JP3419870B2 - Calibration method of optical system for recognition of visual recognition device and visual recognition device using the method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for calibrating a recognition optical system of a visual recognition device and a visual recognition device using the method, and more particularly to an electronic component mounting device for mounting electronic components on a substrate or a substrate. The present invention relates to a method of calibrating a recognition optical system of a visual recognition device suitable for use in a visual recognition device of a cream solder printing apparatus that prints cream solder, and a visual recognition device using the method.

[0002]

2. Description of the Related Art With the recent miniaturization and high functionality of electronic equipment, the number of pins and the number of pins of electronic parts used have been remarkably increased. Therefore, the electronic component device that mounts those electronic components is required to have a mounting capability with higher accuracy. Therefore, the positioning technology for the substrate and the electronic components can perform position correction by visual recognition from the conventional mechanical positioning. It has become common to go and mount electronic components on a substrate.

A conventional electronic component mounting apparatus displays an image of an identification mark provided on a substrate and an electronic component to be mounted by C
A vision sensor such as a CD camera captures it in a vision controller, detects the deviation of the recognition mark and electronic parts in the image, corrects the position of the board / electronic parts according to the detected deviation, and positions it on the board accurately. Electronic components are mounted.

Since the recognition mark and the recognition of the electronic component when mounting the electronic component on the board are performed in the coordinate space provided on the memory of the vision controller, the amount of image shift is also detected in the coordinate space. Is done by the smallest decomposition unit of. However, the actual size of the minimum resolution unit varies depending on the configuration and adjustment of the recognition optical system, and thus the position correction cannot be performed without calibrating the recognition optical system. It was

Therefore, an electronic component mounting apparatus which performs position correction by visual recognition has a conversion constant for converting the minimum resolution unit in the coordinate space in the vision controller of the apparatus into an actual distance, and the visual recognition The detected value in the coordinate space is converted into an actual distance by this constant to correct the position. Generally, obtaining the conversion constant is called calibration, and the conversion constant is called a calibration value.

In an electronic component mounting apparatus that performs position correction by visual recognition, if the calibration value is inappropriate, the correction value obtained will not be accurate no matter how accurate visual confirmation is performed. How to do was an important point.

In the conventional calibration method, a substrate mark or jig whose dimensions are known in advance is used, and the mark or jig is recognized to obtain the calibration value. The conventional calibration method of the recognition optical system will be described below.

As shown in FIG. 8, there is a circular recognition mark 2 having a diameter A on the substrate 1. When the recognition mark 2 is used for calibration, the image is stored in an image memory in the vision controller. It is assumed that the image captured in the coordinate system in space is as shown in FIG. In the coordinate system in the image memory space at that time, if the diameter of the recognition mark 2 is B, the calibration value C can be obtained by the following formula.

In the method of obtaining the value of B, the brightness is quantized for each minimum resolution unit 7 of the captured image, and the brightness of the image is determined by the value of the quantized level to obtain the size. ing.

Generally, when the image of the diameter portion of the recognition mark 2 shown in FIG. 9 is quantized and its distribution is viewed, a graph as shown in FIG. 10 is obtained. Here, the light-dark boundary level is set to level D, and the coordinates of the intersection of the line of level D and the graph are X1 and X2.
Then, the diameter B of the mark can be obtained by the following formula. B = (X2-X1) (Equation 2)

[0011]

In FIG. 10, when the boundary level is changed from D to D1 or D2, the diameter B of the mark is changed.
Also becomes B1 or B2, and the level of the boundary determines the accuracy of the calibration, but there is no criterion for determining the optimum boundary level, and it relied on the sensory sense of the operator. Setting the optimum boundary level was very difficult.

Further, in order to improve the accuracy of the calibration value C, not only the accuracy of the diameter B of the mark needs to be improved by one set, but also the accuracy of the value of the diameter A of the actual circular mark needs to be improved. Accuracy is required for the dimensions of itself.

