JP4306353B2 - Pixel rate calculation method in image recognition apparatus - Google Patents

Description

  The present invention relates to a pixel rate calculation method in an image recognition apparatus that calculates a pixel rate that specifies correspondence between unit pixel size and dimensional information in an acquired image in an image recognition apparatus used for an electronic component mounting apparatus or the like. is there.

Electronic component mounting devices such as electronic component mounting devices that mount electronic components on a board are equipped with an image recognition device that recognizes the position of the board and electronic components, and mounting operations for mounting electronic components on the mounting points of the board, etc. In this work process, alignment is performed based on the position recognition result. When obtaining dimensional information such as a displacement amount in this position recognition, it is necessary to set in advance a pixel rate indicating the relationship between the unit pixel size on the image and the actual dimension. Conventionally, as this pixel rate setting method, a method of obtaining a pixel rate by causing the apparatus to recognize a reference mark or a target part whose dimensions are known in advance is known (see, for example, Patent Document 1).
Japanese Patent Laid-Open No. 11-17395

  By the way, the size of a recognition target such as an electronic component in an electronic component mounting device is various in size. In image recognition targeting a wide size range with the same device, a camera equipped with a zoom mechanism is used to display an image of the target. It is desirable to read in a size suitable for recognition processing. For this reason, when reading an image, an operation of appropriately setting the optical magnification of the camera according to the object is performed.

  However, since the change in the optical magnification is accompanied by the change in the pixel rate described above, when changing the optical magnification from the past, a new pixel rate is used by using a reference mark or target component of a size corresponding to the object. We needed work to calculate. For this reason, it is necessary to prepare a plurality of fiducial marks and target parts at the production site, and it is necessary to perform complicated management depending on the object. Calculation of pixel rates applicable to multiple types of optical magnification can be performed with a simple method. A method was sought.

  Therefore, an object of the present invention is to provide a pixel rate calculation method in an image recognition apparatus that is a simple method and applicable to a plurality of types of optical magnifications.

In order to solve the above problems, a pixel rate calculation method in an image recognition apparatus according to the present invention recognizes an acquired image obtained by imaging a recognition target with a camera having a zoom mechanism that changes an optical magnification. An image recognition unit that outputs predetermined information including dimension information of a recognition target based on a recognition processing result of the acquired image, and a correspondence relationship between a unit pixel size in the acquired image and the dimension information A pixel rate calculation method in an image recognition apparatus for calculating a pixel rate indicating a square frame-shaped calibration mark having a plurality of diagonal points as feature points and known intervals between the feature points . and the jig mounting step of mounting a plurality of the calibration jig which is built more in-similarity shape as the same to the center corresponds to a particular value in the same imaging position and the recognition target, before An image selection acquisition step of capturing a calibration jig with the specified optical magnification by the camera and selectively acquiring an image of a calibration mark corresponding to the optical magnification among the plurality of calibration marks as a desired calibration mark image; A recognition processing step for recognizing the desired calibration mark image by the image processing unit, and the desired number of pixels based on the number of pixels between feature points obtained from the recognition processing result of the recognition processing step and the distance between the feature points. including a pixel rate calculation step of calculating the pixel rate in the calibration mark image.

  According to the present invention, using a calibration jig in which a plurality of calibration marks configured to include a plurality of feature points whose mutual intervals are known are formed corresponding to a plurality of specific values of optical magnification, The jig is imaged at an optical magnification specified by the camera, and a calibration mark image corresponding to the optical magnification is selectively acquired as a desired calibration mark image from among a plurality of calibration marks. Pixels and features in the desired calibration mark image By calculating the pixel rate that specifies the correspondence with the distance between points, the pixel rate can be calculated for a plurality of types of optical magnifications by a simple method.

  Next, embodiments of the present invention will be described with reference to the drawings. 1 is a block diagram showing an overall configuration of an electronic component mounting apparatus according to an embodiment of the present invention, FIG. 2 is a block diagram showing a configuration of a control system of the electronic component mounting apparatus according to an embodiment of the present invention, and FIG. FIG. 4 is a plan view of a calibration mark used in the pixel rate calculation method according to the embodiment of the present invention, and FIG. 4 is an explanatory diagram of data contents stored in the storage unit of the electronic component mounting apparatus according to the embodiment of the present invention. 5 and 6 are explanatory diagrams of a pixel rate calculation method in the image recognition apparatus according to the embodiment of the present invention.

