Resin member identification device and resin member identification method

Abstract

To provide an identification device and an identification method that can reduce the amount of data required for information processing and enable identification in a short time. [Solution] The identification device 20 for resin members 12 comprises an imaging means 26 that images the painted three-dimensional resin member 12, and an illumination means 24 that is positioned on the opposite side of the resin member 12 from the imaging means 26 and irradiates the resin member 12 with light. The image data captured by the imaging means 26 is divided into dark areas 42 and light areas 44, the dark areas 42 are extracted, and the resin member 12 imaged by the imaging means 26 is identified based on reference image data obtained in advance by imaging the resin member 12 to be identified.

Description

[Technical Field]

[0001] The present invention relates to a resin member identifying device and a resin member identifying method. [Background technology]

[0002] A classification device that accurately identifies the type of bread through image recognition is known, as described in Patent Document 1. This classification device divides a color image of bread into a color image of an inner region and a color image of an outer region, and determines from the image of the bread feature amounts related to the contour of the bread, feature amounts related to the texture of the bread, feature amounts within the color space of the inner region of the bread, and feature amounts within the color space of the outer region of the bread, thereby identifying the type of bread. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2011-170745 Summary of the Invention [Problem to be solved by the invention]

[0004] However, color images contain a large amount of information, and when a large number of features are obtained from a color image, such as features related to the contour, features related to the texture, features within the color space of the inner region, and features within the color space of the outer region of the bread, the amount of data required for information processing is large, which poses a problem in that the information processing takes a long time.

[0005] The present invention has been proposed in view of the above-mentioned problems associated with the conventional technology in order to preferably solve these problems, and has as its main object to provide an identification device that can reduce the amount of data required for information processing and perform identification in a short time. Another object is to provide an identification method that can reduce the amount of data required for information processing and perform identification in a short time. [Means for solving the problem]

[0006] A first aspect of the identification device according to the present invention comprises: an imaging means for imaging the painted three-dimensional resin member; an illumination means that is disposed on the opposite side of the resin member from the imaging means and that irradiates the resin member with light; an extraction means for dividing the image data captured by the imaging means into bright regions and dark regions and extracting the dark regions; an identification unit that identifies the resin member imaged by the imaging unit based on the dark region and a dark region of reference image data obtained by previously imaging the resin member to be identified; The gist of the project is that it is equipped with the following:

[0007] A second aspect of the resin member identification device according to the present invention is the first aspect of the identification device, The illumination means may be a surface illuminator having an area larger than the region in which the characteristics of the resin member appear.

[0008] A third aspect of the resin member identification device according to the present invention is the first or second aspect of the identification device, The identification means may identify the resin member imaged by the imaging means by comparing a feature amount obtained by quantifying a feature point of an area in the image data in which the characteristics of the resin member appear with a feature amount of the reference image data.

[0009] A fourth aspect of the resin member identification device according to the present invention is any one of the first to third aspects of the identification device, Further, a conveying means for conveying the resin member is provided, The imaging means may be transported by the transport means to a position in front of the lighting means, and may capture an image when an area in which a characteristic of the resin member appears is positioned in front of the imaging means.

[0010] A first aspect of the method for identifying a resin member according to the present invention includes: an extraction step of dividing image data captured by an imaging means that captures an image of a painted three-dimensional resin member from its front side while irradiating the back side with light into a light region and a dark region, and extracting the dark region; an identification step of identifying the resin member imaged by the imaging means based on the dark region and a dark region of reference image data obtained by previously imaging the resin member to be identified; The gist is that it includes the following. [Effects of the Invention]

[0011] According to the identification device of the present invention, information processing can be performed with a smaller amount of data than when identification is performed using a color image, and identification can be performed in a shorter time. According to the identification method of the present invention, information processing can be performed with a smaller amount of data than when identification is performed using a color image, and identification can be performed in a shorter time. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 2 is a block diagram illustrating electrical connections of the identification device according to the first embodiment. [Figure 2] FIG. 2 is an explanatory diagram illustrating an example of image data. [Figure 3] FIG. 10 is an explanatory diagram illustrating an example of reference image data. [Figure 4] 10 is a flowchart illustrating an example of a resin member identification processing routine. DETAILED DESCRIPTION OF THE INVENTION

[0013] Next, as an example of an embodiment of the present invention, the identification device 20 will be described in detail with reference to the drawings. Note that by describing a method for identifying a resin member 12 using the identification device 20, an example of a method for identifying a resin member 12 using the identification device 20 and an example of a method for using the identification device 20 will also be clarified.

