Component assembly machine, image processing method and computer program

Abstract

A component assembly machine comprises a component holding section configured to hold a component, an imaging device capable of capturing a lateral image of the component held by the component holding section, and an image processing device configured to detect the attitude of the component held by the component holding section based on an image of the component captured by the imaging device. The component has a first side that is orthogonal to the printed circuit board surface when the component is mounted on the printed circuit board surface, and a second side that extends from one end of the first side in a direction that intersects the first side.The image processing device performs a first extraction process in which at least one of several first edges is extracted, which, based on a change in a first optical property along a first scanning line extending in the first direction in the image of the component, represent a boundary between the component and a background, and a second extraction process in which at least one second edge is extracted, which, based on a change in a second optical property along a second scanning line extending in a second direction orthogonal to the first direction in the image of the component, represents a boundary between the component and the background.

Description

Technical field

[0001] The technique disclosed in the present description relates to a technique for capturing the attitude of a component based on an image of the component. State of the art

[0002] Patent literature 1 discloses a placement device comprising a control device that performs image processing on the image data of an image captured by a camera. The control device detects an edge of a lower end of an electronic component using a caliper. Citation list for patent literature

[0003] Patent literature 1: WO 2016 / 075790A Summary of the invention; Technical task

[0004] The present description provides a technique that is able to detect the attitude of a component more accurately based on an image of the component compared to a conventional technique. Solution to the task

[0005] The technology disclosed in the present description is embodied by a component assembly machine for mounting a component onto the surface of a printed circuit board. The component assembly machine has a component holding section configured to hold the component, an imaging device capable of capturing a lateral image of the component held by the component holding section, and an image processing device configured to detect the posture of the component held by the component holding section based on the image of the component captured by the imaging device.The component has a first side that is located on a first directional side of the printed circuit board surface orthogonal to the printed circuit board surface when the component is mounted on the printed circuit board surface, and a second side that extends from an end of the first side in a direction that intersects the first side.The image processing device performs a first extraction process in which at least one of several first edges, which are boundaries between the component and a background, is extracted based on a change in a first optical property along a first scanning line extending in the first direction in the image of the component, and a second extraction process in which at least one second edge, which is a boundary between the component and the background, is extracted based on a change in a second optical property along a second scanning line extending in a second direction orthogonal to the first direction in the image of the component.

[0006] In the component assembly machine described above, the image processing device performs a second extraction process—extracting the second edge in the second direction (parallel to the printed circuit board surface) orthogonal to the first direction—in addition to the first extraction process, which extracts the first edges in the first direction (orthogonal to the printed circuit board surface) from the image obtained by capturing the component from the lateral side. This allows for more accurate detection of the component's orientation based on the image obtained from the lateral side than with conventional techniques that only perform the first extraction process.

[0007] Furthermore, a method for recording the position of a component in the component assembly machine and a computer program for recording the position of a component are also new and useful. Brief description of the drawings Fig. Figure 1 is a side view of a component assembly machine according to an example. Fig. Figure 2 is a configuration view of a control device provided in the component assembly machine. Fig. Figure 3 is an example of an image of a component that was captured by an imaging device. Fig. Figure 4 is a diagram that illustrates a relationship between gray values ​​and positions. Fig. Figure 5 is an example of an image taken by the imaging device of a component that has a curved side on an arc. Fig. Figure 6 is a flowchart of the posture detection processing performed by the control device. Description of embodiments

[0008] The main features of the examples described below are listed. The technical elements described below are each independent technical elements that, independently of one another or in various combinations, have a technical benefit and are not limited to the combinations disclosed in the originally filed claims.

[0009] (Feature 1) In the component assembly machine described above, the image processing device can further perform a computational processing operation to calculate an approximate straight line of the first side using the plurality of first edges extracted by the first extraction processing operation and the at least one second edge extracted by the second extraction processing operation.

