JP6542641B2 - Component mounting machine, component recognition method - Google Patents

Description

  The present invention relates to a component recognition technique for recognizing a component to be sucked by a nozzle based on a captured image obtained by imaging the vicinity of a nozzle capable of sucking a component.

  2. Description of the Related Art A component mounting machine is generally used which mounts components on a substrate by suctioning the components with a nozzle. In addition, as disclosed in Patent Literatures 1 and 2, in such a component mounting machine, the nozzle adsorbs the component based on the result of judging whether the component can be recognized from the captured image obtained by imaging the vicinity of the nozzle by the camera. Check if it is.

Unexamined-Japanese-Patent No. 10-313198 Patent 4255115 gazette

  However, such component recognition processing has the following problems. That is, even though the nozzle does not suck the component, the nozzle erroneously recognizes that the component is sucked because the nozzle in the captured image is erroneously recognized as the component in the component recognition process. there were. Therefore, there has been a demand for a technology capable of accurately determining the presence or absence of a component to be adsorbed by the nozzle.

  Alternatively, even if the nozzle sucks the part, the part may not be recognized from the captured image in the part recognition process due to an improper attitude of the part or the like. Therefore, the failure may be due to the fact that the nozzle does not adsorb the component or the posture of the component to which the nozzle adsorbs, only by the result of the component recognition processing that the component can not be recognized from the captured image. In some cases, effective measures could not be taken. Therefore, there has been a demand for a technology that makes it possible to easily grasp the suction state of parts by the nozzle at the time of part recognition processing that has failed in part recognition.

  The present invention has been made in view of the above problems, and it is an object of the first object to make it possible to accurately determine the presence or absence of a component to be adsorbed by the nozzle and the suction state of components by the nozzle during component recognition processing which failed in component recognition. An object of the present invention is to provide a technique for achieving at least one of the second objects such as easy to grasp.

  A component mounter according to a first aspect of the present invention, in order to achieve the above first object, a component supply unit for supplying components, a mounting head capable of adsorbing components by a nozzle, and components from the component supply unit. An imaging unit that acquires a captured image captured by a camera that brings a nozzle that has tried suction into a field of view, and control that performs component recognition processing that determines whether a component can be recognized from the captured image based on a predetermined recognition criterion And a storage unit that stores a feature of the part that is different from the image of the nozzle, and the control unit determines that the part has been recognized from the captured image by the part recognition processing. In this case, component feature confirmation processing is performed to confirm whether the component feature is included in the captured image, and the presence or absence of a component to be adsorbed by the nozzle is determined based on the result of the component feature confirmation processing.

  A component recognition method according to a first aspect of the present invention includes the steps of acquiring a captured image captured by a camera that brings within a field of view a nozzle that has attempted to suck a component, in order to achieve the first object. In the process of performing component recognition processing that determines whether a component can be recognized from among the above based on a predetermined recognition criterion, and when it is determined in the component recognition processing that a component has been recognized from a captured image, A step of executing component feature confirmation processing for confirming whether a component feature that is different from the image of the nozzle is included in the captured image and determining whether there is a component to be adsorbed by the nozzle based on the result of the component feature confirmation processing And a process.

  According to the first aspect of the present invention (component mounter, component recognition method) configured as described above, the nozzle sucks the component even when it is determined by the component recognition processing that the component has been recognized from the captured image. If it does, it does not immediately stop, but executes part feature confirmation processing. This makes it possible to confirm whether a component feature that is a feature of the component image and that represents a difference from the image of the nozzle is included in the captured image. Then, the presence or absence of a part to be suctioned by the nozzle is determined based on the result of the part feature confirmation process. Therefore, for example, even if it is a case where a nozzle not picking up a part is erroneously recognized as a part in part recognition processing, it is confirmed in the part feature confirmation processing that a part feature is not included in a captured image, It can be accurately judged that there is no In this way, the first object is achieved such that the presence or absence of a component to be adsorbed by the nozzle can be accurately determined.

  In addition, when it is confirmed by the component feature confirmation processing that the component feature is included in the captured image, the control unit determines that the nozzle is sucking the component and causes the mounting head to mount the component on the substrate, You may comprise a component mounting machine. In such a configuration, the component is recognized from the captured image in the component recognition process, and it is confirmed in the component feature confirmation process that the component feature is included in the captured image. This makes it possible to mount the component on the mounting head after accurately determining that the nozzle is sucking the component.

  Further, the control unit may configure the component mounter so as to update the component feature stored in the storage unit based on the result of imaging the component attracted to the nozzle by the camera. By this, even if the feature of the image of the part changes due to, for example, a change in the lot of the part, the part feature confirmation process can be appropriately performed.

  In addition, the storage unit stores a nozzle feature that is a feature of the nozzle image and indicates a difference from the image of the part, and the control unit determines in the part recognition processing that the part can not be recognized from the captured image. In this case, the component mounter is configured to execute a nozzle feature confirmation process to confirm whether the nozzle feature is included in the captured image and to determine the suction state of the component by the nozzle based on the result of the nozzle feature confirmation process. It is good.

  In this configuration, when it is determined in the component recognition processing that the component can not be recognized from the captured image, the nozzle feature confirmation processing is performed. This makes it possible to confirm whether a nozzle feature that is a feature of the nozzle image and that represents a difference from the image of the part is included in the captured image. Then, the suction state of the component by the nozzle is determined based on the result of the nozzle feature confirmation process. In this way, it is possible to easily grasp the suctioned state of the component by the nozzle at the time of the component recognition processing in which the component recognition has failed.

  Therefore, the control unit may configure the component mounter so as to determine that the nozzle does not adsorb a component when the nozzle feature is confirmed by the nozzle feature confirmation processing that the nozzle feature is included in the captured image. As a result, it is easily understood that the cause of the failure in component recognition in the component recognition processing is that the nozzle does not adsorb the component.

  Alternatively, the component mounter may be configured such that the control unit determines that the posture of the component to be adsorbed by the nozzle is not appropriate when the control unit confirms that the nozzle feature is not included in the captured image in the nozzle feature confirmation processing. As a result, it is easily understood that the cause of the failure in component recognition in the component recognition processing is that the posture of the component to be suctioned by the nozzle is not appropriate.

  The display device may further include a display unit capable of displaying information to the worker, and the control unit may configure the component mounter so as to display the determination result of the suction state of the component by the nozzle on the display unit. As a result, the operator can accurately grasp the suctioned state of the component by the nozzle and can take effective measures.

  In addition, the control unit may configure the component mounter so as to update the nozzle feature stored in the storage unit based on the result of imaging by the camera of a nozzle that does not suck a component. As a result, even if the feature of the image of the nozzle changes due to, for example, the adhesion of dirt, the nozzle feature confirmation process can be appropriately performed.

  In addition, the part feature appears in the relationship between the peripheral length and the area of the feature appearing in the brightness in the contour portion of the image of the part, the feature appearing in the focus of the image of the part, the feature appearing in the luminance of the image of the part The component mounter may be configured to include any of the features.

  A component mounter according to a second aspect of the present invention, in order to achieve the second object, a component supply unit for supplying components, a mounting head capable of adsorbing components by a nozzle, and components from the component supply unit. An imaging unit that acquires a captured image captured by a camera that brings a nozzle that has tried suction into a field of view, and control that performs component recognition processing that determines whether a component can be recognized from the captured image based on a predetermined recognition criterion And a storage unit that stores a nozzle feature that is a feature of the image of the nozzle and indicates a difference from the image of the component, and the control unit determines that the component can not be recognized from the captured image. When it is determined, the nozzle feature confirmation process is performed to confirm whether the nozzle feature is included in the captured image, and the suction state of the component by the nozzle is determined based on the result of the nozzle feature confirmation process.

  In the component recognition method according to the second aspect of the present invention, in order to achieve the above second object, the step of acquiring a captured image captured by a camera that puts a nozzle for which suction of the component was tried within the field of view; The process of performing component recognition processing that determines whether a component can be recognized from among the above based on a predetermined recognition criterion, and the characteristic that the image of a nozzle has when judged by component recognition processing that a component can not be recognized from the captured image And a step of executing a nozzle feature confirmation process for confirming whether or not a nozzle feature representing a difference from the image of the part is included in the captured image, and determining an adsorption state of the part by the nozzle based on the result of the nozzle feature confirmation process And a process.

  The second aspect of the present invention (component mounter, component recognition method) configured as described above executes the nozzle feature confirmation processing when it is determined in the component recognition processing that the component can not be recognized from the captured image. This makes it possible to confirm whether a nozzle feature that is a feature of the nozzle image and that represents a difference from the image of the part is included in the captured image. Then, the suction state of the component by the nozzle is determined based on the result of the nozzle feature confirmation process. In this way, the second object is achieved such that the suction state of the part by the nozzle can be easily grasped at the time of the part recognition processing in which the part recognition has failed.

  The display device may further include a display unit capable of displaying information to the worker, and the control unit may configure the component mounter so as to display the determination result of the suction state of the component by the nozzle on the display unit. As a result, the operator can accurately grasp the suctioned state of the component by the nozzle and can take effective measures.

  In addition, the control unit may configure the component mounter so as to update the nozzle feature stored in the storage unit based on the result of imaging by the camera of a nozzle that does not suck a component. As a result, even if the feature of the image of the nozzle changes due to, for example, the adhesion of dirt, the nozzle feature confirmation process can be appropriately performed.

  As described above, according to the present invention, it is possible to easily grasp the suction state of parts by the nozzle at the time of the part recognition processing in which the first purpose and the part recognition failed. At least one of the second objectives, such as enabling, can be achieved.

It is a fragmentary top view showing typically an example of a component mounting machine concerning the present invention. It is a block diagram which shows an example of the electric constitution with which the component mounting machine of FIG. 1 is provided. It is a figure which shows typically an example of the operation | movement which the mounting head with which the component mounter of FIG. 1 is provided performs. It is a figure which shows an example of the parameter | index which defines components characteristics as a table | surface. It is a figure which shows an example of a luminance histogram. It is a figure which shows an example of the result of having calculated | required the value of the one part parameter | index illustrated in FIG. 5 about the image of each of a nozzle and components. It is a figure which shows an example of the multiple types of components which can be mounted by a component mounting machine. It is a flowchart which shows the 1st example of the arithmetic processing regarding components recognition. It is a flowchart which shows the 2nd example of the arithmetic processing regarding components recognition. It is a flowchart which shows the 3rd example of the arithmetic processing regarding components recognition. It is a flow chart which shows an example of control which updates a part feature and a nozzle feature in parallel with a part mounting operation.

  FIG. 1 is a partial plan view schematically showing an example of a component mounter according to the present invention. FIG. 2 is a block diagram showing an example of the electrical configuration of the component mounter of FIG. FIG. 3 is a view schematically showing an example of an operation performed by the mounting head provided in the component mounter of FIG. In FIG. 1, XYZ orthogonal coordinates are shown as appropriate, which are a Z direction parallel to the vertical direction and an X direction and a Y direction parallel to the horizontal direction, respectively. As shown in FIG. 2, the component mounter 1 includes a controller 100 that generally controls the entire apparatus. The controller 100 includes an arithmetic processing unit 110 which is a computer configured of a central processing unit (CPU) and a random access memory (RAM), and a storage unit 120 configured of a hard disk drive (HDD). Furthermore, the controller 100 includes a drive control unit 130 that controls a drive system of the component mounter 1 and an imaging control unit 140 that controls imaging of a nozzle used for component mounting.

