 206, at the first timing of the irradiation period. The spectral image 206 can be obtained by the imaging unit 120 spectrally analyzing and imaging the reflected light from the irradiated portion 140c of the object 140 at the first timing of the irradiation period, when the light irradiation unit 110 is continuously irradiating the irradiated portion 140c of the object 140. The spectrum 306 can be extracted by the extraction unit 132 from specific extraction regions ERs set in the spectral image 206. For example, specific extraction regions ERs can be pre-set in the spectral image 206 to be the central part in the radial direction where the intensity of the reflected light from the irradiated portion 140c is highest.

[0057] Figure 7(b) shows the irradiated portion 140c of the object 140, the spectral image 207 obtained by the imaging unit 120, and the spectrum 307 extracted from the spectral image 207 for the second timing of the irradiation period. The second timing is a timing in the irradiation period that is later than the first timing and when the intensity of the reflected light from the object 140 (irradiated portion 140c) is lower than that of the first timing. The spectral image 207 can be obtained by spectrally analyzing and imaging the reflected light from the irradiated portion 140c using the imaging unit 120 at the second timing of the irradiation period, when the laser light L is continuously irradiated onto the irradiated portion 140c following the first timing. The spectrum 307 can be extracted by the extraction unit 132 from specific extraction regions ERs set in the spectral image 207. Here, the position in the spectral image where the specific extraction regions ERs are set may be the same for the spectral image 206 at the first timing and the spectral image 207 at the second timing.

[0058] The determination unit 133 selects a spectrum from spectra 306 to 307 in which the maximum intensity of reflected light is less than the threshold TH, and determines the type of object to be identified 140 based on the selected spectrum. In the example in Figure 7, the maximum intensity of reflected light in spectrum 307 is less than the threshold TH. Therefore, the determination unit 133 can determine the type of object to be identified 140 based on spectrum 307.

[0059] Here, the determination unit 133 may select a spectrum by further using a selection condition that the difference between the maximum and minimum intensity of reflected light is greater than or equal to a specified value, in addition to the selection condition that the maximum intensity of reflected light is less than the threshold TH. For example, the determination unit 133 may determine the type of object to be identified 140 based on the spectrum with the largest difference between the maximum and minimum intensity of reflected light.

[0060] Furthermore, in this embodiment, spectra were extracted from each of the multiple spectral images 206 to 207, but spectra may also be extracted only from the spectral images 206 to 207 in which the maximum intensity of reflected light is less than the threshold TH. In other words, the type of object to be identified 140 may be determined using the spectral images 206 to 207 in which the maximum intensity of reflected light is less than the threshold TH.

[0061] Furthermore, in this embodiment, the method described in the first embodiment may be applied to each of the multiple spectral images 206 to 207. Specifically, multiple extraction regions ER may be set for the spectral image 206 at the first timing, and spectrum 306 may be extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH. Similarly, multiple extraction regions ER may be set for the spectral image 207 at the second timing, and spectrum 307 may be extracted from the extraction region ER in which the maximum intensity of reflected light is less than the threshold TH.

[0062] As described above, the identification device 100 of this embodiment acquires spectral images by the imaging unit 120 at each of several different timings during the irradiation period in which a laser beam is irradiated onto a portion of the object to be identified 140. Then, the type of object to be identified 140 is determined using the spectral image in which the maximum intensity of reflected light is less than the threshold TH. In this embodiment as well, similar to the first embodiment, the type of object to be identified 140 can be accurately identified even when the intensity of reflected light differs depending on the color of the object to be identified 140. In other words, the type of object to be identified 140 can be accurately identified with simple processing and configuration, while being robust to the color of the object to be identified 140.

[0063] [Differentiation] The following describes a modified version of this embodiment. In this modified version, the object to be identified 140 is transported along a transport path by a transport mechanism such as a belt conveyor, that is, it is moving along a transport path. The optical scanning mechanism 112 of the optical irradiation unit 110 scans the laser beam L in accordance with the movement of the object to be identified 140 so that the laser beam L is continuously irradiated onto a part of the object to be identified 140 (the irradiation target portion 140c) that is being transported along the transport path.

