Identification device, sorting device, identification method, sorting method, and method for manufacturing articles

Abstract

To provide an identification device which achieves both of high S / N and high throughput.SOLUTION: An identification device for identifying a type of an object comprises: an illumination unit which illuminates the object; a sensor which measures reflection light from the illuminated object; and a processing unit which identifies the type of the object on the basis of the measurement result by the sensor. The illumination unit comprises: a scan part which scans illumination light that illuminates the object; and a control part which controls the scan part. The control part controls the scan part on the basis of the position of the object such that the illumination light tracks the moving object and the position of the illumination light on the surface of the object is changed.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention , knowledge relates to a separate device , sorting apparatus, identification method, sorting method, and method for manufacturing articles and pertains thereto.

Background Art

[0002] Conventionally, resin molded parts have been widely used for structural and exterior parts of automobiles and electric appliances from the viewpoints of mass productivity and part cost. In recent years' industrial circles, there has been a pressing need to consider the environment, and these resin molded parts are also recycled and reused. In conventional recycling, after recovering iron, aluminum, etc. in various recycling processes from waste automobiles and waste home appliances, they are crushed to a size of about 10 to 100 mm and recovered as residues containing various types of plastics. Finally, it is often reused as thermal recycling that is burned as fuel.

[0003] Attempts have also been made to perform horizontal recycling in which these plastics as residues are reused as materials for resin molded parts, and resin identification devices have been developed to identify and select specific types of plastics from residues containing various types of plastics.

[0004] As a measurement method for specifying the resin type non-contact by irradiating light, the following techniques for specifying the resin type using Raman scattering are disclosed.

[0005] Patent Document 1 discloses a resin identification device capable of obtaining a strong Raman scattering signal without damaging or denaturing the resin by translating the illumination light when measuring the resin flowing on a belt conveyor by Raman spectroscopy. During the irradiation of the excitation laser light, the light collection optical system is translated parallel to the upper surface of the mounting table in a range larger than the spot diameter of the excitation laser light on the object, preferably in a range of 2 to 3 times or more of the spot diameter. By doing so, the excitation laser light does not concentrate and irradiate at one location of the object, and it becomes a resin identification device in which the object is not damaged or deteriorated.

Prior Art Documents

[0006] [Patent Document 1] Japanese Patent Publication No. 2013-36971 [Overview of the project] [Problems that the invention aims to solve]

[0007] Patent Document 1 describes a method for identifying resin types by Raman measurement without altering the resin even when irradiated with a high-power laser. Generally, Raman measurement requires the measurement of very weak signals, and a high-power laser light source is necessary to configure it as an identification device. However, if an inexpensive LD that can be mounted on the device is considered as the light source, the LD output is limited, so it is necessary to improve the signal-to-noise ratio (S / N). To obtain a Raman signal with a high S / N, for example, the belt speed can be reduced to increase the exposure time at the sensor to increase signal intensity, or noise can be reduced by repeated averaging during measurement.

[0008] On the other hand, since large amounts of resin residue are generated in the recycling process, such as 1 ton or 2 tons per hour, there is a need to increase the throughput of the identification process, and it is desirable to increase the speed of the conveyor belt. However, if the conveyor belt speed is increased to meet the throughput requirements, the time the sample is directly under the illumination light will be shortened, making it impossible to secure sufficient exposure time for the spectrometer for Raman measurement, or to secure sufficient number of repeated averages, making it difficult to ensure a good signal-to-noise ratio (S / N).

[0009] As described above, there was a problem in that it was difficult to sort resins while simultaneously achieving high signal-to-noise ratio and high throughput.

