JP2023089746A - Image recognition device and component mounting machine - Google Patents

Abstract

To provide an image recognition device with which it is possible to perform accurate image recognition of a subject, and a component mounting machine that includes this image recognition device.SOLUTION: An image recognition device 4 comprises a polarized image output unit 42, an image identification unit 43, a recognition condition setting unit 44, and an image recognition unit 45. The polarized image output unit 42 acquires a plurality of captured polarized images G1 differentiated in the direction of polarization by receiving the light reflected by a subject, and outputs a plurality of polarized images GA including the plurality of captured polarized images G1. The image identification unit 43 identifies an image GC of interest from the plurality of polarized images GA, the contrast value C of which is highest. The recognition condition setting unit 44 sets the direction of polarization corresponding to the image GC of interest as a recognition condition DS for the image recognition of the subject. The image recognition unit 45 causes a polarized image conforming to the recognition condition DS to be output as an image GD for recognition, from the polarized image output unit 42, and does image recognition of the subject on the basis of the image GD for recognition.SELECTED DRAWING: Figure 5

Description

The present invention relates to an image recognition device for recognizing an image of a subject and a component mounter equipped with the same.

A component mounter for mounting electronic components on a board such as a printed wiring board is equipped with a head unit that holds the electronic component and mounts the held electronic component on the board at a predetermined position. there is When the head unit performs a component mounting operation, an image of a subject such as an electronic component or a substrate is captured, and image recognition of the subject is performed based on the captured image.

The electronic component has a component body and electrodes with different light reflection characteristics, and the substrate has a substrate body and identification marks with different light reflection characteristics. In other words, the object to be image-recognized during the component mounting operation by the head unit has a plurality of areas with different light reflection characteristics. In order to accurately recognize an image of a subject, it is necessary to increase the contrast between image areas corresponding to a plurality of areas of the subject in an image of the subject.

Japanese Patent Application Laid-Open No. 2002-200002 discloses a technique for image recognition of a substrate as a subject. In this technique, a polarizing filter is placed on the optical path between the light source and the substrate, and a polarizing filter is also placed on the optical path between the substrate and the camera. In this case, each polarizing filter is detachable and rotatable around the optical axis.

Further, Patent Literature 2 discloses a technique for recognizing an image of an electronic component as a subject. With this technology, a polarizing filter is placed on the optical path between the light source and the electronic component to irradiate the electronic component with polarized light in a specific direction, and a polarizing filter is also placed on the optical path between the electronic component and the camera. light in a specific direction and is received by the camera.

JP-A-5-152799 Japanese Patent Publication No. 2005-504986

In the technique disclosed in Patent Document 1, an image of the subject is acquired while adjusting the attachment/detachment and rotation of each polarizing filter, but in order to acquire an image with the best contrast, precise adjustment is required. , which can be difficult to adjust. Further, in the technique disclosed in Patent Document 2, since each polarizing filter is in a fixed state without being rotated, it may not be possible to acquire an image having the best contrast as an image of the subject. Therefore, the techniques disclosed in Patent Literatures 1 and 2 may not be able to accurately recognize the image of the subject.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image recognition apparatus capable of accurately recognizing an image of a subject, and a component mounter having the same.

An image recognition device according to one aspect of the present invention is a device for image recognition of a subject having a plurality of areas with different light reflection characteristics. This image recognition device acquires a plurality of polarized captured images having different polarization directions by receiving light reflected by the subject, and a polarized image output unit that outputs a plurality of polarized images including the plurality of polarized captured images. , an image identifying unit that calculates a contrast value of each of the plurality of polarization images and identifies an image of interest having the highest contrast value; and an image recognition unit that outputs a polarized image according to the recognition condition as an image for recognition from the polarized image output unit, and recognizes the subject based on the image for recognition. Prepare.

According to this image recognition device, the image specifying unit specifies the image of interest with the highest contrast value based on the plurality of polarized images having different polarization directions output from the polarized image output unit. When the image specifying unit specifies the target image from among the plurality of polarized images, the recognition condition setting unit sets the polarization direction corresponding to the target image as the recognition condition for image recognition of the object. Then, the image recognition section causes the polarization image output section to output the polarization image according to the recognition condition automatically set by the recognition condition setting section as the image for recognition. The recognition image is a polarized image that conforms to the recognition condition and is an image in a polarization direction corresponding to the image of interest with the highest contrast value among the plurality of polarized images that can be output from the polarized image output unit. Therefore, the image for recognition is an image of the subject with good contrast. Therefore, the image recognition section can accurately perform image recognition of the subject based on the recognition image according to the recognition conditions.

In the above image recognition device, the polarization image output unit generates a plurality of polarization composite images by synthesizing the plurality of polarization captured images while changing the weighting, and uses the plurality of polarization composite images as the polarization images. It may be configured to output.

In this aspect, the polarization image output unit generates a plurality of polarization combined images by combining a plurality of polarization captured images with different polarization directions while changing the weighting. In this case, the polarization image output section generates a plurality of polarization composite images of different polarization directions subdivided according to the weighting. The polarization image output unit outputs a plurality of polarization images in different polarization directions by outputting the plurality of polarization composite images as polarization images. In this case, the image specifying unit can specify the image of interest with the highest contrast value based on the plurality of subdivided polarization images with different polarization directions. As a result, the image for recognition, which is a polarized image according to the recognition condition indicated by the polarization direction corresponding to the image of interest, becomes an image of the object with more reliable contrast. Therefore, the image recognition section can perform image recognition of the subject more accurately based on the recognition image according to the recognition conditions.

