Component mounting machine, image processing method, and computer program

Abstract

This component mounting machine comprises: a component holding unit that holds a component; an imaging device that can capture an image of the component held by the component holding unit from the side; and an image processing device that acquires the orientation of the component held by the component holding unit on the basis of the image of the component captured by the imaging device. The component has a first side positioned on a first direction side of a substrate surface when the component is mounted on the substrate surface, the first direction being orthogonal to the substrate surface, and a second side extending from an end of the first side in a direction intersecting the first side. The image processing device executes a first extraction process for extracting a plurality of first edges that are boundaries between the component and the background on the basis of a change in a first optical characteristic along a first scanning line extending in the first direction in the image of the component, and a second extraction process for extracting at least one second edge that is the boundary between the component and the background on the basis of a change in a second optical characteristic along a second scanning line extending in a second direction orthogonal to the first direction in the image of the component.

Description

Component mounter, image processing method, and computer program

[0001] The technology disclosed in this specification relates to a technology for acquiring the orientation of a part based on an image of the part.

[0002] Patent Document 1 discloses a placement machine equipped with a control device that processes image data of an image captured by a camera. The control device detects the bottom edge of an electronic component using a caliper tool.

[0003] International Publication No. 2016 / 075790

[0004] This specification provides a technique that can detect the orientation of a part more accurately based on an image of the part than conventional techniques.

[0005] The technology disclosed in this specification is embodied in a component mounter that mounts a component on a surface of a board. The component mounter includes a component holder that holds the component, an imaging device that can image the component held by the component holder from the side, and an image processing device that acquires the orientation of the component held by the component holder based on the image of the component captured by the imaging device. The component has a first side that is located on a first direction side of the board surface that is perpendicular to the board surface when mounted on the board surface, and a second side that extends from an end of the first side in a direction intersecting the first side. The image processing device executes a first extraction process to extract a plurality of first edges that are boundaries between the part and the background based on a change in a first optical property along a first scanning line extending in the first direction in the image of the part, and a second extraction process to extract at least one second edge that is a boundary between the part and the background based on a change in a second optical property along a second scanning line extending in a second direction perpendicular to the first direction in the image of the part.

[0006] In the above-described component mounter, the image processing device performs a first extraction process to extract a first edge in a first direction (a direction perpendicular to the board surface) in an image of the component captured from the side, and also performs a second extraction process to extract a second edge in a second direction (a direction parallel to the board surface) perpendicular to the first direction. This makes it possible to more accurately detect the orientation of the component based on the image of the component captured from the side, compared to conventional techniques that perform only the first extraction process.

[0007] Furthermore, the method for acquiring the component orientation in the component mounter and the computer program for acquiring the component orientation are also novel and useful.

[0008] FIG. 1 is a side view of a component mounter according to an embodiment; FIG. 2 is a configuration diagram of a control device provided in the component mounter; FIG. 3 is an example of an image of a component captured by an imaging device; FIG. 4 is a graph showing the relationship between gray value and position; FIG. 5 is an example of an image of a component captured by an imaging device, the image having a side that curves on an arc; and FIG. 6 is a flowchart of an orientation detection process executed by a control device.

[0009] The main features of the embodiments described below are listed below. Note that the technical elements described below are independent technical elements that exhibit technical utility alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing.

[0010] (Feature 1) In the above-described component mounter, the image processing device may further execute a calculation process to calculate an approximate straight line of the first side by using the plurality of first edges extracted by the first extraction process and the at least one second edge extracted by the second extraction process.

[0011] According to this configuration, by calculating the approximate straight line of the first side using not only the first edge but also the at least one second edge extracted by the second extraction process, it is possible to calculate the approximate straight line of the first side more accurately than the conventional technology in which the approximate straight line of the first side is calculated using only the first edge.

[0012] (Feature 2) In the above-described mounter, the first extraction process may extract the plurality of first edges arranged along the second direction, and the second extraction process may extract the second edge when it is determined that the second optical characteristic along the second scanning line extending through at least one first edge located at an end in the second direction among the plurality of first edges extracted by the first extraction process changes by more than a predetermined threshold. In this case, the calculation process may calculate an approximate straight line of the first side using the remaining first edges obtained by excluding, from the plurality of first edges, the at least one first edge through which the second scanning line passes that corresponds to the second optical characteristic that changes by more than the predetermined threshold.