  Then, the arithmetic processing unit 110 controls the drive control unit 130 according to the program stored in the storage unit 120 to execute component mounting in the sequence defined by the program. At this time, the arithmetic processing unit 110 controls component mounting based on the image captured by the component recognition camera 5 by the imaging control unit 140. In addition, the component mounting machine 1 is provided with a display / operation unit 150, and the arithmetic processing unit 110 displays the status of the component mounter 1 on the display / operation unit 150 or inputs it to the display / operation unit 150. It receives instructions from the selected worker.

  As shown in FIG. 1, the component mounter 1 includes a pair of conveyors 12, 12 provided on a base 11. Then, the component mounter 1 mounts the component on the substrate S carried in from the upstream side in the X direction (substrate transfer direction) by the conveyor 12 to the mounting operation position (the position of the substrate S in FIG. 1). The completed substrate S is unloaded by the conveyor 12 from the mounting operation position to the downstream side in the X direction.

  In the component mounting machine 1, a pair of Y-axis rails 21, 21 extending in the Y-direction, a Y-axis ball screw 22 extending in the Y-direction, and a Y-axis motor My for rotating the Y-axis ball screw 22 are provided. The reference numeral 23 is fixed to the nut of the Y-axis ball screw 22 in a state of being supported by the pair of Y-axis rails 21 and 21 so as to be movable in the Y direction. An X-axis ball screw 24 extending in the X direction and an X-axis motor Mx for rotationally driving the X-axis ball screw 24 are attached to the head support member 23, and the head unit 3 can be moved in the X direction with the head support member 23. While being supported by the X axis ball screw 24 is fixed to the nut of the X axis ball screw 24. Therefore, the drive control unit 130 rotates the Y-axis ball screw 22 by the Y-axis motor My to move the head unit 3 in the Y direction, or rotates the X-axis ball screw 24 by the X-axis motor Mx to rotate the head unit 3 It can be moved in the direction.

  Two component supply units 28 are arranged in the X direction on both sides of the pair of conveyors 12 and 12 in the Y direction. A plurality of tape feeders 281 are detachably mounted side by side in the X direction with respect to each component supply unit 28, and each tape feeder 281 has small pieces of electronic components such as integrated circuits, transistors, capacitors, etc. A reel on which an embossed carrier tape TP (FIG. 3), in which P (FIG. 3) is stored at predetermined intervals, is wound is disposed. Then, the tape feeder 281 supplies components in the embossed carrier tape TP by intermittently feeding the embossed carrier tape TP to the head unit 3 side.

  The head unit 3 has a plurality of (four) mounting heads 4 linearly arranged in the X direction. Each mounting head 4 can move up and down in the Z direction by receiving the driving force of the Z-axis motor Mz, and the drive control unit 130 can raise and lower each mounting head 4 by the Z-axis motor Mz. And each mounting head 4 performs adsorption | suction * mounting of the components P by the nozzle N (FIG. 3) attached to the lower end. Specifically, as shown in the column of "component suction operation" in FIG. 3, the mounting head 4 is moved immediately above the component P in the embossed carrier tape TP fed by the tape feeder 281 (step S11). Then, the mounting head 4 descends until the lower end of the nozzle N contacts the upper end surface of the component P, and generates a negative pressure in the nozzle N (step S12). As a result, the component P is attracted to the nozzle N. Subsequently, the mounting head 4 rises while maintaining the negative pressure in the nozzle N, and takes out the component P from the embossed carrier tape TP (step S13). Thus, when the component suction operation is completed, the mounting head 4 moves above the substrate S at the mounting operation position and mounts the component P on the substrate S. Specifically, the mounting head 4 mounts the component P by lowering the nozzle N until the component P abuts on the substrate S and then generating atmospheric pressure or positive pressure in the nozzle N.

  Further, between the component supply units 28 arranged in the X direction, a component recognition camera 5 for imaging the upper side thereof is disposed. The component recognition camera 5 is configured of a CCD (Charge-Coupled Device) camera or the like, and has a field of view 50 facing upward. When the component suction operation is completed, the mounting head 4 places the nozzle N in the field 50 of the component recognition camera 5 via the upper part of the component recognition camera 5 on the way from the component supply unit 28 to the substrate S. Then, as shown in the column of “imaging operation” in FIG. 3, the imaging control unit 140 causes the component recognition camera 5 to perform imaging in a state where the nozzle N falls within the field of view 50. At this time, the imaging control unit 140 brings the component recognition camera 5 into focus in the Z direction with respect to the position (height) of the component P attracted to the nozzle N. In this manner, a captured image Is captured by the component recognition camera 5 that brings the nozzle N, which has tried to suction the component P from the component supply unit 28, into the field of view 50 is acquired. Here, the trial of suction of the component P indicates that a negative pressure is generated in the nozzle N to raise the nozzle N in a state where the nozzle N is in contact with the component P. Further, the imaging control unit 140 stores the captured image Is acquired by the imaging operation in the storage unit 120. Then, the arithmetic processing unit 110 performs arithmetic processing on the captured image Is using the recognition reference Rf, the component feature Fp, and the nozzle feature Fn stored in the storage unit 120 (component recognition processing, component feature confirmation processing, nozzle feature confirmation By executing the process), the presence or absence and state of suction of the component P by the nozzle N are determined.