In the above image recognition device, the polarization image output unit generates a polarization angle image expressed in grayscale for each polarization direction angle based on the plurality of polarization captured images, and uses the polarization angle image as the polarization image. It may be configured to output.

In this aspect, the polarization image output section outputs the polarization angle image as a polarization image. A polarization angle image is an image based on a plurality of polarization captured images with different polarization directions, and is an image expressed in grayscale for each angle of the polarization direction. Therefore, when the image specifying unit specifies the polarization angle image as the image of interest, the recognition condition setting unit sets the polarization direction with the highest contrast value as the recognition condition based on the polarization angle image specified as the image of interest. can do.

In the image recognition device described above, the polarized image output unit calculates a linear polarization degree for each pixel based on the plurality of polarized light captured images, and linearly polarized light composed of a pixel group having a luminance value corresponding to the linear polarization degree. The configuration may be such that a polarization image is generated and output as the polarization image. In this case, the recognition condition setting unit recognizes that the polarization image output unit generates and outputs the linear polarization degree image when the image specifying unit specifies the linear polarization degree image as the image of interest. Set as a condition.

In this aspect, the polarization image output section outputs the linear polarization degree image as the polarization image. The linear polarization degree image is an image based on a plurality of polarization captured images with different polarization directions, and is an image composed of a pixel group having a luminance value corresponding to the linear polarization degree. When the image specifying unit specifies the linear polarization degree image as the image of interest, the recognition condition setting unit sets the recognition condition that the polarization image output unit generates and outputs the linear polarization degree image. In this case, the image recognition section causes the polarization image output section to output the linear polarization degree image as the recognition image according to the recognition conditions. The image recognition section can accurately recognize the image of the subject based on the recognition image represented by the linear polarization degree image according to the recognition conditions.

In the above image recognition device, the image specifying unit generates a brightness histogram indicating the number of pixels for each brightness value for each of the plurality of polarized images, and based on the brightness histogram, the contrast of the plurality of polarized images. The configuration may be such that each value is calculated.

In this aspect, the image specifying unit can accurately calculate the contrast value of each of the plurality of polarization images based on the brightness histogram. Thereby, the image specifying unit can more accurately specify the image of interest with the highest contrast value from among the plurality of polarized images output by the polarized image output unit.

A component mounter according to another aspect of the present invention includes a substrate transport section that transports a substrate with a predetermined mark to a predetermined position; A head unit that performs a component mounting operation to be mounted on the substrate at a predetermined position; An image recognition device for recognizing an image as the subject, and a head control section for controlling the component mounting operation of the head unit based on a recognition result by the image recognition section of the image recognition device.

According to this component mounter, the image recognition unit recognizes the electronic component or the board based on the recognition image in which the subject is the electronic component held by the head unit or the board on which the electronic component is to be mounted. do. The head control section controls the component mounting operation of the head unit based on the recognition result of the image recognition section. The component mounting operation of the head unit controlled based on the recognition result of the image recognition unit improves the mounting accuracy of the electronic component on the board.

In the component mounter, target holding position information, which is management data referred to when the head control unit controls the head unit, indicates a target holding position when the head unit holds the electronic component. and target mounting position information indicating the target mounting position of the electronic component set on the board. The management data storage unit stores the management data in association with the recognition condition set by the recognition condition setting unit.

In this aspect, the image recognition section can refer to the recognition conditions associated with the management data stored in the management data storage section when performing image recognition of the electronic component or board based on the recognition image. In this case, the image recognition section causes the polarized image output section to output the polarized image according to the recognition condition associated with the management data as the recognition image.

In the component mounter described above, the recognition conditions set by the recognition condition setting unit are component data including parameter data relating to the characteristics of the electronic component, and board data including data relating to the mark attached to the board. are stored in association with the component data or the board data in a predetermined storage storage unit that stores and stores the above.

In this aspect, the recognition condition set by the recognition condition setting unit is accumulated and stored in the predetermined accumulation storage unit in association with the component data or the board data. In this case, the image recognition section can read out the recognition conditions from the accumulation storage section and use them each time a polarized image according to the recognition conditions is output from the polarized image output section as an image for recognition.

In the component mounter described above, the image specifying unit includes, for each of the plurality of polarization images, an electrode image area corresponding to the electrode on the electronic component or a mark image area corresponding to the mark on the substrate. A region of a predetermined range is set as a target region for calculating the contrast value.

When accurately recognizing an image of an electronic component, it is possible to recognize the image of the electronic component based on a recognition image having a high contrast between an electrode image area corresponding to a characteristic electrode of the electronic component and an image area surrounding the electrode image area. desirable. Similarly, when accurately recognizing the image of the substrate, the image of the substrate can be recognized based on the recognition image in which the contrast between the mark image area corresponding to the characteristic mark on the substrate and the image area around the mark image area is high. desirable. In view of such circumstances, the image of interest that the recognition condition setting unit refers to when setting the recognition condition corresponding to the recognition image is the contrast between the electrode image area or mark image area and the surrounding image area. It is desirable that the image has a high

Therefore, for each of the plurality of polarization images, the image specifying unit sets a predetermined area including the electrode image area or the mark image area as the target area for calculating the contrast value. In this case, the image specifying unit specifies, as the image of interest, the polarized image having the highest contrast value within a predetermined range including the electrode image region or the mark image region among the plurality of polarized images output by the polarized image output unit. do. As a result, the image for recognition, which is a polarized image according to the recognition condition corresponding to the image of interest, is an image with high contrast between the electrode image area or mark image area and the surrounding image area. Therefore, the image recognition section can accurately perform image recognition of the electronic component or substrate as the subject based on the recognition image.

As described above, according to the present invention, it is possible to provide an image recognition device capable of accurately recognizing an image of a subject, and a component mounter equipped with the same.

1 is a block diagram of a component mounter to which an image recognition device according to an embodiment of the invention is applied; FIG. FIG. 2 is a plan view showing the configuration of a mounter main body in the component mounter; 3 is a diagram schematically showing a light irradiation section and a polarized image output section provided in the image recognition device; FIG. It is a figure which shows the polarization image output by the polarization image output part. FIG. 4 is a diagram for explaining each process of an image specifying unit, a recognition condition setting unit, and an image recognition unit provided in the image recognition device; It is a figure which shows the modification of a polarization image output part.

DESCRIPTION OF THE PREFERRED EMBODIMENTS An image recognition device and a mounter equipped with the image recognition device according to embodiments of the present invention will be described below with reference to the drawings. In the following, the directional relationship will be explained using XY rectangular coordinates that are orthogonal to each other on the horizontal plane.

A component mounter 1 shown in FIGS. 1 and 2 is a device that mounts (mounts) electronic components on a board PP to produce an electronic circuit board. Examples of electronic components include chip components having electrodes at one end and the other end of a component body, SOP (Small Outline Package), QFP (Quad Flat Package), PLCC (Plastic Leaded Chip Carrier), BGA (Ball Grid Array). ) and other multiple types of parts. The SOP is a component in which a plurality of electrodes are arranged at one end and the other end in the X-axis direction of the component body. The QFP and PLCC are components in which a plurality of electrodes are arranged at one end and the other end of the component body in the X-axis direction, and a plurality of electrodes are arranged at one end and the other end of the component body in the Y-axis direction. A BGA is a component in which a plurality of ball-shaped electrodes are provided on the lower surface of a component body. Some electronic components do not have electrodes exposed to the lower surface of the component body.

The component mounter 1 includes a mounter body 2 , a control device 3 , an image recognition device 4 and a storage device 5 .

The mounter main body 2 constitutes a structural part that performs a component mounting operation for mounting electronic components on the board PP during the production of the electronic circuit board. A solder paste pattern is printed on the substrate PP before the electronic component is mounted by the mounter main body 2 . That is, the mounting machine main body 2 mounts electronic components on the substrate PP on which the solder paste pattern is printed. The mounter body 2 includes a body frame 21 , a board transfer section 22 , a component supply device 23 , a head unit 25 and a board support device 28 .

The main body frame 21 is a structure in which each part constituting the mounting machine main body 2 is arranged, and is formed in a substantially rectangular shape when viewed in a plan view in a direction perpendicular to both the X-axis direction and the Y-axis direction (vertical direction). It is The substrate transfer section 22 is configured by a conveyor and arranged on the body frame 21 so as to extend in the X-axis direction. The substrate transporter 22 transports the substrate PP in the X-axis direction. The board PP transported by the board transport section 22 is positioned by the board supporting device 28 at a predetermined work position (component mounting position where components are mounted on the board PP). The substrate supporting device 28 positions the substrate PP at the component mounting position by supporting the substrate PP from below.

The component supply devices 23 are arranged in respective regions of both ends of the body frame 21 in the Y-axis direction. The component supply method is not particularly limited as long as the component supply device 23 is configured to be able to supply electronic components. Examples of the component supply device 23 include a tape feeder that supplies electronic components using a tape as a carrier, and a tray that supplies electronic components by moving a pallet containing a tray on which electronic components are placed. A feeder, a stick feeder that feeds electronic components stored in a cylindrical stick while pushing them out from the stick, or the like can be adopted.

The head unit 25 is held by a moving frame 27 . A fixed rail 261 extending in the Y-axis direction and a ball screw shaft 262 rotationally driven by a Y-axis servomotor 263 are arranged on the body frame 21 . The moving frame 27 is arranged on a fixed rail 261 , and a nut portion 271 provided on this moving frame 27 is screwed onto a ball screw shaft 262 . A guide member 272 extending in the X-axis direction and a ball screw shaft 273 driven by an X-axis servomotor 274 are arranged on the moving frame 27 . The head unit 25 is movably held by the guide member 272 , and a nut portion provided on the head unit 25 is screwed onto the ball screw shaft 273 . The Y-axis servomotor 263 is operated to move the moving frame 27 in the Y-axis direction, and the X-axis servomotor 274 is operated to move the head unit 25 relative to the moving frame 27 in the X-axis direction. there is That is, the head unit 25 can move in the Y-axis direction as the moving frame 27 moves, and can move along the moving frame 27 in the X-axis direction. The head unit 25 is movable between the component supply device 23 and the substrate PP supported by the substrate support device 28 .