[0013] For example, if the second edge of a component has a curved portion extending in an arc shape from the end of the first edge, the first edge may be extracted at the curved portion of the second edge. The first edge located at the curved portion of the second edge is offset from the extension line of the first edge and moves away from the extension line along the curved portion. Therefore, when an approximation line of the first edge is calculated using the first edge extracted at the curved portion of the second edge, the calculated approximation line inclines relative to the first edge, moving away from the first edge as it approaches the second edge. In other words, when an approximation line of the first edge is calculated using the first edge extracted at the curved portion, an accurate approximation line cannot be calculated. In the above-described component mounter, a second edge is considered to be extracted when it is determined that the second optical characteristic along a second scanning line extending through at least one first edge located at the end in the second direction (i.e., near the second edge) changes by more than a predetermined threshold. As described above, when a second edge is extracted by a second scanning line that extends through a first edge, the component included in the image is considered to have a second side having a curved portion. In the calculation process, an approximate straight line for the first side is calculated using the remaining first edges, excluding the first edge through which the second scanning line from which the second edge was extracted passed. With this configuration, the approximate straight line is calculated using the remaining first edges, excluding the first edge that is estimated to be located on the curved portion. Therefore, compared to a configuration in which an approximate straight line is calculated using all first edges, it is possible to calculate an accurate approximate straight line for, for example, a component having a second side with a curved portion.

[0014] (Feature 3) In the mounter, the image processing device may further execute a specification process for specifying a specific edge of the plurality of first edges that is located closest to the first direction, and a comparison process for comparing the number of one-side edges of the plurality of first edges that are located on the second direction side of the specific edge specified by the specification process with the number of other-side edges of the plurality of first edges that are located on the opposite side of the specific edge in the second direction. In this case, the calculation process may calculate the approximate straight line using the edge with the larger number of the one-side edges or the other-side edges.

[0015] According to this configuration, by calculating the approximate line using the edge with the larger number of one side edges or the other side edges, it is possible to calculate an accurate approximate line.

[0016] (Feature 4) In the mounter, the second scanning line may include a plurality of pixels arranged in the first direction, and the second optical characteristic may be determined based on an average value of luminance of the plurality of pixels. In this case, in the second extraction process, a position where a differential value of the average value of luminance exceeds a predetermined threshold may be extracted as the second edge.

[0017] The second optical characteristic changes significantly at the boundary between the part and the background, so this configuration makes it possible to accurately extract the second edge based on the differential value of the average luminance value of multiple pixels.

[0018] (Feature 5) In the mounter described above, the second scanning line may include a plurality of pixels arranged in the first direction. In this case, the second optical characteristic may be determined based on an average value of luminance of the plurality of pixels, and the number of the plurality of pixels may be determined according to a size of the component.

[0019] With this configuration, a second scan line containing a number of pixels appropriate for the size of the part can be used to extract the second edge.

[0020] (Feature 6) In the component mounter described above, the first direction may be a downward direction, and the first side may be a bottom side of the component.

[0021] However, in another embodiment, the first direction may be the left-right direction, in which case the first side may be a side edge of the component.

[0022] (Example) A component mounter 10 according to an example will be described with reference to the drawings. The component mounter 10 is a device that mounts components 4 on a surface 3 of a circuit board 2. The component mounter 10 is also called an electronic component mounting device or a chip mounter, and the components 4 are electronic components. Typically, the component mounter 10 is installed alongside other board work machines such as a solder printer and a board inspection machine, forming a continuous mounting line.

[0023] 1, the component mounter 10 includes a component feeder unit 12, a feeder holder 14, a head unit 15 composed of a mounting head 16 and a head moving device 18, an imaging device 30, a board conveyor 20, and a control device 22. The component feeder unit 12 has multiple feeders arranged in the +X direction (i.e., the direction into the paper in FIG. 1), and each feeder stores multiple components 4. Each feeder of the component feeder unit 12 is detachably attached to the feeder holder 14 and supplies components 4 to the mounting head 16. The component mounter 10 of this embodiment holds the circuit board 2 along the XY plane.