  The recognition reference Rf is a reference used to recognize the part P from the captured image Is in the part recognition process, and indicates the position of the electrode of the part P (the electrode of the chip part, the lead of the lead part). That is, in the component recognition process, the arithmetic processing unit 110 extracts a region estimated to be the electrode of the component P from the captured image Is. Then, when the position of each extraction area matches the position indicated by the recognition reference Rf, the arithmetic processing unit 110 determines that the part P can be recognized from the captured image Is (success in part recognition), and the position of each extraction area If the position indicated by the recognition reference Rf does not match, it is determined that the component P can not be recognized from the captured image Is (failure in component recognition). When the position of each extraction area is within a predetermined range from the position indicated by the recognition reference Rf, the arithmetic processing unit 110 determines that the position of each extraction area matches the position indicated by the recognition reference Rf.

  The part feature Fp is a reference used to confirm whether or not the feature (part likeness) of the image of the part P is included in the captured image Is in the part feature recognition process, and the part P showing a difference from the image of the nozzle N Feature unique to the image. In particular, the part feature Fp indicates a reference different from the recognition reference Rf, as exemplified later using FIG. Then, as will be described later, in the component feature confirmation process, the arithmetic processing unit 110 confirms whether or not the component feature Fp is included in the captured image Is for which component recognition is successful in the component recognition process. Then, it is evaluated whether the picked-up image Is includes a part likeness, and the validity of the result of the part recognition process is determined.

  The nozzle feature Fn is a reference used to confirm whether the feature (nozzle likeness) of the nozzle N is included in the captured image Is in the nozzle feature recognition process, and an image of the nozzle N showing a difference from the image of the part P Is a unique feature of Then, as described later, in the nozzle feature confirmation process, the operation processing unit 110 confirms whether the nozzle feature Fn is included in the captured image Is for which component recognition has failed in the component recognition process, thereby causing the failure of the component recognition process. Understand

  The part feature Fp and the nozzle feature Fn can be defined based on various indexes as illustrated in FIG. 4. Here, FIG. 4 is a diagram showing an example of an index that defines a part feature as a table. The edge strength is an index indicating the brightness of the contour portion of the object in the image obtained by imaging the object (part P, nozzle N). For example, the background at the edge detected from the captured image Is by edge detection And the average value or the maximum value of the contrast. This edge strength is particularly effective when, for example, the surface of the nozzle N is finished to be matte and the reflection of light at the nozzle N is suppressed. In this case, the image of part P has a high edge strength because it has a bright outline, while the image of nozzle N has a lower edge strength than part P because it has a dark outline. Therefore, by setting the edge strength to be higher than the predetermined edge threshold as the part feature Fp and setting the edge strength to be lower than the edge threshold as the nozzle feature Fn, the captured image Is has any of the part feature Fp and the nozzle feature Fn. It can be identified.

  The focus index is an index indicating the degree to which the captured image Is is in focus, and the degree of focus is calculated using the fact that the contrast is lowered to the extent that the focus is shifted as in the contrast AF (autofocus) of the digital camera. It is The focus index represents the rising sharpness of the edge at the contour portion of the object (part P, nozzle N), and is set to have a larger value as the focus shift is smaller. Specifically, for example, the one obtained by normalizing the difference in luminance between the object and the background can be used as the focus index. At the time of capturing the captured image Is, the focus of the component recognition camera 5 is adjusted to the position of the component P to be attracted to the nozzle N. Therefore, when the component P is attracted to the nozzle N, the focus is high because the captured image Is is in focus. On the other hand, when the component P is not attracted to the nozzle N, the focus index is lowered because the focus of the captured image Is is shifted. Therefore, by setting the focus index to be higher than the predetermined focus threshold as the component feature Fp and setting the focus index to be lower than the focus threshold as the nozzle feature Fn, either of the component feature Fp and the nozzle feature Fn It can be identified.

The interclass variance of the luminance histogram is an index indicating the distribution of luminance in the captured image Is. In order to obtain the interclass variance of the luminance histogram, as shown in FIG. 5, a histogram is created, with the luminance class of the captured image Is as the horizontal axis and the number of pixels having luminances belonging to each class as the vertical axis. Ru. Here, FIG. 5 is a diagram showing an example of the luminance histogram, and in particular, the luminance histogram of the captured image Is including the image of the part P is illustrated. As shown in FIG. 5, when the captured image Is includes the image of the part P, the distribution Dl mainly consisting of the background luminance appears in a range lower than a predetermined threshold Th and the distribution Dh mainly consisting of the luminance of the part P is a threshold Appears in the range higher than Th. Then, the interclass variance σ b is calculated from the following equation σ b ² = ω 1 ··· from the number ω 1 and average m 1 of dark class pixels whose luminance is lower than the threshold Th and the number ω 2 and average m 2 of bright classes whose luminance is higher than the threshold Th. ω ² · (m 1-m ² ) ² / (ω 1 + ω 2) ²
Required based on

  Then, the captured image Is including the image of the part P has high interclass variance σb, while the captured image Is including the image of the nozzle N has low interclass variance σb. Therefore, by setting the inter-class variance σb of the luminance histogram to be higher than a predetermined variance threshold as the component feature Fp, and setting the inter-class variance σb of the brightness histogram to be lower than the variance threshold as the nozzle feature Fn, the captured image Is It is possible to identify which of the part feature Fp and the nozzle feature Fn is included. The interclass dispersion θb of the luminance histogram is particularly effective when reflection of light at the nozzle N is suppressed, as in the case of the edge intensity.

  The shape constant is an index obtained by dividing the peripheral length of the object (part P, nozzle N) by the area, and is particularly effective when the shapes of the contours of the part P and the nozzle N are largely different. For example, if the contour of the part P is rectangular and the contour of the nozzle N is circular, the image of the part P has a high shape constant while the image of the nozzle N has a low shape constant. Therefore, by setting the shape constant to be higher than the predetermined shape threshold as the component feature Fp and setting the shape constant to be lower than the shape threshold as the nozzle feature Fn, the captured image Is has any of the component feature Fp and the nozzle feature Fn. It can be identified.