[0010] Therefore, the present invention aims to provide an identification device that achieves both high signal-to-noise ratio and high throughput. [Means for solving the problem]

[0011] An identification device, as one aspect of the present invention that solves the above problems, is an identification device that identifies the type of object, A recognition unit that acquires information regarding the brightness of the aforementioned object, The processing unit comprises a scanning unit that scans illumination light for illuminating the object, a light receiving unit that receives light from the object illuminated by the illumination light, and a processing unit that identifies the type of object based on the light receiving result from the light receiving unit, wherein the processing unit Based on the aforementioned information regarding brightness, The scanning unit is controlled so that the illumination time for illuminating a first object having a first brightness is different from the illumination time for illuminating a second object having a second brightness different from the first brightness. [Effects of the Invention]

[0012] According to the present invention, it is possible to provide an identification device that achieves both high signal-to-noise ratio and high throughput. [Brief explanation of the drawing]

[0013] [Figure 1] This is a diagram showing the sorting system. [Figure 2] This is a diagram showing the details of the measuring unit 30. [Figure 3] This diagram shows how a galvanometer scanner scans illumination light. [Figure 4] This figure shows the scanning of illumination light within the scanning area 400. [Figure 5] This figure shows the scanning trajectory when measuring multiple samples within a scanning area. [Figure 6] This figure shows the measurement process for a sample outside the scanning area. [Modes for carrying out the invention]

[0014] Preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0015] FIG. 1 is a diagram showing the configuration of a sorting system. The sorting system of the present embodiment has a belt conveyor 1 (moving unit) that moves a sample (object), performs an identification process for identifying the type of the sample conveyed by the belt conveyor 1, and sorts the sample according to the identification result.

[0016] Samples 10 and 11 are metal pieces mainly containing metal, ceramic pieces mainly containing metal oxide crystals, glass pieces containing amorphous metal oxides, resin pieces mainly containing resin, and the like. In many cases, Samples 10 and 11 are crushed through a crushing process as a pre-process in recycling of industrial waste such as household electrical appliances and automobiles, and are crushed into a flat size of about 10 to 100 mm.

[0017] Here, the resin in this specification means polymers of organic substances including thermoplastic resins (plastics), thermosetting resins, rubbers, elastomers, celluloses, papers, and the like.

[0018] Note that Samples 10 and 11 may contain fillers such as glass and fibers, and various additives such as flame retardants and plasticizers.

[0019] The sorting system according to the present embodiment includes identifying the type of resin constituting the resin piece, that is, the material type (material, color, etc.) of the resin piece, and also identifying the presence or absence and type of these additives. >

[0020] In the sorting device of the present embodiment, the sample to be sorted is input to the belt conveyor 1 by a metering cutting device or a vibratory feeder (not shown), and the sample flows on the belt conveyor 1. On the belt conveyor 1, a recognition unit 2 (measurement unit), a measurement unit 3, and a sorting unit 162 are arranged from the upstream side of the flow of the belt conveyor. After the sample 10 passes under the recognition unit 2 and under the measurement unit 3, it is projected from the belt conveyor 1 and passes in front of the sorting unit 162.

[0021] The recognition unit 2 performs image processing on 2D images, including samples acquired by, for example, an area camera capable of acquiring 2D images in bulk, or a line camera capable of reconstructing 2D images. For example, based on the image processing results, the recognition unit 2 calculates the coordinates of the position where sample 10 exists at time t0 as p1(x1,y1) from the shape and size information of the sample.

[0022] As shown in Figure 3, the measurement unit 3 includes a galvanoscanner 41 and a galvanoscanner 42 (scanning unit) that scan the illumination light 50, and an optical spectrometer 30 that measures Raman scattered light (reflected light) from the sample. By controlling the scanning unit to tilt the main ray of the illumination light 50, the illumination light can be tilted as 50a or 50b with respect to the measurement unit 3, making the illumination light (measurement point) movable.

[0023] As shown in Figure 2, the optical spectrometer 30 has a light source 301 and a lens 302 (illumination unit). The optical spectrometer 30 illuminates a mirror configured in the galvanometer scanner with laser light from the light source 301 as illumination light 50 via lenses 302 and 304. A dichroic mirror 303 is positioned between lenses 302 and 304 to transmit the wavelength λ of the light source 301.