  First, the configuration of an electronic component mounting apparatus in which an image recognition device is incorporated will be described with reference to FIG. In FIG. 1, a tray 2 is placed on the component supply unit 1, and chips 3 are held on the tray 2 in a lattice arrangement. The tray 2 can be moved in the Y direction by a tray Y-axis drive mechanism 4, and an arbitrary chip 3 on the tray 2 is mounted by driving the tray Y-axis drive mechanism 4 together with a head XZΘ drive mechanism 6 described later. It can be moved to the take-out position by the head 7.

  A mounting mechanism 5 including a mounting head 7 is disposed above the component supply unit 1. The mounting head 7 is movable in the X direction, the Z direction, and the θ direction (rotation direction around the nozzle shaft holding the chip 3) by the head XZΘ drive mechanism 6, and the chip 3 is removed from the tray 2 of the component supply unit 1. The pickup 3 is picked up and moved above the substrate holding unit 8, and the chip 3 is mounted on the substrate 9 held by the substrate holding unit 8. The substrate holding unit 8 can be moved in the Y direction by the substrate Y-axis drive mechanism 10, and thereby the alignment in the Y direction during mounting by the mounting head 7 is performed.

  A chip recognition camera 13 is disposed on the moving path of the mounting head 7 by the head XZΘ driving mechanism 6 with the imaging direction facing upward. The chip recognition camera 13 images the chip 3 held by the mounting head 7 from below. Imaging by the chip recognition camera 13 is controlled by a chip camera controller 14.

  A substrate recognition camera 11 is disposed above the substrate holding unit 8 with the imaging direction facing downward, and the substrate recognition camera 11 images the substrate 9 placed on the substrate holding unit 8. Imaging by the board recognition camera 11 is controlled by the board camera controller 12. The substrate recognition camera 11 and the chip recognition camera 13 are provided with a zoom mechanism for changing the optical magnification, and the substrate 9 and the chip 3 are imaged at a desired optical magnification by manually operating the zoom adjustment rings 11a and 13a. Is possible. The zoom mechanism may be automatic.

In the chip mounting operation by the mounting head 7, the mounting head 7 is positioned with respect to the substrate 9 based on the recognition result of the substrate 9 by the substrate recognition camera 11 and the recognition result of the chip 3 by the chip recognition camera 13. Here, the chip 3 to be mounted includes a plurality of types having different sizes, and in the imaging of the chip 3 by the chip recognition camera 13 and the imaging of the mounting position of the substrate 9 by the substrate recognition camera 11, the size of the chip 3 is set. An image is captured at a corresponding optical magnification.

  Next, the configuration of the control system will be described with reference to FIG. In FIG. 2, the image data captured by the board camera controller 12 and the chip camera controller 14 is sent to the image recognition unit 21 of the control unit 20, where the image data is recognized and processed, thereby mounting positions on the board 9. And the position of the chip 3 held by the mounting head 7 is recognized. The recognition result is sent to the calculation unit 22 where predetermined calculation processing is performed to detect the mounting position and the amount of displacement of the chip 3. Then, the drive mechanism control unit 23 controls the head XZΘ drive mechanism 6 and the substrate Y-axis drive mechanism 10 based on the detection result, thereby performing alignment in the mounting operation by the mounting head 7.

  The display unit 15 is a display panel that displays images captured by the substrate recognition camera 11 and the chip recognition camera 13 and displays a guidance screen when data and operation commands are input by the input unit 16. When the control unit 20 executes various processes, various programs and data stored in the storage unit 24 are used.

  In the above configuration, the substrate recognition camera 11, the substrate camera controller 12, the chip recognition camera 13, the chip camera controller 14, the image recognition unit 21, and the calculation unit 22 perform position recognition of the substrate 9 and the chip 3 in the electronic component mounting apparatus. The image recognition device includes a camera having a zoom mechanism that changes the optical magnification, and an image recognition unit that recognizes an acquired image obtained by imaging a recognition target with the camera. Provided, and is configured to output predetermined information including dimension information to be recognized based on the recognition processing result of the acquired image. The configuration may include only one of a substrate recognition system and a chip recognition system.

  Next, a pixel rate calculation method in the image recognition apparatus for calculating the pixel rate in this image recognition apparatus will be described. The pixel rate indicates a correspondence relationship between the unit pixel size in the acquired image and dimensional information such as the positional deviation amount of the substrate 9 or the chip 3. As described above, since the substrate recognition camera 11 and the chip recognition camera 13 can arbitrarily change the optical magnification, dimensions such as the size of the object and the amount of displacement are determined based on the imaging results of the substrate recognition camera 11 and the chip recognition camera 13. When obtaining the information, it is necessary to obtain a pixel rate that is a conversion coefficient between the unit pixel size and the mounting dimension in the acquired image. In the present embodiment, a pixel rate corresponding to each optical magnification is calculated using a calibration jig 17 described below.