The head unit 25 has a plurality of holding nozzles 251 . The plurality of holding nozzles 251 are detachably attached to the head unit 25 . Each holding nozzle 251 is, for example, a nozzle capable of sucking and holding an electronic component supplied by the component supply device 23 . In this case, each holding nozzle 251 can communicate with any one of the negative pressure generator, the positive pressure generator, and the atmosphere via the electric switching valve. In other words, the supply of negative pressure to each holding nozzle 251 enables the holding nozzle 251 to suck and hold the electronic component, and then the supply of positive pressure cancels the sucking and holding of the electronic component.

Each holding nozzle 251 can move up and down with respect to the frame of the head unit 25 in a vertical direction (Z-axis direction) orthogonal to both the X-axis direction and the Y-axis direction, and can rotate around the nozzle axis extending in the Z-axis direction. is possible. Each holding nozzle 251 can move up and down along the Z-axis direction between a holdable position capable of holding an electronic component supplied by the component supply device 23 and a retracted position above the holdable position. is. That is, when holding an electronic component supplied by the component supply device 23, each holding nozzle 251 descends from the retracted position toward the holdable position, and holds the electronic component at the holdable position. On the other hand, each holding nozzle 251 after holding the electronic component rises from the holdable position toward the retracted position. Further, each holding nozzle 251 is arranged along the Z-axis direction between a mountable position where the held electronic component can be mounted on a predetermined target mounting position on the substrate PP and the retracted position. It is possible to go up and down.

The head unit 25 performs a component mounting operation for mounting the electronic component P (FIG. 3) held by each holding nozzle 251 on the substrate PP, corresponding to each of a plurality of target mounting positions set on the substrate PP.

Further, as shown in FIG. 2, the head unit 25 is provided with a first imaging camera 252 and a second imaging camera 253 . The first imaging camera 252 and the second imaging camera 253 are cameras having imaging elements such as CMOS (Complementary metal-oxide-semiconductor) and CCD (Charged-coupled device).

The first imaging camera 252 acquires an image by capturing an image of the mark M from above in order to recognize the image of the mark M attached to the upper surface of the board PP transported to the component mounting position by the board transport unit 22. do. By recognizing the mark M on the substrate PP based on the image captured by the first imaging camera 252, the positional deviation amount of the substrate PP with respect to the origin coordinates is detected. Image recognition of the marks M provided on the substrate PP may be performed by an image recognition device 4 described later.

The second imaging camera 253 captures an image of the supply position of the electronic component P in the component supply device 23 from obliquely above. The image captured by the second imaging camera 253 is referred to when recognizing the attitude of the electronic component P supplied to the component supply position by the component supply device 23 .

The storage device 5 includes a management data storage unit 51 that stores management data D1 referred to by the control device 3, and an accumulation storage unit 52 that accumulates and stores component data D2 and board data D3 referred to by the image recognition device 4. and including.

The management data D<b>1 stored in the management data storage unit 51 is data composed of various kinds of information required for control of the component mounting process and the like by the control device 3 . Information constituting the management data D1 includes, for example, component information, component supply information, nozzle information, head information, target holding position information, target mounting position information, and the like. Component information is information for identifying the type of electronic component P. FIG. The component supply information is information for specifying the component supply device 23 . The nozzle information is information for identifying the type of the held nozzle 251 . The head information is information for specifying the head unit 25 . The target holding position information is information indicating the target holding position when the holding nozzle 251 holds the electronic component P. FIG. The target mounting position information is information indicating the target mounting position of the electronic component P set on the board PP.

As shown in FIG. 1, the management data storage unit 51 stores the management data D1 in association with the recognition condition DS set by the recognition condition setting unit 44 of the image recognition device 4 described later.

The component data D2 and the board data D3 accumulated and stored in the accumulation storage unit 52 will be described with reference to FIG. The component data D2 is composed of data relating to the characteristics of the electronic component P. FIG. Data on the characteristics of the electronic component P include the external dimensions of the electronic component P, the size of the electrode P2 (FIG. 3) in the electronic component P, the shape of the electrode P2, and the plurality of electrodes P2 for the component body P1 (FIG. 3). In the case of the electronic parts P arranged in a predetermined arrangement direction, each data such as the pitch between the electrodes P2 can be mentioned. The board data D3 is composed of data relating to the characteristics of the board PP. Data relating to the characteristics of the substrate PP include data such as the shape of the mark M provided on the substrate PP and the position of the mark M with respect to the substrate PP.

The accumulation storage unit 52 accumulates and stores component data D2 for each type of electronic component P, and accumulates and stores board data D3 for each type of board PP. As shown in FIG. 1, the accumulation storage unit 52 stores the component data D2 or the board data D3 in association with the recognition condition DS set by the recognition condition setting unit 44 of the image recognition device 4 described later. Note that the accumulation storage unit 52 is not limited to being included in the storage device 5 provided in the mounter 1 . For example, the accumulation storage section 52 may be incorporated in a server device that is separate and independent from the mounter 1 .