[0024] The placement head 16 has a nozzle 6 that picks up a component 4. That is, the nozzle 6 holds the component 4. The nozzle 6 is detachably attached to the placement head 16. The placement head 16 can move the nozzle 6 along the Z axis (i.e., up and down on the paper surface of FIG. 1 ) to move the nozzle 6 toward and away from each feeder of the component feeder unit 12 or the board surface 3 of the circuit board 2. That is, the placement head 16 moves the nozzle 6 in a direction perpendicular to the board surface 3. The placement head 16 can pick up components 4 from the feeders using the nozzle 6 and place the components 4 picked up by the nozzle 6 onto the circuit board 2. Note that the placement head 16 is not limited to having a single nozzle 6, and may have multiple nozzles 6.

[0025] The head moving device 18 moves the mounting head 16 and a fixed member 29 (described later) between each feeder of the component feeder unit 12 and the circuit board 2. As an example, the head moving device 18 in this embodiment is an XY robot that moves a moving base 18a along the XY plane, and the mounting head 16 is fixed to the moving base 18a. Note that the mounting head 16 does not have to be fixed to the moving base 18a, but may be detachably attached to the moving base 18a.

[0026] The imaging device 30 is fixed to the movable base 18a by a fixing member 29 and moves integrally with the movable base 18a. The imaging device 30 includes a camera 32, an illumination light source (not shown), and a prism (not shown). The camera 32 captures images of the side surface of the component 4 (see FIG. 4 (hereinafter also referred to as the side surface of the component 4)) held by the nozzle 6 and the lower end of the nozzle 6 from the side (i.e., the right side of the paper in FIG. 1 ). The camera 32 may be, for example, a CCD camera. The illumination light source is composed of an LED and illuminates the imaging surface of the component 4. The prism aligns the optical axis of the camera 32 with the imaging target. The illumination light source illuminates the side surface of the component 4 and the lower end of the nozzle 6, and the reflected light is reflected by the prism and directed to the camera 32, allowing the camera 32 to capture images of the side surface of the component 4 and the lower end of the nozzle 6. Image data of the images captured by the camera 32 is transmitted to the control device 22. The camera 32 is not limited to one that captures images of the side surfaces of the component 4, but may also capture images of the underside of the component 4, or may selectively capture images of the side surfaces and the underside of the component 4 (including the case where both are captured). Furthermore, the camera 32 may capture images of multiple side surfaces of the component 4. In this case, the nozzle 6 may be configured to be rotatable about the longitudinal axis of the component 4 while holding the component 4 by suction.

[0027] The board conveyor 20 is a device that carries in, positions, and carries out the circuit board 2. As an example, the board conveyor 20 of this embodiment includes a pair of belt conveyors and a support device (not shown) that supports the circuit board 2 from below.

[0028] The configuration of the control device 22 will be described with reference to FIG. 2 . The control device 22 controls the operation of the component mounter 10. The control device 22 is configured to be able to communicate with, for example, the head unit 15 and the imaging device 30. The control device 22 includes a memory 24 and a CPU 28. The memory 24 is configured with a volatile memory and a non-volatile memory. The memory 24 stores a production program 26, a threshold angle θth, a threshold differential value Dth, and a pixel number table T1. The production program 26 is a program for mounting components 4 on the circuit board 2 and includes, for example, information including the type and size of the component 4, mounting position data indicating the mounting position of the component 4, and path data indicating the path of the mounting head 16 to the mounting position. The threshold angle θth is a threshold angle indicating the orientation of the component 4 picked up by the nozzle 6 and is used to determine whether the component 4 will be properly mounted on the board surface 3 of the circuit board 2. The threshold differential value Dth is a threshold value for the differential value of the average gray value, and is a threshold value for detecting the boundary of the component 4 in the image captured by the imaging device 30. The pixel number table T1 is a table for determining the number of pixels of a scanning line (described later) based on the size of the component 4. The CPU 28 executes various processes in accordance with a program (e.g., the production program 26) stored in the memory 24. In this embodiment, the CPU 28 functions as, for example, a first extraction unit 29F, a second extraction unit 29S, and a calculation unit 29C.

[0029] Next, the processing executed by the control device 22 to detect the orientation of the part 4 will be described with reference to FIGS. 3 to 6. FIG. 3 shows an image 50A including the part 4 captured by the camera 32 of the imaging device 30. FIG. 4 shows a graph illustrating the relationship between the average gray value G1 and the differential value D1 of the average gray value G1 in the image 50A and the positions at which they are detected. FIG. 5 shows an image 50B including a part 40 with four corners curved like arcs. FIG. 6 shows a flowchart of the orientation detection processing executed by the control device 22.