[0010] According to such a configuration, by calculating the approximate straight line of the first side using at least one second edge extracted by the second extraction processing, in addition to the first edges, the approximate straight line of the first side can be calculated more accurately than with the conventional technique, where the approximate straight line of the first side is calculated using only the first edges.

[0011] (Feature 2) In the component assembly machine described above, in the first extraction process, the plurality of first edges arranged along the second direction can be extracted, and in the second extraction process, the second edge can be extracted if it is found that the second optical property along the second scanning line extending through at least one first edge located at an end section in the second direction changes beyond a predetermined limit among the plurality of first edges extracted by the first extraction process.In this case, during the calculation process, the approximately straight line of the first side can be calculated using the remaining first edges, with the exception of the at least one first edge through which the second scanning line, corresponding to the second optical property that changes beyond the predetermined limit, passes from the multitude of first edges.

[0012] If, for example, the second side of the component has a curved section extending in an arc from one end of the first side, the first edge can be extracted from the curved section of the second side. The first edge located on the curved part of the second side is offset relative to an extension line of the first side and separated from the extension line along the curved part. Therefore, when the approximately straight line of the first side is calculated using the first edge extracted from the curved part of the second side, the calculated approximately straight line is inclined relative to the first side, such that it moves away from the first side as the approximately straight line approaches the second side.This means that if the approximately straight line of the first side is calculated using the first edge extracted from the curved part, an exact, approximately straight line cannot be calculated. In the component assembly machine described above, the second edge is assumed to be extracted when it is determined that the second optical property changes beyond the specified limit along the second scanning line, which extends through the at least one first edge at the end section in the second direction (i.e., near the second side). If, as described above, the second edge is extracted by the second scanning line extending through the first edge, it is assumed that the component included in the image has the second side that contains the curved section.During the computational processing, the approximately straight line of the first side is calculated using the remaining first edges, except for the first edge through which the second scan line, from which the second edge is extracted, passed. In such a configuration, the approximately straight line is calculated using the remaining first edges, except for the first edge assumed to be on the curved part. Therefore, compared to a configuration where the approximately straight line is calculated using all first edges, an exact approximately straight line can be calculated, for example, for the component that has the second side with the curved section.

[0013] (Feature 3) In the component assembly machine described above, the image processing device can further perform a determination process to identify a specific edge that is closest to the side with the first direction among the plurality of first edges, and a comparison process to compare the plurality of first side edges located on the side with the second direction of the specific edge, determined by the determination process among the plurality of first edges, with the number of second side edges located on a side opposite the second direction of the specific edge among the plurality of first edges. In this case, the approximately straight line can be calculated using edges with a larger number of edges among the first side edges and the second side edges.

[0014] According to such a configuration, it is possible to calculate an accurate approximately straight line by calculating the approximately straight line using the edges with the larger number of edges among the first side edges and the second side edges.

[0015] (Feature 4) In the component assembly machine described above, the second scanning line can have a plurality of pixels arranged in the first direction, and the second optical property can be determined based on an average luminance value across the plurality of pixels. In this case, during the second extraction process, a position where a difference value of the average luminance exceeds a predefined threshold can be extracted as the second edge.

[0016] The second optical property changes significantly at the boundary between the component and the background. Therefore, according to such a configuration, the second edge can be accurately extracted based on the difference in the mean luminance value across a multitude of pixels.

[0017] (Feature 5) In the component assembly machine described above, the second scanning line can have a plurality of pixels arranged in the first direction. In this case, the second optical property can be determined based on an average luminance value across the plurality of pixels, and the number of pixels in the plurality can be determined depending on the size of the component.

[0018] According to such a configuration, the second edge can be extracted using the second scan line, which has the appropriate number of pixels for the size of the component.

[0019] (Feature 6) In the component assembly machine described above, the first direction can be a downward direction, and the first side can be a bottom side of the component.