[0014] The embodiments and drawings described below illustrate one example of an embodiment of the present invention and are not intended to limit the scope of the invention. Furthermore, the novel embodiments described below can be implemented in various other forms, and various omissions, substitutions, modifications, etc. can be made without departing from the spirit of the invention. Furthermore, the following embodiments and their modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as set forth in the claims. Corresponding components in each drawing are designated by the same or similar reference numerals.

[0015] 1, an identification device 20, which is an example of an embodiment of the present invention, includes a transport means 22 that transports a resin member 12 to be identified to a predetermined position, an illumination means 24 that irradiates light onto the resin member 12 transported to the predetermined position, an imaging means 26 that images the resin member 12 transported to the predetermined position, and an information processing device 30 that identifies the resin member 12 based on image data 40 (see FIG. 2) captured by the imaging means 26. The identification device 20 identifies the resin member 12 imaged by the imaging means 26 based on the image data 40 (see FIG. 2) captured by the imaging means 26 and reference image data 35 (see FIG. 3) obtained by previously capturing an image of the resin member 12 to be identified.

[0016] Here, the resin member 12 is, for example, a vehicle part, a three-dimensional resin molded product with a painted surface and an uneven shape, and is a non-transparent resin member. Specifically, the resin member 12 is a resin member with a wide variety of uneven shapes, colors, and surface smoothness, such as a vehicle exterior part such as a spoiler. In these cases, light incident on the resin member 12 may be reflected in many directions due to the uneven shape, or the intensity of the reflected light may change depending on the color and surface smoothness.

[0017] 1, the conveying means 22 is a known conveying means for conveying the resin member 12 to a position in front of the imaging means 26, such as a movable hanger that conveys the resin member 12 by hanging it from the movable hanger, or a belt conveyor that conveys the resin member 12. The resin member 12 is conveyed to the front position by the conveying means 22, and is imaged by the imaging means 26.

[0018] The front position of the imaging means 26 means a position where the center of the light irradiation surface of the illumination means 24 is aligned with the center of the lens (not shown) of the imaging means 26. When the characteristic region of the resin member 12 is located at this front position, light irradiated from the center of the illumination means 24 is irradiated onto the characteristic region of the resin member 12. In this way, light traveling straight from near the center of the illumination means 24 to near the center of the imaging means 26 is irradiated onto the outline of the characteristic region, which reduces the possibility of the outline being unclear when the imaging means 26 captures an image, compared to when light irradiated from near the center of the illumination means 24 to near the outer edge of the imaging means 26 is irradiated onto the outline.

[0019] The characteristic region of the resin member 12 is a region in which a characteristic shape of the resin member 12 appears, and is a region in which various types of resin members 12 can be distinguished from each other by their contours. The characteristic region is any predetermined region. In this embodiment, the characteristic region is a region including the leftmost corner of the resin member 12, and is the region shown enlarged in FIG. 2.

[0020] As shown in FIG. 1 , the illumination unit 24 is a surface illuminator having a known light source such as an LED or fluorescent lamp, with a diffuser (not shown) disposed in front of the light source. The illumination unit 24 is a highly directional light source that emits highly linear parallel light (light whose rays are nearly parallel) in a direction perpendicular to the light irradiation surface, and is disposed on the opposite side of the imaging unit 26 across the resin member 12. In the following description, the side of the resin member 12 that is imaged by the imaging unit 26 is referred to as the front side (surface side), and the side that is irradiated with light from the illumination unit 24 is referred to as the back side (rear side). The illumination unit 24 is disposed away from the back of the resin member 12, which is located at a predetermined position (the front position of the imaging unit 26). The distance between the illumination unit 24 and the resin member 12 is not particularly limited, but is, for example, approximately 300 mm to 2500 mm. The distance between the illumination unit 24 and the resin member 12 at the predetermined position makes it difficult for the light emitted from the illumination unit 24 to wrap around to the front side of the resin member 12, allowing the contours of the characteristic regions to be clearly imaged.