[0024] The processing unit 140 has a control unit that controls the galvanometer scanner. The processing unit 140 calculates the position coordinates p1(x1, y1 + V(t1-t0)) of the moving sample at time t1 from the pre-adjusted or measured speed V of the belt conveyor 1 and the time t0 when the recognition unit 2 recognizes the sample. The processing unit 140 then outputs a control command to the galvanometer scanner so that the illumination light 50 illuminates the sample 10 at the time when the sample 10 reaches the scanning area 400 in which the illumination light 50 can be scanned by the galvanometer scanners 41 and 42. The processing unit 140 then controls the galvanometer scanner so that the illumination light 50 follows the moving sample 10. By using two galvanometer scanners, 41 and 42, the illumination light can be scanned in the y-axis direction, which is the direction of movement of the belt conveyor 1 (sample), and in the x-axis direction, which is perpendicular to the y-axis.

[0025] Specifically, as shown in Figure 4, when the sample 10 reaches the upstream boundary of the belt conveyor 1 in the scanning area 400 (10a), the illumination light is scanned to the position of the illumination light 50a and measurement begins. As the sample 10 moves at a speed V in accordance with the movement of the belt conveyor 1, the sample 10 is illuminated to trace a trajectory 60 on the sample, and Raman measurement is performed. The measurement is completed when the sample 10 reaches the downstream boundary of the belt conveyor 1 in the scanning area 400 (10b). Here, the trajectory from illumination light 50a to 50b indicates the average position of the scanning illumination light 50.

[0026] Let B be the magnitude of the belt movement direction of the scanning area 400. The position of the illumination light 50 changes in the same direction as the belt movement direction, and the illumination light follows the sample 10 for a time of T = B / V, and during illumination, the illumination does not remain at a single point on the surface of the sample 10.

[0027] Therefore, the energy of the illumination light is not concentrated on a single point on the sample, and Raman measurements can be performed without altering the resin. Moreover, since the belt conveyor speed can be increased, even if a large number of samples are sent onto the belt conveyor to increase the identification throughput, they can be arranged in a spatially separated manner.

[0028] In the conventional method, where the illumination light 50 is not scanned, the sample 10 will be illuminated for a time of t = a / V, where a is the magnitude of the belt movement of the sample 10.

[0029] As an example, we will describe the case where the belt conveyor speed is 1 m / s, the size of sample 10 is assumed to be 10 mm, which is the minimum size of a sample in the recycling process, and the scan area of ​​B is set to 200 mm.

[0030] In conventional methods with a fixed illumination light, the illumination time t is 10 ms, whereas when the illumination light 50 is scanned and tracks the sample during measurement, there is approximately 200 ms of measurement time. Therefore, for black resins with low reflectivity, the S / N ratio of weak Raman signals can be improved by extending the exposure time or by averaging 20 sets of data acquired with a 10 ms exposure time.

[0031] On the other hand, in this embodiment, the illumination light traces a zigzag trajectory, such as the trajectory 60, so that the illumination light 50 does not remain in one place while the sample 10 is being measured. Therefore, the energy of the illumination light is dispersed over a wide area of ​​the sample 10, and the resin does not denature during measurement.

[0032] In the optical spectrometer 30, as described above, the illumination light 50 is reflected by the mirror sections 41m and 42m of the galvanoscanner 41 and galvanoscanner 42, and then irradiated onto the sample. After being reflected by the sample, the Raman scattered light is reflected isotropically and again by the mirrors 41m and 42m of the galvanometer scanner, returning to the optical spectrometer 30. The Raman scattered light undergoes a Raman shift depending on the material of the sample. Therefore, it is then reflected by the dichroic mirror 303, and only the component that has undergone Raman scattering is transmitted by the bandpass filter 305, which acts as an excitation light cut filter, and is spectrally analyzed by the spectroscopic element 306 (spectroscopic unit). The light beam diffracted at different diffraction angles for each wavelength spectrally analyzed by the spectroscopic element 306 becomes focused light again by the lens 307, and the light for each wavelength that has undergone the Raman shift is received by the photodetector elements configured in the sensor array 308. The signal output from the photodetector elements is sent to the processing unit 140 as a characteristic signal corresponding to the material of the sample, and is compared with, for example, a material-specific waveform obtained by measuring a sample of a known material in advance, thereby identifying the material and identifying the type of sample.