[0020] In another embodiment, however, the first direction can also be a left-right direction. In this case, the first side can be a lateral side of the component. (Example)

[0021] The following describes an example component assembly machine 10 with reference to the drawings. Component assembly machine 10 is a device that mounts components 4 onto the printed circuit board surface 3 of the printed circuit board 2. Component assembly machine 10 is also referred to as an electronic component assembly device or a chip placement device, and component 4 is an electronic component. Typically, component assembly machine 10 is provided together with other printed circuit board processing machines, such as a solder printing machine and a printed circuit board testing machine, to configure a series of assembly lines.

[0022] As in Fig. As shown in Figure 1, the component assembly machine 10 comprises a component feeding unit 12, a component holding section 14, a head unit 15 with an assembly head 16 and a head movement device 18, an imaging device 30, a printed circuit board transport device 20, and a control device 22. The component feeding unit 12 has several feeding devices that extend in a +X direction (i.e., in the depth direction of the paper surface). Fig. 1) are arranged, and each feeder holds several components 4. Each feeder of the component feeder unit 12 is detachably attached to a component holding section 14 and feeds the components 4 to the assembly head 16. The component assembly machine 10 of the present example holds the printed circuit board 2 along an XY plane.

[0023] The mounting head 16 has a nozzle 6 that receives the component 4. That is, the nozzle 6 holds the component 4. The nozzle 6 is detachably attached to the mounting head 16. The mounting head 16 is able to move the nozzle 6 along a Z-axis (i.e., a vertical direction of the paper surface). Fig. 1) The assembly head 16 moves the nozzle 6 back and forth to each feeder of the component feeder 12 or to the printed circuit board surface 3 of the printed circuit board 2. That is, the assembly head 16 moves the nozzle 6 in a direction orthogonal to the printed circuit board surface 3. The assembly head 16 can pick up the component 4 from the feeder through the nozzle 6 and mount the component 4 picked up by the nozzle 6 onto the printed circuit board 2. The assembly head 16 is not limited to an assembly head with a single nozzle 6, but can also be an assembly head with a plurality of nozzles 6.

[0024] The head movement device 18 moves the assembly head 16 and the fastening element 29 (described later) between each feeder of the component feeder unit 12 and the printed circuit board 2. In the present example, the head movement device 18 is, for instance, an XY robot that moves the movable base 18a in the XY plane, and the assembly head 16 is attached to the movable base 18a. The assembly head 16 is not limited to being permanently attached to the movable base 18a, but can be a detachable assembly head.

[0025] The imaging device 30 is attached to the movable base 18a by the fastening element 29 and moves together with the movable base 18a. The imaging device 30 comprises a camera 32, an illumination light source (not shown), and a prism (not shown). The camera 32 captures an image of a side surface of component 4, which is captured by the nozzle 6 (see Fig. 4 (hereinafter also referred to as the side surface of component 4)), and a lower end section of the nozzle 6 from a lateral side (i.e. from the right side of the paper surface in Fig. 1) For example, a CCD camera is used as camera 32. The illumination light source has an LED and illuminates an imaging surface of component 4. The prism aligns an optical axis of camera 32 with an imaging target. The side surface of component 4 and the lower end section of nozzle 6 are illuminated by the illumination light source, and the reflected light is reflected by the prism and directed to camera 32, so that camera 32 captures an image of the side surface of component 4 and the lower end section of nozzle 6. The image data of the image captured by camera 32 is transmitted to the control device 22.The camera 32 is not limited to a camera that captures an image of the side surfaces of component 4, but can also be a camera that captures an image of the bottom surface of component 4, or a camera that optionally captures an image of both the side surfaces and the bottom surface of component 4 (including the case where both images are captured). Furthermore, the camera 32 can capture images of multiple side surfaces of component 4. In this case, the nozzle 6 can be configured to rotate about a longitudinal axis of component 4 when component 4 is being captured.

[0026] The printed circuit board transport device 20 is a device for inserting, positioning and removing printed circuit boards 2. The printed circuit board transport device 20 of the present example has, for example, a pair of conveyor belts and a support device (not shown) that supports the printed circuit board 2 from below.