[0064] For example, the device including the identification device 100 and the transport mechanism may be provided with a position detection unit that detects the position of the object to be identified 140 on the transport path, and a state detection unit (e.g., an encoder) that detects the state of the transport path, such as the transport speed of the object to be identified 140 and the position of the belt. In this case, the control unit 130 can control the scanning of the laser beam L by the optical scanning mechanism 112 so that the laser beam L is irradiated onto the target area 140c in accordance with the movement of the object to be identified 140, based on the detection results of the position detection unit and the state detection unit. This operation (control) is sometimes called tracking, and by repeating tracking for each of the multiple objects to be identified 140 that are sequentially transported by the transport mechanism, it is possible to identify the type of each object to be identified 140 with high throughput.

[0065] Figure 8 shows a method for acquiring spectral images at each of multiple timings in this modified example. In this modified example, spectral images are acquired by the imaging unit 120 at each of multiple timings Ta to Td during the period of tracking a single object 140 (i.e., the irradiation period). Timing Ta is earlier than timing Tb, timing Tb is earlier than timing Tc, and timing Tc is earlier than timing Td. The irradiation intensity of the laser light L from the light irradiation unit 110 and the imaging time of the imaging unit 120 are the same for all of the multiple timings Ta to Td.

[0066] Timing Ta may include the start of tracking on the object to be identified 140. When obtaining a spectral image at this timing Ta, in the early part of the imaging time of the imaging unit 120, the laser beam L is in a standby state irradiated onto the transport path, and the laser beam L has not yet irradiated the object to be identified 140. Then, in the later part of the imaging time of the imaging unit 120, the laser beam L begins to irradiate the object to be identified 140, and tracking is started in which the laser beam L is irradiated onto the target area 140c in accordance with the movement of the object to be identified 140. At timing Ta, the imaging unit 120 obtains a spectral image 208, and the extraction unit 132 extracts the spectrum 308 from the spectral image 208.

[0067] Each of the timings Tb to Td is a timing within the period during which tracking is performed on the object to be identified 140, that is, within the irradiation period during which the laser light L is continuously irradiated onto the irradiation target portion 140c of the object to be identified 140. At timing Tb, the imaging unit 120 obtains a spectral image 209, and the extraction unit 132 extracts the spectrum 309 from the spectral image 209. At timing Tc, the imaging unit 120 obtains a spectral image 210, and the extraction unit 132 extracts the spectrum 310 from the spectral image 210. At timing Td, the imaging unit 120 obtains a spectral image 211, and the extraction unit 132 extracts the spectrum 311 from the spectral image 211.

[0068] As mentioned above, in Raman spectroscopy, if a portion of the object to be identified 140 (the irradiated portion 140c) is continuously irradiated with laser light L, the reflected light from that portion may decrease in proportion to the irradiation time of the laser light L. In the example in Figure 8, the spectra 309-310 extracted from the spectral images 209-210 at timings Tb-Tc contain a saturation region where the maximum intensity of the reflected light reaches the threshold TH. In contrast, the spectrum 311 extracted from the spectral image 211 at timing Td, which is later than timings Tb-Tc, contains no saturation region because the maximum intensity of the reflected light is less than the threshold TH.

[0069] The determination unit 133 selects a spectrum from spectra 309 to 311 in which the maximum intensity of reflected light is less than the threshold TH, and determines the type of object to be identified 140 based on the selected spectrum. In the example in Figure 8, the maximum intensity of reflected light in spectrum 311 is less than the threshold TH. Therefore, the determination unit 133 can determine the type of object to be identified 140 based on spectrum 311.

[0070] Here, the spectrum 308 extracted from the spectral image 208 at timing Ta does not contain a saturated portion because the maximum intensity of the reflected light is less than the threshold TH. Therefore, the determination unit 133 may determine the type of object to be identified 140 based on the spectrum 308. However, at timing Ta, the laser light L is irradiated onto the transport path (belt) in the early part of the imaging time, and the spectral image 208 at timing Ta includes the influence of reflected light from the transport path. Therefore, the determination unit 133 may select the spectrum 311 extracted from the spectral image 211 at timing Td, rather than the spectrum 308 extracted from the spectral image 208 at timing Ta, and determine the type of object to be identified 140 based on the spectrum 311.