As another method of calibrating the recognition optical system, the vision sensor is manually moved between two points on the jig whose dimensions are known in advance, and the actual movement distance of the vision sensor and the vision controller There is a method of obtaining the calibration value from the moving distance in the memory space.

This method also requires the dimensional accuracy of the jig and the movement of the vision sensor is manually performed, so that the movement distance of the vision sensor is likely to include an error and it is difficult to perform accurate calibration. Met.

From the above, even if an attempt is made to obtain a more accurate correction value at the time of calibration that affects the accuracy when the position of the board and the electronic component is corrected by visual recognition, there are various restrictions as described above. There is a problem that it is difficult to obtain an accurate value.

Further, in the cream solder printing apparatus,
The recognition mark on the board and the recognition mark on the screen mask are loaded into the vision controller by a vision sensor such as a CCD camera, the deviation of the recognition mark in the image is detected, and the position of the board / screen mask is corrected according to the detected deviation. Then, the cream solder is printed at the correct position on the board.

Even in the visual recognition system including the CCD camera of this cream solder printing apparatus, it is necessary to calibrate the recognition optical system, which causes the above-mentioned problems.

An object of the present invention is to provide a method of calibrating a recognition optical system of a visual recognition device capable of accurately obtaining a calibration value, and a visual recognition device using the method.

[0019]

According to a first aspect of the present invention, there is provided a method of calibrating a recognition optical system of a visual recognition device for detecting a position of a recognition target by using the recognition optical system. And the recognition optical system are moved relative to each other in a predetermined direction, and the distance at which the recognition target object has moved within the field of view of the recognition optical system and the actual relative movement distance between the recognition target object and the recognition optical system. The calibration value of the recognition optical system for the predetermined direction is calculated from the above. According to a second aspect of the present invention, in the calibration method of the recognition optical system of the visual recognition device that detects the position of the recognition target using the recognition optical system, the recognition target and the recognition optical system are in the X direction. The optical system for recognition in the X direction based on the distance moved by the object to be recognized within the field of view of the optical system for recognition at this time and the actual relative moving distance between the object to be recognized and the optical system for recognition. Of the calibration object, the recognition object and the recognition optical system are moved in the Y direction, and the distance the recognition object has moved within the field of view of the recognition optical system at this time and the recognition object and the recognition optical system. The calibration value of the recognition optical system in the Y direction is calculated from the actual relative movement distance of the system.

According to a third aspect of the present invention, there is provided a method of calibrating a recognition optical system of a visual recognition device for detecting the position of a recognition target using the recognition target optical system, wherein the recognition target and the recognition optical system are combined. Is moved relative to the X-Y direction in an oblique direction, and the distance that the recognition target object has moved within the visual field of the recognition optical system at this time and the actual relative movement distance between the recognition target object and the recognition optical system. The calibration value of the recognition optical system for the X direction and the Y direction is calculated from the above.

According to a fourth aspect of the present invention, in the visual recognition device for detecting the position of the recognition target object using the recognition optical system, the recognition target object and the recognition optical system are relatively moved in a predetermined direction. The driving mechanism for driving the recognition optical system and the distance for moving the recognition target within the visual field of the recognition optical system due to this relative movement and the actual relative moving distance between the recognition target and the recognition optical system for recognition optical in the predetermined direction. And a control device for calculating a calibration value of the system.

[0022]

According to the present invention, when the recognition target and the recognition optical system are moved relative to each other in a predetermined direction, the movement distance of the recognition target in the visual field of the recognition optical system and the recognition of the recognition target are recognized. A calibration value of the recognition optical system in a predetermined direction is calculated from the actual relative movement distance of the usage optical system.