[0021] The light irradiation surface of the illumination means 24 has an area larger than the feature region of the resin member 12. Therefore, when the feature region of the resin member 12 is imaged by the imaging means 26 in front of the imaging means 26, the light is irradiated onto the entire feature region from the rear side. In other words, because the light is irradiated onto the entire surface of the feature region, the outline of the feature region can be clearly imaged.

[0022] The lighting means 24 is set so that when irradiating the resin member 12 with light, the illuminance on the surface on the rear side of the characteristic region of the resin member 12 is 2000 lux or more, or 3000 lux, and preferably 5000 lux or more brighter than the illuminance on the surface on the front side of the characteristic region of the resin member 12. The lighting means 24 is set so that the illuminance on the surface on the rear side of the characteristic region of the resin member 12 is 7100 lux or more, for example, depending on the conveying speed of the conveying means 22 and the imaging time of the imaging means 26, and is placed in a position where the center of the light irradiation surface and the center of a lens (not shown) of the imaging means 26 are aligned in a straight line. By doing this, by capturing an image when the characteristic region of the resin member 12 is located on a straight line connecting the center of the light irradiation surface of the illumination means 24 and the center of the imaging means 26, it is possible to capture an image in a state where the most direct light of the illumination means 24 passes through the contour of the characteristic region and is focused on the imaging means 26, thereby making it possible to accurately capture an image of the contour of the resin member 12 that appears in the characteristic region while suppressing the effects of the uneven shape, color, and surface smoothness on the front surface of the resin member 12.

[0023] The imaging unit 26 is a known imaging unit using a known monochrome image sensor, such as a monochrome CMOS image sensor or a monochrome CCD image sensor, and captures an image of the resin member 12 to obtain image data 40 (see FIG. 2 ). In this manner, the imaging unit 26 captures an image of the resin member 12 from the front side, irradiated with light from the rear side by the illumination unit 24, thereby obtaining image data 40 with a clear outline of the resin member 12. This image data 40 is transmitted to the information processing device 30 via a communication unit 28 provided in the imaging unit 26. The imaging unit 26 is disposed away from the front of the resin member 12, which is located at a predetermined position (the front position of the imaging unit 26). It is preferable to set the distance between the illumination unit 24 and the resin member 12 to be greater than the distance between the imaging unit 26 and the resin member 12. This configuration makes it difficult for light irradiated from the illumination unit 24 to reach the front side of the resin member 12, allowing the imaging unit 26 to clearly capture the outline of the characteristic region.

[0024] The imaging means 26 captures an image of the resin member 12 when the characteristic region of the resin member 12 is located at a predetermined position (front position of the imaging means 26) and acquires image data 40. In this embodiment, two-dimensional image data 40 is acquired. Specifically, for example, an operator may visually confirm that the resin member 12 is located at the predetermined position and press an imaging button (not shown) provided on the information processing device 30, or the image may be captured when the resin member 12 is transported to the predetermined position according to the transport speed of the transport means 22, or the image may be captured when a sensor (not shown) detects that the resin member 12 is located at the predetermined position.

[0025] The communication unit 28 is a known communication unit that receives a control signal that controls the imaging of the imaging unit 26 and outputs image data 40 output from the imaging unit 26. The image data 40 transmitted from the communication unit 28 is temporarily stored in the storage unit 34 of the information processing device 30.