[0071] The first to third embodiments described above illustrate one embodiment of the present invention, and naturally, the present invention is not limited to the configurations, conditions, settings, etc., described in the above embodiments.

[0072] In the above embodiment, the object to be identified 140 was exemplified as crushed plastic material fragments (plastic pieces) that were crushed to a length and width of several tens of millimeters. Naturally, the present invention is useful regardless of the size of the object to be identified 140 and can be applied regardless of the size, color, condition, etc., of the plastic pieces. Furthermore, in actual plastic recycling processes, the object to be identified 140 may contain substances other than plastic, such as metal, wood, and paper, as impurities. Even when these substances are included in the object to be identified 140, the effect that the present invention has on identifying the type of object to be identified 140 does not change in any way.

[0073] Furthermore, as shown in the above embodiment, the object to be identified 140 may be placed (installed) in a fixed position such as on a stage, or it may be transported and moved by a transport mechanism such as a belt conveyor. In this case, only one object to be identified 140 may be placed / transported, or multiple (many) objects to be identified 140 may be placed / transported. Also, if the position and orientation of the object to be identified 140 are known to some extent, predetermined specific coordinates may be used as coordinates representing the irradiation position of the laser beam L on the object to be identified 140. On the other hand, if the position and orientation of the object to be identified 140 are uncertain, the coordinates representing the irradiation position of the laser beam L on the object to be identified 140 may be determined using a detection unit that detects the position and orientation of the object to be identified 140. As long as the objective of identifying the irradiation position of the laser beam L on the object to be identified 140 can be achieved, the position and orientation of the object to be identified 140 are arbitrary, and the method for determining the irradiation position of the laser beam L on it is also arbitrary.

[0074] In the above embodiment, an example was described in which a laser diode is used as the light source unit 111 of the light irradiation unit 110. However, a solid-state laser or a laser made of another medium material may be used as the light source unit 111. Furthermore, the wavelength of the laser light L can be any wavelength, such as visible light or near-infrared light. Any light source of laser light L that can irradiate the object to be identified 140 and provide information that allows for the identification of its material is acceptable. In short, the type, wavelength, output, and other specifications of the light source unit 111 are arbitrary.

[0075] In the above embodiment, an example was described in which the optical scanning mechanism 112 of the light irradiation unit 110 is configured by a scan head device including orthogonal two-axis galvanometer mirrors. However, for example, if it is assumed that laser light L is to be irradiated to a specific part of the object to be identified 140, and the position of the object to be identified 140 is fixed, the optical scanning mechanism 112 may simply be configured with only a mirror that reflects the laser light L toward the fixed position. Also, if the object to be identified 140 has a certain thickness and height, and height control is also required for irradiation of the laser light L, the optical scanning mechanism 112 may have a mechanism to adjust the height at which the laser light L is irradiated and the focal position of the laser light L. As long as it has a mechanism for irradiating the laser light L to a desired position on the object to be identified 140, the configuration and various specifications of the optical scanning mechanism 112 are arbitrary.

[0076] In the above embodiment, an example was described in which a one-dimensional transmission type diffraction element is used as the diffraction element of the spectrometer 121. However, the diffraction element of the spectrometer 121 may also be a reflection type diffraction element, and the type and arrangement of the diffraction elements are arbitrary as long as the objective of spectrally separating the reflected light (scattered light) from the object to be identified 140 and sending it to the image sensor 122 is achieved.

[0077] In the above embodiment, an example was described in which a camera including a photoelectric conversion element such as a CMOS sensor or a CCD sensor is used as the image sensor 122. However, the image sensor 122 may be composed of a camera including other types of sensors. Furthermore, there are no particular restrictions on the specifications of the image sensor 122, such as the pixel size or frame rate. As long as the objective of capturing light reflected by the object to be identified 140 and spectrally separated by the spectrometer 121, and obtaining a spectral image that can be used to identify the type of object to be identified 140, is achieved, the type, specifications, and performance of the image sensor 122 are arbitrary.