[0027] A configuration of the control device 22 is described with reference to Fig. 2 described. The control device 22 controls the operation of the component assembly machine 10. The control device 22 is configured to communicate, for example, with the head unit 15 and the imaging device 30. The control device 22 has a memory 24 and a CPU 28, wherein the memory 24 comprises volatile memory and non-volatile memory. The memory 24 stores the production program 26, the limit angle θth, the limit differential value Dth, and the pixel count table T1. The production program 26 is a program for mounting the component 4 onto the printed circuit board 2 and includes, for example, information about the type and size of the component 4, mounting position data specifying a mounting position for the component 4, and path data specifying a path for the mounting head 16 to the mounting position.The limiting angle θth is a limit value for an angle indicating the orientation of component 4 as picked up by the nozzle 6, and is a limit value for determining whether component 4 is properly mounted on the printed circuit board surface 3 of the printed circuit board 2. The limiting differential value Dth is a limit value for the differential of a mean value of grayscale values, and is a limit value for detecting a boundary of component 4 in the image captured by the imaging device 30. The pixel counting table T1 is a table for determining the number of pixels of a later-described scan line based on the size of component 4. The CPU 28 performs various types of processing according to a program stored in memory 24 (e.g., the production program 26). In the present example, the CPU 28 functions, for example, as the first extraction section 29F, the second extraction section 29S, and the computation section 29C.

[0028] Next, the processing performed by the control device 22 to detect the position of component 4 is described with reference to the Fig. 3 to 6 described. Fig. Figure 3 represents an image 50A with component 4, which was taken by the camera 32 of the imaging device 30. Fig. Figure 4 presents a diagram showing the relationship between the mean G1 of the grey values ​​and the differential D1 of the mean G1 in Figure 50A, as well as the positions where these values ​​are recorded. Fig. Figure 5 represents an image 50B that has a component 40 whose four corners are curved in an arc. Fig. Figure 6 shows a flowchart of the posture detection processing performed by the control device 22.

[0029] As in Fig. Figure 3 shows, in Figure 50A, side surfaces of component 4, a lower end of the nozzle 6, and the background 60. Component 4 has a rectangular shape and features the bottom 4B, the left side 4L, and the right side 4R. As can be seen from the comparison of the Fig. 3 and Fig. 5 emerges, the four corners of the in Fig. Component 4, as shown in section 3, is essentially rectangular in shape.

[0030] Upon receiving image 50A from the imaging device 30, the control device 22 detects the orientation of component 4 picked up by the nozzle 6 based on image 50A. In this example, the control device 22 determines the orientation angle θ1 between the underside 4B of component 4 and the reference line BL, which is located in the +X direction (i.e., in the left direction of the paper surface), based on image 50A. Fig. 3) extends parallel to the printed circuit board surface 3.

[0031] Upon receiving image 50A from the imaging device 30, the control device 22 first defines a plurality of first scanning lines 51A for component 4 in image 50A. As shown in Fig. As shown in Figure 3, a multitude of first scanning lines 51A extend in the +Z direction (i.e., a top surface of the paper in Fig. 3) To facilitate understanding of the drawing, the following are included: Fig. 3. Of the multitude of first scanning lines 51A, only the first scanning lines 51A at the left and right ends and the first scanning line 51A in the middle are marked with a reference sign, while the reference sign "51A" is omitted for the other first scanning lines. The same applies to Fig. 5 in relation to the reference signs of the first scanning lines.

[0032] As shown in a partially enlarged view of the first scanning line 51A in the lower part of Fig. As shown in Figure 3, the first scan line 51A is subdivided into a multitude of pixels. For example, in the first scan line 51A, which is located in the middle, five pixels arranged in the -X direction are arranged in the +Z direction across a multitude of rows. The first scan line 51A has columns C1 to C5 and a multitude of rows (for example, rows L1 to L4).