The control device 3 includes a CPU (Central Processing Unit), a ROM (Read Only Memory) for storing control programs, a RAM (Random Access Memory) used as a work area for the CPU, and the like. The control device 3 controls the operation of each component of the mounting machine body 2 by the CPU executing a control program stored in the ROM. The control device 3 controls the operation of each component according to the management data D1 stored in the management data storage unit 51. FIG. As shown in FIG. 1, the control device 3 includes a substrate transfer control section 31, a component supply control section 32, and a head control section 33 as main functional components.

The board transfer control section 31 controls the transfer operation of the board PP by the board transfer section 22 . The component supply control unit 32 controls the supply operation of the electronic components P by the component supply device 23 according to the component information and the component supply information of the management data D1. The head control unit 33 controls the holding nozzles 251 by controlling the head unit 25 according to the component information, nozzle information, head information, target holding position information, and target mounting position information of the management data D1. Accordingly, the head control unit 33 causes the holding nozzle 251 to perform a holding operation for holding the electronic component P corresponding to each of the plurality of target mounting positions set on the substrate PP, and moves the head unit 25. to execute the component mounting operation.

The image recognition device 4 is a device for image recognition of a subject having a plurality of areas with different light reflection characteristics. The image recognition device 4 will be described with reference to FIGS. 3 to 6 in addition to FIGS. 1 and 2. FIG.

The electronic component P and the circuit board PP are examples of subjects for image recognition by the image recognition device 4 . The electronic component P has a component body P1 and an electrode P2 having different light reflection characteristics. In the electronic component P, the electrode P2 made of metal has a reflection characteristic of mirror-reflecting incident light, and the component body P1 made of non-metal such as ceramics or resin has a reflection characteristic of diffusely reflecting incident light. . Further, the substrate PP has a substrate main body having different light reflection characteristics and a recognition mark M. As shown in FIG. The image recognition device 4 recognizes the image of the electronic component P held by the holding nozzle 251 of the head unit 25 or the substrate PP on which the electronic component P held by the holding nozzle 251 is mounted as a subject. Below, the image recognition device 4 in the case where the electronic component P is used as the subject will be described.

The image recognition device 4 performs image recognition of the electronic component P held by the holding nozzle 251 of the head unit 25 before the electronic component P held by the holding nozzle 251 of the head unit 25 is mounted on the substrate PP. The image recognition device 4 includes a light irradiation section 41 , a polarized image output section 42 , an image specifying section 43 , a recognition condition setting section 44 and an image recognition section 45 . The light irradiation unit 41 and the polarized image output unit 42 are arranged on the body frame 21 (see FIG. 2). On the other hand, the image specifying unit 43, the recognition condition setting unit 44, and the image recognition unit 45 may be configured by a microcomputer that is independent from the control device 3, or may be integrated into the control device 3. may

As shown in FIG. 3, the light irradiation unit 41 irradiates the electronic component P held by the holding nozzle 251 of the head unit 25 with non-polarized light L1 obliquely from below. When the light L1 emitted from the light irradiation unit 41 is incident on the electronic component P, it is diffusely reflected by the component main body P1 and specularly reflected by the electrode P2.

The polarized image output unit 42 is arranged below the electronic component P held by the holding nozzle 251, and the light L2 reflected by the electronic component P in response to the incident light L1 emitted from the light irradiation unit 41 is generated. receive light. Thereby, the polarized image output unit 42 acquires a plurality of polarized captured images G1 having different polarization directions shown in FIG. The polarization image output unit 42 outputs a plurality of polarization images GA including a plurality of polarization captured images G1. Each polarized captured image G1, which is an image of the electronic component P, includes a component body image area AR1 corresponding to the component body P1 and an electrode image area AR2 corresponding to the electrode P2.

In this embodiment, the polarization image output unit 42 is configured by a polarization camera equipped with a polarizer stacked imaging element 421 in which polarizers 4212 with different polarization directions are stacked on each pixel of an imaging element 4211 such as a CMOS or CCD. be. Each pixel constituting the polarizer-stacked imaging element 421 is provided with a polarizer 4212 that functions as an optical filter that transmits only light polarized in a specific direction. An imaging device 4211 that receives light transmitted through the polarizer 4212 is provided below the polarizer 4212 .

The polarizer 4212 set in each pixel constituting the polarizer-stacked imaging device 421 has a plurality of pixels (for example, 4 pixels) as a unit, and these plurality of pixels (4 pixels) transmit only light in different polarization directions. It is configured to let In the case of the polarizer-stacked imaging element 421 in which the polarizers 4212 each having four pixels as one unit are stacked, the polarization directions of the four pixels a, b, c, and d of the polarizer-stacked imaging element 421 are, for example, as follows. is set to That is, the polarization direction of pixel a is the direction of 0 degrees in the horizontal direction, and in this case, pixel a receives only polarized light in the direction of 0 degrees. The polarization direction of the pixel b is the 45-degree direction oblique to the upper right, and in this case, the pixel b receives only polarized light in the direction of 45 degrees. The polarization direction of pixel c is 90 degrees in the vertical direction, and in this case, pixel c receives only polarized light in the direction of 90 degrees. The polarization direction of the pixel d is the direction of 135 degrees oblique to the lower right, and in this case, the pixel d receives only polarized light in the direction of 135 degrees.