[0078] In the above embodiment, an example was described in which the control unit 130 is configured as an information processing device (computer), but any type of device is acceptable as long as it can implement the functions of the setting unit 131, extraction unit 132, and determination unit 133. Furthermore, although the control unit 130 is connected to the imaging unit 120 by a cable and acquires (receives) spectral images from the imaging unit 120 via the cable, spectral images may also be acquired from the imaging unit 120 via wireless communication such as a wireless LAN instead of a cable. As long as the control unit 130 acquires spectral images from the imaging unit 120 and executes the processing of the setting unit 131, extraction unit 132, and determination unit 133, the method and mechanism for acquiring and processing spectral images are arbitrary.

[0079] In the first embodiment described above, multiple spectra were acquired by setting the extraction target region in a single spectral image. In the second embodiment described above, multiple spectra were acquired by obtaining a spectral image for each of several parts of the object to be identified 140. Furthermore, in the third embodiment described above, multiple spectra were acquired by obtaining a spectral image for each of several timings during the irradiation period of the laser light L onto the object to be identified 140. In each embodiment, 2 to 4 spectra were acquired, but it is also possible to set it to acquire 5 or more spectra.

[0080] Furthermore, the methods described in the first to third embodiments above can be combined in any way. For example, different spectral images may be obtained from multiple parts of the same object 140, as in the second embodiment, and multiple spectra may be extracted from each spectral image, as in the first embodiment. Similarly, multiple spectra may be acquired by combining the first to second embodiments, or by combining the first to third embodiments. The setting unit 131 is for setting the acquisition of multiple spectra, and as long as this objective is achieved, it may be used by using some of the methods described in the first to third embodiments, or by using a combination of them.

[0081] In the above embodiment, an example was described in which a spectrum is obtained by averaging the intensity for each wavelength in the extraction region set in the spectral image. However, this averaging process is not essential, and the spectrum may be obtained using a single pixel in the radial direction of the spectral image. Even when the averaging process is omitted in this way, it is possible to obtain a spectrum having shape characteristics for identifying the types of objects to be identified 140, and the method described in the present invention can be applied.

[0082] Furthermore, in the above embodiment, the extraction region for extracting spectra was defined as a region located at a predetermined position on the spectral image, regardless of whether one spectrum or multiple spectra were extracted from a single spectral image. This is because, as explained in the first embodiment, the extraction region for extracting spectra from the spectral image 200 is known from information indicating the positional relationship between the spectrometer 121 (diffractive element) and the image sensor 122. However, as in the first embodiment, when spectra are extracted from multiple predetermined extraction regions for a single spectral image, it is possible that all spectra may contain saturated portions. To make better use of the first embodiment, the extraction region from which spectra are extracted from the spectral image may be determined (calculated). For example, by identifying a region in the spectral image that does not contain any saturated portions in the spectral direction and adding that region to the extraction region, the inclusion of saturated portions in all spectra can be reduced. The purpose of the extraction unit 132 is to acquire multiple spectra based on the conditions set in the setting unit 131, and the setting of the spectral extraction region is arbitrary.

[0083] In the above embodiment, the determination unit 133 demonstrated only the minimal function of determining the type of object to be identified 140 based on the spectrum extracted from a portion of the spectral image. Specifically, this function is achieved by calculating and comparing the similarity between the shape of a representative spectrum of various pre-registered plastic materials and the shape of the spectrum extracted from a portion of the spectral image. Furthermore, the accuracy of identifying the type of object can be improved by adding processing that focuses not only on similarity but also on the number of peaks in the spectrum, or by adding other criteria besides similarity. In addition to improving identification accuracy, numerous techniques can also be considered to improve processing time.