[0033] The control device 22 detects the luminance (e.g., the gray value) of each pixel along the projection direction P1 for each of the rows L1 to L4. Furthermore, the control device 22 calculates an average of the gray values ​​of the pixels for each of the rows L1 to L4.

[0034] Here, a relationship is established between the mean value G1 of the gray values ​​of each row and the differential value D1 of the mean value G1, with reference to Fig. 4 described. The mean G1 is calculated, for example, by averaging the gray values ​​of the pixels in each column of a row to be calculated. The differential D1 of the mean G1 is a value that indicates a change in the mean G1 of adjacent rows and is calculated, for example, by applying a primary differential filter to the mean G1. In the diagram in Fig. 4 represents a vertical axis, the mean value G1 and the differential value D1, and a horizontal axis represents a position of the scanning line in the direction in which the scanning line extends (i.e., an upward direction of the paper surface in Fig. 3) This position corresponds to the row of the sampling line (for example, rows L1 to L4). In this example, the mean value G1 between positions 0 and 20 is kept below 50. A mean value G1 of less than 50 corresponds to a background gray value of 60, which is shown in Figure 50A. Since the mean value G1 does not change between positions 0 and 20, the differential value D1 is kept at zero.

[0035] In Figure 50A, the gray value of a pixel located within component 4 is higher than the gray value of a pixel located in background 60. As can be seen from the mean value G1 of positions 20 to 25, the mean value G1 therefore increases rapidly at the boundary between component 4 and background 60, and the differential value D1 also increases rapidly. Consequently, the differential value D1 exceeds the limit differential value Dth near position 25.

[0036] The control device 22 extracts a position at which the differential value D1 exceeds the limit differential value Dth as the first edge E10 (see Fig. 3), which represents the boundary between component 4 and the background 60. More precisely, the control device 22 extracts an inter-row boundary as the first edge E10 (a boundary between row L2 and row L3 in the enlarged view of the first scan line 51A in Fig. 3), where the differential value D1 exceeds the limit differential value Dth. The first edge E10 indicates a position of the bottom surface 4B of component 4 in a Z-direction. The first edge E10 indicates a contour of component 4. As described above, the control device 22 extracts the first edge based on the differential value D1 of the mean value G1 for each of the plurality of first scan lines 51A. Consequently, the first edge E11 is extracted from the first scan line 51 that is located furthest in the -X direction (i.e., in the right direction of the paper surface in Fig. 3) is located, and the first edge E19 is extracted from the first scanning line 51, which is located furthest in the X direction (i.e. in the left direction of the paper surface in Fig. 3). In other first scan lines 51, the first edges are extracted in a similar way. Fig. However, the reference symbols of the first edges are omitted, with the exception of the first edges E10, E11 and E19.

[0037] The control device 22 calculates the approximately straight line AL1 using a plurality of first edges (e.g., first edges E10, E11, and E19) extracted by means of several first scanning lines 51A. That is, the approximately straight line AL1 approximates the bottom surface 4B of component 4. Furthermore, the control device 22 calculates the angle θ1 between the calculated approximately straight line AL1 and the reference line BL, which extends in the +X direction (i.e., in a direction parallel to the printed circuit board surface 3 of printed circuit board 2).

[0038] The control device 22 compares the calculated holding angle θ1 with the limiting angle θth stored in memory 24. If the holding angle θ1 is smaller than the limiting angle θth, component 4 is correctly picked up by the nozzle 6. In this case, the control device 22 mounts component 4 onto the printed circuit board surface 3 of the printed circuit board 2. However, if the holding angle θ1 exceeds the limiting angle θth, component 4 is not correctly picked up by the nozzle 6. In this case, there is a risk that component 4, which is not correctly picked up by the nozzle 6, will not be properly mounted onto the printed circuit board surface 3 of the printed circuit board 2. Therefore, if the holding angle θ1 exceeds the limiting angle θth, the control device 22 corrects the holding of component 4.The posture correction consists, for example, of throwing component 4 into a disposal box (not shown) and then extracting a new component 4 using nozzle 6. In a modified form, the posture correction can also consist of temporarily stopping the intake of component 4 by nozzle 6 and then re-intake the component 4 to correct its posture.