[0026] As shown in FIG. 1 , the information processing device 30 is, for example, a computer. It includes: a control unit 32 that executes various programs, including a resin component identification processing routine that transmits control signals to control the imaging of the imaging means 26 and identifies the resin component 12 based on image data 40 transmitted from the imaging means 26; a memory unit 34 that stores various information, such as the resin component identification processing routine, the image data 40 transmitted from the imaging means 26, and reference image data 35; and a communication unit 36 ​​that transmits and receives various information, such as control signals to control the imaging of the imaging means 26 and the image data 40 transmitted from the imaging means 26. The control unit 32, the memory unit 34, and the communication unit 36 ​​are electrically connected via a bus 38. The information processing device 30 identifies the resin component 12 captured by the imaging means 26 based on the received image data 40. The information processing device 30 may be on-premise or cloud-based. In the cloud-based configuration, the above-described functions and processes may be provided, for example, through cloud computing.

[0027] The control unit 32 is, for example, a central processing unit (CPU), and realizes various functions of the information processing device 30 by reading various programs stored in the storage unit 34. In other words, information processing by software stored in the storage unit 34 is specifically realized by the control unit 32, which is an example of hardware, and can be executed as each functional unit included in the control unit 32. Note that the control unit 32 is not limited to being a single unit, and may be implemented with multiple control units 32 for each function, or a combination thereof. This control unit 32 processes and controls the overall operations related to the information processing device 30 and executes various steps of the resin component identification processing routine. Note that the control unit 32 corresponds to the extraction means and identification means of the present invention.

[0028] The memory unit 34 is, for example, a storage device such as a solid state drive (SSD) or a hard disk drive (HDD) that stores various programs executed by the control unit 32, or a memory such as a random access memory (RAM) that temporarily stores image data 40. In addition, as shown in Fig. 3, the memory unit 34 stores a plurality of reference image data 35 in association with the labels of the captured images of the resin members 12.

[0029] The communication unit 36 ​​is a known communication unit that transmits a control signal for controlling the imaging of the imaging unit 26 and receives image data 40 output from the communication unit 28 of the imaging unit 26. The image data 40 received by the communication unit 36 ​​is temporarily stored in the storage unit 34 via the bus 38.

[0030] Next, an operation performed when identifying the resin member 12 based on the image data 40 will be described using a resin member identification processing routine executed by the control unit 32 as an example. This resin member identification processing routine is repeatedly executed every time the information processing device 30 receives the image data 40 via the communication unit 36, and identifies the resin member 12 in the image data 40 captured by the imaging means 26. Note that FIG. 4 is a flowchart showing an example of the resin member identification processing routine.

[0031] First, the control unit 32 of the information processing device 30 extracts the dark region 42 from the image data 40 received via the communication unit 36 ​​and temporarily stores it in the storage unit 34 (step S110). Specifically, as shown in Fig. 2, the image data 40 is divided into two regions: a dark region 42 indicating the region of the resin member 12 and a light region 44 indicating the region other than the resin member 12, and only the dark region 42 indicating the region of the resin member 12 is extracted and temporarily stored in the storage unit 34. By doing so, the boundary between the dark region 42 and the light region 44 indicates the outline of the resin member 12, so that the outline of the resin member 12 can be extracted.

[0032] Next, the control unit 32 reads out the dark region 42 from the storage unit 34, extracts feature quantities from the dark region 42, and temporarily stores the feature quantities in the storage unit 34 (step S120). Specifically, the control unit 32 quantifies the feature quantities in a predetermined feature region from the dark region 42 read out from the storage unit 34. For example, as shown in FIG. 2, if the resin member 12 has a tip, the tip is defined as a feature region, and its angle is defined as a feature quantity in advance. Specifically, for example, the coordinate (A) located at the leftmost position of the dark region 42 in the image data 40 is set as the origin. Then, the slopes of the contour line (B) extending upward from the coordinate (A) of the dark region 42 and the contour line (C) extending downward from the coordinate (A) of the dark region 42 are extracted, and the angle formed by these contour lines is extracted as a feature quantity value and temporarily stored in the storage unit 34.

[0033] Next, the control unit 32 reads out the reference image data 35 from the storage unit 34 and extracts feature values ​​from the pre-stored reference image data 35 (step S130). Note that the method for extracting feature values ​​from the reference image data 35 is similar to the method for extracting feature values ​​from the image data 40, and therefore a description thereof will be omitted here.

[0034] Next, the control unit 32 compares the feature values ​​of the resin member 12 extracted in step S120 with the feature values ​​of the reference image data 35 extracted in step S130 (step S140). If it determines that the two values ​​match, the control unit 32 notifies the identification result of the resin member 12 by displaying a label stored in association with the reference image data 35 on a display unit (not shown) (step S150), and ends this routine. In this way, the resin member 12 can be identified based on the feature values ​​of the resin member 12. Note that a match in feature values ​​here includes not only a complete match in numerical values ​​of the feature values, but also a case where the feature values ​​are within a predetermined numerical range (for example, a case where the feature value of the resin member 12 is within ±10% of the feature value of the reference image data 35).

[0035] On the other hand, if the control unit 32 determines in step S140 that the values ​​of both feature quantities do not match, it determines whether or not other reference image data 35 is stored in the storage unit 34 (step S160), and if it determines that other reference image data 35 is not stored in the storage unit 34, it ends this routine. In such a case, it is found that there is no reference image data 35 that matches the image data 40 obtained from the resin member 12 imaged by the imaging means 26.

[0036] On the other hand, if the control unit 32 determines in step S160 that other reference image data 35 is stored in the storage unit 34, step S130 is executed again. In this way, by executing step S130 and step S140 for all of the reference image data 35 stored in the storage unit 34, it is possible to identify the resin member 12 imaged by the imaging means 26.

[0037] According to the identification device 20 of the embodiment described above, the illumination means 24 irradiates the painted three-dimensional resin member 12 with light from the rear side, and the imaging means 26 images the resin member 12 from the front side. This makes it possible to acquire image data 40 while suppressing the effects of the paint color, uneven shape, and surface smoothness of the resin member 12. The image data 40 acquired by the imaging means 26 is divided into dark regions 42 and light regions 44 in a resin member identification processing routine executed by the information processing device 30, and the resin member 12 is identified based on the feature amount of the dark region 42 (step S140). At this time, since the image data 40 is monochrome image information, the amount of information is smaller than in the case of color image information, and identification can be performed in a short time.

[0038] In addition, since the resin member 12 comes in a variety of colors, and the brightness of the entire image may vary depending on the color, when a color image is binarized, the contour may become unclear when binarized depending on the color, even if the resin member 12 has the same shape. Also, the three-dimensional shape of the resin member 12 may result in inaccurate contour extraction. However, by using the image data 40, which is monochrome image information, the contour can be extracted more accurately than when a color image is binarized.

[0039] Because the illumination means 24 is a surface illumination having an area larger than the feature region, which is the region where the features of the resin member 12 appear, it is possible to irradiate light of sufficient brightness from the back side, including the outline of the resin member 12, toward the illumination means 24. By doing so, the outline of the resin member 12 appears clearly as the boundary between the dark region 42 and the light region 44, so it is possible to accurately extract the outline of the resin member 12 and identify the resin member 12 with high precision. For example, if the outline of the resin member 12 has irregularities, the irregularities may cause diffuse reflection of light, making the outline unclear, but this possibility can be reduced.

[0040] In the resin component identification processing routine, the resin component 12 is identified by comparing feature amounts obtained by digitizing feature points in feature areas of the dark regions 42 extracted from the resin component 12 with feature amounts of pre-stored reference image data 35 (step S140). This makes it possible to identify the resin component 12 using the feature amounts in the feature areas. In other words, compared to when the entire dark region 42 of the resin component 12 is used as the comparison target, the amount of information required for identification can be reduced, while the processing speed and identification accuracy can be improved by using the feature amounts in the feature areas.

[0041] The imaging means 26 is transported to the front of the illumination means 24 by the transport means 22, and images the resin member 12 at a position where the center of the light irradiation surface of the illumination means 24 and the center of the lens (not shown) of the imaging means 26 are aligned on a straight line with respect to the characteristic region of the resin member 12. At this time, light emitted from the center of the illumination means 24 is irradiated onto the characteristic region of the resin member 12. By doing so, while the resin member 12 is transported by the transport means 22, light traveling straight from near the center of the illumination means 24 to near the center of the imaging means 26 is irradiated onto the outline of the characteristic region. This reduces the possibility of the outline being unclear when the imaging means 26 captures an image, compared to when light irradiated from near the center of the illumination means 24 to near the outer edge of the imaging means 26 is irradiated onto the outline.