[0023]

[0024]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration of an electronic component mounting apparatus according to a first embodiment. In FIG. 1, reference numeral 4 denotes a CCD camera, which is an identification mark 2 printed on a substrate 1 and an electronic device mounted on the substrate 1. It is a vision sensor for capturing the image of a part and detecting the recognition mark and the center of gravity of the electronic part. 5 is a vision controller, CCD camera 4
The image processing apparatus is capable of processing the recognition mark and the image data of the electronic component taken in to calculate the position of the center of gravity of the recognition mark 2 and the electronic component. Reference numeral 6 denotes a control device, which controls the vision controller 5 as a whole. 8 is X
A Y robot, which is a recognition optical system driving mechanism that moves the CCD camera 4 in the X and Y directions according to a command from the control device 6. Reference numeral 10 denotes an XY robot, which moves the board setting jig 9 for setting the board 1 according to a command from the control device 6 in X and Y directions.
It is a substrate setting jig drive mechanism that can be moved in any direction. In addition, the control device 6 uses the XY robots 8 and 10.
And the vision controller 5 are collectively controlled.

The operation of this embodiment will be described with reference to FIG. Since the position of the recognition mark provided on the substrate 1 is taught to the control device 6 in advance, the CC above the recognition mark 2 provided on the substrate 1 is instructed by the control device 6.
The D camera 4 moves, captures the image of the recognition mark 2 into the vision controller 5, and detects the center of gravity of the recognition mark 2. After that, the recognition mark 2 from the field of view of the CCD camera 4
The CCD camera 4 is moved in the X or Y direction by a certain distance by the XY robot 8 to the extent that it is not in the range, and the centroid of the recognition mark 2 is detected again.

As a result, the calibration value is calculated from the distance in the image memory space of the vision controller 5 between the two centers of gravity of the recognition mark 2 obtained and the actual movement distance of the XY robot 8. You can

Specifically, as shown in FIG. 2B, the center of gravity of the recognition mark 2 moves in the memory space of the vision controller 5 when the CCD camera 4 moves by dX in the X direction and dY in the Y direction. If the distances are A and B, respectively, the calibration values in the X and Y directions can be obtained by the following equation.

According to this embodiment, there is a margin in the outer dimension accuracy of the recognition mark 2 and the jig on the substrate 1 and the setting width of the boundary level of light and dark at the time of recognition, and moreover, there is a human-like sensory adjustment element. Since it is eliminated, it is possible to obtain an excellent effect that a stable calibration value can be obtained more easily and accurately than in the past.

In this embodiment, the CCD camera 4 is moved by the XY robot 8 in order to move the recognition mark 2 in the field of view of the CCD camera 4, but it is shown in FIG. As described above, the XY robot 10 on the substrate side moves the substrate 1 on the substrate setting jig 9 in the X and Y directions to move the recognition mark 2 in the field of view of the CCD camera 4. , A similar effect can be obtained.

Specifically, as shown in FIG. 3B, the substrate 1
Let C and D be the distances of movement of the center of gravity of the mark in the memory space of the vision controller when C moves by dX1 in the X direction and dY1 in the Y direction, the calibration values in the X and Y directions are Can be found at.

In addition, the XY robots 8 and 10 use CCDs.
In both cases of moving the camera 4 or the substrate 1, a method of individually calibrating in each of the X and Y directions to obtain a calibration value, and a method of simultaneously calibrating in both the X and Y directions and moving in two directions at a time. A method of obtaining a calibration value can be considered, but it goes without saying that the above effect can be obtained in any case.

Next, a second embodiment of the present invention will be described with reference to FIG. In FIG. 4A, 3 is QFP (Quad).
Flat Package) IC. QFPIC
Since the lead of 3 is made with the same accuracy as jigs,
The center of gravity of the two adjacent leads is detected, the distance (lead pitch) between the detected centers of gravity is found in the image memory space of the vision controller 5, and the value of the lead pitch and the actual QFPIC3 taught in advance. Calculate the calibration value from the lead pitch value of. As a result, there is a margin in the setting range of the light and dark boundary level at the time of recognition, and since there is no human sensory adjustment element,
It is possible to obtain an excellent effect that a stable calibration value can be obtained more easily and more accurately than before.