[0030] As shown in Fig. 3, the image 50A includes the side surface of the component 4, the bottom end of the nozzle 6, and a background 60. The component 4 has a rectangular shape and includes a bottom side 4B, a left side 4L, and a right side 4R. As can be particularly seen by comparing Fig. 3 with Fig. 5, the four corners of the component 4 shown in Fig. 3 are formed at approximately right angles.

[0031] When the control device 22 receives the image 50A from the imaging device 30, the control device 22 detects the attitude of the component 4 picked up by the nozzle 6 based on the image 50A. In this embodiment, the control device 22 detects, based on the image 50A, the attitude angle θ1 between the bottom side 4B of the component 4 and a reference line BL that extends in the +X direction (i.e., the leftward direction on the paper in FIG. 3 ) parallel to the board surface 3.

[0032] When the control device 22 receives the image 50A from the imaging device 30, the control device 22 first sets a plurality of first scanning lines 51A for the component 4 in the image 50A. As shown in FIG. 3 , the plurality of first scanning lines 51A extend in the +Z direction (i.e., upward in the plane of FIG. 3 ). To facilitate understanding of the figure, in FIG. 3 , only the first scanning lines 51A located at both the left and right ends and the first scanning line 51A located in the center are labeled with reference numerals, and the reference numeral "51A" is omitted for the other first scanning lines. The same is true for FIG. 5 .

[0033] As shown in the enlarged partial view of the first scanning line 51A shown at the bottom of Figure 3, the first scanning line 51A is divided into multiple pixels. For example, in the first scanning line 51A located in the center, five pixels arranged in the -X direction are arranged in multiple rows in the +Z direction. The first scanning line 51A is composed of columns C1 to C5 and multiple rows (e.g., rows L1 to L4).

[0034] The control device 22 obtains the luminance (e.g., gray value) of each pixel along the projection direction P1 for each of the rows L1 to L4, and then calculates the average gray value of the pixels for each of the rows L1 to L4.

[0035] Here, with reference to FIG. 4 , the relationship between the average value G1 of the gray values ​​of each row and the derivative value D1 of the average value G1 will be described. The average value G1 is calculated, for example, by averaging the gray values ​​of the pixels in each column of the row to be calculated. The derivative value D1 of the average value G1 indicates the change in the average value G1 of adjacent rows and is calculated, for example, by applying a first-order derivative filter to the average value G1. In the graph of FIG. 4 , the vertical axis represents the average value G1 and the derivative value D1, and the horizontal axis represents the position on the scanning line in the direction in which the scanning line extends (i.e., the upward direction on the paper in FIG. 3 ). This position corresponds to the row of the scanning line (e.g., rows L1 to L4). In this embodiment, the average value G1 is maintained below 50 between positions 0 and 20. An average value G1 below 50 corresponds to the gray value of the background 60 included in image 50A. Because there is no change in the average value G1 between positions 0 and 20, the derivative value D1 is maintained at zero.

[0036] In image 50A, the gray values ​​of pixels located inside part 4 are higher than the gray values ​​of pixels located in background 60. Therefore, as shown by the average value G1 at positions 20 to 25, the average value G1 rises sharply at the boundary between part 4 and background 60, and the differential value D1 also rises sharply. As a result, the differential value D1 exceeds the threshold differential value Dth near position 25.

[0037] The control device 22 extracts the position where the differential value D1 exceeds the threshold differential value Dth as a first edge E10 (see FIG. 3 ), which is the boundary between the component 4 and the background 60. Specifically, the control device 22 extracts the space between the lines where the differential value D1 changes and exceeds the threshold differential value Dth (the boundary between lines L2 and L3 in the enlarged view of the first scanning line 51A in FIG. 3 ) as the first edge E10. The first edge E10 indicates the position of the bottom side 4B of the component 4 in the Z direction. The first edge E10 indicates the outline of the component 4. In this way, the control device 22 extracts the first edge for each of the multiple first scanning lines 51A based on the differential value D1 of the average value G1. As a result, the first edge E11 is extracted from the first scanning line 51 located furthest in the −X direction (i.e., toward the right in FIG. 3 ), and the first edge E19 is extracted from the first scanning line 51 located furthest in the X direction (i.e., toward the left in FIG. 3 ). First edges are similarly extracted from the other first scanning lines 51. However, in FIG. 3 , the symbols for the first edges other than the first edges E10, E11, and E19 are omitted.