[0084] Furthermore, in a modified example of the fourth embodiment, it was explained that the type of object to be identified 140 may be identified based on a spectrum that includes the influence of reflected light from a conveying mechanism (belt conveyor) made of a different material from the object to be identified 140. As explained in this modified example, when the influence of reflected light from the conveying mechanism is relatively small, there is a high probability that the type of object to be identified 140 will be identified successfully, but when the influence of reflected light from the conveying mechanism is relatively large, there is a possibility that the type of object to be identified 140 will not be identified. Therefore, the shape of the spectrum relating to the material of the conveying mechanism may be registered in advance, and the similarity between the shape of the spectrum relating to the material of the conveying mechanism and the shape of the spectrum extracted from the spectral image may be calculated and compared. In this case, if the similarity is greater than or equal to a predetermined value, the influence of reflected light from the conveying mechanism may be considered significant and excluded from the list of spectra to be used for identifying the type of object to be identified 140.

[0085] The purpose of the determination unit 133 is to determine (identify) the type of object to be identified 140 based on the spectrum extracted from the spectral image, and it is possible to apply the numerous known techniques for improving identification accuracy and processing time as described above.

[0086] In the above embodiment, the light irradiation unit 110 (light source unit 111, light scanning mechanism 112) was described as a component of the identification device 100, but some or all of the light irradiation unit 110 does not have to be a component of the identification device 100. The components and arrangement of the identification device 100 are arbitrary as long as the identification process according to the present invention can be realized.

[0087] Furthermore, a computer program that implements each of the functions in the above embodiment may be supplied to the identification device 100 (control unit 130) via a network or various storage media. The identification device 100 (control unit 130) may then read and execute the program. In this case, the program and the storage medium storing the program constitute the present invention.

[0088] <Embodiment of a Classification Device> A classification device 400 (classification system) according to one embodiment of the present invention will be described. The classification device 400 is a device for classifying objects being transported along a transport path. Specifically, the classification device 400 identifies the type of object being transported along the transport path and sorts the object based on the identification result.

[0089] In this embodiment, crushed plastic material is used as an example of the object being transported along the transport path. The plastic material is, for example, a thermoplastic or thermosetting plastic, and is crushed into pieces of about several tens to several hundreds of millimeters in size before being transported along the transport path. Hereinafter, the crushed plastic material pieces may be referred to as "plastic pieces". Multiple plastic pieces may be transported along the transport path. Furthermore, the sorting device 400 of this embodiment includes the identification device 100 described above, which identifies the type of plastic piece (object to be identified 140) being transported along the transport path. The sorting device 400 identifies the type of plastic piece using the identification device 100 described above and sorts them using an air jet or the like.

[0090] Figure 9 shows a schematic diagram illustrating an example configuration of the sorting device 400 of this embodiment. The sorting device 400 may include a belt conveyor 410, a detection unit 420, a measurement unit 430, a control unit 440, and a sorting unit 450. In the following, directions are indicated using an XYZ coordinate system in which the surface (upper surface 411) on which the multiple plastic pieces SP are placed on the belt conveyor 410 is the XY plane, and the direction in which the multiple plastic pieces SP are transported by the belt conveyor 410 is described as the Y direction.

[0091] The belt conveyor 410 is a conveying mechanism that transports multiple plastic pieces SP along a transport path, and the transport path is defined by the upper surface 411 of the belt conveyor 410. Multiple plastic pieces SP are randomly fed onto the upper surface 411 of the belt conveyor 410 via a quantitative cutting device (crusher) or a vibratory feeder (not shown). Here, the belt conveyor 410 moves a belt made of rubber, resin, or metal at a predetermined speed, and the size of the belt and the moving speed can be set according to the processing capacity of the sorting device 400.

[0092] The detection unit 420 detects the size (size) of each of the multiple plastic pieces SP in the detection area 420' of the transport path. For example, the detection unit 420 includes an imaging unit 421 that uses a part of the upper surface 411 of the belt conveyor 410 as the detection area 420' (imaging field of view), and a processing unit 422 that determines the size of each plastic piece SP by performing predetermined image processing on the image obtained by the imaging unit 421. The detection area 420' is set to a shape that extends in the width direction (X direction) of the belt conveyor 410, and the imaging unit 421 may be configured to continuously image each plastic piece SP being transported by the belt conveyor 410. As a result, the processing unit 422 can determine the size (XY direction) of each plastic piece SP based on the multiple images continuously obtained by the imaging unit 421.