[0042] For example, if the contour of the resin member 12 is uneven, light irradiated from a position where the center of a lens (not shown) of the imaging means 26 and the center of the light irradiation surface of the lighting means 24 are not aligned, will be diffused by the unevenness. When this diffused light is focused on the imaging means 26, the contour of the resin member 12 may become unclear. However, by aligning the center of the lens (not shown) of the imaging means 26 and the center of the light irradiation surface of the lighting means 24, the possibility of the contour becoming unclear due to the unevenness can be reduced.

[0043] (Example of change) The present invention is not limited to the above-described matters, and may be implemented, for example, as follows: It goes without saying that the present invention is not limited to the above-described embodiment and the following modified examples, and can be implemented in various forms as long as they fall within the technical scope of the present invention. (1) In the above-described embodiment, the illumination means 24 is a surface illumination provided with a diffusion plate, but the diffusion plate may be omitted. (2) In the above-described embodiment, the light irradiation surface of the lighting means 24 is larger than the characteristic region of the resin member 12, but it may be larger than the resin member 12. (3) In the above-described embodiment, the feature amounts of the reference image data 35 are extracted in step S130. However, the feature amounts of the reference image data 35 may be stored in advance in the storage unit 34 in association with the reference image data 35. This can shorten the processing time for extracting the feature amounts of the reference image data 35. (4) In the above-described embodiment, an angle value is used as a feature amount, but this is not limiting, and various values ​​such as the area or length of a specific region can be used as a feature amount. In these cases, the same effects as those of the above-described embodiment can be obtained. (5) In the above-described embodiment, the feature amounts of image data 40 extracted in step S120 are compared with the feature amounts of reference image data 35 extracted in step S130 in step S140. However, steps S130 and S140 may be omitted, and the dark areas 42 of image data 40 may be compared with the reference image data 35 in step S150. This can shorten the processing time required to extract the feature amounts. [Industrial Applicability]

[0044] As shown in the above-described embodiment, the present invention can be used in the field of image capture, particularly as an identification device for resin members. [Explanation of symbols]

[0045] 12...resin member, 20...identification device, 22...transport means, 24...illumination means, 26...imaging means, 35...reference image data, 40...image data, 42...dark area, 44...bright area

Claims

1. an imaging means for imaging the painted three-dimensional resin member; an illumination means that is disposed on the opposite side of the resin member from the imaging means and that irradiates the resin member with light; an extraction means for dividing the image data captured by the imaging means into bright regions and dark regions and extracting the dark regions; an identification unit that identifies the resin member imaged by the imaging unit based on the dark region and a dark region of reference image data obtained by previously imaging the resin member to be identified; A resin member identification device comprising:

2. 2. The resin member identifying device according to claim 1, wherein the illumination means is a surface illumination having an area larger than a region in which the characteristics of the resin member appear.

3. 3. The resin member identification device according to claim 1, wherein the identification means identifies the resin member imaged by the imaging means by comparing a feature quantity obtained by quantifying a feature point of an area in the image data in which the characteristics of the resin member appear with a feature quantity of the reference image data.

4. Further, a conveying means for conveying the resin member is provided, 4. The resin member identification device according to claim 3, wherein the imaging means is transported by the transport means to a position in front of the lighting means, and captures an image when an area in which a characteristic of the resin member appears is located in front of the imaging means.

5. an extraction step of dividing image data captured by an imaging means that captures an image of a painted three-dimensional resin member from its front side while irradiating the back side with light into a light region and a dark region, and extracting the dark region; an identification step of identifying the resin member imaged by the imaging means based on the dark region and a dark region of reference image data obtained by previously imaging the resin member to be identified; A method for identifying a resin member, comprising:

Images