[0038] The control device 22 calculates an approximate straight line AL1 using a plurality of first edges (e.g., first edges E10, E11, E19) extracted using a plurality of first scanning lines 51A. That is, the approximate straight line AL1 approximates the bottom side 4B of the component 4. Furthermore, the control device 22 calculates an attitude angle θ1 between the calculated approximate straight line AL1 and a reference line BL extending in the +X direction (i.e., a direction parallel to the board surface 3 of the circuit board 2).

[0039] The control device 22 compares the calculated attitude angle θ1 with the threshold angle θth stored in the memory 24. If the attitude angle θ1 is smaller than the threshold angle θth, the component 4 is properly picked up by the nozzle 6. In this case, the control device 22 mounts the component 4 on the surface 3 of the circuit board 2. On the other hand, if the attitude angle θ1 exceeds the threshold angle θth, the component 4 is not properly picked up by the nozzle 6. In this case, a component 4 that is not properly picked up by the nozzle 6 may not be accurately mounted on the surface 3 of the circuit board 2. Therefore, if the attitude angle θ1 exceeds the threshold angle θth, the control device 22 executes an attitude correction process to correct the attitude of the component 4. The attitude correction process is, for example, a process of discarding the component 4 in a disposal box (not shown) and then picking up a new component 4 with the nozzle 6. Note that in a modified example, the attitude correction process may be a process of temporarily stopping the nozzle 6 from picking up the component 4 to correct the orientation of the component 4, and then picking up the component 4 again.

[0040] In this way, the control device 22 detects the orientation of the component 4 by calculating the orientation angle θ1 of the component 4 based on the image 50A, and determines whether to mount the component or execute orientation correction processing by comparing the detected orientation with the threshold angle θth. This allows the mounter 10 to properly mount the component 4 on the board surface 3 of the circuit board 2.

[0041] 5, a case will be described in which the orientation of component 40 is detected based on image 50B. Component 40 is the same type of electronic component as component 4. However, unlike component 4 in FIG. 3, component 40 has four corners that are curved in an arc shape. Component 40 has a bottom side 40B, a left side 40L, and a right side 40R. Both ends of left side 40L and right side 40R of component 40 are curved in an arc shape. For example, when component 40 is manufactured, curved portions that are curved in an arc shape are formed at the four corners of component 40.

[0042] 3, in image 50B of component 40, multiple first scanning lines 51B are used to extract, for example, first edges E30, E31, E32, E38, and E39 indicating bottom side 40B. In component 40, for example, first edge E39 is extracted in the curved portion located below left side 40L by the leftmost first scanning line 51.

[0043] As shown in Fig. 5, the first edge E39 extracted in the curved portion of the left side 40L is located in the +Z direction (i.e., upward in the plane of the paper in Fig. 5) relative to the approximated line AL1 extending along the bottom side 40B. Therefore, if the approximated line AL2 is calculated taking into account all first edges including the first edge E39, the orientation angle θ2 between the approximated line AL2 and the reference line BL may be larger than the orientation angle θ1 between the approximated line AL1 and the reference line BL, as shown in Fig. 5. In other words, if the first edge E39 extracted in the curved portion is used in calculating the approximated line, it may not be possible to accurately detect the orientation of the part 40.

[0044] For this reason, the control device 22 of this embodiment further sets a second scanning line 52B after all first edges have been extracted. Unlike the first scanning line 51B, the second scanning line 52B extends in a direction perpendicular to the first scanning line 51B (i.e., the −X direction). The second scanning line 52B has five pixels arranged in the −Z direction, each of which is arranged in multiple rows in the −X direction. The second scanning line 52B is composed of columns C1 to C5 and multiple rows (e.g., rows L1 to L7). The control device 22 calculates the average gray value for each of rows L1 to L7 of the second scanning line 52B, and extracts a second edge at a position where a differential value D1 of the average gray value G1 in an adjacent row exceeds a threshold differential value Dth.

[0045] As described above, the first edge E39 extracted in the curved portion of the left side 40L is positioned in the +Z direction along the curved portion relative to the bottom side 40B. Therefore, as shown in the partially enlarged view of the second scanning line 52B at the bottom of FIG. 5 , the second scanning line 52B passing through the first edge E39 is set to a position that includes the component 40 over the entire range in the +Z direction. As a result, the average gray value of the second scanning line 52B, which includes the curved portion of the left side 40L, changes significantly. As described in FIG. 4 , when the average gray value G1 of multiple pixels changes, the differential value D1 also changes. As a result, the second edge E40 is extracted, for example, at the boundary between rows L1 and L2 of the second scanning line 52B. In this way, the mounter 10 of this embodiment can accurately extract the second edge E40 based on the differential value D1 of the average gray value G1.

[0046] In contrast, the first edge E19 extracted from the straight portion of the base side 4B shown in Fig. 3 is located on the base side 4B. Therefore, as shown in the partially enlarged view of the second scanning line 52A at the bottom of Fig. 3, on the second scanning line 52A that passes through the first edge E19, the component 4 is not included in the pixels of columns C4 and C5, which are arranged on the -Z direction side. As a result, for example, on row L2 of the second scanning line 52A, the average gray value of the pixels of columns C1 to C5 is smaller than the average gray value of the pixels of columns C1 to C5 of row L2 of the second scanning line 52B.

[0047] Therefore, in the second scanning line 52A in FIG. 3 , the differential value of the average value does not exceed the threshold differential value. As a result, a second edge is not extracted in the second scanning line 52A. By extracting the second edge in this manner, it is possible to determine whether a first edge is extracted on a curved portion. When the second edge E40 is extracted, the control device 22 calculates the approximate straight line AL1 using the remaining first edges (e.g., first edges E31 and E38) excluding the first edge E39 through which the second scanning line 52 passes. This allows for accurate calculation of the approximate straight line AL1 without calculating the approximate straight line AL2 taking the first edge E39 into account.

[0048] The orientation detection process executed by the control device 22 will be described with reference to Fig. 6. The process in Fig. 6 is triggered by the control device 22 acquiring an image (e.g., image 50B in Fig. 5) from the imaging device 30. Hereinafter, a part included in the image will be referred to as a "target part."

[0049] In S2, the control device 22 acquires the size of the side of the target part (for example, the length in the Z direction and the length in the X direction). The size of the target part is stored in the part information included in the production program 26.

[0050] In S4, the control device 22 uses the pixel number table T1 (see FIG. 2) to determine the number of pixels based on the size of the target part acquired in S2. Specifically, the control device 22 determines the number of columns (e.g., five columns) and rows of the scan lines. By determining the number of pixels based on the size of the target part in this way, edges can be extracted using scan lines containing a number of pixels appropriate for the size of the target part. Furthermore, by determining the number of pixels based on the size of the target part, it is possible to prevent the setting of scan lines that are unnecessarily long compared to the size of the target part, thereby reducing the load of the orientation detection process.

[0051] In S10, the control device 22 sets a plurality of first scanning lines for the bottom side of the target part. The control device 22 estimates the position of the bottom side of the target part using the lower end of the nozzle 6 as a reference, based on the size of the target part included in the production program 26. The control device 22 sets a number of first scanning lines corresponding to the length of the bottom side of the target part in the X direction, for a position in the Z direction in the image where the bottom side of the target part is estimated to be located.

[0052] In S12, the control device 22 extracts a plurality of first edges using a plurality of first scanning lines, as described with reference to FIGS.

[0053] In S14, the control device 22 identifies the bottom edge (for example, the first edge E31 in FIG. 5) located at the bottom among the plurality of first edges extracted in S12.

[0054] Next, in S20, the control device 22 compares the number of one-side edges located on one side of the bottom edge (e.g., the +X direction side) with the number of other-side edges located on the other side of the bottom edge (e.g., the -X direction side).

[0055] In S22, the control device 22 specifies the direction in which the edge with the larger number of one side edges or the other side edges is located as the measurement direction (for example, the +X direction in FIG. 5).