[0033] The signals (measurement results) from the sensor array 308 are transmitted to the processing unit 140 at, for example, 300 fps. Measurement begins when the sample 10 reaches the upstream interface of the belt movement in the scanning area 400 (10a), and the processing unit 140 identifies the material by extracting the signal corresponding to the measurement from the group of signals transmitted from the sensor array 308 at 300 fps. The method for extracting the corresponding signal can be determined, for example, based on the time the operation command was issued.

[0034] If the processing unit 140 identifies the type of sample and determines that sample 10 is the material to be sorted, an open signal is sent to the sorting unit 162 via the sorting controller 161 at time t2 when sample 10 passes through the sorting unit 162. Then, air is ejected from the part of the sorting unit 162 located at the position corresponding to the coordinate x1 of sample 10 measured by the recognition unit 2, and sample 10 is blown away by the air and collected in the collection box 72. If sample 10 is not the material to be sorted, sample 10 is projected from the belt conveyor 1 at speed V and falls, and is collected in the collection box 71.

[0035] Once the measurement unit 3 has finished measuring sample 10, it will then measure sample 11, and subsequently measure the samples that come in succession (not shown). The measurement unit 3 moves to the starting point for measuring sample 11 by rotating the galvanoscanners 41 and 42 to change the inclination angle of the illumination light 50. Since the galvanoscanners can scan the illumination light at high speed, the travel time between measuring the next resin is short, and samples flowing in succession can be measured. As a result, resin sorting can be performed with high throughput by using Raman measurement with a long measurement time and high signal-to-noise ratio, and by measuring a large number of samples.

[0036] In this embodiment, the recognition unit 2 is described as obtaining a two-dimensional image from an area sensor or line camera and determining the coordinates of the sample through image processing. However, this is not the only option; for example, a method that recognizes objects from height information, such as a light section sensor, can also be employed. Furthermore, in this embodiment, a zigzag trajectory was used for the trajectory of the illumination light on the resin so that the illumination light does not remain at a single point on the sample. However, it goes without saying that the same effect can be obtained with circular motion, spiral motion, or drawing a square, as long as the illumination light does not remain at the same point on the resin.

[0037] Furthermore, in the configuration of this embodiment, the recognition unit 2 acquires a two-dimensional image including the sample, and the two-dimensional image contains information about the brightness of the sample, i.e., diffuse reflectance. As those skilled in the art have found, the light intensity of Raman scattering correlates with diffuse reflectance, and the Raman scattered light from a bright, white sample is significantly greater than the Raman scattered light from a dark, black sample. In other words, the measurement time required to obtain the same signal-to-noise ratio can be less for a white sample than for a dark sample.

[0038] Based on the above facts, in order to improve throughput and more efficiently identify samples, it is also effective to set the measurement time for white resins to be shorter than that for black resins. In this case, as shown in Figure 5, for the black resin 10, the illumination light is scanned under scanning conditions that measure along the trajectory 61 which moves the entire region 400 from position 10a to position 10b.

[0039] On the other hand, the illumination light is scanned under scanning conditions that result in a measurement time, for example, one-third, compared to that of the black resin. For example, the trajectory of the illumination light 50 is scanned in a step-like trajectory 62. This makes it possible to measure all three samples 12, 13, and 14, which are aligned side by side and have high brightness, as they move from position 12a to 12b, from position 13a to 13b, and from position 14a to 14b. Of course, in this case as well, the trajectories 61 and 62 indicate the average position of the scanning illumination light 50, and it goes without saying that the average position of the illumination light 50 moves along each trajectory while simultaneously drawing the trajectory 60 on the resin. Thus, the processing unit has a control unit that controls the scanning unit to change the scanning conditions of the illumination light based on information about the object.

[0040] By the way, the samples that flow through vary in size and color. As already mentioned, the higher the brightness of the sample, the stronger the Raman scattering intensity, and the lower the brightness, the weaker the Raman scattering light. In other words, the higher the brightness, the higher the reliability of identifying the resin type, and the lower the brightness, the lower the reliability of identifying the resin type. It goes without saying that the higher the purity of the recovered resin of the desired type, the more valuable it is for practical use, but it is preferable to recover a large amount of resin with high brightness because this increases the average reliability of identification, and as a result the purity of the recovered desired resin also increases.

[0041] On the other hand, since the samples flowing over region 400 actually flow along the conveyor belt 1 in a random arrangement, there is a certain probability that a state of high density will occur, and it is conceivable that there may not be enough time for the illumination light 50 to scan all of the high-density resin groups in order. In this case, the processing unit 140 can issue a command to the galvanometer scanners 41 and 42 to skip the measurement of one of the samples. In other words, when multiple objects are present within the scannable range of the illumination light, the processing unit 140 determines the object to be measured based on the information of the objects.

[0042] At this time, the purity of the ultimately recovered sample can be improved by prioritizing the measurement of samples with higher brightness based on the brightness information of each sample obtained by the recognition unit 2. Furthermore, it is also effective to increase the amount of resin recovered in the end by prioritizing the measurement of larger samples based on the shape information of the samples obtained by the recognition unit 2.

[0043] In this embodiment, the speed of the belt conveyor 1 is assumed to be a pre-measured or adjusted speed V, and the operation of the measurement unit 3 to track the illumination light and the sorting operation in the sorting unit 162 are configured to issue commands based on time. However, for example, a laser Doppler velocometer or an image correlation type displacement meter that directly measures the belt speed can be placed around the recognition unit 2 or the measurement unit 3, and the measured values ​​regarding the amount of belt movement from these measuring instruments can be input to the processing unit 140. This makes it possible to issue operation commands to the galvanometer scanners 41 and 42 based on the measured amount of belt movement, and more accurate tracking of the illumination light and sorting operation in the sorting unit 162 are effective in increasing the amount of recovered material. In this case, the method for extracting the signal when the sample is being illuminated from the signal group transmitted from the sensor array 308 can be determined using the measured value regarding the amount of belt movement at the time each signal was generated. This makes it possible to extract the signal when the illumination light 50 is staying on the resin with greater accuracy.

[0044] Regarding signal extraction, it is also effective to pre-measure a sample (specimen) that exhibits specific known characteristics (features) not expected during normal measurement, such as a sample with high reflectivity that causes measurement saturation, just before moving to the measurement start point or before a specified time. For example, as shown in Figure 6, a fluorescent sample 20 that exhibits high fluorescence and causes measurement saturation in the sensor array 308 is placed outside the scanning area 400 but at a position where it can be measured by the measurement unit 3. Then, the illumination light 50 is changed to an angle of 50c to illuminate the fluorescent sample 20 just before the sample 10 reaches the upstream interface of the scanning area 400's belt movement, or before a specified time, so that the fluorescent sample 20 can be measured. This simplifies the extraction of measurement data in the subsequent processing unit 140. The processing unit 140 can extract and process the data that should be used for determining the material of the sample from the data continuously transmitted from the sensor array 308, specifically the data immediately following the appearance of data where the level of each element is at its maximum value, or the data after a specified number of data points.

[0045] Samples exhibiting distinctive data are not limited to this example; they can also include light traps with extremely low reflected light, or alloy resins that simultaneously exhibit the characteristics of multiple resins. In short, any sample that exhibits known characteristics not expected in normal measurements will suffice.

[0046] As explained above, based on the resin information recognized by the recognition unit 2, the scanning unit, using a galvanometer scanner as an example, changes the angle of the main ray of the illumination light so that the average position follows the direction of belt movement. Furthermore, by scanning the illumination light so that it does not remain in the same position on the resin during measurement, point measurement Raman spectroscopy can be performed while tracking the sample even at high belt speeds, enabling resin identification by Raman spectroscopy with a high signal-to-noise ratio and high throughput.

