Information processing device and method for updating model data

Abstract

In this information processing device, wherein a plurality of components scattered on a stage are imaged in advance and model data for identifying the plurality of components scattered on the stage is updated on the basis of correlation values acquired by comparing the images of the plurality of components imaged in advance and a plurality of model data candidates, when one model data candidate from among the plurality of model data candidates is compared with the images of the plurality of components and correlation values are acquired, the one model data candidate is excluded from the candidates for updating the model data if the number of component images among the plurality of component images for which it is determined that the correlation value is low exceeds a fixed number.

Description

Information processing device and method for updating model data

[0001] The present invention relates to an information processing device or the like that updates model data for identifying a plurality of parts scattered on a stage.

[0002] The following patent document describes a technique for comparing model data with image data of parts and the like.

[0003] Japanese Patent Publication No. 2008-170331 Japanese Patent Publication No. 10-50786

[0004] An object of the present invention is to appropriately update model data for identifying a plurality of parts scattered on a stage.

[0005] In order to solve the above-mentioned problems, this specification discloses an information processing device that updates model data for identifying a plurality of parts scattered on a stage based on a correlation value obtained by comparing a plurality of pre-imaged images of the plurality of parts with a plurality of model data candidates, and that, when comparing one of the plurality of model data candidates with an image of the plurality of parts to obtain a correlation value, excludes the one model data candidate from the candidates for updating the model data if the number of images of the parts determined to have a low correlation value exceeds a certain number among the images of the plurality of parts.

[0006] The present specification also discloses a method for updating model data for identifying a plurality of parts scattered on a stage, which includes an image acquisition step of capturing an image of a plurality of parts scattered on a stage to acquire images of the plurality of parts, and a step of comparing the images of the plurality of parts acquired in the image acquisition step with a plurality of model data candidates to acquire a correlation value between the images of the plurality of parts and the plurality of model data candidates, wherein, in the step of acquiring the correlation value, when one model data candidate among the plurality of model data candidates is compared with the image of the plurality of parts to acquire the correlation value, the one model data candidate for which the number of images of parts determined to have a low correlation value among the images of the plurality of parts exceeds a certain number is excluded from the candidates for updating the model data.

[0007] According to the present disclosure, when one of multiple model data candidates is compared with multiple component images to obtain a correlation value, a model data candidate for which the number of component images determined to have a low correlation value exceeds a certain number is excluded from candidates for updating model data, thereby making it possible to appropriately update model data for identifying multiple components scattered on a stage.

[0008] 17 is a perspective view of a component mounter. FIG. 18 is a perspective view of a component mounting device of the component mounter. FIG. 19 is a perspective view of a bulk component supply device. FIG. 19 is a perspective view of a component supply unit. FIG. 19 is a see-through view of a component supply unit. FIG. 19 is a see-through view of a component supply unit. FIG. 19 is a perspective view of a component scattering device. FIG. 19 is a perspective view of a component scattering device. FIG. 19 is a perspective view of a component holding head. FIG. 19 is a view of a component receiving member in which electronic circuit components are stored. FIG. 19 is a block diagram of a control device of a component mounter. FIG. 19 is a view of a stage on which lead components of the same shape are scattered. FIG. 19 is a view of model data used for pattern matching. FIG. 19 is a schematic view of a two-dimensional camera that images a plurality of lead components placed in different positions on a stage. FIG. 19 is a schematic view of a two-dimensional camera that images lead components illuminated with light from an LED light disposed on the side. FIG. 19 is a schematic view of a two-dimensional camera that images lead components illuminated with light from an LED light disposed on the side. FIG. 19 is a view showing lead components placed in the same orientation at five locations on a stage. FIG. 19 is a view showing lead components placed in an orientation different from that of FIG. 19 is a diagram showing lead components placed at five locations on a stage in orientations different from those in FIGS. 17 and 18. FIG. 19 is a diagram showing lead components placed at five locations on a stage in orientations different from those in FIGS. 17 to 19. FIG. 20 shows 20 types of lead components for which pattern matching is performed when creating model data. FIG. 21 shows a flowchart for creating model data using a conventional method. FIG. 22 shows correlation values ​​when pattern matching is performed using model data candidates for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage. FIG. 23 shows average correlation values ​​when pattern matching is performed using 16 types of model data candidates. FIG. 24 shows correlation values ​​when pattern matching is performed using model data candidates for a lead component having an edge level of 100 and an orientation of 270° at the center of the stage. FIG. 25 shows a flowchart for creating model data using the method of the present invention. FIG. 26 shows a flowchart for creating model data using the method of the present invention.1 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage. FIG. 2 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 100 and an orientation of 0° at the center of the stage. FIG. 3 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 50 and an orientation of 90° at the center of the stage. FIG. 4 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 100 and an orientation of 90° at the center of the stage. FIG. 5 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 50 and an orientation of 180° at the center of the stage. FIG. 6 is a diagram showing imaging data of a lead component when only one pair of leads is recognized. FIG. 7 is a diagram showing correlation values ​​when pattern matching is performed using candidate model data for a lead component having an edge level of 100 and an orientation of 180° at the center of the stage. 10A and 10B are diagrams showing correlation values ​​when pattern matching is performed using candidate model data of a lead component having an edge level of 50 and an orientation of 270° at the center of the stage, respectively. 10B and 10C are diagrams showing correlation values ​​when pattern matching is performed using candidate model data of a lead component having an edge level of 100 and an orientation of 270° at the center of the stage.

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings as modes for carrying out the present invention.

[0010] 1 shows a component mounter 10. The component mounter 10 is an apparatus for performing the operation of mounting components on a circuit board 12. The component mounter 10 includes an apparatus main body 20, a substrate transport and holding device 22, a component mounting device 24, imaging devices 26 and 28, a component supply device 30, a bulk component supply device 32, and a control device (see FIG. 11) 34. The circuit board 12 may be a circuit board or a substrate with a three-dimensional structure, and the circuit board may be a printed wiring board or a printed circuit board.

[0011] The device main body 20 is composed of a frame 40 and a beam 42 suspended from the frame 40. The substrate transport and holding device 22 is disposed in the center of the frame 40 in the front-to-rear direction and includes a transport device 50 and a clamp device 52. The transport device 50 transports the circuit board 12, and the clamp device 52 holds the circuit board 12. As a result, the substrate transport and holding device 22 transports the circuit board 12 and securely holds the circuit board 12 at a predetermined position. In the following description, the transport direction of the circuit board 12 is referred to as the X direction, the horizontal direction perpendicular to that direction is referred to as the Y direction, and the vertical direction is referred to as the Z direction. In other words, the width direction of the component mounter 10 is the X direction, and the front-to-rear direction is the Y direction.

[0012] The component mounting device 24 is mounted on the beam 42 and includes two work heads 60, 62 and a work head moving device 64. Each work head 60, 62 has a suction nozzle 66 (see FIG. 2) that holds a component. The work head moving device 64 also includes an X-direction moving device 68, a Y-direction moving device 70, and a Z-direction moving device 72. The X-direction moving device 68 and the Y-direction moving device 70 move the two work heads 60, 62 together to any position on the frame 40. As shown in FIG. 2, each work head 60, 62 is detachably attached to a slider 74, 76, and the Z-direction moving device 72 moves the slider 74, 76 individually in the vertical direction. That is, the work heads 60, 62 are moved individually in the vertical direction by the Z-direction moving device 72.

[0013] The imaging device 26 is attached to the slider 74 facing downward, and is movable in the X, Y, and Z directions together with the work head 60. This allows the imaging device 26 to capture an image of any position on the frame 40. The imaging device 28, as shown in FIG. 1, is disposed on the frame 40 between the substrate material conveying and holding device 22 and the component supply device 30 facing upward. This allows the imaging device 28 to capture an image of a component held by the suction nozzle 66 of the work heads 60, 62.

[0014] The component supply device 30 is disposed at one end of the frame 40 in the front-to-rear direction. The component supply device 30 has a tray-type component supply device 78 and a feeder-type component supply device (not shown). The tray-type component supply device 78 is a device that supplies components placed on a tray. The feeder-type component supply device is a device that supplies components using a tape feeder (not shown) or a stick feeder (not shown).

[0015] The bulk component supply device 32 is disposed at the other end of the frame 40 in the front-to-rear direction. The bulk component supply device 32 aligns a plurality of components scattered loosely and supplies the aligned components. In other words, the bulk component supply device 32 aligns a plurality of components in any orientation into a predetermined orientation and supplies the components in the predetermined orientation. The configuration of the component supply device 32 is described in detail below. Components supplied by the component supply device 30 and the bulk component supply device 32 include electronic circuit components, solar cell components, and power module components. Electronic circuit components include components with leads and components without leads.

[0016] As shown in FIG. 3, the bulk component supply device 32 includes a main body 80, a component supply unit 82, an imaging device 84, and a component delivery device 86.

[0017] The component supply unit 82 includes a component supply device 88, a component scattering device (see FIG. 4) 90, and a component returning device (see FIG. 4) 92, and these component supply devices 88, component scattering device 90, and component returning device 92 are integrally configured. The component supply units 82 are detachably attached to a base 96 of the main body 80, and in the bulk component supply device 32, five component supply units 82 are arranged in a row in the X direction.

[0018] The component supply device 88 has a generally rectangular box shape and is disposed so as to extend in the Y direction as shown in Figures 4 and 5. The Y direction is referred to as the front-to-rear direction of the component supply device 88, and the direction toward the side of the component supply unit 82 where the component returning device 92 is disposed is referred to as the front, and the direction toward the side where the component supply device 88 is disposed is referred to as the rear.

[0019] The component supply device 88 has openings on the top and front, with the opening on the top serving as a component inlet 97 and the opening on the front serving as a component outlet 98. In the component supply device 88, an inclined plate 104 is disposed below the inlet 97. The inclined plate 104 is disposed so as to slope downward from the rear end face of the component supply device 88 toward the center.

[0020] 5, a conveyor device 106 is disposed in front of the inclined plate 104. The conveyor device 106 is disposed so as to slope upward from the front end of the inclined plate 104 toward the front of the component supply device 88. The conveyor belt 112 of the conveyor device 106 rotates counterclockwise in FIG. 5. In other words, the conveying direction of the conveyor device 106 is diagonally upward from the front end of the inclined plate 104 toward the front.

[0021] An inclined plate 126 is disposed below the front end of the conveyor 106. The inclined plate 126 extends from the front end face of the component supply device 88 toward the bottom of the conveyor 106, with its rear end slanted diagonally downward. Furthermore, an inclined plate 128 is disposed below the inclined plate 126. The inclined plate 128 is inclined from below the center of the conveyor 106 toward the discharge port 98 of the component supply device 88, with its front end positioned downward.

[0022] 4, a pair of side frames 130 are attached to the base 96. The pair of side frames 130 are erected facing each other and parallel to each other, extending in the Y direction. The distance between the pair of side frames 130 is slightly larger than the width dimension of the component supplier 88, and the component supplier 88 is detachably mounted between the pair of side frames 130.

[0023] The component scattering device 90 includes a component support member 150 and a component support member moving device 152. The component support member 150 is composed of a stage 156 and a pair of side walls 158. The stage 156 has a generally longitudinal plate shape and is disposed so as to extend forward from below the component supply device 88, which is mounted between a pair of side frames 130. The upper surface of the stage 156 is generally horizontal and, as shown in FIG. 5, is disposed with a slight clearance from the front end of the inclined plate 128 of the component supply device 88. The pair of side walls 158 are fixed in an upright position on both longitudinal sides of the stage 156, as shown in FIG. 4, and the upper ends of the side walls 158 extend above the upper surface of the stage 156.

[0024] The component support member moving device 152 slides the component support member 150 in the Y direction by operating an air cylinder (see FIG. 11) 166. During this movement, the component support member 150 moves between a stored state (see FIG. 6) in which it is stored below the component feeder 88 and an exposed state (see FIG. 5) in which it is exposed from below the component feeder 88.

[0025] As shown in FIG. 7 , the component returning device 92 includes an object storage container 180 and a container swinging device 181. The object storage container 180 is generally box-shaped with an arc-shaped bottom. The object storage container 180 is swingably held at the front end of the stage 156 of the component support member 150 and swings by the operation of the container swinging device 181. In this case, the object storage container 180 swings between a storage position (see FIG. 7 ) in which the opening faces upward and a return position (see FIG. 8 ) in which the opening faces the upper surface of the stage 156 of the component support member 150.

[0026] As shown in FIG. 3 , the imaging device 84 includes a two-dimensional camera 290, a camera moving device 292, and an LED light (see FIG. 11 ) 294. The camera moving device 292 includes a guide rail 296 and a slider 298. The guide rail 296 is fixed to the main body 80 above the component supply device 88 and extends in the width direction (X direction) of the bulk component supply device 32. The slider 298 is slidably attached to the guide rail 296 and slides to any position by the operation of an electromagnetic motor (see FIG. 11 ) 299. The two-dimensional camera 290 is attached to the slider 298 while facing downward. Note that the two-dimensional camera 290 is not a three-dimensional camera capable of detecting height information, depth information, etc. of a three-dimensional object, nor is it equipped with a three-dimensional image processing device. It is a camera that captures two-dimensional images that cannot detect height information, depth information, etc. of a three-dimensional object. In other words, the two-dimensional camera 290 is a camera that captures a two-dimensional image of a three-dimensional component projected onto the XY plane. Furthermore, the two-dimensional camera 290 does not use a telecentric lens in which the aperture stop is at the focal position of the lens, but rather uses a general lens, i.e., a lens with a non-zero angle of view. Furthermore, the LED light 294 is the light source for the two-dimensional camera 290, and is disposed on the side of the two-dimensional camera 290, on the part delivery device 86 side. In other words, the two-dimensional camera 290 captures a two-dimensional image of an object illuminated by the LED light 294 from the side.

[0027] As shown in FIG. 3, the component delivery device 86 includes a component holding head moving device 300, a component holding head 302, and two shuttle devices 304.

[0028] The component holding head moving device 300 includes an X-direction moving device 310, a Y-direction moving device 312, and a Z-direction moving device 314. The Y-direction moving device 312 has a Y-slider 316 disposed above the component supply unit 82 so as to extend in the X direction. The Y-slider 316 is moved to any position in the Y direction by driving an electromagnetic motor (see FIG. 11) 319. The X-direction moving device 310 has an X-slider 320 disposed on a side of the Y-slider 316. The X-slider 320 is moved to any position in the X direction by driving an electromagnetic motor (see FIG. 11) 321. The Z-direction moving device 314 has a Z-slider 322 disposed on a side of the X-slider 320. The Z-slider 322 is moved to any position in the Z direction by driving an electromagnetic motor (see FIG. 11) 323.

[0029] As shown in FIG. 9 , the component holding head 302 includes a head main body 330, a suction nozzle 332, a nozzle swiveling device 334, and a nozzle rotating device 335. The head main body 330 is integrally formed with the Z-slider 322. The suction nozzle 332, which holds a component, is detachably attached to the lower end of a holder 340. The holder 340 is bendable about a support shaft 344, and the nozzle swiveling device 334 bends the holder 340 upward by 90 degrees. As a result, the suction nozzle 332 attached to the lower end of the holder 340 rotates 90 degrees and is positioned at a swiveled position. In other words, the suction nozzle 332 rotates between a non-swiveled position and a swiveled position by the operation of the nozzle swiveling device 334. Of course, it is also possible to position and stop the suction nozzle 332 at an angle between the non-swiveled position and the swiveled position. The nozzle rotating device 335 rotates the suction nozzle 332 around its axis.

[0030] 3, each of the two shuttle devices 304 includes a component carrier 388 and a component carrier moving device 390, and is fixed to the main body 80, side-by-side in the lateral direction, on the front side of the component supply unit 82. Five component receiving members 392 are attached to the component carrier 388, lined up in a row in the lateral direction, and components are placed on each of the component receiving members 392.

[0031] The bulk component supply device 32 is capable of supplying a variety of components, and various component receiving members 392 are prepared according to the shape of the components. Here, as an example of an electronic circuit component supplied by the bulk component supply device 32, a component receiving member 392 corresponding to a lead component 410 having leads, as shown in Figure 10, will be described. The lead component 410 is composed of a block-shaped component body 412 and two leads 414 protruding from the bottom surface of the component body 412.

[0032] The component receiving member 392 is also formed with a component receiving recess 416 shaped to accommodate the lead component 410. The component receiving recess 416 is a stepped recess, and is composed of a main body receiving recess 418 that opens on the top surface of the component receiving member 392, and a lead receiving recess 420 that opens on the bottom surface of the main body receiving recess 418. The lead component 410 is inserted into the component receiving recess 416 with the leads 414 facing downward. As a result, the leads 414 are inserted into the lead receiving recess 420, and the lead component 410 is placed inside the component receiving recess 416 with the component body 412 inserted into the main body receiving recess 418.

[0033] 3, the component carrier moving device 390 is a plate-shaped longitudinal member disposed in front of the component supply unit 82 so as to extend in the front-rear direction. A component carrier 388 is disposed on the upper surface of the component carrier moving device 390 so as to be slidable in the front-rear direction, and is driven by an electromagnetic motor 430 (see FIG. 11) to slide to any position in the front-rear direction. When the component carrier 388 slides toward the component supply unit 82, it slides to a component receiving position located within the range of movement of the component holding head 302 by the component holding head moving device 300. On the other hand, when the component carrier 388 slides away from the component supply unit 82, it slides to a component supply position located within the range of movement of the work heads 60, 62 by the work head moving device 64.

[0034] 11 , the control device 34 includes an overall control device 450, a plurality of individual control devices (only one of which is shown in the figure), and an image processing device 454. The overall control device 450 is configured mainly by a computer and is connected to the substrate conveying and holding device 22, the component mounting device 24, the imaging device 26, the imaging device 28, the component supply device 30, and the bulk component supply device 32. As a result, the overall control device 450 controls the substrate conveying and holding device 22, the component mounting device 24, the imaging device 26, the imaging device 28, the component supply device 30, and the bulk component supply device 32 in an overall manner. The plurality of individual control devices 452 are also configured mainly by a computer and are provided corresponding to the substrate conveying and holding device 22, the component mounting device 24, the imaging device 26, the imaging device 28, the component supply device 30, and the bulk component supply device 32 (only the individual control device 452 corresponding to the bulk component supply device 32 is shown in the figure).

[0035] The individual control device 452 of the bulk component supply device 32 is connected to the imaging device 84, the component scattering device 90, the component returning device 92, the component holding head moving device 300, the component holding head 302, and the shuttle device 304. As a result, the individual control device 452 of the bulk component supply device 32 controls the imaging device 84, the component scattering device 90, the component returning device 92, the component holding head moving device 300, the component holding head 302, and the shuttle device 304. The image processing device 454 is also connected to the two-dimensional camera 290 and processes image data captured by the two-dimensional camera 290. The image processing device 454 is also connected to the individual control device 452 of the bulk component supply device 32. As a result, the individual control device 452 of the bulk component supply device 32 acquires the image data captured by the two-dimensional camera 290.

[0036] The bulk component supply device 32 also has a storage device 458. The storage device 458 is connected to the individual control device 452, and stores various information in accordance with commands from the individual control device 452.

[0037] The component mounter 10, with the above-described configuration, performs a component mounting operation on the circuit board 12 held by the substrate conveying and holding device 22. Specifically, the circuit board 12 is conveyed to the work position and fixedly held at that position by the clamping device 52. Next, the imaging device 26 moves above the circuit board 12 and captures an image of the circuit board 12. This provides information regarding an error in the holding position of the circuit board 12. Furthermore, the component supply device 30 or the bulk component supply device 32 supplies components at a predetermined supply position. Note that the supply of components by the bulk component supply device 32 will be described in detail later. Then, one of the work heads 60, 62 moves above the component supply position and holds the component with the suction nozzle 66. Next, the work head 60, 62 holding the component moves above the imaging device 28, and the imaging device 28 captures an image of the component held by the suction nozzle 66. This provides information regarding an error in the holding position of the component. Then, the work heads 60, 62 holding the components move above the circuit board 12, and the components they are holding are mounted on the circuit board 12 after correcting any errors in the holding position of the circuit board 12, errors in the holding position of the components, etc.

[0038] Next, we will explain the supply of components by the bulk component supply device 32. In the bulk component supply device 32, a worker inserts lead components 410 into the inlet 97 of the component supply device 88. The component supply unit 82 and the component delivery device 86 then operate to supply the inserted lead components 410 placed on the component receiving members 392 of the component carrier 388.

[0039] Specifically, the operator loads a plurality of lead components 410 of the same type, i.e., a plurality of lead components 410 of the same shape, through the loading port 97 on the top surface of the component supply device 88. At this time, the component support member 150 has been moved to a position below the component supply device 88 by the operation of the component support member moving device 152, and is in a stored state (see FIG. 6). Note that when the component support member 150 is stored, the component storage container 180 disposed at the front end of the component support member 150 is positioned in front of the component supply device 88, with the opening of the component storage container 180 facing upward (the storage position).

[0040] The lead components 410 fed through the inlet 97 of the component supply device 88 fall onto the inclined plate 104 of the component supply device 88 and roll down to the lower end of the front side of the inclined plate 104. At this time, the lead components 410 that have rolled down to the lower end of the front side of the inclined plate 104 are piled up between the lower end of the front side of the inclined plate 104 and the lower end of the rear side of the conveyor device 106. Then, the conveyor belt 112 of the conveyor device 106 rotates counterclockwise in FIG. 6. As a result, the lead components 410 piled up between the inclined plate 104 and the conveyor belt 112 are transported diagonally upward by the conveyor belt 112.

[0041] Then, the lead components 410 transported by the conveyor belt 112 drop from the upper front end of the conveyor device 106 onto the inclined plate 126. The lead components 410 that have fallen onto the inclined plate 126 roll backward along the inclined plate 126 and drop onto the inclined plate 128. The lead components 410 that have fallen onto the inclined plate 128 roll forward and are discharged from the discharge port 98 on the front side of the component supply device 88.

[0042] As a result, the lead components 410 discharged from the discharge port 98 of the component supply device 88 are accommodated inside the component storage container 180. Then, when a predetermined amount of lead components 410 has been discharged from the component supply device 88, that is, when the conveyor device 106 has operated a certain amount, the conveyor device 106 stops. Next, the component support member 150 is moved forward from the stored state by operation of the component support member moving device 152.

[0043] Then, when the component support member 150 moves forward a predetermined distance from the stored state, the container swinging device 181 of the component returning device 92 is activated, swinging the component housing container 180. As a result, the orientation of the component housing container 180 changes vigorously from an orientation in which the opening faces upward (storage orientation) to an orientation in which the opening faces the stage 156 (return orientation). At this time, the lead components 410 accommodated in the component housing container 180 are expelled vigorously toward the stage 156. As a result, lead components 410 of the same shape are scattered from the component housing container 180 onto the stage 156, as shown in FIG. 12 .

[0044] The lead components 410 are scattered on the stage 156 in four different orientations. Specifically, in a first orientation, the lead components 410 are scattered with the extending surfaces of the leads 414 facing sideways, with two of the leads 414 aligned generally horizontally. In a second orientation, the lead components 410 are scattered with the extending surfaces of the leads 414 facing sideways, with two of the leads 414 aligned generally vertically. In a third orientation, the lead components 410 are scattered with the extending surfaces of the leads 414 facing upward. In a fourth orientation, the lead components 410 are scattered with two or more lead components 410 overlapping each other. The four positions in which the lead components 410 are scattered will be referred to as lead components 410a in a first position, lead components 410b in a second position, lead components 410c in a third position, and lead components 410d in a fourth position.

[0045] When the lead components 410 are scattered on the stage 156 as described above, the two-dimensional camera 290 of the imaging device 84 is moved above the component support member 150 by the operation of the camera moving device 292. The two-dimensional camera 290 then captures images of the lead components 410 of the same shape scattered on the stage 156. Because the viewing angle, i.e., the imaging range, of the two-dimensional camera 290 is wider than that of the stage 156, the two-dimensional camera 290 captures an image of the entire stage 156, i.e., all of the lead components 410 scattered on the stage 156, all at once. In other words, the two-dimensional camera 290 captures an image of all of the lead components 410 scattered on the stage 156 all at once, with the image captured within the imaging range. When the two-dimensional camera 290 captures an image of the lead components 410 on the stage 156, the LED light 294 irradiates light onto the top surface of the stage 156. Then, based on the image data captured by the two-dimensional camera 290, the individual control device 452 identifies the lead component to be picked up by pattern matching.

[0046] Specifically, the individual control device 452 identifies the outline of the lead component 410 based on the image data of the lead component 410 captured by the two-dimensional camera 290, and calculates the shape of the upper surface of the lead component 410, that is, the shape of the lead component 410 as viewed from above. Meanwhile, the storage device 458 stores model data corresponding to the outline of the lead component 410a in the first orientation, as shown in FIG.

[0047] The individual control device 452 then calculates the shape of the top surface of the lead component 410 based on the image data of the lead component 410 captured by the two-dimensional camera 290, and determines whether the calculated shape of the top surface of the lead component 410 matches the model data stored in the storage device 458. If the calculated shape of the top surface of the lead component 410 matches the model data, the individual control device 452 sets the component 410 whose shape matches the top surface of the lead component 410 as the component to be picked up.

[0048] That is, lead component 410a in the first position is set as the component to be picked up, while lead component 410b in the second position, lead component 410c in the third position, and lead component 410d in the fourth position are not set as components to be picked up. This is because the area of ​​the top surface of lead component 410b in the second position is small, making it impossible for suction nozzle 332 to properly hold lead component 410b. Also, lead component 410c in the third position has leads 414 on its top surface, which get in the way and make it impossible for suction nozzle 332 to properly hold lead component 410c. Also, lead component 410d in the fourth position cannot be properly held by suction nozzle 332 because the top surface of lead component 410d is not horizontal, for example.

[0049] The individual control device 452 then calculates the position information of the lead component 410 that has been set as the component to be picked up based on the image data. Next, based on the calculated position information of the component to be picked up, the component holding head 302 is moved above the component to be picked up by the component holding head moving device 300, and the suction nozzle 332 sucks and holds the component to be picked up. Note that when the suction nozzle 332 sucks and holds the component to be picked up, the suction nozzle 332 is located in a non-rotating position.

[0050] Next, after the suction nozzle 332 holds the lead component 410, which is the component to be picked up, the component holding head 302 moves above the component carrier 388. At this time, the component carrier 388 has been moved to the component receiving position by the operation of the component carrier moving device 390. Also, when the component holding head 302 moves above the component carrier 388, the suction nozzle 332 pivots to the pivot position. Note that the suction nozzle 332 pivots by the operation of the nozzle rotation device 335 so that the leads 414 of the lead component 410 held by the suction nozzle 332 in the pivot position face vertically downward.

[0051] When the component holding head 302 moves above the component carrier 388, the component holding head 302 descends and inserts the lead component 410, with its leads 414 facing vertically downward, into the component receiving recess 416 of the component receiving member 392. As a result, the lead component 410 is placed on the component receiving member 392 with its leads 414 facing vertically downward, as shown in FIG.

[0052] Then, when the lead components 410 are placed on the component receiving member 392, the component carrier 388 is moved to the component supply position by operation of the component carrier moving device 390. Because the component supply position is located within the movement range of the work heads 60, 62, the bulk component supply device 32 supplies the lead components 410 to the component mounter 10 at this position. In this way, in the bulk component supply device 32, the individual control device 452 identifies a component to be picked up from among the multiple lead components 410 scattered on the stage 156 based on one piece of model data, and the component holding head 302 holds the lead component 410 identified as the component to be picked up and places it on the component receiving member 392, thereby supplying the lead component 410.

[0053] For this reason, it is necessary to appropriately identify, based on a single piece of model data, the component to be picked up from among the multiple lead components 410 scattered on the stage 156. However, in the bulk component supply device 32, the component to be picked up is identified based on the image data captured by the two-dimensional camera 290, and therefore there is a risk that the component to be picked up may not be appropriately identified due to the parallax of the two-dimensional camera 290.