[0056] In S24, a second scanning line (e.g., second scanning line 52B in FIG. 5) is set, which extends in the −X direction and passes through a first edge (e.g., first edge E39 in FIG. 5) located at the end of the measurement direction identified in S22 among the multiple first edges extracted in S12.

[0057] In S30, the control device 22 determines whether or not a second edge has been extracted using the second scanning line set in S24. If a second edge has been extracted (YES in S30, see FIG. 5), the control device 22 proceeds to S32. If a second edge has not been extracted (NO in S30, see FIG. 3), the control device 22 skips S32 and proceeds to processing in S40.

[0058] In S32, the control device 22 deletes the first edge to which the second scanning line is set from among the multiple first edges extracted in S12, from the first edges located on the measurement direction side identified in S22.

[0059] Furthermore, in S40, the control device 22 calculates the approximate line using multiple first edges located closer to the measurement direction than the bottom edge. This allows for an accurate calculation of the approximate line using multiple first edges located closer to the measurement direction. Furthermore, for example, as shown in FIG. 5, the first edge E32 located on the curved portion of the right side 40R on the −X direction side of the bottom edge (first edge E31) is not used in calculating the approximate line. This allows for a more accurate calculation of the approximate line.

[0060] In S42, the control device 22 calculates the attitude angle θ between the approximation line calculated in S40 and the reference line. As described above, the control device 22 determines whether to continue mounting the target component or to perform attitude correction processing based on the calculated attitude angle θ.

[0061] (Effects of this embodiment) In this way, the mounter 10 of this embodiment executes a process of extracting a second edge (S30) in addition to a process of extracting a first edge using the first scanning line (S12 in FIG. 6). Therefore, even if the target component includes a curved portion on the left side, such as component 40 in FIG. 5, the posture angle θ can be accurately calculated.

[0062] Furthermore, the mounter 10 determines whether or not a curved portion is included on the left side of the target component by using a second edge (e.g., the second edge E40 in FIG. 5 ) in addition to the first edge, and then calculates an approximate line. This allows for more accurate calculation of the approximate line compared to conventional techniques that calculate an approximate line by using only the first edge.

[0063] (Correspondence) The nozzle 6 is an example of a "component holder." The control device 22 is an example of an "image processing device." The gray value is an example of an "optical characteristic." The bottom edge is an example of a "specific edge."

[0064] Here are some points to note regarding the mounter 10 described in the embodiment. The circuit board 2 may be held along the ZY plane. In this case, the mounting head 16 may move the nozzle 6 along the X axis. In this case, for example, the +X direction is an example of the "first direction," and the +Z direction is an example of the "second direction."

[0065] The mounter 10 may be provided with a chuck that clamps both sides of the component 4, instead of the nozzle 6. In this modification, the chuck is an example of a "component holder."

[0066] 5, the control device 22 may not calculate the approximate straight line AL1 of the bottom side 40B of the component 40. In this case, the control device 22 may extract, for example, a first edge located at the bottommost position on the bottom side 40B and a second edge located at the leftmost position on the left side 40L.

[0067] Furthermore, when the second edge is extracted using the second scanning line (YES in S30) instead of deleting the first edge in S32, the control device 22 may calculate the attitude angle θ in S42 and then correct the attitude angle θ using a correction coefficient stored in advance for the attitude angle θ. This makes it possible to correct the attitude angle θ when it is determined that a curved portion is included in the side surface of the part.

[0068] Furthermore, the control device 22 may not need to identify the bottom edge in S14. In that case, for example, the control device 22 may set second scanning lines for both the left edge 40L and the right edge 40R of the component 40. In this case, the control device 22 may delete the first edges located at both ends in the left-right direction from among the first edges in S32.

[0069] The process of S4 can be omitted. In this case, the control device 22 may set scanning lines including a predetermined number of columns and rows stored in advance, regardless of the component size.

[0070] The technical elements described in this specification or drawings exhibit technical utility either alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. Furthermore, the technologies illustrated in this specification or drawings simultaneously achieve multiple objectives, and achieving one of those objectives is itself technically useful.

[0071] For example, this specification also discloses the technical idea of ​​changing "the component mounter according to claim 1" to "the component mounter according to any one of claims 1 to 4" in claim 5. Similarly, the technical idea of ​​changing "the component mounter according to claim 1" to "the component mounter according to any one of claims 1 to 5" in claim 6, and the technical idea of ​​changing "the component mounter according to claim 1" to "the component mounter according to any one of claims 1 to 6" in claim 7 are also disclosed.

Claims

1. A component mounter for mounting components on a surface of a board, comprising: a component holding unit for holding the components; an imaging device capable of imaging the components held by the component holding unit from the side; and an image processing device for acquiring a posture of the components held by the component holding unit based on an image of the components captured by the imaging device, wherein the components have a first side located on a first direction side of the board surface perpendicular to the board surface when mounted on the board surface, and a second side extending from an end of the first side in a direction intersecting the first side, and the image processing device executes a first extraction process for extracting a plurality of first edges that are boundaries between the components and a background based on a change in a first optical characteristic along a first scanning line that extends in the first direction in the image of the components; and a second extraction process for extracting at least one second edge that is a boundary between the components and a background based on a change in a second optical characteristic along a second scanning line that extends in a second direction perpendicular to the first direction in the image of the components. Component mounting machine.

2. The component mounter according to claim 1, wherein the image processing device further executes a calculation process for calculating an approximate straight line of the first side by using the multiple first edges extracted by the first extraction process and the at least one second edge extracted by the second extraction process.

3. The component mounter according to claim 2, wherein in the first extraction process, the multiple first edges arranged along the second direction are extracted, and in the second extraction process, when it is determined that the second optical characteristic along the second scanning line extending through at least one first edge located at an end in the second direction among the multiple first edges extracted by the first extraction process changes beyond a predetermined threshold, the second edge is extracted, and in the calculation process, an approximation straight line of the first side is calculated using the remaining first edges obtained by excluding from the multiple first edges the at least one first edge through which the second scanning line corresponding to the second optical characteristic that changes beyond the predetermined threshold passes.

4. The component mounter of claim 2 or 3, wherein the image processing device further executes: an identification process for identifying a specific edge among the plurality of first edges that is located furthest on the first direction side; and a comparison process for comparing the number of one-side edges among the plurality of first edges that are located on the second direction side of the specific edge identified by the identification process, with the number of other-side edges among the plurality of first edges that are located on the opposite side of the specific edge in the second direction; and wherein the calculation process calculates the approximate straight line using the larger number of edges among the one-side edges and the other-side edges.

5. The component mounter according to claim 1, wherein the second scanning line includes a plurality of pixels arranged in the first direction, the second optical characteristic is determined based on an average value of luminance in the plurality of pixels, and in the second extraction process, a position where a differential value of the average value of luminance exceeds a predetermined threshold is extracted as the second edge.

6. The component mounter according to claim 1, wherein the second scanning line includes a plurality of pixels arranged in the first direction, the second optical characteristic is determined based on an average value of luminance in the plurality of pixels, and the number of the plurality of pixels is determined according to a size of the component.

7. The component mounter according to claim 1, wherein the first direction is a downward direction, and the first side is a bottom side of the component.

8. An image processing method for acquiring the posture of a component based on an image taken from the side of the component held by a component holding unit of a component mounter that mounts the component on a surface of a board, wherein the component has a first side located on a first direction side of the board surface perpendicular to the board surface when mounted on the board surface, and a second side extending from an end of the first side in a direction intersecting the first side, the image processing method including: a first extraction step for extracting a plurality of first edges that are boundaries between the component and a background based on a change in a first optical characteristic along a first scanning line that extends in the first direction in the image of the component; and a second extraction step for extracting at least one second edge that is a boundary between the component and the background based on a change in a second optical characteristic along a second scanning line that extends in a second direction perpendicular to the first direction in the image of the component.

9. A computer program for acquiring the posture of a component based on an image taken from the side of the component held by a component holding unit of a mounter that mounts the component on a surface of a board, wherein the component has: a first side located on a first direction side of the board surface perpendicular to the board surface when mounted on the board surface; and a second side extending from an end of the first side in a direction intersecting the first side, the computer program causing a computer to function as: a first extraction unit that extracts a plurality of first edges that are boundaries between the component and a background based on a change in a first optical characteristic along a first scanning line that extends in the first direction in the image of the component; and a second extraction unit that extracts at least one second edge that is a boundary between the component and the background based on a change in a second optical characteristic along a second scanning line that extends in a second direction perpendicular to the first direction in the image of the component.

Images