In the above description, the horizontal direction, upper right oblique direction, lower right oblique direction, and vertical direction are the directions with respect to the polarization camera that constitutes the polarization image output unit 42, and the direction perpendicular to the optical axis of the polarization camera is the horizontal direction. Direction, the direction parallel to the optical axis of the polarization camera is taken as the vertical direction. Therefore, the polarization direction of each pixel changes according to the tilt of the polarization camera.

The polarized image output unit 42 configured by a polarization camera equipped with a polarizer-stacked imaging device 421 in which polarizers 4212 each having 4 pixels as a unit are stacked, responds to the incidence of the light L1 emitted from the light irradiation unit 41. By receiving the light L2 reflected by the electronic component P, the polarization captured images G1 in the polarization directions of 0 degrees, 45 degrees, 90 degrees, and 135 degrees are obtained. By configuring the polarized image output unit 42 with a polarized camera, four polarized captured images G1 having different polarization directions can be acquired by one imaging from the lower side of the electronic component P held by the holding nozzle 251. can be done. Thereby, the polarization image output unit 42 can shorten the time required to acquire the plurality of polarization captured images G1.

Further, as shown in FIG. 4, the polarization image output unit 42 generates a plurality of polarization synthesized images G2 by synthesizing a plurality of polarization captured images G1 with different polarization directions while changing the weighting. In this case, the polarization image output unit 42 generates a plurality of polarization composite images G2 in different polarization directions subdivided according to the weighting. The polarization image output unit 42 outputs a plurality of segmented polarization images GA in different polarization directions by outputting the plurality of polarization composite images G2 as the polarization images GA. Note that, similarly to the polarization image G1, the polarization composite image G2 includes a component body image area AR1 corresponding to the component body P1 and an electrode image area AR2 corresponding to the electrode P2.

The polarization image output unit 42 also generates a polarization angle image G3 expressed in gray scale for each angle of the polarization direction based on the plurality of polarization captured images G1, and outputs the polarization angle image G3 as a polarization image GA. Note that the polarization angle image G3 includes a component body image area AR1 corresponding to the component body P1 and an electrode image area AR2 corresponding to the electrode P2, similarly to the polarization captured image G1.

Furthermore, the polarized image output unit 42 calculates the degree of linear polarization (DoLP) for each pixel based on the plurality of polarized captured images G1, and extracts from the pixel group having the luminance value according to the degree of linear polarization. A linear polarization degree image G4 is generated. In this case, the polarization image output unit 42 outputs the linear polarization degree image G4 as the polarization image GA. Note that the linear polarization degree image G4 includes a component body image area AR1 corresponding to the component body P1 and an electrode image area AR2 corresponding to the electrode P2, similarly to the polarization image G1.

As described above, the polarization image output unit 42 outputs the plurality of captured polarization images G1, the plurality of polarization composite images G2, the polarization angle image G3, and the linear polarization degree image G4 as the polarization image GA.

Further, the polarization image output unit 42 may be configured to generate a non-polarization image GB by synthesizing a plurality of polarization captured images G1 with the same weighting, and to output the non-polarization image GB. The non-polarized image GB includes a component body image area AR1 corresponding to the component body P1 and an electrode image area AR2 corresponding to the electrode P2, similarly to the polarized captured image G1. Note that the non-polarized image GB may be acquired by an ordinary imaging camera having an imaging device such as a CMOS or CCD, instead of being generated by the polarized image output unit 42 that is a polarization camera.

As shown in FIG. 5 , the plurality of polarized images GA and non-polarized images GB output by the polarized image output section 42 are input to the image specifying section 43 . The image specifying unit 43 calculates the contrast values C of the plurality of polarized images GA and the non-polarized images GB, respectively, and specifies the image of interest GC having the highest contrast value C. FIG. Specifically, the image specifying unit 43 generates a brightness histogram HG indicating the number of pixels for each brightness value for each of the plurality of polarized images GA and non-polarized images GB. Then, the image specifying unit 43 calculates contrast values C of the plurality of polarized images GA and non-polarized images GB according to the following formula (1) based on each luminance histogram HG.
Contrast value C=(max−min)/(max+min) (1)

In the above formula (1), regarding the luminance value of each pixel constituting each of the polarized image GA and the non-polarized image GB, the highest luminance value is indicated as "max" and the lowest luminance value is indicated as "min". .

The image specifying unit 43 can accurately calculate the contrast value C of each of the plurality of polarized images GA and non-polarized images GB based on the brightness histogram HG. Accordingly, the image specifying unit 43 can more accurately specify the image of interest GC having the highest contrast value C from among the plurality of polarized images GA and non-polarized images GB output by the polarized image output unit 42. .

The recognition condition setting unit 44 sets a recognition condition DS for image recognition of the electronic component P based on the image of interest GC specified by the image specifying unit 43 . When the image specifying unit 43 specifies the image of interest GC from among the plurality of captured polarization images G1, the plurality of combined polarization images G2, and the polarization angle image G3, the recognition condition setting unit 44 selects the polarization direction corresponding to the image of interest GC. is set as the recognition condition DS. When the image specifying unit 43 specifies the linear polarization degree image G4 as the image of interest GC, the recognition condition setting unit 44 sets the recognition condition DS that the polarization image output unit 42 generates and outputs the linear polarization degree image G4. do. When the image specifying unit 43 specifies the non-polarized image GB as the image of interest GC, the recognition condition setting unit 44 sets the recognition condition DS that the polarized image output unit 42 generates and outputs the non-polarized image GB.