[0054] 14, two-dimensional camera 290 is fixedly disposed at a predetermined position at a predetermined height above stage 156, for example, at a position above the center of stage 156. Two-dimensional camera 290 is moved to any position in the X direction at the predetermined height by camera moving device 292, but when imaging a plurality of lead components scattered on the stage, it stops above the center of stage 156 to be imaged, and therefore is fixedly positioned at a position above the center of stage 156 at the predetermined height above stage 156. Two-dimensional camera 290 then images a plurality of lead components 410 scattered at multiple positions on stage 156 all at once. In this case, for example, if two-dimensional camera 290 simultaneously captures images of lead component 410A and lead component 410B, the image of lead component 410A captures surface 460 from which lead 414 extends, while the image of lead component 410B captures surface 462 opposite surface 460 from which lead 414 extends. Therefore, the images of lead component 410A and lead component 410B have different shapes. Thus, when two-dimensional camera 290 positioned at a fixed location captures images of multiple objects at different positions, the images of the objects have different shapes. However, the differences between the images of the objects at different positions are not large but relatively small. However, even if the differences are relatively small, there is a risk that the component to be picked up cannot be properly identified from among the multiple lead components 410 scattered at different positions on stage 156.

[0055] Furthermore, LED light 294, which is used as illumination when two-dimensional camera 290 captures an image of an object, emits light from the side of two-dimensional camera 290 toward stage 156, and the light reflected from the light emitted by LED light 294 differs depending on the orientation of lead components 410. Specifically, for example, as shown in FIG. 15 , when two-dimensional camera 290 captures lead components 410C scattered on stage 156 with leads 414 facing away from LED light 294, it captures reflected light 470, which is light emitted by LED light 294 and reflected by lead components 410C. On the other hand, for example, as shown in FIG. 16 , when two-dimensional camera 290 captures lead components 410D scattered on stage 156 with leads 414 facing toward LED light 294, it captures reflected light 480, which is light emitted by LED light 294 and reflected by lead components 410D. As described above, the reflected light 470 reflected by lead component 410C is different from the reflected light 480 reflected by lead component 410D, and therefore the image captured of reflected light 470 reflected by lead component 410C and the image captured of reflected light 480 reflected by lead component 410D will have different shapes. Thus, when imaging objects with different orientations using light irradiated from the side as a light source, the images captured of the objects will have different shapes. However, the differences between the images captured of the objects with different orientations are not large, but relatively small. However, even if the differences are relatively small, there is a risk that the component to be picked up will not be properly identified from among the multiple lead components 410 scattered in different orientations on stage 156.

[0056] One possible solution is to lower the criteria for identifying components to be picked up by pattern matching from among the multiple lead components 410 scattered on the stage 156. Specifically, in pattern matching, the individual control device 452 identifies the outline of the lead component 410 from the image data of the lead component 410 captured by the two-dimensional camera 290 and calculates a correlation value between the outline of the lead component 410 and the outline of the model data. The correlation value is the rate of match between the outline of the lead component 410 and the outline of the model data. When the outline of the lead component 410 and the outline of the model data match perfectly, the correlation value is 100. When the outline of the lead component 410 and the outline of the model data do not match at all, the correlation value is 0. In other words, the higher the correlation value, the more similar the outline of the lead component 410 and the outline of the model data are, and the lower the correlation value, the less similar the outline of the lead component 410 and the outline of the model data are. Therefore, for example, if the correlation value between the outline of the lead component 410 to be determined as a target for pickup and the outline of the model data is 90 or greater, the individual control device 452 identifies the lead component 410 as a target for pickup because they are very similar. However, as described above, the differences between images of multiple imaged objects at different positions are relatively small, and the differences between images of multiple imaged objects at different postures are also relatively small. Therefore, for example, by lowering the correlation value threshold, it is possible to identify a target for pickup while allowing for relatively small differences. Specifically, for example, the individual control device 452 can identify a lead component 410 with a correlation value of 70 or greater as a target for pickup. However, if the correlation value threshold is lowered and pattern matching is performed, there is a risk that a lead component 410 to be held that is slightly tilted and therefore cannot be held by the suction nozzle 332 may be identified as a target for pickup. Therefore, it is not particularly desirable to perform pattern matching of components with a low correlation value threshold.

[0057] To address this issue, the mounter 10 creates a single piece of model data based on image data of multiple lead components 410 placed in the same first orientation at multiple positions on the stage 156 and image data of multiple lead components 410 placed in an orientation different from the first orientation at multiple positions on the stage 156. Specifically, as shown in FIG. 17 , a worker places five lead components 410 in the same orientation at five locations on the stage 156 (stage center, lower right, lower left, upper right, and upper left). At this time, the worker places the five lead components 410 on the stage in the same orientation, with the surfaces from which the leads 414 of the lead components 410 extend facing the same direction. In other words, the worker places the five lead components 410 on the stage in the same orientation, with the leads 414 of the lead components 410 extending in the same direction. Note that in FIG. 17 , the direction in which the leads 414 of the lead components 410 extend is defined as 0°. Then, the two-dimensional camera 290 simultaneously captures images of the five lead components 410 placed on the stage with the leads 414 extending at 0°.

[0058] 18 , an operator places five lead components 410 at five locations on stage 156 (stage center, lower right, lower left, upper right, and upper left) with the leads 414 extending in a direction rotated 90° counterclockwise from 0°. Note that the direction in which the leads 414 extend is rotated 90° counterclockwise from 0° is defined as 90°. Then, two-dimensional camera 290 simultaneously captures images of the five lead components 410 placed on stage 156 with the leads 414 extending at 90°.

[0059] 19 , an operator places five lead components 410 at five locations on stage 156 (stage center, lower right, lower left, upper right, and upper left) with the leads 414 extending in a direction rotated 180° counterclockwise from 0°. Note that the direction in which the leads 414 extend is defined as 180° when the direction is rotated 180° counterclockwise from 0°. Then, two-dimensional camera 290 simultaneously captures images of the five lead components 410 placed on stage 156 with the leads 414 extending at 180°.

[0060] 20 , an operator places five lead components 410 at five locations on stage 156 (stage center, lower right, lower left, upper right, and upper left) with the leads 414 extending in a direction rotated 270° counterclockwise from 0°. Note that the direction in which the leads 414 extend is defined as 270° when rotated 270° counterclockwise from 0°. Then, two-dimensional camera 290 simultaneously captures images of the five lead components 410 placed on stage 156 with the leads 414 extending at 270°.

[0061] Hereinafter, the lead component 410 placed on the stage at the center of the stage 156 in Fig. 17 with the leads 414 extending at a 0° angle will be referred to as a lead component with a 0° angle at the center of the stage. Also, the lead component 410 placed on the stage at the lower right of the stage 156 in Fig. 18 with the leads 414 extending at a 90° angle will be referred to as a lead component with a 90° angle at the lower right of the stage. Also, the lead component 410 placed on the stage at the lower left of the stage 156 in Fig. 19 with the leads 414 extending at a 180° angle will be referred to as a lead component with a 180° angle at the lower left of the stage. Also, the lead component 410 placed on the stage at the upper right of the stage 156 in Fig. 20 with the leads 414 extending at a 270° angle will be referred to as a lead component with a 270° angle at the upper right of the stage.

[0062] In this way, when two-dimensional camera 290 captures images of a plurality of lead components in the same orientation placed on stage 156 shown in Figures 17 to 20, imaging data of 20 types of lead components 410 is created as shown in Figure 21. The imaging data of these 20 types of lead components 410 is imaging data of lead components 410 that differ in at least one of the orientation of lead components 410 and the placement position of lead components 410.

[0063] The individual control device 452 also generates model data candidates based on the image data of a lead component with an orientation of 0° at the center of the stage, a lead component with an orientation of 90° at the center of the stage, a lead component with an orientation of 180° at the center of the stage, and a lead component with an orientation of 270° at the center of the stage. In this case, the individual control device 452 generates four types of model data candidates with edge levels of 50, 100, 200, and 300. Specifically, since the model data of the lead components is data corresponding to the outline of the lead components, when the individual control device 452 generates the model data, it identifies the outline of the lead components based on the image data of the lead components. In this case, the individual control device 452 identifies the boundary line between the stage 156 and the lead component as the outline of the lead component. In this case, the individual control device 452 identifies a location in the image data of the lead component where the difference in brightness between adjacent pixels is equal to or greater than the edge level as the boundary line between the stage 156 and the lead component, which is the outline of the lead component. For example, if the edge level is 50, the individual control device 452 identifies, as the outline of the lead component, any location in the imaging data of the lead component where the difference in brightness between adjacent pixels is 50 or more. If the edge level is 100, the individual control device 452 identifies, as the outline of the lead component, any location in the imaging data of the lead component where the difference in brightness between adjacent pixels is 100 or more. If the edge level is 200, the individual control device 452 identifies, as the outline of the lead component, any location in the imaging data of the lead component where the difference in brightness between adjacent pixels is 200 or more. If the edge level is 300, the individual control device 452 identifies, as the outline of the lead component, any location in the imaging data of the lead component where the difference in brightness between adjacent pixels is 300 or more. The individual control device 452 then creates the identified outline of the lead component as a candidate for model data.

[0064] In this way, the individual control device 452 generates model data candidates with edge levels of 50, 100, 200, and 300 for each of the four types of lead component orientations based on the imaging data of a lead component with an orientation of 0° at the center of the stage, a lead component with an orientation of 90° at the center of the stage, a lead component with an orientation of 180° at the center of the stage, and a lead component with an orientation of 270° at the center of the stage. In other words, the individual control device 452 generates 16 (= 4 × 4) types of model data candidates. These 16 types of model data candidates are component identification data used when performing pattern matching. The individual control device 452 performs pattern matching with each of the 20 types of lead components 410 shown in FIG. 21 based on each model data candidate.

[0065] Specifically, the individual control device 452 performs pattern matching between each model data candidate and each of the 20 types of lead components 410 in accordance with the flowchart shown in Fig. 22. Therefore, in S10, the individual control device 452 first reads, as model data candidates, model data candidates for lead components having an edge level of 50 and an orientation of 0° at the center of the stage. Next, in S12, the individual control device 452 reads image data for a lead component 410 having an orientation of 0° at the center of the stage, among the 20 types of lead components 410 having different orientations from the placement positions shown in Fig. 21. Then, in S14, the individual control device 452 performs pattern matching between the image data for the lead component 410 read in S12, the model data candidates for lead components having an edge level of 50 and an orientation of 0° at the center of the stage, and the image data for the lead component having an orientation of 0° at the center of the stage. As a result, in S16, the individual control device 452 calculates that the correlation value between the candidate model data for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage and the image data for the lead component having an orientation of 0° at the center of the stage is 96, as shown in FIG. After calculating the correlation value, the individual control device 452 determines in S18 whether pattern matching between the candidate model data and the image data for all 20 types of lead components 410 has been completed. At this point, pattern matching has only been performed with the image data for one type of lead component 410. Therefore, the individual control device 452 determines NO in S18 because pattern matching with all 20 types of image data has not been completed, and returns to S12. Next, in S12, the individual control device 452 reads the image data for the lead component 410 having an orientation of 90° at the center of the stage from among the 20 types of lead components 410. Next, in S14, the individual control device 452 performs pattern matching between the candidate model data of a lead component having an edge level of 50 and an orientation of 0° at the center of the stage and the image data of the lead component having an orientation of 90° at the center of the stage that has been read in, and as a result, in S16, calculates that the correlation value is 89, as shown in Figure 23.In this way, the individual control device 452 repeatedly executes the processes of S12 to S18, performing pattern matching between the model data candidates and the imaging data of each of the 20 types of lead components 410. When all pattern matching is complete, the individual control device 452 determines YES in S18 because pattern matching is complete. The individual control device 452 then calculates the average of the correlation values ​​between the model data candidates and the imaging data of the 20 types of lead components 410. The average of the correlation values ​​calculated by the individual control device 452 is 88.15, as shown in FIG. 23.

[0066] Next, in S22, the individual control device 452 determines whether pattern matching has been completed for all edge levels of the model data candidates loaded in S10. However, currently, pattern matching has only been performed for the model data candidates for lead components loaded in S10 with an edge level of 50 and an orientation of 0° at the center of the stage. Therefore, the individual control device 452 determines NO in S22 because pattern matching has not been completed for all edge levels, and changes the edge level of the model data candidates to another edge level of 100 in S24. Specifically, the individual control device 452 loads the model data candidates whose edge level has been changed to 100 in S10. Next, the individual control device 452 repeatedly executes the processes of S12 to S18 to perform pattern matching between the model data candidates and the imaging data of each of the 20 types of lead components 410, and calculates the average of each correlation value in S20. In this way, the individual control device 452 repeatedly executes the processes of S22, S24, and S10 to S20, thereby performing pattern matching between each of the candidate model data for all edge levels, that is, the candidate model data for lead components with an orientation of 0° at the center of the stage, and the imaging data for each of the above 20 types of lead components 410, and calculates the average of each correlation value.

[0067] After calculating all the averages of the correlation values, the individual control device 452 determines YES in S22 because pattern matching with all edge level candidates has been completed for the model data candidate for a lead component with an orientation of 0° at the center of the stage. Then, in S26, the individual control device 452 determines whether pattern matching has been completed for all model data candidates. At this point, pattern matching has only been performed for model data candidates for lead components with an orientation of 0° at the center of the stage. Therefore, since pattern matching has not been completed for all model data candidates, the individual control device 452 determines NO in S26 and changes the model data candidate to a different type of model data candidate with a different orientation. Specifically, in S10, the individual control device 452 reads the model data candidate for a lead component with an edge level of 50 and an orientation of 90° at the center of the stage. Next, the individual control device 452 repeatedly executes the processes of S12 to S18 to perform pattern matching between the candidate model data for lead components having an edge level of 50 and an orientation of 90° at the center of the stage and the imaging data for each of the 20 types of lead components 410, and calculates the average of each correlation value in S20. Furthermore, the individual control device 452 repeatedly executes the processes of S22, S24, and S10 to S20 to perform pattern matching between each of the candidate model data for lead components having an orientation of 90° at the center of the stage at all edge levels and the imaging data for each of the 20 types of lead components 410, and calculates the average of each correlation value.

[0068] Next, the individual control device 452 repeatedly executes the processes of S26, S28, and S10 to S24 for the model data candidates for lead components having an orientation of 180° at the center of the stage and the model data candidates for lead components having an orientation of 270° at the center of the stage.The individual control device 452 then performs pattern matching between the image data of each of the 20 types of lead components 410 and each of the model data candidates for lead components having an orientation of 0° at the center of the stage, lead components having an orientation of 90° at the center of the stage, lead components having an orientation of 180° at the center of the stage, and lead components having an orientation of 270° at the center of the stage, with the model data candidates having edge levels of 50, 100, 200, and 300 (i.e., 4 x 4 = 16 types of model data candidates), and calculates the average of each correlation value.

[0069] Figure 24 shows the average correlation values ​​from pattern matching performed based on each of these 16 types of model data candidates. As can be seen from this figure, when pattern matching was performed using candidate model data for lead components with an edge level of 100 and an orientation of 270° at the center of the stage with image data for each of the 20 types of lead components 410, the correlation value was 92, the highest of all correlation values. Therefore, when pattern matching was performed using candidate model data for lead components with an edge level of 100 and an orientation of 270° at the center of the stage, the probability of recognizing the outlines of the 20 types of lead components 410 was highest. Note that, as shown in Figure 25, when pattern matching was performed between candidate model data for lead components with an edge level of 100 and an orientation of 270° at the center of the stage with image data for the 20 types of lead components 410, the correlation values ​​were all 90 or higher. Therefore, since it is possible to recognize the outlines of all of the above 20 types of lead components 410, in S30 the individual control device 452 registers in the memory device 458 as model data a candidate for model data of a lead component having an edge level of 100 and an orientation of 270° at the center of the stage.

[0070] By performing pattern matching using the model data registered in storage device 458 in this way, it is possible to recognize the outlines of lead components placed in four different orientations at five positions on stage 156 shown in Figures 17 to 20, that is, all 20 types of lead components shown in Figure 21. Therefore, when individual control device 452 identifies a component to be picked up from among the multiple lead components scattered on stage 156, it performs pattern matching using the model data registered in storage device 458.

[0071] Specifically, first, two-dimensional camera 290 captures an image of a plurality of lead components scattered on stage 156. Then, individual control device 452 calculates a correlation value between the image data of each of the plurality of lead components captured by two-dimensional camera 290 and the model data registered in storage device 458. At this time, individual control device 452 identifies lead components with a correlation value of 90 or greater as components to be picked up, and does not identify lead components with a correlation value of less than 90 as components to be picked up. In this way, by performing pattern matching using the model data registered in storage device 458, it becomes possible to appropriately identify components to be picked up from among a plurality of lead components 410 scattered in a plurality of different positions and with a plurality of different orientations on stage 156, without lowering the correlation value threshold.

[0072] However, as described above, in order to identify model data from among 16 types of model data candidates, pattern matching must be performed for each of the 16 types of model data candidates with all 20 types of lead components shown in FIG. 21 . In other words, pattern matching must be performed 16 x 20 = 320 times. Furthermore, for each pattern matching, the individual control device 452 must calculate the correlation value between the model data candidate and the imaging data. Therefore, it takes a long time to complete 320 pattern matchings. Therefore, the individual control device 452 reduces the number of pattern matchings and identifies model data from among the 16 types of model data candidates.

[0073] 11 stored in the individual control device 452 is executed in accordance with the flowcharts shown in Figures 26 to 28. Specifically, first, in S50, the individual control device 452 reads one model data candidate out of the 16 types of model data candidates, for example, a model data candidate for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage. Next, in S52, the individual control device 452 reads imaging data for a lead component 410 having an orientation of 0° at the center of the stage out of the 20 types of lead components 410 shown in Figure 21. Then, in S54, the individual control device 452 performs pattern matching between the read model data candidate and the read imaging data of the lead component. As a result, in S56, as shown in Figure 29, the individual control device 452 calculates that the correlation value between the candidate model data for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage and the image data for the lead component having an orientation of 0° at the center of the stage is 96, and also calculates the X, Y, and Q coordinates of the lead component based on the image data for the lead component. The X, Y coordinates of the lead component indicate the position of the lead component, and the Q coordinate of the lead component indicates the rotation angle of the lead component. Note that, as shown in Figure 29, based on the image data for the lead component having an orientation of 0° at the center of the stage, the individual control device 452 calculates the X, Y coordinates of the lead component to be (31, 30) and the Q coordinate to be 1°.

[0074] Then, in S58, the individual control device 452 executes a subroutine to forcibly terminate pattern matching. In the subroutine to forcibly terminate pattern matching, as shown in FIG. 27, in S80, the individual control device 452 determines whether the correlation value calculated in S56 was less than 90. More specifically, the individual control device 452 determines whether the calculated correlation value was a low value that did not satisfy the criterion for pattern matching. Because the correlation value calculated in S56 was 96, the individual control device 452 determines NO in S80 because the correlation value is not less than 90, and proceeds to S88. In S88, the individual control device 452 determines whether multiple parts have been recognized based on the image data. The processing of S88 will be described in detail later, so the description of S88 will be skipped and the description of S90 will be proceeded to.

[0075] In S90, individual control device 452 determines whether the X, Y, and Q coordinates of the lead component calculated in S56 differ from the position and rotation angle of the lead component placed on the stage by the operator, which are input to the individual control device and stored in the memory device. The X, Y coordinates and rotation angle of the lead component placed by the operator in the center of stage 156 at a 0° orientation are input by the operator as (30, 30) and 0°, respectively, and memory device 458 stores these input coordinates and angles. Therefore, individual control device 452 compares the X, Y coordinates of the lead component (30, 30) and the rotation angle of the lead component (0°) stored in memory device 458 with the X, Y coordinates of the lead component (31, 30) and the rotation angle of the lead component (1°) calculated based on the image data captured by the individual control device of the lead component placed in the center of stage 156 at a 0° orientation. As a result of the comparison by the individual control device, the difference in the XY coordinates and rotation angle of the lead component is small, and the individual control device 452 considers it to be an error. In S90, the individual control device 452 determines NO because the component placement position and component placement direction are not different, and the subroutine that forcibly terminates pattern matching ends.

[0076] Next, in S60 of the routine shown in FIG. 26 , the individual control device 452 determines whether pattern matching has been completed for the image data of all 20 types of lead components 410 with the candidate model data for lead components having an edge level of 50 and an orientation of 0° at the center of the stage. At this point, the one model data candidate has only been pattern matched with the image data of one lead component 410 out of the 20 types of lead components 410, so the individual control device 452 determines NO in S60 and returns to S52. Then, in S52, the individual control device 452 reads in image data for another of the 20 types of lead components 410, that is, image data for a lead component 410 having an orientation of 90° at the center of the stage. Next, in S54, the individual control device 452 performs pattern matching between the candidate model data for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage and the image data of the lead component that was just read. As a result, in S56, the individual control device 452 calculates that the correlation value is 89, as shown in FIG. 29 , and calculates that the XY coordinates of the lead component are (30, 30) and the Q coordinate is 91° based on the image data of the lead component positioned at the center of the stage and oriented at 90°. Then, in S80 of the subroutine for forcibly terminating pattern matching, as shown in FIG. 27 , the individual control device 452 determines whether the correlation value calculated in S56 is less than 90. Since the correlation value calculated in S56 is 89, which is less than 90, the individual control device 452 determines YES. Next, in S82, the individual control device 452 adds 1 to the count N, which counts correlation values ​​less than 90, i.e., correlation values ​​that do not satisfy the pattern matching criteria. The initial value of the count N is 0. Then, in S84, the individual control device 452 determines whether the count N exceeds the threshold M. As will be explained in detail later, the initial value of the threshold value M is set to a value greater than 20, for example, 21.Therefore, the individual control device 452 determines NO in S84 because the count number of N, 1, does not exceed the value of the threshold M, and determines NO in the processes of S88 and S90, thereby terminating the subroutine that forcibly terminates pattern matching.The individual control device 452 then repeatedly executes the processes of S52 to S58, and when it has completed pattern matching between the candidate model data for a lead component with an edge level of 50 and an orientation of 0° at the center of the stage and all of the imaging data for the 20 types of lead components 410, it determines YES in S60, assuming that pattern matching is complete.The individual control device 452 then executes S62, a subroutine that updates the model data.

[0077] In the subroutine for updating model data, as shown in FIG. 28, the individual control device 452 determines in S100 whether N, the number of counts of those whose correlation values ​​do not satisfy the pattern matching criteria, is less than the threshold value M. When pattern matching is performed with the image data of each of the 20 types of lead components 410, the count number N for those with correlation values ​​less than 90 is 8, as shown in FIG. 29. Therefore, the number of lead components with correlation values ​​of 90 or greater among the 20 types of lead components 410 is 20-8=12. In other words, when pattern matching is performed using candidate model data for lead components with an edge level of 50 and an orientation of 0° at the center of the stage, 12 types of lead components among the 20 types of lead components 410 can be recognized. Here, the individual control device 452 determines YES in S100 because the count number N, 8, is less than the threshold value M, 21, and updates the model data in S102. The individual control device 452 updates the default model data set as the initial model data to candidate model data for a lead component with an edge level of 50 and an orientation of 0° at the center of the stage. When updating the model data, the individual control device 452 also calculates the average correlation value of the updated model data. Here, as shown in FIG. 29, the average correlation value of the updated model data is calculated to be 88.15. After updating the model data in S102, the individual control device 452 changes the value of the threshold value M set in S104 from 21 to 8, which is the minimum number for N. Next, the individual control device 452 resets the count number for N to 0 in S106 and ends the subroutine for updating the model data.

[0078] When the individual control device 452 ends the subroutine for updating model data in S62, it determines in the following S64 whether or not pattern matching has been completed for model data candidates of all edge levels by performing pattern matching this time with the model data candidates read in S50. At this stage, because only model data candidates with an edge level of 50 were read in S50, the individual control device 452 determines NO because pattern matching has not been completed for model data candidates of all edge levels, and in the following S66 changes the edge level of the model data candidates for lead components from 50 to 100. The individual control device 452 then returns to S50 and newly reads model data candidates for lead components with an attitude of 0° at the center of the stage with the edge level changed to 100.

[0079] Subsequently, the individual control device 452 executes the processes of S52 to S60. Specifically, as with the previously loaded model data candidates, the individual control device 452 performs pattern matching between the model data candidates loaded in S50 this time and the image data of the lead component 410 of the 20 types of lead components 410 that is positioned at the center of the stage and has an orientation of 0°. As a result of the calculation as shown in FIG. 30, the correlation value between them is 88. Therefore, in S82, N, which is the number of lead components with a correlation value of less than 90, is changed from the initial value of 0 to 1. In this way, pattern matching is performed sequentially with the image data of the 20 types of lead components. When pattern matching is performed with the image data of the lead component 410 positioned at the upper right of the stage and has an orientation of 90°, which is the 14th image data, the correlation value becomes 87, as shown in FIG. 30. Therefore, in S82, the count number of N is accumulated to 9.

[0080] Since the count number of N, 9, is greater than the number of threshold values ​​M, 8, set in S104, the individual control device 452 determines YES in S84. Subsequently, the individual control device 452 resets the count number of N to 0 in S86, ends the subroutine of S58, and returns to the main routine shown in Fig. 26, proceeding from S62 to S64. In other words, as shown in Fig. 30, the individual control device 452 performs pattern matching between the candidate model data of a lead component having an edge level of 100 and an attitude of 0° at the center of the stage and the image data of the first to fourteenth lead components of the 20 types of lead components 410, but does not perform pattern matching with the image data of the fifteenth to twentieth lead components.

[0081] Specifically, when the individual control device 452 performs pattern matching between a candidate model data for a lead component having an edge level of 100 and an orientation of 0° at the center of the stage and the image data for the first through fourteenth lead components, N, which is the count number of lead components with a correlation value of less than 90, is 9. Therefore, the correlation values ​​of five types of lead components among the image data for the first through fourteenth lead components are 90 or greater. Therefore, if pattern matching is performed with the image data for the fifteenth through twentieth lead components and the correlation values ​​of all of these lead components are 90 or greater, the number of lead components with a correlation value of 90 or greater among the 20 types of lead components shown in FIG. 30 is 5 + 6 = 11. Meanwhile, in S102, the model data is updated to a candidate model data for a lead component having an edge level of 50 and an orientation of 0° at the center of the stage. Then, when pattern matching is performed with the updated candidate model data against the above 20 types of lead components, N, which is the count number of lead components with a correlation value of less than 90, is 8, as shown in FIG. 29. Therefore, when pattern matching is performed between the above 20 types of lead components and candidate model data for lead components having an edge level of 50 and an orientation of 0° at the center of the stage, 12 of the 20 types of lead components can be recognized. On the other hand, when pattern matching is performed between the above 20 types of lead components and candidate model data for lead components having an edge level of 100 and an orientation of 0° at the center of the stage, only 11 of the 20 types of lead components can be recognized. In other words, in terms of the recognition rate of lead components through pattern matching, candidate model data for lead components having an edge level of 100 and an orientation of 0° at the center of the stage does not exceed candidate model data for lead components having an edge level of 50 and an orientation of 0° at the center of the stage.Therefore, the individual control device 452 sequentially performs pattern matching with the above 20 types of lead components, and when it performs pattern matching with the image data of the 14th lead component and the count number of N exceeds 8, which is the value of the threshold M, it does not perform pattern matching with the image data of the 15th or subsequent lead components. At the same time, in S102, the model data candidate for a lead component with an edge level of 100 and an attitude of 0° at the center of the stage is not updated as model data, and is excluded from the target model data.

[0082] In S102, a subroutine of S62, when the candidate model data for a lead component having an edge level of 100 and an orientation of 0° at the center of the stage is excluded from the target model data, the individual control device 452 returns to the main routine shown in Fig. 26 and executes the processing from S64 onwards. The individual control device 452 then repeatedly executes the processing of S64, S66, and S50 to S62, and performs pattern matching with the above 20 types of lead components using the candidate model data for lead components having an edge level of 200 and an edge level of 300 and an orientation of 0° at the center of the stage. As a result, the count number of N exceeds the value of the threshold M, which is 8, and therefore the two candidate model data are excluded from the target model data.

[0083] In this way, when pattern matching using model data candidates for lead components with an orientation of 0° at the center of the stage at all edge levels is completed, the individual control device 452 determines YES in S64. Subsequently, in S68, the individual control device 452 determines whether pattern matching has been completed for all types of model data candidates. At this point, pattern matching has only been performed with model data candidates for lead components 410 with an orientation of 0° at the center of the stage, so the individual control device 452 determines NO in S68 because pattern matching has not been completed for all model data candidates, and changes to another type of model data candidate in S70. Then, in S50, the individual control device 452 reads model data candidates for lead components with an orientation of 90° at the center of the stage. Subsequently, the individual control device 452 repeatedly executes the processes of S52 to S60, performs pattern matching between the newly read model data candidates and each of the image data for the 20 types of lead components 410, and calculates the count number N. 31, the result of the calculation of the count number of N by the individual control device 452 is 4. Therefore, the individual control device 452 judges YES in S60, and executes a subroutine to update the model data in S62.

[0084] In the model data updating subroutine, as shown in FIG. 28 , the individual control device 452 determines in S100 whether the count number N, which is the number of correlation values ​​that do not satisfy the pattern matching criteria, is less than the threshold value M. Since the result of counting N is 4, which is less than the threshold value 8, the individual control device 452 determines YES in S100 and updates the model data candidate to a model data candidate for a lead component with an edge level of 50 and a 90° orientation at the center of the stage in S102. Furthermore, as shown in FIG. 31 , the individual control device 452 calculates the average correlation value of the updated model data candidate to 90.8. Then, in S104, the individual control device 452 changes the threshold value M to 4, which is the minimum number for N. Next, in S106, the individual control device 452 resets the count number N to 0 and ends the model data updating subroutine.

[0085] When the individual control device 452 finishes the subroutine for updating the model data, it determines NO in the following S64 because pattern matching has not been completed for all edge level model data candidates, and in the following S66 it changes the edge level of the model data candidate to 100. The individual control device 452 then returns to S50 and reads in the model data candidate for a lead component with an attitude of 90° at the center of the stage for which the edge level has been changed to 100.

[0086] The individual control device 452 then executes the processes of S52 to S60. Specifically, the individual control device 452 performs pattern matching between the currently loaded model data candidate with an edge level of 100 and the image data of each of the 20 types of lead components 410. As a result, the individual control device 452 determines YES in S90 because the component positions are different in the subroutine of S58. Specifically, as shown in FIG. 32, the XY coordinate values ​​calculated based on the image data acquired by imaging the lead component 410 at a 90° orientation at the lower right of the stage are (10, 2). Meanwhile, the XY coordinates of the lead component input to the individual control device and stored in the memory device 458 are (50, 10). Therefore, the individual control device 452 determines YES in S90 because the calculated lead component position and the lead component position stored in the memory device 458 are significantly different, and resets the count value of N to 0 in S86. The individual control device 452 then terminates the subroutine for forcibly terminating pattern matching and proceeds to S64 of the main routine shown in FIG. 26. That is, the individual control device 452 terminates pattern matching using the currently loaded model data candidate for a lead component having an edge level of 100 and a 90° orientation at the center of the stage. That is, the individual control device 452 performs pattern matching between the currently loaded model data candidate and the image data of the sixth lead component of the 20 types of lead components listed in FIG. 32, and if it determines in S90 that the component positions differ, it does not perform pattern matching with the image data of the seventh and subsequent lead components. This is because if the position of the lead component calculated based on the image data differs significantly from the position of the lead component stored in the storage device 458, the position of the lead component to be held cannot be recognized based on the image data, and the model data candidate used in pattern matching cannot be optimal model data. Therefore, in S102, the model data candidate for a lead component having an edge level of 100 and a 90° orientation at the center of the stage is not updated as model data and is excluded from the target model data.

[0087] The individual control device 452, which has excluded from the target model data the candidate model data of a lead component with an edge level of 100 and an orientation of 90° at the center of the stage, executes S64 and subsequent steps of the main routine shown in Fig. 26. At this time, the individual control device 452 repeatedly executes the processes of S64, S66, and S50 to S62, thereby also performing pattern matching on the candidate model data whose edge levels have been changed to 200 and 300. Note that when the individual control device 452 performs pattern matching using these candidate model data whose edge levels have been changed, the count number N exceeds the threshold value of 4. Therefore, once the count number N exceeds 4, the candidate model data of a lead component with an orientation of 90° at the center of the stage whose edge levels have been changed to 200 and 300 is excluded from the target model data.

[0088] In this way, the individual control device 452 completes pattern matching and determines YES in S64. Since the individual control device 452 has not completed pattern matching using all model data candidates, it determines NO in S68 and changes the type of model data candidate to a model data candidate for a lead component with an orientation of 180° at the center of the stage in S70. The individual control device 452 then reads the model data candidate changed in S50. Next, the individual control device 452 repeatedly executes the processes of S52 to S60 to perform pattern matching between the model data candidate changed in S50 and the image data obtained by capturing an image of each of the 20 types of lead components 410. Then, it performs pattern matching between the model data candidate changed in S50 and the image data obtained by capturing an image of a lead component 410 with an orientation of 270° at the center of the stage, and if multiple components are recognized based on the image data of the lead component 410, the individual control device 452 determines YES in S88 of the subroutine of S58.

[0089] For example, Figure 34 shows an image 510 obtained by capturing an image of a lead component 410 positioned at a 270° angle in the center of the stage. In this image 510, the pair of leads 414 of the lead component 410 can be recognized, but the component body 412 cannot. Note that the dotted and dashed lines in image 510 are imaginary lines. From this image 510, the individual control device 452 recognizes the pair of leads 414 and estimates that the component body 412 is located at the position of the imaginary dotted line 512, and also at the position of the imaginary dashed line 514. In other words, the individual control device 452 estimates that two lead components 410 are present in image 510. In this way, if the individual control device 452 recognizes multiple components based on the captured image data and determines YES in S88, it resets the count N to 0 in S86. The individual control device 452 then terminates the subroutine for forcibly terminating pattern matching shown in FIG. 27 and proceeds to S64 of the main routine shown in FIG. 26. That is, if the individual control device 452 determines that multiple components have been recognized based on the image data acquired by imaging the lead component 410 at the center of the stage with an orientation of 270°, it terminates pattern matching using candidate model data for a lead component with an edge level of 50 and an orientation of 180° at the center of the stage. That is, if the individual control device 452 performs pattern matching with the image data for the fourth lead component shown in FIG. 33 and determines that multiple components have been recognized, it does not perform pattern matching with the image data for the fifth or subsequent lead components. This is because, if multiple components are recognized based on the image data, the position of the lead component to be held cannot be appropriately recognized based on the image data, and therefore the candidate model data used in pattern matching cannot be optimal model data. Therefore, in S102, the candidate model data for the lead component with an edge level of 50 and an attitude of 180° at the center of the stage is not updated as model data and is excluded from the target of model data.

[0090] The individual control device 452, which has excluded from the target model data the candidate model data for a lead component with an edge level of 50 and an attitude of 180° at the center of the stage, determines NO in S64 of the main routine shown in Fig. 26 because pattern matching has not been completed at all edge levels, and in S66 changes the edge level of the above model data candidate to another edge level, 100. The individual control device 452 then returns to S50 and newly reads in the candidate model data for a lead component with an attitude of 180° at the center of the stage whose edge level has been changed to 100.

[0091] Next, the individual control device 452 repeatedly executes the processes of S52 to S60, performing pattern matching between the newly loaded model data candidates and the image data of each of the 20 types of lead components 410. As a result, when the individual control device 452 executes the subroutine of S58 and performs pattern matching with the image data of the lead component 410 at the bottom right of the stage with an orientation of 0°, it determines that the component orientations are different, resulting in a YES determination in S90. Specifically, as shown in FIG. 35 , the Q coordinate value calculated based on the image data acquired by imaging the lead component 410 at the bottom right of the stage with an orientation of 0° is 205°. Meanwhile, the lead component rotation angle input to the individual control device and stored in the memory device 458 is 0°. Therefore, since the calculated lead component rotation angle and the lead component rotation angle stored in the memory device 458 are significantly different, the individual control device 452 determines a YES determination in S90 and resets the count value of N to 0 in S86. The individual control device 452 then terminates the subroutine shown in FIG. 27 and returns to the main routine shown in FIG. 26, proceeding to S64. In other words, if the rotation angle of the lead component calculated based on the imaging data is significantly different from the rotation angle of the lead component stored in the memory device 458, the individual control device 452 terminates pattern matching using the newly read candidate model data of a lead component with an edge level of 100 and a posture of 180° at the center of the stage. Therefore, if the individual control device 452 performs pattern matching with the imaging data of the fifth lead component shown in FIG. 35 and determines in S90 that the component orientations are different, it does not perform pattern matching with the imaging data of the sixth or subsequent lead components. This is because if the rotation angle of the lead component stored in the memory device 458 is significantly different from the Q coordinate of the lead component calculated based on the imaging data, the angle of the lead component to be held cannot be properly recognized based on the imaging data, and the candidate model data used in pattern matching cannot be optimal model data.Therefore, in S102, the candidate model data for the lead component with an edge level of 100 and an attitude of 180° at the center of the stage is not updated as model data, and is excluded from the target of model data.

[0092] The individual control device 452, which has excluded from the target model data candidates of lead components with an edge level of 100 and an orientation of 180° at the center of the stage, executes S64 and subsequent steps of the main routine shown in Fig. 26. At this time, the individual control device 452 repeatedly executes the processes of S64, S66, and S50 to S62, thereby also performing pattern matching on the model data candidates with edge levels changed to 200 and 300. Note that when the individual control device 452 performs pattern matching using these model data candidates with changed edge levels, the count number of N exceeds the threshold value of M, which is 4. Therefore, when the count number of N exceeds 4, the model data candidates of lead components with an orientation of 180° at the center of the stage with edge levels changed to 200 and 300 are excluded from the target model data.

[0093] In this way, the individual control device 452 completes pattern matching and determines YES in S64. Since the individual control device 452 has not completed pattern matching using all model data candidates, it determines NO in S68 and changes the type of model data candidate to a model data candidate for a lead component with a 270° orientation at the center of the stage in S70. The individual control device 452 then reads the model data candidate changed in S50. Next, the individual control device 452 repeatedly executes the processes of S52 to S60 to perform pattern matching between the model data candidate changed in S50 and the image data of each of the 20 types of lead components 410, and calculates the count N. As shown in FIG. 36, the result of calculating the count N is 0. Next, the individual control device 452 completes pattern matching between the model data candidate changed in S50 and the image data of each of the 20 types of lead components 410, determines YES in S60, and executes S62, a subroutine for updating the model data.

[0094] In the subroutine S62 shown in FIG. 28 , the individual control device 452 determines in S100 whether the count number of N is less than the threshold value M. Since the count number of N is 0, which is less than the threshold value M of 4, the individual control device 452 determines YES in S100 and updates the model data candidate from one with an edge level of 50 and a 90° orientation at the center of the stage to one with an edge level of 50 and a 270° orientation at the center of the stage in S102. The individual control device 452 also calculates the average of the correlation values ​​between the updated model data candidate and the imaging data of the 20 types of lead components 410 as shown in FIG. 36 to be 91.3. After updating the model data, the individual control device 452 then changes the threshold value M to 0, which is the count number of N, in S104, and subsequently resets the count number of N to 0 in S106, thereby terminating the subroutine S62 for updating the model data shown in FIG. 28 .

[0095] Next, the individual control device 452 determines NO in S64 of the main routine shown in FIG. 26 because pattern matching has not been completed at all edge levels, and changes the edge level to 100 in S66. The individual control device 452 then reads model data candidates for a lead component with an orientation of 270° at the center of the stage for which the edge level was changed to 100 in S50. Next, the individual control device 452 repeatedly executes the processes of S52 to S60, performs pattern matching between the changed model data candidates and the image data for each of the 20 types of lead components 410, and calculates the count N. As shown in FIG. 37, the calculated count N is 0. Therefore, the individual control device 452 determines YES in S60 because pattern matching with the image data for each of the 20 types of lead components 410 has been completed using the changed model data candidates, and executes a subroutine to update the model data in S62.

[0096] 28, the individual control device 452 determines whether the count number N is less than the threshold value M in S100, which is a subroutine of S62. As a result, the individual control device 452 determines NO in S100 because the count number N, 0, is not less than the threshold value M, 0. Then, in S108, the individual control device 452 determines whether the count number N is the same as the threshold value M. As a result, the individual control device 452 determines YES in S108 because the count number N and the threshold value M are both 0, and then, in S110, determines whether the average correlation value of the updated model data candidate, with the edge level changed to 100, is higher than the average correlation value of the current model data. As a result, as shown in FIG. 37, the individual control device 452 calculates the average correlation value of 92 between the newly loaded candidate model data for a lead component with an orientation of 270° at the center of the stage, with the edge level changed to 100. 36 , the average correlation value is 91.3 when using candidate model data for lead components with an edge level of 50 before the edge level change. Therefore, the individual control device 452 determines YES in S110 because the average correlation value using candidate model data with an edge level changed to 100 is higher than the average correlation value using model data before the edge level change. Therefore, it is determined that the candidate model data with an edge level of 100 and an orientation of 270° at the center of the stage is more optimal as model data at this point in time. Therefore, in S102, the individual control device 452 updates the model data candidate with an edge level of 50 and an orientation of 270° at the center of the stage to the candidate model data with an edge level of 100. Then, in S104, the individual control device 452 sets the number of threshold values ​​M to 0, and in S106, it resets the count N to 0, thereby terminating the subroutine for updating model data.

[0097] After completing the subroutine for updating the model data, the individual control device 452 repeatedly executes the processes of S64, S66, and S50 to S62, thereby also executing pattern matching on the model data candidates whose edge levels have been changed to 200 and 300. Note that when the individual control device 452 executes pattern matching using these model data candidates whose edge levels have been changed, the count number of N exceeds 0, which is the value of the threshold M. Therefore, when the count number of N exceeds 0, the model data candidates for lead components whose attitude is 270° at the center of the stage whose edge levels have been changed to 200 and 300 are excluded from the target model data.

[0098] In this way, the individual control device 452 completes pattern matching using all model data candidates and determines YES in S68. Then, in S72, the individual control device 452 registers, as model data, in the storage device 458, the model data candidate with an edge level of 100 and an orientation of 270° at the center of the stage, which was updated in S102 of the subroutine that updates the model data.

[0099] In this way, the individual control device 452 performs pattern matching between any one model data candidate and each of the imaging data of the 20 types of lead components, and excludes from the target model data any model data candidate whose count number of N exceeds the threshold value M. This makes it possible to reduce the number of times pattern matching is performed using model data candidates that have no chance of becoming model data, and ultimately to shorten the time required to identify model data.

[0100] Furthermore, the individual control device 452 executes processing for optimizing the model data while reducing the number of times pattern matching is performed, according to the program 500. This allows the model data to be optimized without increasing the burden on the worker. When new model data is to be registered in the storage device 458, or when model data already registered in the storage device 458 is to be replaced with new model data, the worker can operate the program 500, which causes the individual control device 452 to optimize the model data.

[0101] Furthermore, pattern matching is performed between any one of the model data candidates and each of the image data of the 20 types of lead components, and if the count number of N and the value of the threshold M are the same and the average correlation value of the one of the model data candidates is higher than the average correlation value of the current model data, the one of the model data candidates is updated as new model data, thereby making it possible to further optimize the model data.

[0102] Furthermore, the individual control device 452 performs pattern matching between any one model data candidate and each of the 20 types of lead component imaging data, and if at least one of the component position and component orientation differs, excludes that one model data candidate from the target model data. This makes it possible to exclude from the target model data model data candidates that cannot properly recognize the position or orientation of lead components on the stage, and ultimately makes it possible to properly identify lead components scattered on the stage based on the model data.

[0103] Furthermore, the individual control device 452 performs pattern matching between any one model data candidate and the imaging data of the 20 types of lead components, and if multiple components are recognized based on the imaging data, it excludes that one model data candidate from the target model data. This makes it possible to exclude model data candidates when a single lead component cannot be properly recognized from the target model data, and ultimately makes it possible to properly identify lead components scattered on the stage based on the model data.

[0104] Incidentally, the stage 156 is an example of a stage, the lead component 410 is an example of a component, and the individual control device 452 is an example of an information processing device.

[0105] The present invention is not limited to the above-described embodiment, and various modifications or improvements can be made based on the knowledge of those skilled in the art. Specifically, for example, the individual control device 452 identifies model data by automatically updating the model data candidates multiple times in accordance with the program 500. On the other hand, for example, when updating the model data candidates, the individual control device 452 may update the model data by displaying on a monitor whether or not an update has occurred, and the operator may perform an operation via an input device to update the model data candidates. In other words, the model data may be updated in response to an operation by the operator.

[0106] Furthermore, in the above embodiment, when a model data candidate is updated, the count number of N is reset to the number of threshold value M, but the number of threshold value M may be set in advance to a specific numerical value. For example, by setting threshold value M in advance to 0, pattern matching is performed between one model data candidate and the image data of the above 20 types of lead components, and if the correlation values ​​are not all less than 90, that is, if the correlation values ​​are all 90 or greater, the one model data candidate is updated as model data.

[0107] In the above embodiment, the individual control device 452 creates the model data, but the model data may be created by an information processing device different from the individual control device 452. In this way, the model data created by an information processing device different from the individual control device 452 is stored in the storage device 458.

[0108] In the above embodiment, the present invention is applied to the lead component 410, but it can be applied to various types of components. Specifically, the present invention can be applied to components of solar cells, components of power modules, electronic circuit components without leads, etc.

[0109] 156: Stage 410: Lead component (component) 452: Individual control device (information processing device)

Claims

1. An information processing device that updates model data for identifying multiple parts scattered on a stage based on correlation values ​​obtained by comparing images of the multiple parts scattered on a stage taken in advance and comparing the images of the multiple parts taken in advance with multiple model data candidates, and that, when comparing one of the multiple model data candidates with the images of the multiple parts to obtain a correlation value, excludes the one model data candidate from the candidates for updating the model data if the number of images of the multiple parts determined to have a low correlation value exceeds a certain number.

2. The information processing device according to claim 1, wherein the certain number is the smallest number of part images among the images of the plurality of parts that are judged to have a low correlation value when obtaining a correlation value obtained before the correlation value obtained by comparing the one model data candidate with the images of the plurality of parts.

3. The information processing apparatus according to claim 1 or 2, wherein the model data is automatically updated multiple times based on correlation values ​​obtained by comparing the images of the multiple parts with the multiple model data candidates.

4. A method for updating model data for identifying a plurality of parts scattered on a stage, by performing an image acquisition step of capturing an image of the plurality of parts to acquire images of the plurality of parts, and a step of comparing the images of the plurality of parts acquired in the image acquisition step with a plurality of model data candidates to acquire a correlation value between the images of the plurality of parts and the plurality of model data candidates, wherein, in the step of acquiring the correlation value, when one model data candidate among the plurality of model data candidates is compared with the images of the plurality of parts to acquire the correlation value, the one model data candidate for which the number of images of the plurality of parts determined to have a low correlation value exceeds a certain number is excluded from the candidates for updating the model data.

Images