[0093] Furthermore, the detection unit 420 can detect the position of each plastic piece SP on the upper surface 411 (on the conveying path) of the belt conveyor 410. The processing unit 422 can determine the position of each plastic piece SP in the conveying direction (Y direction) and the width direction (X direction) of the belt conveyor 410 based on the image obtained by the imaging unit 421.

[0094] The measurement unit 430 irradiates a plastic piece SP being transported on the belt conveyor 410 with laser light L in a measurement area 430' downstream of the detection area 420' in the transport path, and measures the reflected light from the plastic piece SP. The measurement unit 430 may include a light irradiation unit 431 that irradiates the plastic piece SP with laser light L in the measurement area 430', and an imaging unit 432 that spectrally analyzes and images the reflected light from the plastic piece SP.

[0095] The control unit 440 is composed of a computer (information processing device) having, for example, a processor such as a CPU (Central Processing Unit) and a storage unit such as memory, and controls the classification process in the classification device 400. In this embodiment, the control unit 440 causes the measurement unit 430 to measure the reflected light from the plastic piece SP and performs identification processing to identify the type of plastic piece SP based on the measurement result. The control unit 440 then controls the sorting unit 450, described later, to classify the plastic piece SP based on the result of the identification processing.

[0096] Here, the measurement unit 430 and the control unit 440 can constitute the identification device 100. Specifically, the light irradiation unit 431 of the measurement unit 430 functions as the light irradiation unit 110 of the identification device 100, and the imaging unit 432 of the measurement unit 430 functions as the imaging unit 120 of the identification device 100. In addition, at least a part of the control unit 440 functions as the control unit 130 of the identification device 100.

[0097] The sorting unit 450 (sorting mechanism) is located near the end of the belt conveyor 410 and sorts each of the multiple plastic pieces SP conveyed by the belt conveyor 410 according to type. For example, the sorting unit 450 has multiple openings 451 (air nozzles) arranged along the width direction of the belt conveyor 410 and is configured as an air jet that releases compressed air from each opening 451 at an independent timing. This allows the drop position of each plastic piece SP conveyed by the belt conveyor 410 to be adjusted.

[0098] As an example, the sorting unit 450, under the control of the control unit 440, releases compressed air from at least one opening 451 towards the plastic pieces SP identified as a specific type by the identification device 100. As a result, the plastic pieces SP identified as a specific type are collected (stored) in the collection box 462, while the other plastic pieces SP are collected (stored) in the collection box 461 by free fall. In other words, each plastic piece SP transported by the belt conveyor 410 is classified according to the type identified by the identification device 100. Note that the collection box 462 may be positioned further away from the belt conveyor 410 than the collection box 461.

[0099] <Embodiment of Article Manufacturing Method> The classification device (identification device) according to the above embodiment can be used in a method for manufacturing articles such as plastic products. This method for manufacturing articles includes, for example, a classification step of classifying plastic pieces using the above classification device (identification device), and a manufacturing step of manufacturing articles by processing the plastic pieces classified in the classification step. The processing of the plastic pieces may include, for example, at least one of melting, molding, cutting, assembly, inspection, etc. The method for manufacturing articles according to this embodiment is advantageous compared to conventional methods in terms of article performance, quality, productivity, and production cost.

[0100] <Other Embodiments> The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that implements one or more functions.

[0101] <Summary of Embodiments> The disclosures herein include at least the following identification devices, classification devices, methods for manufacturing articles, identification methods, and programs. (Item 1) An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a portion of the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The identification device is characterized in that the control unit determines the type of object based on the spectrum extracted from a region where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum. (Item 2) The identification device according to item 1, characterized in that the imaging unit obtains a spectral image in which the light intensity distribution in the second direction is represented for each wavelength, as a two-dimensional image defined by the first direction in which the wavelength changes and the second direction intersecting the first direction in the cross-section of the reflected light. (Item 3) The identification device according to item 2, characterized in that the plurality of regions have different positions in the second direction in the spectral image. (Item 4) The identification device according to any one of items 1 to 3, characterized in that the control unit extracts the spectrum for each of the plurality of regions and determines the type of object based on the spectrum in which the difference between the maximum intensity and the minimum intensity of the reflected light is greater than or equal to a specified value. (Item 5) The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than the threshold. The identification device according to any one of items 1 to 4, characterized in that the parts of the object irradiated with the laser light are different from each other. (Item 6) An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device is characterized in that the parts of the object irradiated with the laser light are different from each other. (Item 7) The identification device according to item 5 or 6, characterized in that the plurality of conditions include a condition in which the laser light is irradiated onto a first portion of the object, and a condition in which the laser light is irradiated onto a second portion of the object that is different from the first portion. (Item 8) The identification device according to item 7, characterized in that the first part includes the center of gravity of the object. (Item 9) The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device according to any one of items 1 to 4, characterized in that the timing at which the imaging unit captures the reflected light during the period in which the laser light is irradiated onto a part of the object is different from that of the other multiple conditions. (Item 10) An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device is characterized in that the multiple conditions described above are such that the timing at which the imaging unit captures the reflected light during the period in which the laser light is irradiated onto a portion of the object is different from that of the other. (Item 11) The identification device according to item 9 or 10, characterized in that, during the aforementioned period, the reflected light from the object decreases in proportion to the irradiation time of the laser light. (Item 12) The identification device according to item 11, characterized in that the plurality of conditions include a condition in which the imaging unit images the reflected light at a first timing during the period, and a condition in which the imaging unit images the reflected light at a second timing during the period that is later than the first timing and when the intensity of the reflected light is lower than that of the first timing. (Item 13) An identification device according to any one of items 5 to 12, characterized in that the intensity of the laser light irradiated onto the object and the imaging time by the imaging unit are set to be the same under the multiple conditions. (Item 14) A classification device for classifying objects, A transport mechanism for transporting the object along a transport path, An identification device according to any one of items 1 to 13, which identifies the type of object being transported along the transport path by the transport mechanism, A sorting mechanism that sorts the objects transported by the transport mechanism according to their type based on the identification result of the identification device, A classification device characterized by comprising the following features. (Item 15) A classification process in which objects are classified using the classification device described in item 14, A manufacturing process for producing articles by processing the objects classified in the aforementioned classification process, A method for manufacturing articles, characterized by including the following: (Item 16) A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a portion of the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, The identification method is characterized in that, in the determination step, the type of object is determined based on the spectrum extracted from a region where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum. (Item 17) A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, In the determination step, the type of object is determined by using a spectral image in which the maximum intensity of the reflected light is less than a threshold, from among a plurality of spectral images obtained by imaging the reflected light under each of a plurality of conditions in the imaging step. The identification method is characterized in that the parts of the object irradiated with the laser light are different from each other, according to the aforementioned multiple conditions. (Item 18) A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, In the determination step, the type of object is determined by using a spectral image in which the maximum intensity of the reflected light is less than a threshold, from among a plurality of spectral images obtained by imaging the reflected light under each of a plurality of conditions in the imaging step. The identification method is characterized in that the timing of imaging the reflected light during the period in which the laser light is irradiated onto a part of the object is different from that of the other conditions. (Item 19) A program that causes a computer to perform the identification method described in any one of items 16 through 18.

[0102] The technologies described herein may contribute to the realization of a sustainable society, such as a decarbonized / circular economy.

[0103] The invention is not limited to the embodiments described above, and various modifications and variations are possible without departing from the spirit and scope of the invention. Accordingly, claims are attached to disclose the scope of the invention. [Explanation of Symbols]

[0104] 100: Identification device, 110: Light irradiation unit, 120: Imaging unit, 130: Control unit, 140: Object to be identified, 400: Classification device

Claims

1. An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a portion of the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The identification device is characterized in that the control unit determines the type of object based on the spectrum extracted from a region where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum.

2. The identification device according to claim 1, characterized in that the imaging unit obtains a spectral image in which the light intensity distribution in the second direction is represented for each wavelength, as a two-dimensional image defined by the first direction in which the wavelength changes and the second direction intersecting the first direction in the cross-section of the reflected light.

3. The identification device according to claim 2, characterized in that the plurality of regions have different positions in the second direction in the spectral image.

4. The identification device according to claim 1, characterized in that the control unit extracts the spectrum for each of the plurality of regions and determines the type of object based on the spectrum in which the difference between the maximum intensity and the minimum intensity of the reflected light is greater than or equal to a specified value.

5. The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than the threshold. The identification device according to claim 1, characterized in that the parts of the object irradiated with the laser light are different from each other.

6. An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device is characterized in that the parts of the object irradiated with the laser light are different from each other.

7. The identification device according to claim 5 or 6, characterized in that the plurality of conditions include a condition in which the laser light is irradiated onto a first portion of the object and a condition in which the laser light is irradiated onto a second portion of the object that is different from the first portion.

8. The identification device according to claim 7, characterized in that the first part includes the center of gravity of the object.

9. The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device according to claim 1, characterized in that the timing at which the imaging unit captures the reflected light during the period in which the laser light is irradiated onto a part of the object is different from that of the other multiple conditions.

10. An identification device for identifying the type of object, An imaging unit obtains a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing and imaging the reflected light from the object to which a laser beam is partially irradiated; A control unit that determines the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Equipped with, The control unit determines the type of object using a spectral image from among a plurality of spectral images obtained by the imaging unit capturing the reflected light under each of a plurality of conditions, in which the maximum intensity of the reflected light is less than a threshold. The identification device is characterized in that the multiple conditions described above are such that the timing at which the imaging unit captures the reflected light during the period in which the laser light is irradiated onto a portion of the object is different from that of the other.

11. The identification device according to claim 9 or 10, characterized in that during the aforementioned period, the reflected light from the object decreases in proportion to the irradiation time of the laser light.

12. The identification device according to claim 11, characterized in that the plurality of conditions include a condition in which the imaging unit images the reflected light at a first timing during the period, and a condition in which the imaging unit images the reflected light at a second timing during the period that is later than the first timing and when the intensity of the reflected light is lower than that of the first timing.

13. The identification device according to any one of claims 5, 6, 9, and 10, characterized in that the intensity of the laser light irradiated onto the object and the imaging time by the imaging unit are set to be the same under the multiple conditions.

14. A classification device for classifying objects, A transport mechanism for transporting the object along a transport path, An identification device according to any one of claims 1, 6, and 10, which identifies the type of object being transported along the transport path by the transport mechanism, A sorting mechanism that sorts the objects transported by the transport mechanism according to their type based on the identification result of the identification device, A classification device characterized by comprising the following features.

15. A method for manufacturing an article, comprising: a classification step of classifying objects using a classification device described in claim 14; and a manufacturing step of manufacturing an article by processing the objects classified in the classification step.

16. A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a portion of the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, The identification method is characterized in that, in the determination step, the type of object is determined based on the spectrum extracted from a region where the maximum intensity of the reflected light is less than a threshold, among a plurality of regions set in the spectral image for extracting the spectrum.

17. A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, In the determination step, the type of object is determined by using a spectral image in which the maximum intensity of the reflected light is less than a threshold, from among a plurality of spectral images obtained by imaging the reflected light under each of a plurality of conditions in the imaging step. The identification method is characterized in that the parts of the object irradiated with the laser light are different from each other, according to the aforementioned multiple conditions.

18. A method for identifying the type of object, An imaging step is to obtain a spectral image in which the light intensity distribution of the reflected light is shown for each wavelength by spectrally analyzing the reflected light from the object to which a laser beam is partially irradiated and imaging the reflected light, A determination step of determining the type of object based on a spectrum extracted from a specific region in the spectral image as the relationship between the wavelength and intensity of the reflected light, Includes, In the determination step, the type of object is determined by using a spectral image in which the maximum intensity of the reflected light is less than a threshold, from among a plurality of spectral images obtained by imaging the reflected light under each of a plurality of conditions in the imaging step. The identification method is characterized in that the timing of imaging the reflected light during the period in which the laser light is irradiated onto a part of the object is different from that of the other conditions.

19. A program for causing a computer to execute the identification method described in any one of claims 16 to 18.

Images