  The calibration jig 17 will be described with reference to FIG. The calibration jig 17 has a structure in which a plurality of square frame-like marks M1, M2, M3, M4, and M5 are vapor-deposited on a transparent glass plate, and the similarity ratio of each mark is a chip on the substrate 9. This corresponds to a specific optical magnification (see FIG. 4B) set based on the size of the mounting point and the size of the chip 3. The mark portion may be vapor-deposited and the background transparent, or the mark portion may be transparent and the background vapor-deposited. Of course, you may form a mark by methods other than vapor deposition.

  The diagonal points of each mark (for example, the diagonal points A1, B1, C1, and D1 in the mark M1) are the characteristic points of the mark, and the dimensions (characteristic points) between the diagonal points A1 and B1 and between B1 and C1. Distance) is known. In other words, the calibration jig 17 has a configuration in which a plurality of calibration marks each including a plurality of feature points whose mutual intervals are known are formed in a substantially similar shape corresponding to a plurality of specific values of optical magnification. It has become.

Next, program data stored in the storage unit 24 will be described with reference to FIG. The program storage unit 25 stores a mounting operation program 25a, a substrate / chip recognition program 25b, and a pixel rate calculation program 25c. The mounting operation program 25 a picks up the chip 3 from the component supply unit 1 by the mounting head 7, and the substrate 9 on the substrate holding unit 8.
It is a program for mounting operation to be mounted on.

  The substrate / chip recognition program 25 b is a program for image recognition processing in which the image recognition unit 21 recognizes the substrate 9 and the chip 3. The pixel rate calculation program 25c is a program for calculating the pixel rate by recognizing the acquired image obtained by capturing the calibration jig 17 by the image recognition unit 21 and calculating the recognition result by the calculation unit 22. .

  The data storage unit 26 stores mounting data 26a and calibration mark data 26b. The mounting data 26a is data such as component data such as the type and size of the chip 3, data on the mounting position on the substrate 9, and sequence data indicating the mounting order. The calibration mark data 26b stores the optical magnification and the distance between feature points for each of a plurality of marks formed on the calibration jig 17. That is, as shown in FIG. 4B, specific values of optical magnifications of the substrate recognition camera 11 and the chip recognition camera 13 (in the example shown here, 0.3, 0.5, 1.0, 2.5 and 6). ., Five distances between feature points L1, L2, L3, L4, and L5 are stored for the marks M1, M2, M3, M4, and M5.

  Next, an example of calculating the pixel rate will be described with reference to FIGS. In the electronic component mounting apparatus shown in FIG. 1, when the size of the target chip 3 is different, the optical magnifications of the substrate recognition camera 11 and the chip recognition camera 13 are set by the zoom magnification adjustment rings 11a and 13a. It is necessary to change according to the size of the point or the size of the chip 3. When the optical magnification is changed, it is necessary to newly set the pixel rate each time. By making the camera recognize the dummy substrate / dummy chip in which the calibration jig 17 is mounted on the dummy to be recognized by each camera, the pixel is set. Rate calculation is performed. Here, a pixel rate calculation example of the substrate recognition camera 11 will be described.

  First, the dummy substrate on which the calibration jig 17 is mounted is mounted at the imaging position of the substrate recognition camera 11 on the substrate holding unit 8 (jig mounting step). Next, the calibration jig 17 is imaged by the substrate recognition camera 11 at a designated magnification (here, an example in which an optical magnification of 1 is designated) is shown. Thereby, as shown in FIG. 5, images of a plurality of marks are located in the imaging field 30.

  At this time, among the marks M1 and M2 corresponding to a magnification smaller than the designated magnification 1, the optical image of the mark M1 is larger than the imaging visual field size and therefore does not appear in the imaging visual field 30, and the mark M2 is the first one. Only the part appears in the imaging visual field 30. The marks M4 and M5 corresponding to a magnification larger than the designated magnification appear inside the mark M3.

  Thereafter, after capturing an image in the imaging field of view 30, a masking process is performed in order to selectively acquire an image of only the desired calibration mark M3 corresponding to the designated optical magnification. That is, as shown in FIG. 5, the mask range 31 is set so that only the mark M3 is positioned in the image capture range 32, that is, the marks M4 and M5 are excluded, and an image of only the mark M3 is captured. In this process, the calibration jig 17 is imaged by the substrate recognition camera 11 at a designated optical magnification, and a calibration mark image corresponding to the designated optical magnification among a plurality of calibration marks is selectively selected as a desired calibration mark image. It is an image selection acquisition process to be acquired.