  FIG. 6 is a view showing an example of a result of obtaining values of some of the indexes illustrated in FIG. 5 for the images of the nozzle and the part. From the figure, it can be seen that the image of the nozzle N and the image of the part P have significantly different values for each of the edge strength, the focus index and the interclass variance. Then, the controller 100 measures the values of each of the four types of indexes in FIG. 4 for the nozzle N in advance and stores them as the nozzle feature Fn in the storage unit 120, and the values of each of the four types of indexes in FIG. In advance are stored in the storage unit 120 as part features Fp. In the case where a plurality of different types of components P can be mounted on the substrate S in the component mounting machine 1, as shown in FIG. 7, the controller 100 sets component characteristics Fp for each type of components P1 to P4. It measures and memorizes in storage part 120. Here, FIG. 7 is a view showing an example of plural types of components mountable by the component mounting machine.

  And the controller 100 which comprises this structure can selectively perform the arithmetic processing regarding component recognition of FIGS. 8-10 mentioned later, for example. Subsequently, a specific flow of arithmetic processing relating to component recognition will be described.

  FIG. 8 is a flow chart showing a first example of arithmetic processing related to component recognition. In step S101, the captured image Is is acquired by execution of the “imaging operation” of FIG. Then, a component recognition process is performed to determine whether the component P can be recognized from among the captured image Is based on the recognition reference Rf (step S102), and it is determined whether the component recognition process is successful (step S103). If it is determined that the component recognition process fails because the component P can not be recognized from the captured image Is ("NO" in step S103), the display / operation unit 150 indicates that the recognition of the component P has failed. And the worker is notified of recognition failure of the part P (step S104). Then, the flowchart of FIG. 8 ends.

  On the other hand, when it is determined that the component P can be recognized from the captured image Is and the component recognition processing is successful (in the case of "YES" in step S103), the feature amount of the captured image Is, that is, the index shown in FIG. The value is measured. Here, among the four types of indices, one type of index selected by the operator by the operation of the display / operation unit 150 is used. Then, in step S106, based on the measurement result in step S105, it is determined whether the captured image Is satisfies the component feature Fp (component feature confirmation processing).

  When it is determined that the captured image Is does not satisfy the part feature Fp (in the case of “No” in step S106), the part recognition process is performed on the part P in the part recognition process despite the absence of the part feature Fp in the captured image Is. It means that the result of having succeeded in recognition came out. Therefore, it is determined that false recognition has occurred, in fact, that there is no part P to which the nozzle N sucks, and the result that the part recognition processing is successful is not valid (step S107). Then, the occurrence of the erroneous recognition in the component recognition process is displayed on the display / operation unit 150, and the worker is notified of the occurrence of the erroneous recognition in the component recognition process (step S108). Then, the flowchart of FIG. 8 ends.

  On the other hand, when it is determined that the captured image Is satisfies the part feature Fp (in the case of “YES” in step S106), recognition of the part P based on the recognition reference Rf is succeeded in part recognition processing, and part feature confirmation processing In the above, a component feature Fp corresponding to a reference different from the recognition reference Rf is confirmed from the captured image Is. Therefore, it is determined that the result that the component recognition processing is successful is valid and that the nozzle N is sucking the component P (step S109). Further, the mounting head 4 mounts the component P to be suctioned to the nozzle N on the substrate S in response to the determination in step S109.

  FIG. 9 is a flowchart showing a second example of the arithmetic processing related to component recognition. In steps S201 and S202, as in the case of steps S101 and 102 in FIG. 8, the captured image Is is acquired (step S201), and component recognition processing is executed (step S202). Then, in step S203, it is determined whether this part recognition process has failed (step S203). If it is determined that the component recognition processing has succeeded (in the case of “NO” in step S203), it is determined that the nozzle N is sucking the component P, and the flowchart of FIG. 9 ends. Further, the mounting head 4 mounts the component P to be suctioned to the nozzle N on the substrate S in response to the determination in step S204.

  On the other hand, when it is determined that the component recognition processing has failed (in the case of “YES” in step S203), the feature value of the captured image Is, that is, the value of the index shown in FIG. 4 is measured. Here, among the four types of indices, one type of index selected by the operator by the operation of the display / operation unit 150 is used. Then, in step S206, it is determined whether the captured image Is satisfies the nozzle feature Fn based on the measurement result in step S205 (nozzle feature confirmation processing).

  If it is determined that the captured image Is does not satisfy the nozzle feature Fn (in the case of “NO” in step S206), the nozzle N is not exposed to the component recognition camera 5, so It can be judged that Therefore, it is determined that the cause of the failure in the component recognition process is that the posture of the component P to be absorbed by the nozzle N is not appropriate because it is inclined (step S207), and the flowchart of FIG. 9 ends. In step S207, the operator is notified by the display / operation unit 150 that the component recognition processing has failed because the posture of the component P is not appropriate.

  On the other hand, when it is determined that the captured image Is satisfies the nozzle feature Fn (in the case of “YES” in step S206), since the nozzle N is exposed to the component recognition camera 5, the nozzle in the captured image Is It can be determined that N is reflected. Therefore, it is determined that the cause of the failure in the component recognition processing is that the nozzle N is not sucking the component P (step S208), and the flowchart of FIG. 9 ends. Further, in step S208, the worker is notified by the display / operation unit 150 that the component recognition processing has failed because the nozzle N does not suck the component P.