[0039] As described above, the control device 22 detects the position of component 4 by calculating the attitude angle θ1 of component 4 based on Figure 50A and determines, by comparing the detected attitude with the limit angle θth, whether the component is to be mounted or whether attitude correction processing is to be carried out. Accordingly, the component assembly machine 10 can mount component 4 appropriately onto the printed circuit board surface 3 of the printed circuit board 2.

[0040] Here, a case is described in which the attitude of component 40 is recognized based on image 50B, with reference to Fig. 5. Component 40 is an electronic component of the same type as component 4. Unlike component 4 in Fig. However, the four corners of component 40 are curved in an arc. Component 40 has a bottom side 40B, a left side 40L, and a right side 40R. Both ends of the left side 40L and the right side 40R of component 40 are curved in an arc. During the manufacturing of component 40, for example, curved sections are formed at the four corners of component 40.

[0041] In image 50B of component 40, for example, the first edges E30, E31, E32, E38 and E39, which show the bottom side 40B, are extracted by a multitude of first scan lines 51B, similar to component 4 in Fig. 3. In component 40, for example, the first edge E39 on a curved section below the left side 40L is extracted by the first scan line 51, which is located on the leftmost side.

[0042] As in Fig. As shown in Figure 5, the first edge E39, which was extracted from the curved section of the left side 40L, is located in the +Z direction (i.e., at the top of the paper surface in Fig. 5) with respect to the approximately straight line AL1 extending along the bottom 40B. Thus, when the approximately straight line AL2 is calculated taking into account all first edges, including the first edge E39, there is a case where the attitude angle θ2 between the approximately straight line AL2 and the reference line BL is greater than the attitude angle θ1 between the approximately straight line AL1 and the reference line BL, as shown in Fig. Figure 5 shows that if the first edge E39, extracted from the curved section, is used to calculate the approximate straight line, there is a risk that the orientation of component 40 cannot be accurately determined.

[0043] Therefore, the control device 22 of the present example further defines the second scan line 52B after all first edges have been extracted. Unlike the first scan line 51B, the second scan line 52B extends in one direction orthogonal to the first scan line 51B (i.e., in the -X direction). In the second scan line 52B, five pixels arranged in the -Z direction are arranged in the -X direction over a plurality of rows. The second scan line 52B has columns C1 to C5 and a plurality of rows (for example, rows L1 to L7). The control device 22 calculates a mean value of the gray values ​​for each of the rows L1 to L7 of the second scan line 52B and extracts a second edge at a position where the differential value D1 of the mean value G1 in the adjacent row exceeds the limit differential value Dth.

[0044] As described above, the first edge E39, extracted from the curved section of the left side 40L, is located in the +Z direction along the curved section with respect to the bottom 40B. Therefore, as shown in a partially magnified view of the second scan line 52B in a lower part in Fig. As shown in Figure 5, the second scan line 52B, which passes through the first edge E39, is fixed at a position that includes component 40 across its entire range in the +Z direction. This significantly alters the mean gray value in the series that includes the curved section of the left side 40L of the second scan line 52B. As shown in Fig. As described in section 4, a change in the mean value G1 of the gray values ​​across the multitude of pixels also changes the differential value D1. Consequently, for example, the second edge E40 is extracted at a boundary between rows L1 and L2 of the second scanning line 52B. As described above, in the component assembly machine 10 of the present example, the second edge E40 can be accurately extracted based on the differential value D1 of the mean value G1 of the gray values.

[0045] The first edge E19, which is located at a point in Fig. The linear section of the underside 4B shown in Figure 3, which was extracted, is located on the underside 4B. As shown in a partially enlarged view of the second scanning line 52A in the lower part of Fig. As shown in Figure 3, the component 4 in the pixels of columns C4 and C5, located on the side of the -Z direction, is not included in the second sampling line 52A, which passes through the first edge E19. Consequently, for example, in row L2 of the second sampling line 52A, the mean of the gray values ​​of the pixels in each of columns C1 to C5 is smaller than, for example, the mean of the gray values ​​of the pixels in each of columns C1 to C5 in row L2 of the second sampling line 52B.