The recognition condition DS set by the recognition condition setting unit 44 is stored in the management data storage unit 51 in association with the management data D1, and is stored in the accumulation storage unit 52 in association with the part data D2. When the image recognition device 4 recognizes the substrate PP as an object, the recognition condition DS is stored in the accumulation storage unit 52 in association with the substrate data D3.

The image recognizing unit 45 causes the polarized image output unit 42 to output the polarized image GA or the non-polarized image GB according to the recognition condition DS automatically set by the recognition condition setting unit 44 as the recognition image GD. When the recognition condition DS indicates the polarization direction, the image recognition unit 45 outputs the polarization image GA of the polarization direction indicated by the recognition condition DS from the polarization image output unit 42 as the recognition image GD. When the recognition condition DS indicates that the linear polarization degree image G4 is to be generated and output, the image recognition section 45 causes the polarization image output section 42 to output the linear polarization degree image G4 as the recognition image GD. When the recognition condition DS indicates that the non-polarized image GB is to be output, the image recognition section 45 causes the polarized image output section 42 to output the non-polarized image GB as the recognition image GD.

The recognition image GD is a polarized image GA or a non-polarized image GB according to the recognition condition DS, and the contrast value C of the plurality of polarized images GA and non-polarized images GB that can be output from the polarized image output unit 42 is an image corresponding to the image of interest GC with the highest . Therefore, the recognition image GD is an image of the electronic component P with good contrast. Therefore, the image recognition unit 45 can accurately perform image recognition of the electronic component P based on the recognition image GD according to the recognition condition DS.

The image recognition unit 45 can refer to the recognition condition DS associated with the management data D1 stored in the management data storage unit 51 when performing image recognition of the electronic component P based on the recognition image GD. In this case, the image recognition unit 45 causes the polarized image output unit 42 to output the polarized image GA or the non-polarized image GB according to the recognition condition DS associated with the management data D1 as the recognition image GD.

Further, since the recognition condition DS is associated with the part data D2 and stored in the accumulation storage unit 52, the image recognition unit 45 uses the polarized image GA or the non-polarized image GB according to the recognition condition DS as the recognition image GD. The recognition condition DS can be read from the accumulation storage unit 52 and used each time the polarization image output unit 42 outputs the image.

The image recognition unit 45 outputs the polarized image GA or the non-polarized image GB according to the recognition condition DS set by the recognition condition setting unit 44 as the recognition image GD for image recognition of the electronic component P to the polarized image output unit 42. output from Then, the image recognition unit 45 recognizes the electronic component P held by the holding nozzle 251 based on the component data D2 stored in the accumulation storage unit 52 and the recognition image GD output from the polarization image output unit 42. image recognition. Specifically, the image recognition unit 45 recognizes the position of the electronic component P with respect to the holding nozzle 251 as the holding state of the electronic component P with respect to the holding nozzle 251 based on the recognition image GD while referring to the component data D2. At the same time, the posture of the electronic component P held by the holding nozzle 251 is recognized.

Recognition result information DSS indicating the recognition result by the image recognition unit 45 is input to the head control unit 33 . The head control unit 33 controls the component mounting operation by the head unit 25 based on the recognition result information DSS. The component mounting operation of the head unit 25 controlled based on the recognition result information DSS by the image recognition unit 45 improves the mounting accuracy of the electronic component P on the substrate PP.

When accurately recognizing the image of the electronic component P, based on the recognition image GD in which the contrast between the electrode image area AR2 corresponding to the characteristic electrode P2 of the electronic component P and the surrounding component body image area AR1 is high. Image recognition of the electronic component P is desirable.

For this reason, as shown in FIG. 5, the image specifying unit 43 designates a predetermined range including the electrode image area AR2 for each of the plurality of polarized images GA and non-polarized images GB as a target for calculating the contrast value C. Set as area AR3. In this case, the image specifying unit 43 determines that, among the plurality of polarized images GA and non-polarized images GB output by the polarized image output unit 42, the contrast value C within the predetermined range area AR3 including the electrode image area AR2 is the highest. A polarized image GA or a non-polarized image GB is specified as an image of interest GC. As a result, the recognition image GD, which is the polarized image GA or the non-polarized image GB according to the recognition condition DS corresponding to the image of interest GC, has a contrast between the electrode image area AR2 and the surrounding component body image area AR1. High image. Therefore, the image recognition section 45 can accurately perform image recognition of the electronic component P as a subject based on the recognition image GD.

When the image recognition device 4 recognizes the image of the substrate PP as an object, the image specifying unit 43 sets the area of a predetermined range including the mark image area corresponding to the mark M attached to the substrate PP to the contrast value C is set as the target area for calculation. In this case, the image specifying unit 43 specifies, as the image of interest GC, the polarized image GA or the non-polarized image GB having the highest contrast value C within a predetermined range including the mark image area. As a result, the recognition image GD, which is the polarized image GA or the non-polarized image GB according to the recognition condition DS corresponding to the target image GC, is an image with high contrast between the mark image area and its surrounding areas. Therefore, the image recognition unit 45 can accurately perform image recognition of the substrate PP as a subject based on the recognition image GD.