  FIG. 10 is a flowchart showing a third example of the arithmetic processing related to component recognition. In steps S301 and S302, as in steps S101 and 102 in FIG. 8, the captured image Is is acquired (step S301), and component recognition processing is performed (step S302). In step S303, the feature amount of the captured image Is is measured as preparation for the subsequent steps S305 and S309. The measurement at this time is performed based on one type of index selected by the operator by the operation of the display / operation unit 150.

  In step S304, it is determined whether the component recognition process for the captured image Is has succeeded. When the component recognition process is successful (in the case of “YES” in step S304), steps S305 to S308 are performed in the same manner as steps S106 to S109 of FIG. 8. That is, when the captured image Is does not satisfy the part feature Fp, the worker is notified of the occurrence of the erroneous recognition in the part recognition process. On the other hand, when the captured image Is satisfies the component feature Fp, it is determined that the nozzle N adsorbs the component P, and the mounting head 4 mounts the component P adsorbed to the nozzle N on the substrate S.

  On the other hand, when the component recognition process fails (in the case of “NO” in step S304), steps S309 to S311 are executed as in steps S206 to S208 of FIG. That is, when the captured image Is does not satisfy the nozzle feature Fn, it is determined that the component recognition processing has failed because the posture of the component P is not appropriate, and the worker is notified of that. On the other hand, when the captured image Is satisfies the nozzle feature Fn, it is determined that the component recognition processing has failed because the nozzle N does not adsorb the component P, and the worker is notified of the failure.

  As described above, in the first example and the third example of the arithmetic processing related to component recognition, the nozzle N is a component even when it is determined by the component recognition processing that the component P can be recognized from the captured image Is. If the suction of P is not performed, the part feature confirmation process (steps S106 and S305) is executed instead of immediately stopping. As a result, it is confirmed whether a component feature Fp that is a feature of the image of the component P and that represents a difference from the image of the nozzle N is included in the captured image Is. And the presence or absence of the component P which the nozzle N adsorbs | sucks is judged based on the result of a component characteristic confirmation process (step S107, S109, S306, 308). Therefore, for example, even if it is a case where a nozzle N which does not suction a part P is erroneously recognized as a part P in the part recognition process, it is confirmed that the part feature Fp is not included in the captured image Is in the part feature confirmation process. It can be accurately determined that there is no part P to be adsorbed by the nozzle N. In this way, it is possible to accurately determine the presence or absence of the component P to be adsorbed by the nozzle N.

  In addition, when the controller 100 confirms that the component feature Fp is included in the captured image Is in the component feature confirmation process (steps S106 and S305), it determines that the nozzle N adsorbs the component P, and the mounting head 4 mount the component P on the substrate S. In such a configuration, the component P is recognized from the captured image Is in the component recognition process, and it is confirmed that the component feature Fp is included in the captured image Is in the component feature confirmation process. Thus, the component P can be mounted on the mounting head 4 after accurately determining that the nozzle N is sucking the component P.

  Further, in the second and third examples of the arithmetic processing related to component recognition, when the component recognition processing determines that the component P can not be recognized from among the captured image Is, the nozzle feature confirmation processing (steps S206 and S309) is executed. Ru. As a result, it is confirmed whether a nozzle feature Fn that is a feature of the image of the nozzle N and that represents a difference from the image of the part P is included in the captured image Is. Then, the suction state of the component P by the nozzle N is determined based on the result of the nozzle feature confirmation process (steps S207, S208, S310, and S311). In this way, it is possible to easily grasp the suction state of the component P by the nozzle N at the time of the component recognition processing in which the component recognition has failed.

  Specifically, when it is confirmed in the nozzle feature confirmation processing that the component feature Fp is included in the captured image Is, the controller 100 determines that the nozzle N does not adsorb the component P. Thus, it can be easily understood that the cause of the failure in component recognition in the component recognition processing is that the nozzle N is not sucking the component P.

  Further, when the controller 100 confirms in the nozzle feature confirmation processing that the component feature Fp is not included in the captured image Is, the controller 100 determines that the posture of the component P suctioned by the nozzle N is not appropriate. By this, it can be easily understood that the cause of the failure in component recognition in the component recognition processing is that the posture of the component P to be suctioned by the nozzle N is not appropriate.

  Moreover, the display / operation unit 150 display part which can display information with respect to a worker is comprised. Then, the controller 100 causes the display / operation unit 150 to display the determination result of the suction state of the component P by the nozzle N. As a result, the operator can accurately grasp the suction state of the component P by the nozzle N, and can take effective measures.

  By the way, the component feature Fp and the nozzle feature Fn stored in the storage unit 120 can be measured and stored in advance before the component mounting operation. Furthermore, the component mounter 1 can be configured to appropriately update the component feature Fp and the nozzle feature Fn in parallel with the component mounting operation.

  FIG. 11 is a flow chart showing an example of control for updating component features and nozzle features in parallel with the component mounting operation. In step S401, the production setup in which the tape feeder 281 containing the predetermined component P is mounted at the predetermined position is executed by the operator, and in step S402, the component mounting operation of sequentially mounting the plurality of components P on the substrate S is stored. In accordance with the production program stored in unit 120, automatic operation to be performed automatically is started.

  Then, the controller 100 causes the mounting head 4 to pass the upper side of the component recognition camera 5 before the component P is sucked from the tape feeder 281, and causes the component recognition camera 5 to capture an image of the nozzle N (step S403). Furthermore, the controller 100 measures the nozzle feature Fn defined by each of the four types of indices shown in FIG. 4 based on the imaging result in step S403, and updates the nozzle feature Fn stored in the storage unit 120 to the measurement value. (Step S404).