[0046] Therefore, in the second scanning line, 52A exceeds Fig. 3. The differential value of the mean is not the limit differential value. Consequently, the second edge is not extracted from the second scan line 52A. In this way, it is possible to determine whether the first edge at the curved section has been extracted by extracting the second edge. When the second edge E40 is extracted, the control device 22 calculates the approximately straight line AL1 using the remaining first edges (e.g., the first edges E31 and E38), excluding the first edge E39, through which the second scan line 52 passed. Accordingly, it is possible to calculate an exact approximately straight line AL1 without calculating the approximately straight line AL2 taking into account the first edge E39.

[0047] The posture detection processing performed by the control device 22 is based on Fig. 6 described. The processing of Fig. 6 is executed when the control device 22 displays an image (e.g., image 50B in Fig. 5) is received as a trigger from the imaging device 30. A component included in the image is subsequently referred to as the “target component”.

[0048] In S2, the control device 22 detects a dimension (e.g., a length in the Z-direction and a length in the X-direction) of a side surface of the component. The dimension of the target component is stored in the component information contained in the production program 26.

[0049] In S4, the control device 22 determines the number of pixels based on the pixel counting table T1 (see Fig. 2) based on the size of the target component detected in S2. Specifically, the control device 22 determines the number of columns (e.g., five columns) and the number of rows of a scanning line. In this way, by determining the number of pixels based on the size of the target component, it is possible to extract an edge using a scanning line that has the number of pixels suitable for the size of the target component. Furthermore, determining the number of pixels based on the size of the target component prevents the creation of an unnecessarily long scanning line relative to the size of the target component and reduces the load on the posture detection processing.

[0050] In S10, the control device 22 defines a plurality of first scanning lines for a bottom surface of the target component. The control device 22 estimates the position of the bottom surface of the target component relative to the lower end of the nozzle 6 based on the size of the target component as defined in the production program 26. The control device 22 defines the number of first scanning lines corresponding to a length in the X-direction of the bottom surface of the target component relative to a position in the Z-direction where the bottom surface of the target component is estimated to be located in the image.

[0051] In S12, the control device 22 extracts, as with reference to Fig. 3 and Fig. As described in section 4, a multitude of first edges are obtained using multiple first scanning lines.

[0052] In S14, the control device 22 determines a lower end edge from the multitude of first edges extracted in S12, which is located on the bottom side (e.g., the first edge E31 in Fig. 5).

[0053] Next, the control device 22 in S20 compares the number of first side edges located on a first side (e.g., the side in the +X direction) of the lower end edge with the number of second side edges located on a second side (e.g., the side in the -X direction) of the lower end edge.

[0054] In S22, the control device 22 determines the measuring direction (e.g., the +X direction in Fig. 5) a direction in which the edges with the greater number of edges are located below the first and second side edges.

[0055] In S24, a second scanning line (e.g., second scanning line 52B) is selected from the multitude of first edges extracted in S12. Fig. 5) specified, which passes through the first edge located at an end section in the measuring direction determined in S22 (e.g. first edge E39 in Fig. 5) and extends in the -X direction.

[0056] In S30, the control device 22 determines whether the second edge has been extracted using the second scan line specified in S24. If the second edge has been extracted (YES in S30, see Fig. 5), the control device 22 continues with S32, and if the second edge was not extracted (NO in S30, see Fig. 3), skips control device 22 S32 and continues processing in S40.

[0057] In S32, the control device 22 deletes from the plurality of first edges extracted in S12 the first edge in which the second scanning line is defined, from the first edges that are on the side of the measuring direction determined in S22.