Although the image recognition device 4 and the mounter 1 having the same according to the embodiments of the present invention have been described above, the present invention is not limited to this, and the following modified embodiments may be employed. be able to.

In the above-described embodiment, the polarization image output unit 42 of the image recognition device 4 is configured by a polarization camera equipped with the polarizer stack imaging element 421, but the configuration is not limited to this.

In a modified embodiment, the polarized image output section 42 may be a camera having the structure shown in FIG. In the example shown in FIG. 6, the polarized image output unit 42 includes a polarizing filter 423 rotatable around an axis 4231 parallel to the optical axis and an imaging element 422 such as a CMOS or CCD with a predetermined distance. It consists of cameras arranged side by side on an axis. In this case, the polarizing filter 423 is divided into a plurality of regions (for example, four regions) that pass light with different polarization directions. For example, the polarizing filter 423 is divided into four regions that pass light with polarization directions of 0 degrees, 45 degrees, 90 degrees, and 135 degrees. In the polarized image output unit 42 having such a configuration, the polarizing filter 423 is rotated by a rotation angle of 45 degrees each time the imaging device 422 takes an image. In this case, it is possible to acquire four polarization captured images G1 with different polarization directions by capturing images four times with the image sensor 422 . As a result, the polarization image output unit 42 takes a longer time to acquire the plurality of polarization captured images G1 than when the polarization camera is configured as described above, but the polarized image output unit 42 can obtain the polarized captured images G1 with high resolution. can be obtained.

1 component mounting machine 2 mounting machine main body 25 head unit 3 control device 33 head control section 4 image recognition device 41 light irradiation section 42 polarized image output section 43 image specifying section 44 recognition condition setting section 45 image recognition section 5 storage device 51 management data Storage unit 52 Accumulation storage unit

Claims (9)

An image recognition device for image recognition of a subject having a plurality of areas with different light reflection characteristics,
a polarized image output unit that acquires a plurality of polarized captured images having different polarization directions by receiving light reflected by the subject and outputs a plurality of polarized images including the plurality of polarized captured images;
an image identifying unit that calculates contrast values of the plurality of polarization images and identifies an image of interest with the highest contrast value;
a recognition condition setting unit that sets a polarization direction corresponding to the image of interest as a recognition condition for image recognition of the subject;
an image recognition unit that outputs a polarized image according to the recognition condition from the polarized image output unit as a recognition image, and recognizes the subject based on the recognition image. 2. The polarization image output unit according to claim 1, wherein the polarization image output unit generates a plurality of polarization composite images by synthesizing the plurality of polarization captured images while changing weights, and outputs the plurality of polarization composite images as the polarization images. The described image recognition device. 2. The polarization image output unit generates a polarization angle image expressed in gray scale for each polarization direction angle based on the plurality of polarization captured images, and outputs the polarization angle image as the polarization image. 3. The image recognition device according to 2. The polarization image output unit calculates the degree of linear polarization for each pixel based on the plurality of polarization captured images, generates a linear polarization degree image composed of a pixel group having a luminance value according to the linear polarization degree, and output as a polarized image,
The recognition condition setting unit sets, as the recognition condition, that when the image specifying unit specifies the linear polarization degree image as the image of interest, the polarization image output unit generates and outputs the linear polarization degree image. The image recognition device according to any one of claims 1 to 3, wherein wherein the image specifying unit generates a brightness histogram indicating the number of pixels for each brightness value for each of the plurality of polarized images, and calculates contrast values of the plurality of polarized images based on the brightness histogram. 5. The image recognition device according to any one of Items 1 to 4. a substrate transport unit that transports a substrate with a predetermined mark to a predetermined position;
a head unit that holds an electronic component having electrodes and performs a component mounting operation of mounting the held electronic component on the substrate at the predetermined position;
6. The image recognition apparatus according to any one of claims 1 to 5, wherein the electronic component held by the head unit or the board on which the electronic component is mounted is image-recognized as the subject. an image recognition device;
and a head control section that controls the component mounting operation of the head unit based on a recognition result by the image recognition section of the image recognition device. management data referred to when the head control unit controls the head unit, the target holding position information indicating a target holding position when the head unit holds the electronic component; a management data storage unit for storing management data including target mounting position information indicating a target mounting position of the electronic component,
7. The component mounter according to claim 6, wherein said management data storage section stores said management data in association with said recognition condition set by said recognition condition setting section. The recognition conditions set by the recognition condition setting section are stored by accumulating and storing component data including parameter data relating to the characteristics of the electronic component and board data including data relating to the mark attached to the board. 8. The component mounter according to claim 6, wherein said component data or said board data is stored in a predetermined accumulation storage section in association with said component data or said board data. For each of the plurality of polarized images, the image specifying unit selects an area of a predetermined range including an electrode image area corresponding to the electrode on the electronic component or a mark image area corresponding to the mark on the substrate. 9. The component mounter according to any one of claims 6 to 8, which is set as a target area for calculating a contrast value.

Images