Specifically, as shown in FIG. 4B, the CCD camera 4 captures an image of the QFPIC 3 having the lead pitch P, and the captured image of the QFPIC 3 is stored in the memory space of the vision controller 5. Then, the lead pitches E and F of the QFPIC 3 in the X and Y directions in the image memory space of the vision controller 5 are calculated, and based on the lead pitches E and F, the calibration values in the X and Y directions are calculated by the following equation. Can be obtained by

In the present embodiment, the center of gravity of the two leads of the QFPIC 3 is detected, and the lead pitch in the image memory space of the vision controller 5 is used.
The center of gravity of a plurality of leads C3 may be detected, and the average of the distances between the centers of gravity may be used as the lead pitch in the image memory space of the vision controller 5.

Next, the configuration of the cream solder printing apparatus according to the third embodiment of the present invention is shown in FIG. CCD camera 1 for capturing an image of the screen mask 23 and its recognition mark 24 and the cream solder 13 printed on the substrate 11.
4, a vision controller 15 for processing the captured image data, an XY robot 18 for moving the CCD camera 14 in the X and Y directions, and a controller 16 for controlling the XY robot 18 and the vision controller 15. It consists of

The operation of this embodiment will be described with reference to FIG. First, the CCD camera 14 moves onto the previously recognized recognition mark 24, the image of the recognition mark 24 is taken into the vision controller 15, and the center of gravity of the recognition mark 24 is detected. After that, the CCD camera 14 is moved a certain distance in the X or Y direction so that the recognition mark 24 does not protrude from the field of view of the CCD camera 14, and the center of gravity of the recognition mark 24 is detected again.

The resulting distance between the two centers of gravity in the memory space of the vision controller 15 and the actual X
The calibration value can be calculated from the movement distance of the Y robot 18.

Specifically, as shown in FIG. 6, the CCD camera 1
Assuming that the moving distances of the centers of gravity of the recognition marks 24 in the memory space of the vision controller 15 when 4 is moved by dX in the X direction and dY in the Y direction are A and B, respectively, the calibration values in the X and Y directions are , Equation 3 and Equation 4 can be used.

According to this embodiment, the screen mask 23
Since there is a margin in the outer dimension accuracy of the recognition mark 24 and the jig and the setting width of the boundary level of the light and dark at the time of recognition, and since there is no human sensational adjustment element, it is easier and more accurate than before. There is an excellent effect that the calibration value can be obtained. Therefore, it is possible to obtain an excellent effect that high-precision printing can be performed.

In the above third embodiment, the CCD camera 14 is moved by the XY robot 18 in order to move the recognition mark 24 within the field of view of the CCD camera 14, but the configuration shown in FIG. Like the XY robot 2
Even if the recognition mark 24 is moved by moving the screen mask 23 side in the X-Y direction by 5, it is possible to obtain the same effect.

Specifically, as shown in FIG. 7, the moving distance of the center of gravity of the recognition mark 24 in the memory space of the vision controller 15 when the screen mask 23 moves by dX1 in the X direction and dY1 in the Y direction, respectively. If C and D are used, the values can be obtained by the equations 5 and 6.

Further, by the XY robots 18 and 25,
In any case in which the CCD camera 14 or the screen mask 23 is moved, a method of individually calibrating each in X and Y directions to obtain calibration values respectively, and a method of simultaneously moving in both X and Y directions in two directions at a time. A method of obtaining a calibration value can be considered, but it goes without saying that the above effects can be obtained in any case.

[0043]

According to the present invention, the calibration value can be accurately obtained.

[0044]

[Brief description of drawings]

FIG. 1 is a configuration diagram of an electronic component mounting apparatus according to a first embodiment of the present invention.

FIG. 2 is an operation explanatory view of the electronic component mounting apparatus according to the first embodiment of the present invention.