  When the mounting head 4 sucks the component P from the tape feeder 281 by the nozzle N (step S405), the controller 100 executes the flowchart of FIG. 10 (step S406). Then, in the flowchart of FIG. 10, it is checked whether it is determined that the nozzle N is sucking the component P (step S407). If it is not determined that the nozzle N is sucking the component P (in the case of “NO” in step S407), the flowchart of FIG. 11 is ended. On the other hand, when it is determined that the nozzle N is sucking the component P (in the case of “YES” in step S407), the storage unit 120 is used as the measurement value obtained by measuring the component feature Fp in step S303 of the flowchart of FIG. The part feature Fp stored in is updated (step S408).

  Then, when the mounting head 4 mounts the component P on the substrate S (step S409), the controller 100 determines whether the mounting of all the components P specified in the production program is completed and the operation is ended (step S410). Then, steps S403 to S409 are repeatedly executed until the mounting of all the parts P is completed ("YES" in step S410).

  As described above, the controller 100 updates the component feature Fp stored in the storage unit 120 based on the result of imaging the component P attracted to the nozzle N by the component recognition camera 5. As a result, even if the feature of the image of the part P changes due to, for example, a change in the lot of the part P, the part feature confirmation process can be appropriately executed based on the appropriate part feature Fp.

  In addition, the controller 100 updates the nozzle feature Fn stored in the storage unit 120 based on the result of imaging the nozzle N not sucking the component P with the component recognition camera 5. Thus, even if the feature of the image of the nozzle N changes due to, for example, the adhesion of dirt, the nozzle feature confirmation process can be appropriately performed based on the appropriate component feature Fp.

  As described above, in the present embodiment, the component mounter 1 corresponds to an example of the “component mounter” of the present invention, and the component supply unit 28 corresponds to an example of the “component supply unit” of the present invention. P corresponds to an example of the "part" of the present invention, the mounting head 4 corresponds to an example of the "mounting head" of the present invention, and the nozzle N corresponds to an example of the "nozzle" of the present invention Corresponds to an example of the “camera” of the present invention, the component recognition camera 5 and the controller 100 cooperate to function as the “imaging unit” of the present invention, and the controller 100 corresponds to an example of the “control unit” of the present invention The storage unit 120 corresponds to an example of the "storage unit" of the present invention, the display / operation unit 150 corresponds to an example of the "display unit" of the present invention, and the part feature Fp corresponds to the "part feature" of the present invention. The nozzle feature Fn corresponds to an example, and the “nozzle feature” of the present invention. Each of steps S102, S202, and S302 corresponds to an example of the "part recognition process" of the present invention, and each of steps S106 and S305 corresponds to an example of the "part feature confirmation process" of the present invention. The substrate S corresponds to an example of the “substrate” in the present invention, and steps S206 and S309 each correspond to an example of the “nozzle feature confirmation process” in the present invention.

  The present invention is not limited to the above-described embodiment, and various modifications can be made to the above-described one without departing from the scope of the invention. For example, the component mounter 1 does not have to be able to execute all of the first to third examples shown above, and may be able to execute any of the first to third examples. At this time, when only the first example is to be executed, it is not necessary to store the nozzle feature Fn in the storage unit 120, and when only the second example is to be executed, the component feature Fp is stored in the storage unit 120. There is no need to keep it.

  Also, various changes can be made to the flowchart of FIG. Therefore, for example, when only the first example of FIG. 8 is executed, the first example of FIG. 8 is executed in step S406, and steps S403 and 404 for acquiring the nozzle feature Fn are omitted. Also good. Alternatively, when only the second example of FIG. 9 is executed, the second example of FIG. 9 may be executed in step S406, and step S408 for acquiring the component feature Fp may be omitted.

  In the second and third examples described above, when it is determined that the captured image Is does not satisfy the nozzle feature Fn in the nozzle feature confirmation process, the failure of the component recognition process that the orientation of the component P is not appropriate Judging that it is the cause, it was displayed on the display / operation unit 150 (steps S207 and S310). However, if the worker makes a mistake in the type of the component P loaded into the tape feeder 281, the component recognition processing may fail even if the component P attracted to the nozzle N is not inclined. Therefore, in steps S207 and S310, it may be determined that the cause or failure of the component recognition processing is that the posture or type of the component P is not appropriate, and this may be displayed on the display / operation unit 150.

  Further, the specific content of the recognition reference Rf used in the part recognition process is not limited to the position of the electrode described above. Therefore, for example, a pattern indicating the outer shape of the part P may be stored in the storage unit 120 as the recognition reference Rf, and in the part recognition process, the part P may be recognized from the captured image Is by pattern matching.

  In addition, four types are prepared in the storage unit 120 as an index for defining the part feature Fp and the nozzle feature Fn. However, it is not necessary to prepare all these indicators in the storage unit 120. Alternatively, a different type of indicator than the one described above can be used.

DESCRIPTION OF SYMBOLS 1 ... Component mounting machine 28 ... Component supply part 4 ... Mounting head 5 ... Component recognition camera 100 ... Controller 120 ... Memory | storage part 150 ... Display / operation unit P ... Component N ... Nozzle Fp ... Component feature Fn ... Nozzle feature S ... Board S

Claims (17)