[0058] Furthermore, the control device 22 in S40 calculates an approximately straight line using the plurality of first edges located on the side of the measuring direction of the lower end edge. Accordingly, an exact approximately straight line can be calculated using the plurality of first edges located on the side of the measuring direction. Furthermore, for example, as in Fig. Figure 5 shows that the first edge E32, located on the curved section of the right side 40R on the -X-direction side of the lower end edge (first edge E31), is not used to calculate the approximate straight line. This allows the approximate straight line to be calculated more accurately.

[0059] In S42, the control device 22 calculates the attitude angle θ between the approximate straight line calculated in S40 and a reference line. As described above, the control device 22 determines, based on the calculated attitude angle θ, whether the assembly of the target component should continue or the attitude correction processing should be carried out. (Effects of the present example)

[0060] As described above, the component assembly machine 10 of the present example performs the processing (S30) of extracting the second edge in addition to the processing (S12 in Fig. 6) of extracting the first edges using the first scan lines and can therefore accurately calculate the attitude angle θ even if the left side of the target component has a curved section, as is the case with component 40 in Fig. 5.

[0061] Furthermore, the component assembly machine 10 calculates the approximately straight line after determining whether the curved section is contained in the left side of the target component by examining the second edge (e.g., the second edge E40 in Fig. 5) in addition to the first edge. This allows the approximately straight line to be calculated more accurately than with the conventional technique, where the approximately straight line is calculated only using the first edges.

[0062] (Correspondence relationship) The nozzle 6 is an example of a "component holding section". The control device 22 is an example of an "image processing device". The gray value is an example of an "optical property". The lower end edge is an example of a "specific edge".

[0063] The following section discusses the component assembly machine 10, which is described in the example. The printed circuit board 2 can be held along a ZY plane. In this case, the assembly head 16 can move the nozzle 6 along an X-axis. In this case, for example, the +X direction is an example of a "first direction" and the +Z direction is an example of a "second direction".

[0064] The component assembly machine 10 can have a clamping device that clamps both side surfaces of component 4 instead of the nozzle 6. In this modification, the clamping device is an example of a "component holding section".

[0065] The control device 22 does not need to follow an approximately straight line AL1 of the underside 40B of component 40 based on, for example, Figure 50B in Fig.5. In this case, the control device 22 can, for example, extract a first edge located on the bottom side of the underside 40B and a second edge located on the left side 40L.

[0066] Furthermore, instead of deleting the first edge in S32, the control device 22 can correct the attitude angle θ using a correction coefficient that was previously stored in the attitude angle θ after its calculation in S42 when the second edge is extracted using the second scan line (YES in S30). Thus, if it is determined that the side face of the component has a curved section, the attitude angle θ can be corrected.

[0067] Furthermore, the control device 22 in S14 does not need to determine the lower end edge. In this case, the control device 22 can, for example, define a second scanning line for both the left side 40L and the right side 40R of component 40. In this case, the control device 22 in S32 can delete the first edges that are located at both ends in the left-right direction below the first edges.

[0068] The processing of S4 can be omitted. In this case, the control device 22 can define a scanning line that has a predetermined number of columns and rows, stored in advance, regardless of the component size.

[0069] The technical elements described in this description or the drawings have technical benefits, whether independently or in various combinations, and are not limited to the combinations described in the originally filed claims. Furthermore, the technology illustrated by way of example in this description or the drawings simultaneously achieves a multitude of purposes, and the achievement of any one of these purposes itself has a technical benefit.

[0070] For example, the present description also discloses a technical idea in which "the component assembly machine according to claim 1" is amended in claim 5 to "the component assembly machine according to any one of claims 1 to 4". Likewise, a technical idea is disclosed in which "the component assembly machine according to claim 1" is amended in claim 6 to "the component assembly machine according to any one of claims 1 to 5", and a technical idea in which "the component assembly machine according to claim 1" is amended in claim 7 to "the component assembly machine according to any one of claims 1 to 6". QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] WO 2016 / 075790A

[0003]

Images