Claims

1. An identification device for identifying the type of object, A recognition unit that acquires information regarding the brightness of the aforementioned object, A scanning unit that scans the illumination light that illuminates the aforementioned object, A light receiving unit that receives light from the object illuminated by the aforementioned illumination light, The processing unit includes a unit that identifies the type of object based on the light received by the light receiving unit, The identification device is characterized in that the processing unit controls the scanning unit to make the illumination time for illuminating a first object having a first brightness different from the illumination time for illuminating a second object having a second brightness different from the first brightness, based on the information regarding brightness.

2. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit to illuminate the object while moving the irradiation position of the illumination light on the surface of the object.

3. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit so that the illumination light does not remain at a single point on the surface of the object.

4. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit to change the scanning conditions of the illumination light based on the brightness of the object.

5. The identification device according to claim 1, characterized in that the processing unit determines the object to be identified based on the brightness of each of the multiple objects when multiple objects are present within the scannable range of the illumination light.

6. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit such that the illumination time for illuminating the first object having the first brightness is longer than the illumination time for illuminating the second object having a second brightness which is higher than the first brightness.

7. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit based on information relating to the brightness of the object and at least one of the shape and size of the object.

8. The identification device according to claim 1, characterized in that the processing unit controls the scanning unit based on information regarding the brightness of the object and the position of the object.

9. The identification device according to claim 1, characterized in that the processing unit identifies the type of object using data obtained by receiving light from a sample that has been previously illuminated by the illumination light using the light receiving unit.

10. The identification device according to claim 1, characterized in that the recognition unit measures the position of the object.

11. The identification device according to claim 1, further comprising a spectral unit for spectrally separating the light from the object illuminated by the aforementioned illumination light.

12. The identification device according to claim 1, characterized in that the scanning unit is a galvanometer scanner.

13. The aforementioned object contains resin, The identification device according to claim 1, characterized in that the processing unit identifies the type of resin.

14. The identification device according to claim 1, characterized in that the light receiving unit receives reflected light from the object illuminated by the illumination light.

15. The identification device according to claim 1, characterized in that the light receiving unit receives Raman scattered light from the object illuminated by the illumination light.

16. The identification device according to claim 1, characterized in that the light receiving unit receives scattered light from the object illuminated by the illumination light.

17. The identification device according to claim 1, characterized in that the first object and the second object are moving objects.

18. The identification device according to claim 1, characterized in that the scanning unit scans the illumination light in the direction of movement of the object and in a direction perpendicular to the direction of movement.

19. The identification device according to claim 17, characterized in that the processing unit controls the scanning unit based on information relating to the brightness of the object and at least one of the transport direction and transport speed of the moving object.

20. The identification device according to claim 1, characterized in that the recognition unit measures at least one of the position and velocity of the moving object.

21. The identification device according to claim 17, further comprising a moving part for moving the aforementioned object.

22. An identification device according to any one of claims 1 to 21, A sorting device characterized by having a sorting unit that sorts objects based on the results identified by the identification device.

23. An identification method for identifying the type of object, An acquisition step to acquire information regarding the brightness of the aforementioned object, A scanning step of scanning illumination light that illuminates the aforementioned object, A light receiving step of receiving light from the object illuminated by the aforementioned illumination light, The process includes an identification step that identifies the type of object based on the light reception result in the light reception step, The identification method is characterized in that, in the scanning step, the illumination light is scanned such that, based on the information regarding brightness, the illumination time for illuminating a first object having a first brightness is different from the illumination time for illuminating a second object having a second brightness different from the first brightness.

24. The identification step described in claim 23, A sorting method characterized by comprising: a sorting step of sorting the objects based on the identification result of the identification step.

25. The sorting process described in claim 24, A method for manufacturing an article, comprising a processing step of manufacturing an article by processing the objects selected by the sorting step.

Images