A parts supply unit for supplying parts;
A mounting head capable of suctioning the component by a nozzle;
An imaging unit that acquires a captured image captured by a camera that puts in the field of view the nozzle that has tried to suck the component from the component supply unit;
A control unit that executes component recognition processing that determines whether the component can be recognized from among the captured images based on a predetermined recognition standard;
A storage unit storing a feature of the image of the part, the feature of the part indicating a difference from the image of the nozzle;
The control unit executes component feature confirmation processing to confirm whether the component feature is included in the captured image, when it is determined in the component recognition processing that the component has been recognized from the captured image. The part stored in the storage unit based on the result of imaging the part adsorbed by the nozzle by the camera while determining the presence or absence of the part adsorbed by the nozzle based on the result of the part feature confirmation process Component mounter that updates features . A parts supply unit for supplying parts;
A mounting head capable of suctioning the component by a nozzle;
An imaging unit that acquires a captured image captured by a camera that puts in the field of view the nozzle that has tried to suck the component from the component supply unit;
A control unit that executes component recognition processing that determines whether the component can be recognized from among the captured images based on a predetermined recognition standard;
A storage unit storing a feature of the image of the part, the feature of the part indicating a difference from the image of the nozzle;
The control unit executes component feature confirmation processing to confirm whether the component feature is included in the captured image, when it is determined in the component recognition processing that the component has been recognized from the captured image. Determining the presence or absence of the part adsorbed by the nozzle based on the result of the part feature confirmation process ;
The storage unit stores a nozzle feature that is a feature of the image of the nozzle and indicates a difference from the image of the part.
The control unit executes a nozzle feature confirmation process for confirming whether the nozzle feature is included in the captured image, when it is determined in the component recognition process that the component can not be recognized from the captured image. The component mounting machine which judges the adsorption | suction state of the said components by the said nozzle based on the result of the said nozzle characteristic confirmation process . The component mounter according to claim 2 , wherein the control unit determines that the nozzle does not adsorb the component when it is confirmed in the nozzle feature confirmation processing that the nozzle feature is included in the captured image. Wherein, when the nozzle characteristics were confirmed by the nozzle, wherein the confirmation processing is not included in the captured image, according to claim 2 or 3 wherein the nozzle is determined that is not appropriate orientation of the components to be adsorbed Part mounting machine. It further comprises a display unit that can display information to the workers,
The component mounter according to any one of claims 2 to 4 , wherein the control unit causes the display unit to display a determination result of a suction state of the component by the nozzle. The component mounting according to any one of claims 2 to 5 , wherein the control unit updates the nozzle feature stored in the storage unit based on a result of imaging the nozzle not sucking the component by the camera. Machine. The control unit determines that the nozzle adsorbs the component when the component feature is confirmed in the component feature confirmation process that the component feature is included in the captured image, and the mounting head is configured to mount the component on the substrate The component mounting machine according to any one of claims 1 to 6, which is mounted on the The part mounter according to any one of claims 1 to 7 , wherein the part feature includes a feature appearing in brightness in a contour portion of an image of the part. The part mounter according to any one of claims 1 to 8 , wherein the part feature includes a feature appearing in a focus of an image of the part. The component mounter according to any one of claims 1 to 9 , wherein the component feature includes a feature appearing in luminance of an image of the component. The part mounter according to any one of claims 1 to 10 , wherein the part feature includes a feature that appears in the relation between the perimeter and the area of the image of the part. A parts supply unit for supplying parts;
A mounting head capable of suctioning the component by a nozzle;
An imaging unit that acquires a captured image captured by a camera that puts in the field of view the nozzle that has tried to suck the component from the component supply unit;
A control unit that executes component recognition processing that determines whether the component can be recognized from among the captured images based on a predetermined recognition standard;
And a storage unit that stores a nozzle feature that is a feature of the image of the nozzle and indicates a difference from the image of the part.
The control unit executes a nozzle feature confirmation process for confirming whether the nozzle feature is included in the captured image, when it is determined in the component recognition process that the component can not be recognized from the captured image. The component mounting machine which judges the adsorption | suction state of the said components by the said nozzle based on the result of the said nozzle characteristic confirmation process. It further comprises a display unit that can display information to the workers,
The component mounter according to claim 12 , wherein the control unit causes the display unit to display a determination result of a suction state of the component by the nozzle. The component mounter according to claim 12 , wherein the control unit updates the nozzle feature stored in the storage unit based on a result of imaging the nozzle that does not adsorb the component with the camera. Acquiring a captured image captured by a camera that brings within its field of view a nozzle that has attempted to suck a component;
Performing a component recognition process of determining whether the component can be recognized from the captured image based on a predetermined recognition standard;
If it is determined in the component recognition processing that the component has been recognized from the captured image, whether the component feature which is the feature of the image of the component and represents the difference from the image of the nozzle is included in the captured image A step of executing a part feature confirmation process to be confirmed;
Determining the presence or absence of the part adsorbed by the nozzle based on the result of the part feature confirmation process ;
Updating the part feature stored in the storage unit based on the result of imaging the part adsorbed by the nozzle with the camera . Acquiring a captured image captured by a camera that brings within its field of view a nozzle that has attempted to suck a component;
Performing a component recognition process of determining whether the component can be recognized from the captured image based on a predetermined recognition standard;
If it is determined in the component recognition processing that the component has been recognized from the captured image, whether the component feature which is the feature of the image of the component and represents the difference from the image of the nozzle is included in the captured image A step of executing a part feature confirmation process to be confirmed;
Determining the presence or absence of the part adsorbed by the nozzle based on the result of the part feature confirmation process ;
When it is determined in the component recognition processing that the component can not be recognized from the captured image, a nozzle feature that is a feature of the image of the nozzle and shows a difference from the image of the component is the captured image A nozzle characteristic confirmation process for confirming whether or not it is included, and determining an adsorption state of the component by the nozzle based on the result of the nozzle characteristic confirmation process . Acquiring a captured image captured by a camera that brings within its field of view a nozzle that has attempted to suck a component;
Performing a component recognition process of determining whether the component can be recognized from the captured image based on a predetermined recognition standard;
If it is determined in the component recognition processing that the component can not be recognized from the captured image, whether or not the captured image includes a nozzle feature which is a feature of the nozzle image and which is different from the image of the component Performing a nozzle feature confirmation process for confirming
Determining a suction state of the component by the nozzle based on a result of the nozzle feature confirmation process.