Next, the acquired desired calibration mark image is recognized by the image recognition unit 21 (recognition processing step). As a result, the positions of the diagonal points A3 and B3 of the mark M3 shown in FIG. 6 are recognized, and the pixel number N3 between the feature points is obtained from the recognition result. Based on the number N3 of pixels between feature points and the distance L3 between feature points of the calibration mark data 26b, the calculation unit 22 calculates a pixel rate R3 (pixel rate corresponding to the optical magnification 1) as R3 = L3 / N3. Calculated by (
Pixel rate calculation step). That is, here, the pixel rate in the desired calibration mark image is calculated based on the recognition processing result of the recognition processing step and the interval between the feature points.

  This pixel rate calculation is similarly executed for the chip recognition camera 13. That is, the dummy chip on which the calibration jig 17 is mounted is held on the mounting head 7, the calibration jig 17 is imaged by the chip recognition camera 13, and the pixel rate is obtained by the same method.

  In the above-described image selection and acquisition process, instead of masking the acquired image after the image in the imaging field of view 30 is captured, the image can be captured in advance using a camera such as a CMOS type that can arbitrarily set the image reading range. You may make it acquire the image of only the inclusion range 32. FIG.

  In the chip 3 mounting operation started thereafter, the above-described pixel rate is applied to the substrate / chip recognition processing performed by the image recognition unit 21 and the calculation unit 22. When a chip having a different size is designated as a chip to be subjected to the next operation, processing for newly obtaining a pixel rate corresponding to the optical magnification corresponding to the size is performed.

  In this way, in the case of targeting parts of different sizes by the same electronic component mounting apparatus, the same calibration jig is applied to all the parts by using the pixel rate calculation method shown in the present embodiment. Compared with the conventional method in which it is necessary to use a plurality of target marks separately, complicated management becomes unnecessary.

  Further, in the pixel rate calculation method of the present invention, since it is not necessary to move the dummy substrate or the dummy chip relative to the camera, the substrate 9 is Y with respect to the substrate recognition camera 11 as shown in the present embodiment. Even if the chip 3 is configured to move only in the X direction with respect to the chip recognition camera 13, it can be recognized and processed by imaging the calibration jig having the above configuration. The pixel rate can be calculated easily.

  The pixel rate calculation method in the image recognition apparatus of the present invention has an effect that the pixel rate can be calculated for a plurality of types of optical magnifications by a simple method, and is used for an electronic component mounting apparatus or the like. In the recognition apparatus, it is useful for the purpose of calculating the pixel rate that specifies the correspondence between the unit pixel size and the dimension information in the acquired image.

The block diagram which shows the whole structure of the electronic component mounting apparatus of one embodiment of this invention The block diagram which shows the structure of the control system of the electronic component mounting apparatus of one embodiment of this invention The top view of the calibration mark used in the pixel rate calculation method of one embodiment of this invention Explanatory drawing of the data content memorize | stored in the memory | storage part of the electronic component mounting apparatus of one embodiment of this invention Explanatory drawing of the pixel rate calculation method in the image recognition apparatus of one embodiment of this invention Explanatory drawing of the pixel rate calculation method in the image recognition apparatus of one embodiment of this invention

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Component supply part 7 Mounting head 8 Board | substrate holding part 9 Board | substrate 11 Board | substrate recognition camera 13 Chip | tip recognition camera 17 Calibration jig | tool 21 Image recognition part 22 Calculation part

Claims (1)

A camera having a zoom mechanism for changing the optical magnification; and an image recognition unit for recognizing an acquired image obtained by imaging the recognition target by the camera, based on a recognition processing result of the acquired image. In the image recognition apparatus that outputs predetermined information including dimension information, a pixel rate calculation method in the image recognition apparatus that calculates a pixel rate indicating a correspondence relationship between a unit pixel size in the acquired image and the dimension information,
Mutual spacing of the feature points and feature point multiple diagonal points is the built more in-similarity shape center corresponds as the same to a plurality of specific values of the calibration mark is the optical magnification of the known square Wakugata A jig mounting step for mounting the calibration jig at the same imaging position as the recognition target, and imaging the calibration jig with the specified optical magnification, and the optical of the plurality of calibration marks. An image selection acquisition step of selectively acquiring a calibration mark image corresponding to the magnification as a desired calibration mark image, a recognition processing step of recognizing the desired calibration mark image by the image processing unit, and recognition of the recognition processing step and wherein the early days including the pixel rate calculation step of calculating the pixel rate in the desired calibration mark image on the basis the number of pixels between the obtained feature points obtained from the processing results and on the distance of the feature point each other Pixel rate calculation method in that the image recognition apparatus.

Images