FIG. 3 is an operation explanatory view of the electronic component mounting apparatus according to the first embodiment of the present invention.

FIG. 4 is an operation explanatory view of the electronic component mounting apparatus according to the second embodiment of the present invention.

FIG. 5 is a configuration diagram of a cream solder printing apparatus according to a third embodiment of the present invention.

FIG. 6 is an operation explanatory view of the cream solder printing apparatus according to the third embodiment of the present invention.

FIG. 7 is an operation explanatory view of the cream solder printing apparatus according to the third embodiment of the present invention.

FIG. 8 is a diagram showing circular marks on a substrate.

FIG. 9 is a diagram showing a coordinate system in an image memory space when an image of a circular mark is captured in the vision controller.

FIG. 10 is a diagram showing the relationship between the quantization level and the coordinates in the image memory space when the diameter portion of the circular mark is quantized.

[Explanation of symbols]

1 ... Substrate, 2 ... Recognition mark,
3 ... QFPIC, 4 ... CCD camera, 5 ... Vision controller, 6 ... Control device, 7 ... Minimum resolution unit of captured image, 8 ... XY robot, 9 ... Substrate setting jig, 10 ... XY robot. 11 ... substrate,
12 ... Recognition mark, 13 ... Cream solder, 14 ... CCD camera, 15 ... Vision controller, 16 ... Control device, 18 ... XY robot, 19
... substrate setting jig, 20 ... XY robot 23 ... screen mask, 24 ... recognition mark, 2
5 ... XY robot.

─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A 1-219501 (JP, A) JP-A 4-15883 (JP, A) JP-A 6-249615 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) G01B 11/00 G06T 1/00 305 G01T 1/00 400 G06T 1/00 420

Claims (4)

(57) [Claims] 1. A method of calibrating a recognition optical system of a visual recognition device for detecting the position of a recognition target by using the recognition optical system, wherein the recognition target and the recognition optical system are relatively moved in a predetermined direction. Then, the calibration of the recognition optical system with respect to the predetermined direction is performed based on the distance traveled by the recognition object in the visual field of the recognition optical system and the actual relative movement distance between the recognition object and the recognition optical system. A method for calibrating a recognition optical system of a visual recognition device, which comprises calculating a value. 2. A calibration method for a recognition optical system of a visual recognition device for detecting the position of a recognition target using the recognition target optical system, wherein the recognition target and the recognition optical system are relatively moved in the X direction. ,
The calibration value of the recognition optical system in the X direction is calculated from the distance traveled by the recognition object in the field of view of the recognition optical system and the actual relative movement distance between the recognition object and the recognition optical system. Calculate and move the recognition target and the recognition optical system in the Y direction, and the distance that the recognition target moves within the field of view of the recognition optical system at this time and the actual relative distance between the recognition target and the recognition optical system. A method of calibrating a recognition optical system of a visual recognition device, characterized in that a calibration value of the recognition optical system in the Y direction is calculated from a specific moving distance. 3. A calibration method for a recognition optical system of a visual recognition device for detecting the position of a recognition target object using the recognition target optical system, wherein the recognition target object and the recognition optical system are arranged in the XY direction. And relative movement in an oblique direction, and based on the distance that the recognition target object has moved within the field of view of the recognition optical system and the actual relative movement distance between the recognition target object and the recognition optical system, the X direction and the Y direction.
A method for calibrating a recognition optical system of a visual recognition device, which comprises calculating a calibration value of a recognition optical system with respect to a direction. 4. A visual recognition apparatus for detecting the position of a recognition target object using a recognition optical system, comprising: a drive mechanism for relatively moving the recognition target object and the recognition optical system in a predetermined direction; The calibration value of the recognition optical system in the predetermined direction is calculated from the distance moved by the movement of the recognition object within the field of view of the recognition optical system and the actual relative movement distance between the recognition object and the recognition optical system. A visual recognition device, comprising: