JPH02124404A - Apparatus for detecting position of part - Google Patents

Abstract

PURPOSE:To improve detecting accuracy and simplify the mechanism by employing two pickup devices, whereby the position of a board is detected by the first pickup device and, the position of a part is detected by the second pickup device. CONSTITUTION:When an electronic part 10 is irradiated from below by a reflecting plate 27, a correct silhouette image showing an outline of a lead wire of the part 10 can be picked up. After the position of the part 10 is detected through processing of an input image of a TV camera 30 which is provided for detection of the position of a part, the reflecting plate 27 is retracted to an original position. The deviation in a rotating direction of the part 10 is corrected by rotating a suction head 25, while the deviation in an (x, y) direction is corrected by moving a movable table 20 which moves a printed circuit board. Accordingly, the electronic part is mounted at a predeter mined position on the printed circuit board. A TV camera 34 for detecting the position of the printed circuit board picks up an image of the surface of the printed circuit board radiated by a projector 37, via an objective lens 35 and an eye lens 36. The position of a cross mark on the printed circuit board is detected by the camera 34 the view field of which is moved by the table 20.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a component position detection device, and more particularly to a component position detection device that can automatically mount a component having a plurality of lead wires at a predetermined position on a board with high positional accuracy.

When mounting a component at a predetermined position on a board in a determined posture, if the error in supplying the component to the component holding device that carries the component to the mounting position is sufficiently smaller than the tolerance range of the mounting position, the part The purpose can be achieved by a simple control operation that moves the holding device to the target position according to a predetermined value after supplying the holding device. However, for example, IC,
When assembling an electronic component such as an LSI that has many thin lead wires spaced at extremely narrow intervals, each lead must Since the range of positional deviation allowed between the wire and the wiring terminal is extremely small, it is extremely difficult to suppress the component feeding error from the feeding device to the holding device within a range that allows the simple control described above. . Especially LSIs that have multiple lead wires on each side of the package.
When mounting on a printed circuit board, the permissible range of positional deviation in the longitudinal direction of the lead wire is restricted by the width of the lead wire extending in the direction perpendicular to this, so the variation allowed in the supply of parts to the holding device is limited. The range is extremely strict, and if this is to be achieved, the parts feeding device and holding device will be complicated and expensive. Also, according to this method, the dimensions,
It is difficult to adapt to automatic assembly of many types of parts with different shapes.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a component position detection device that allows components to be mounted at predetermined positions on a board with high positional accuracy.

In order to achieve such an object, the present invention provides a component moving means for moving a component having a plurality of lead wires to a predetermined position on a board on which it is to be mounted, and a component moving means for moving a component having a plurality of lead wires to a predetermined position on a board to be mounted, and a component moving means for moving a component having a plurality of lead wires to a predetermined position on a board to be mounted, and a light beam from above the board to the lead wires of the moved component. an irradiation means for irradiating the light, an imaging means for imaging at least two component areas of the lead wire of the component with transmitted light of the light irradiated by the irradiation means, and a tip of the lead wire of the component. storage means for storing reference position information of at least two specific areas to be detected; and means for moving the imaging area of the imaging means in accordance with the reference position information of the specific areas to be detected read from the storage means; After the movement, the imaging region imaged by the imaging means is divided based on the size or spacing of the lead wires of the component, and one of the lead wires located at the tip of the lead wires existing in the five divided regions is It is characterized by comprising means for detecting the position of the component pattern as seen from the imaging means.

Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings. This embodiment shows an example in which the present invention is applied to an automatic assembly device that sequentially mounts a plurality of electronic components of different sizes at predetermined positions on a printed circuit board. It is something.

FIG. 1 is a diagram showing a schematic configuration of a printed circuit board 1, which is a surface on which electronic components are mounted, and the area surrounded by broken lines shows the area where a fine wiring pattern is formed. In this wiring area, a group of wiring-side terminals connected to each lead terminal is formed in accordance with the arrangement of the lead terminals of the electronic component.

Each rectangle shown at 2°3.4 in Figure 1 is a large IC.
(hereinafter referred to as LSI parts), small ICs (hereinafter referred to as SSI parts),
) and resistance elements (hereinafter referred to as R components). In this example, 18 LSI components and 36 SSI components are mounted on the printed circuit board 1.
parts and 72 R parts are installed. Reference numeral 5 designates through holes for pins for fixing the printed circuit board 1 on a printed circuit board moving table of an automatic assembly device to be described later, and the diagonally shaded portions 68 to 6b at the four corners of the printed circuit board are holes for fixing the printed circuit board 1 on the printed circuit board moving table of an automatic assembly device to be described later. The area in which a cross mark is formed is shown to provide the reference position of the printed circuit board when correcting the positional deviation of the printed circuit board. The positions of these four cross marks and the mounting positions of the three types of electronic components mentioned above are stored in advance in a data processing device that controls the operation sequence of the assembly device, and the data processing device numerically controls the board movement table. This allows each of the above parts to be moved to a predetermined position.

Figures 2, 3, and 4 are LSI parts and SSI parts, respectively.
The shape of the part and R part is shown. L used in this example
The SI component and the SSI component each have a plurality of lead wires 11 along each side of the package 10 containing the integrated circuit, while the R component has a plurality of lead wires 11 each along the two sides of the package 1o. The structure includes lead wires 11. For example, the package size of an LSI component is 21.6 x 21.6 mm, and the lead wire has a width of 0.6 mm.
3 mm, and the interval is 0.762 mm. In addition, the SSI parts have lead wires with a radius of 0.43 mm arranged at 1.27 mm intervals in a 10.4 mm square package, and the R parts have 17.5 mm square leads.
Lead wires are arranged at 1.90 mm intervals in a 4.1 nm square package with Q, 3 mm intervals. These electronic components are held by a vacuum suction head as will be described later, and after positional displacement is corrected by pattern recognition processing using an imaging device, they are mounted at predetermined positions on the printed circuit board.

FIG. 5 shows the state of the lead wire tip of an electronic component mounted on a printed circuit board. 7 is a wiring layer formed on the surface of the printed circuit board 1, 8 is a solder layer, 9 is a paste applied to the surface of the solder layer, ] 2 is a half E11 layer formed on the top surface of the tip of the Riet wire 11 It is. Each component aligned on the printed circuit board is temporarily bonded onto the solder layer 8 of the terminal on the printed circuit board side via the paste 9 by pressing from the top surface of the package as the suction head descends, and all components are mounted. The printed circuit board is removed from the assembly apparatus, placed in a heating furnace, and heat-treated at the melting temperature of solder J58, J12.

FIG. 6 shows the configuration of a mechanical part of an assembly apparatus according to the present invention that automatically performs the assembly work of the above-mentioned printed circuit board and electronic components. In the figure, 20 is a board moving table for moving the printed circuit board 1 in the X+Y axis directions and rotating it around two axes, and 22 is a board moving table for moving the printed circuit board 1 in the X+Y axis directions and rotating it around two axes. A component supply device 23 receives the components supplied from the supply device 22 at its tip 24, and transfers the components to the hair vacuum suction head 2.
This is a work pallet that can be reciprocated in the y-axis direction for conveying to the position n directly below 5. The illustrated work pallet has a structure in which different electronic components can be handled by rotating around the y-axis. The vacuum suction head 25 is moved by a positioning device (not shown) that enables movement in the Z-axis direction and rotation around two axes, and after correcting the rotational deviation around the Z-axis of the parts picked up from the work pallet 24, This is pudding! (It operates to descend and load onto the surface of the plate 1. Reference numeral 27 is a reflector plate attached to the tip of the moving device 28 in the Z-axis direction. After retracting to the position of the supply device 22, it is inserted between the suction head 25 and the printed circuit board 1, and the light from the projector 29 that enters from the side is reflected upward to illuminate the electronic component 10 from below.

Reference numeral 30 denotes a television camera for detecting the position of the electronic component 10 held by the suction head 25. It is located above.

As explained in FIG. 5, since the top surface of the tip of the lead wire 11 of the electronic component is filled with pick-up solder 12, when the illumination light is irradiated from above the electronic component, the illumination light is reflected by the curved surface of the solder surface. Unable to image the correct outline of the lead wire. On the other hand, if the electronic component 10 is irradiated from below using the reflector 27 as shown in FIG. 6, a correct silhouette image of the lead wire outline can be captured. Incidentally, when the input image of the television camera 30 is processed and the position detection of the electronic component 10 is completed, the reflection plate 27 is retracted to the original position, and the position shift in the rotational direction of the electronic component 10 is caused by the immediate rotation of the suction head 25. After the positional deviation in the X+V direction is corrected by the movement of the printed circuit board moving table 20, the suction head 25 descends along the two axes to mount the electronic component at a predetermined position on the printed circuit board. In addition, when one component mounting is completed, the substrate moving table 2 is moved so that the next component mounting position is directly below the suction head 25.
0 is written. 34 is a television camera for detecting the position of the printed circuit board, and an objective lens 35. An image is taken through the eyepiece lens 36. The positions of the cross marks 6a to 6b on the printed circuit board described above are detected by this television camera 34, and the field of view is moved by the board moving table 20.

As mentioned above, this automatic assembly equipment is intended for electronic components with lead wires of extremely minute dimensions, and is required to accurately align these lead wires with the wiring terminals on the printed circuit board. . Therefore, an image magnified by the lens system is input to the television camera. in this case,
When an entire image of an electronic component is captured, the resolution decreases due to the sampling interval of the video a signal for pattern recognition, making it impossible to obtain the high degree of position detection accuracy required for positioning the lead wires.

Therefore, in the apparatus of the present invention, the position of the component is detected by greatly enlarging and capturing an image of the lead wire portion of the electronic component and performing pattern recognition processing in a plurality of imaging fields of view. Area 1 indicated by diagonal lines in Figures 2 to 4
3a, 13b, 13C1 and 13a, 13b' indicate pattern recognition processing areas for detecting positional deviation of the component held by the suction head. In the case of this example, an LSI component with lead wires on four sides of the package and an SS
For the I component, three specific imaging areas 13a, 1
Lead wires are detected at 3b and 13e.

The imaging areas 13a and 13b are set at positions including the endmost part 118° 11b of the lead wires arranged on two sides in the X-axis direction, and the other imaging area 13c is set at a position including the lead wire 11a at the end thereof. , llb is set at the center of the side in the y-axis direction on the side closer to llb. Lead wire 11a detected in imaging regions 13a and 13b.

The amount of positional deviation of the component in the X-axis direction and rotational direction is determined from the position 11b, and the positional displacement of the component in the y-axis direction is determined from the Y-coordinate of the lead wire detected in the imaging area 13c. On the other hand, for R parts that have lead wires only on two sides in the y-axis direction, a relatively large alignment error in the x-axis direction can be tolerated, so as shown in Figure 4, the imaging area is 1
3a13b' and the X-axis from the position coordinates of the endmost lead wires 11a and llb detected in each area.

Find the positional deviation in the y-axis and rotational direction.

The imaging field of view may be changed by moving either the suction head 25 or the television camera 30 including the optical system in the xy plane, but in the automatic assembly device shown in FIG. A method is adopted in which the objective lens 31 is fixed and only the objective lens 31 is moved in the X-axis direction.

Figure 7 shows electronic components 10. Objective lens 31. Eyepiece lens 3
2 and the imaging surface 30' of the television camera 30. Light 29' from the projector 29 is reflected by the reflector 27, illuminating the electronic component 10 from below. When the objective lens 31 is at the position indicated by the solid line, a real image of the lead wire in the imaging area A is formed at point P at the same size. ′. When the objective lens 31 moves to the position indicated by the dotted line, an enlarged image of the lead wire in area B is captured. If the imaging field of view is switched by moving the objective lens 31 in this manner, the movement distance of the objective lens can be r = -R for moving the field of view by a distance R. The distance between the subject and the objective lens is a1, the distance between the objective lens and the image plane is bi, and the distance between the image plane and the eyepiece is F.
12. If the distance between the eyepiece and the imaging surface is b2, then from the equation of simple similar triangles, R=r (ax + bx) / bx. Here, since the objective lens is the same magnification, b1/a1
=1, and R=2r. Therefore, compared to a method in which the television camera 31 or the suction head 25 is moved, the time required to change the imaging field of view can be shortened, and a decrease in positional accuracy due to movement of the field of view can be prevented.

FIG. 8 is an overall configuration diagram of the control circuit system of the automatic assembly device, where 40 is a data processing device that performs various arithmetic operations for controlling the operation sequence of the entire system and aligning electronic components according to a program, and 41 is the data processing device described above. A memory device connected to the data processing device; 42 is an input/output device for operator operation connected to the data processing device via a bus 43; 50 is a pattern detection circuit; Two television cameras 3
0.34, data processing 'JAj
[Perform pattern matching processing between the standard pattern given from 40 and the input video from the selected television camera. The pattern matching results for one screen are recorded in the memory built into the pattern detection circuit 50, and the contents of this memory are read into the data processing device and used for calculations to detect the position of the child part. . Details of the pattern detection circuit 50 will be described later with reference to FIGS. 9 to 12.

51 is a moving circuit of the printed circuit board moving table 20, 52
is the movement circuit of the parts supply device 22 and the work pallet 23,
53 is a moving circuit for the reflection plate moving device 28, 54 is a circuit for controlling the vacuum suction operation of the suction head 25, 55 is an immediate action circuit for the positioning device 25' for moving the suction head 25, and 56 is a circuit for the television camera. This is a quick-acting circuit of a positioning device 31' that moves the objective lens 31. These circuits 51~
56 is connected to the data processing bag γt40 via the information path 43, and immediately moves the controlled object in a predetermined direction in response to a command given from the data processing device.

FIG. 9 shows the overall configuration of the pattern detection circuit g50. This pattern detection circuit 50 successively compares the standard pattern given from the data processing device 40 and the partial patterns cut out at each position in the video input from the television camera, and selects the partial pattern showing the highest degree of matching. There is a type of method that determines the coordinates of the cutout position, but the first-order method is used to accurately detect the position of each of a plurality of lead wires that are projected in the same shape within the imaging field of view.

That is, (1) by changing the sampling interval of the video signal, the size of the video area cut out as a partial pattern can be varied, and coarse pattern matching and highly accurate pattern matching can be selectively performed;
) Setting a pattern search area of any size in the video and validating only the pattern matching results within the search area; (3) dividing the imaging screen into multiple sections and standard The positional coordinates of the partial pattern closest to the pattern are determined for each section. The sampling interval, the position and size of the search area, the size of the sections to be divided, etc. are all commanded from the data processing device 40 in accordance with the steps of the automatic assembly operation sequence.

In FIG. 9, 60 is a timing control circuit for generating a camera synchronization signal 60a, various timing signals 60b to 60f, and a position coordinate signal 60xy of the screen scanning point of the television camera, the details of which are shown in FIG. This will be explained later. The television cameras 30 and 34 scan the screen in synchronization with the horizontal and vertical synchronization signals 60a output from the timing signal generation circuit 6°, and generate the video signal 30.
s.

34s respectively. One of these video signals is selected by a switch SW which is changed over by a control signal S from the data processing device 40, and the video signal is sent to the binarization circuit 61.
is input. The binarization circuit 61 compares the input video signal with a predetermined threshold specified by the control signal 44f from the data processing device 40, and converts this into 111N, “OI
It is converted into a binary signal 61s of I and supplied to the video memory 62.

Reference numeral 62 denotes a video memory for storing two-dimensional images, which stores images corresponding to gn-1 scans while sequentially shifting them, and stores n pictures arranged vertically on the imaging screen. Outputs raw signals in parallel. Reference numeral 63 denotes a partial pattern cutting circuit which receives the output signal from the video memory 62, converts it into a two-dimensional partial pattern of nXn' bits, and outputs it. The video memory 62 and the partial pattern cutout circuit 63 are operated by clock signals 60b and 60c output from the timing control circuit 60, and by changing the period of these clock signals, the sampling interval can be changed and partial patterns can be cut out one after another. It is configured as follows. Its specific configuration will be described later with reference to FIG.

Reference numeral 64 is a register for holding a standard pattern consisting of information of nXn' picture elements to be compared with the above partial pattern. Contents of this register and partial pattern extraction circuit 6
The output of No. 3 is compared and verified for each corresponding bit by a verification and addition circuit 65, and the total number of bits that match in content is a signal 65s indicating the degree of matching between the partial pattern and the standard pattern.
Since the nine circuits 65 output as 9 operate in synchronization with the extraction circuit 63, the coincidence degree signals 65s are output one after another in parallel with the scanning of the imaging screen.

65' is a register for storing matching coordinates, 66 is a matching degree storage register, and 67 is a matching degree storing register, and the content of the distal 6a is compared with the matching degree signal 65s, and - the matching degree signal 65g is greater than the value of the register 66. This is a comparator circuit that outputs a pulse signal 67s.

The second pulse signal 67s is an A signal that is opened by the coincidence degree comparison instruction signal 60s output from the timing control circuit 60.
It is input to the ND gate 68, and the comparison instruction signal 60s
The signal passes through the AND gate 68 only during the output period of , and is inputted to the registers 65' and 66 as a pulse 68s that enables data updating.

The register 65' for storing coincidence coordinates stores the position coordinates 60xy outputted from the timing control circuit 160 when the pulse 68S is input, and the register 66 stores the coincidence degree signal 65s.

The position coordinate signal 60xy is the XY of the scanning point on the imaging screen.
It shows the coordinates and has a specific relationship with the position of the partial pattern cut out at @93.63. Therefore, register 6
5' and 66 store the positional coordinates of the partial pattern closest to the standard pattern detected within the area specified by the matching degree comparison instruction signal 60s and the degree of matching between the two.

A data processing device 70 sets a pattern search area on both sides of the image based on instructions from the data processing unit 40, and divides the search area into a plurality of sections, and when the current scanning point is within the search area, This is an area dividing circuit that generates a coincidence comparison instruction signal 70s and an address 70xy (70x, y) of the section to which the scanning point belongs, and its specific configuration will be described later with reference to FIG. The address signal 70xy output from this circuit is given to the coincidence degree memo 2 and the coordinate memory 73 via the address switching circuit 71.

-m degree memory 72 has a storage area corresponding to the address signal 70xy, and is capable of storing the maximum degree of coincidence between the partial pattern and the analysis standard pattern up to the present moment for each section corresponding to the above address in the search area. There is. That is, the contents of the addressed storage area of memory 72 are
2s, and is input to the comparison circuit 74 together with the coincidence degree signal 65s successively outputted from the verification and addition circuit 65. The comparison circuit 74 outputs a pulse signal 74s if the newly determined degree of coincidence 65s is greater. This pulse signal 74g is input to an AND gate 75 whose opening/closing is controlled by the coincidence degree comparison instruction signal 70s, and is output from the AND gate 75 only during the output period of the comparison instruction signal 70s, making it possible to update the data in the memories 72 and 73. The pulse becomes 75S. Therefore, the coincidence degree memory 72 can store the new coincidence degree given by the signal 65s in the storage area corresponding to the address signal 70xy in response to the pulse signal 75s.

On the other hand, the coordinate memory 73, like the coincidence degree memory 72, has a coordinate storage area corresponding to the address signal 70xy, and when the pulse signal 75s is applied, the timing control circuit 60 stores the coordinate memory area in the addressed storage area. Store the output coordinate data 6°xy.

Since screen scanning is performed repeatedly in the X direction while shifting the position in the Y direction, the addresses of the sections within the search area also change one after another according to the above screen scanning, and at the end of one screen scan, the standard for all sections is The maximum degree of matching between the pattern and the partial pattern and the position coordinates of the partial pattern are stored in the memories 72 and 73.

FIG. 10 shows an example of a specific circuit configuration of the video memory 62 and the partial pattern cutting circuit 63. Here, the video memory 62 consists of 0 to 1 shift registers 62-2 to 62-n that are cascade-connected to each other. The input of the first stage shift register 62-2 is the output 61s of the binarization circuit,
This signal is directly input to the partial pattern cutout circuit 63 as the first manual input signal 1s, and the output of each shift register is inputted to the cutout circuit as the second to nth input signals 2s=ns. These shift registers 62-2
~62~n each has a memory capacity capable of storing information for one scanning line, and the outputs 1s~ns are n picture element signals arranged vertically on the imaging screen. It is configured.

The partial pattern extraction circuit 63 has n shift registers 63-1 to 63-n each having a length of n' bits.
Each shift register converts the signal input serially from the video memory 62 into an n'-bit parallel signal 63s.
and outputs.

Each shift register of the video memory 62 and the partial pattern cutout circuit 63 performs an information shifting operation in response to a clock signal 60b or 60c output from the timing control circuit 60. Therefore, for example, if each shift register of the video memory 62 is given a clock signal 60b with an interval of one picture element, and each shift register of the cutout circuit 63 is given a clock signal 60c with an interval of m picture elements, for every wAm-th scan, A compressed two-dimensional partial pattern consisting of n×n' picture elements sampled at an interval of m picture elements in the vertical and horizontal directions can be obtained, and each scanning line receives clock signals 60b and 6 at an interval of one picture element.
If 0c is given, a two-dimensional partial pattern of a normal size can be obtained from the cutout circuit 63.

FIG. 11 shows an example of a specific circuit configuration of the timing control circuit 60. This circuit includes a first circuit section 60A that generates scanning point coordinates 60xy and a synchronization signal 60a for a television camera, and a data processing device! ! The search range is set according to the parameters given from 40, and the clock signals 60b, 60c,
The second circuit section 60B generates a match comparison instruction signal 60g and a search end signal 60f.

The first circuit section 60A turns on an oscillator 80 that generates a reference clock signal 60d at a period corresponding to one pixel interval, and advances an X-coordinate counter 81 using this clock. The period of the counter 81 is set in the constant setter 82, and -
When the number output circuit 83 detects that the re-counter 81 has reached its cycle, the counter 81 is reset. In this way, the counter 81 repeatedly counts at a predetermined period.

- Output 60eli Y coordinate counter 84 of number output circuit 83
Constant setter 8 that is added as a clock to give the period
5 and the Ichitsu Kenbatake circuit 86, the counter 8
4 performs repeated counting at a predetermined period, 87.88
are horizontal and vertical synchronization signal generation circuits, respectively,
These synchronization signal generation circuits each have X and Y coordinate counters 8.
It is detected that the contents of 1 and 84 have reached predetermined values, respectively, and a synchronization signal 60a having a constant timing and width is transmitted.

The second circuit section 60B receives a signal 44 from the data processing device 40.
X coordinates XS of the left and right ends of the search range given as a
, registers 90 and 91 that hold XE, and registers 95 and 9 that hold the Y coordinates YS and YE of the upper and lower ends of the search range.
6. Register 10 that holds the vertical sampling interval
1, and a register 105 that holds the horizontal sampling interval. Reference numerals 92 and 93 denote one-number ratio devices, which output a pulse when the contents of the registers 90 and 91 match the X coordinate 60x of the scanning point given from the first circuit section, set the flip-flop 94, and set the flip-flop 94. Reset. Therefore, flip-flop 94 is turned on only when the X coordinate is within the specified range. Similarly for the Y coordinate, the upper end.

registers 95, 96, -m detectors 97.9 giving the lower edge;
8 sets and resets the flip-flop 99 to produce a signal that is turned on only within a specified range. still,
- The output of the several bases on base 98 becomes a signal 60f indicating the end of the search.

100 is a counter that operates with the same clock 60d as the X-coordinate counter 81 in the first circuit section, and when its contents match the sampling period held in the register 101, it is reset by the output 102g of the output 102 of the -number on base 102. The contents return to zero. Therefore, the counter 100 repeatedly performs counting at the period held by the register 101. 103 is a circuit for detecting that the contents of the counter 100 are zero, and the output is turned on only for one clock time in one period. Similarly, for the Y direction, a counter 104, a period register 105, a number 8 unit 106,
The zero detector 107 produces a signal that turns on only one horizontal scanning line in one period.

111-112 are AND gates, and clock signal 6
0b, 60c and a match comparison instruction signal 60s are generated. That is, the AND gate 110 outputs the reference clock 60d for one horizontal scanning period at each specified sampling interval in the Y direction, and this clock becomes the shift clock 60b of the video memory 62. The AND gate 111 further outputs a clock sampled in the X direction, and this clock becomes the shift clock 60c of the partial pattern extraction circuit 63. The AND gate 112 outputs the clock 760c only while the flip-flop 113 is in the set state and the scanning point is within the search range in the X and Y directions.

This becomes the coincidence degree comparison instruction signal 60g. Incidentally, since the flip-flop 113 is set by the start instruction signal 44b given from the data processing device 40 and reset by the search end signal 60f, matching between the standard pattern and the partial pattern is performed when the data processing bag W40 instructs the start. This will be executed only during one screen scanning period.

FIG. 12 shows an example of a specific circuit configuration of the area dividing circuit 70.

This circuit consists of an X address control section 70A and a Y address control section 70A.
0B, 120X and 120Y are registers for holding the coordinates Xs+Ys of the starting point of the search, and 121X and 121Y are the registers
122X and 122Y are registers for holding the direction, dimensions d in the Y direction, d1, and 122X and 122Y are the number n of sections in the X and Y directions.
The parameters held in each of these registers are sent from the data processing device 40 as a signal 44c. Also, 123X
, 123Y, 124X, 124Y.

125X, 125Y are coincidence detection circuits, 126X.

126Y, 127X, 127Y are counters, 128X,
128Y, 129 are flip-flops, 130X, 1
30Y, 131, 134 are AND gates, 132X, 1
32Y, 133X.

133Y indicates an OR gate.

First, to explain the operation starting from the X address control unit 70A, the -number output circuit 123x compares the X coordinate 60x of the scanning point output from the timing control circuit 60 with the locus jIAXs of the search start point held in the register 120X. , -4, a pulse signal 140X is output. This pulse signal sets flip-flop 128X and O
Counter 126 via R gate 132X, 133X
X.

127X values are each reset to zero. Flip
When flop 128X becomes set.

The output opens the AND gate 130X, and the basic clocks 60d are successively input to the counter 126X to start counting.

- The multiple output circuit 124X outputs a pulse signal 141X when the value of the counter 126X matches the width d of one divided section held in the register 121X. This pulse signal 141X is input to the counter 127X, increments the count value by 1, and is applied to the reset terminal of the counter 126X via the OR gate 132X. Therefore, the content of the counter 127X is a value indicating which section in the horizontal direction is being scanned, and this value is output as a signal 70x indicating the X address of the section.

The coincidence detection circuit 125X uses the value of the counter 127X and the designated number n of sections in the X direction held in the register 122X.
8, and when they match, a pulse signal 142X is output. The pulse signal 142X is.

It is input to the reset terminal of counter 127X via OR gate 133X, returns its value to zero, resets flip-flop 128x, and outputs AND gate 130
X is closed to block input of the basic clock 60d. Since these operations are repeated for each horizontal scanning line, the signal 70x indicating the X address of the divided section within the search area is repeatedly output.

Next, Y address system #70B will be explained.

In the Y address control section, when the instruction signal 44d to start the detection operation is output from the data processing device 40, the flip
Flop 129 is set and AND gate 134 is opened. - The several output circuit 123Y receives the Y coordinate 60y of the scanning point output from the timing control circuit 60 and the register 1.
Compare the coordinate Ys of the search m starting point held at 20Y,
When they match, a pulse signal 140Y is output. This pulse signal 140Y resets the counters 126Y and 127Y via OR gates 132Y and 133Y, and if A
If ND gate 134 is open, it sets flip-flop 128Y.

As a result, the AND gate 130Y is opened, and the carry signal 60e output from the timing control section #&60 for each horizontal scanning line is inputted one after another to the counter 126Y, and a counting operation is started.

- The number configuration circuit 124Y outputs a pulse signal 141Y when the value of the counter 126Y matches the vertical @d of one divided section held in the register 121Y. This pulse signal is input to the counter 127Y to increment the count value by 1, and is also input to the reset terminal of the counter 126Y via the OR gate 132Y to reset the value.

Therefore, the counter 126Y repeats the counter operation with a counter value equal to the vertical width of the divided section at a cycle, and causes the counter 127Y to perform a counting operation every time the scanning point moves from one section in the Y direction to the pine section. The content of the counter 127Y is a value indicating which section in the vertical direction is being scanned, and is output as a signal 70y indicating the Y address of the section having this value. Signal 70y
- Soft memory 7 along with signal 70x indicating the X address
2. It is applied to the coordinate memory 73.

- Number of occurrences g125Y is the value of the counter 127Y and the designated number n of sections in the vertical direction held in the register 122Y.
1, and when it is -M, a pulse signal 142Y is output. This pulse signal 142Y resets the counter 127Y via the OR gate 133Y, and at the same time resets the flip-flops 128Y and 129. Further, the pulse signal is sent to the data processing device ff140 as a completion signal 70e of the designated pattern detection process.

Since the flip-flop 128Y is in the ON state for exactly one scanning period of the search area, by taking out the AND ratio of its output signal and the output signal of the AND gate 130x of the X address control section from the AND gate 131. , - A softness comparison instruction signal 70g is obtained.

Next, the program operation of the data processing device that controls the automatic assembly apparatus of the present invention will be explained.

Figure 13 shows the basic operating procedure for mounting multiple types of electronic components (LSI, SSI, R) on one printed circuit board 1 in a predetermined order. The routine 300 is started when the operator fixes the printed circuit board 1 on the moving table 20 and inputs an assembly start command from the input device 42. routine 30
Reference numeral 2 is for correcting the position of the printed circuit board, and there is no misalignment due to play between the hole 5 formed in the printed circuit board and the alignment bin provided on the moving table 20, or due to manufacturing error. The positional deviation of the printed wiring is detected using the first television camera 37, and the position and direction of the printed wiring are made to coincide with the reference x and y axes. When this position correction is completed, the routine proceeds to routine 310, where various parameters necessary for processing the first part (for example, the LS1 part) are set. The parameter set is to enable the assembly process of multiple types of electronic components to be executed by a common subprogram, and includes, for example, the type and number of electronic components, the type of standard pattern to be used, the positional relationship of the imaging field of view, Variables indicating the search area, area division, etc. are associated with the first part. 311 is a routine that determines whether or not all the first parts have been mounted. If there are still first parts to be processed, the process proceeds to routine 312; if all the first parts have been completed, the process proceeds to routine 320. . In routine 312, the first component is supplied from the work pallet 23 to the suction head, and the board table 20 is moved so that the terminal group on the printed circuit board side connected to the first component is located under the suction head.
move. The amount of movement of the substrate movement table is stored in advance in the memory according to the mounting order for each type of electronic component. When this is completed, the routine proceeds to routine 313, which involves detecting the position of the first component held by the suction head and correcting the posture. Next, in routine 314, the substrate moving table 20 is moved to correct the positional deviation of the component, and in routine 315, the suction head is lowered to mount the electronic component on the printed circuit board. When the routine 315 ends, the process returns to the determination routine 311.
The above operations are repeated until a predetermined number of first parts have been mounted.

Routines 320-325 and routines 330-33
5 shows similar processing for the second part (for example, SSI part) and the third part (for example, R part), respectively,
When the mounting of all the third parts is completed, the process proceeds to an end routine 340, and a signal is output to the input/output device 42 to notify the operator of the end of the process. If an abnormality is detected in each of the above-mentioned routines, the program sequence proceeds to the end routine 340, and the input/output device 42 notifies the operator of the occurrence of the abnormality and its details.

The details of the printed circuit board position correction routine 302 are explained in the 14th section.
This will be explained with reference to FIGS. The position correction of the printed circuit board 1 is performed by sequentially detecting the positions of the cross marks 6a to 6d formed at the four corners of the printed circuit board with the first television camera 34, and calculating the position relative to the board xy coordinate system from the position coordinates of each cross mark. This is done by determining the amount of positional deviation of the printed circuit board.

FIG. 14 shows an enlarged view of one cross mark.

The cross mark portion 6 consists of a metal layer printed on the substrate and appears as a bright area on a television camera. Data indicating the standard position t of the center 7 of each cross mark with respect to the origin of the substrate moving table 20 is stored in advance, and the substrate moving table 20 is moved based on the above position data by a command from the data processing device 40. As a result, the imaging field of view at each stop position can be set to the position indicated by the two-dot chain line 00. Therefore, as shown in Figure 1, two narrow search areas x1 to Xx+'lx are separated by a predetermined distance from the center T of the standard loss mark
~y2 is set, and the standard pattern 201 shown in FIG. 15(A) is used in the first search area (Xyu~X2), and the standard pattern 201 shown in FIG. 15(B) is used in the second search area (y□~yz). If pattern matching is performed using the standard pattern 202 shown in , the X and Y coordinates of the actual mark center position can be calculated from the coordinates of the partial pattern cutting position showing the maximum softness with respect to the standard pattern detected in each search area. can be found. In this case, each search area can be set to a range where there is only one pattern with the same characteristics as the standard pattern, so
A high-resolution pattern that is not sampled can be applied to the target patterns 201 and 202.

FIG. 16(A) is a detailed program flowchart of a routine 302 for sequentially detecting the four cross marks 6a to 6dt by the method described above and correcting the positional deviation of the printed circuit board. In this program, routine 40
In step 2, the printed circuit board is moved to the position of the mark 6a at the lower left end, and in routine 404, the mark position is detected. When this is completed, the detection result is determined in a determination routine 406. If the mark detection has ended normally, the process proceeds to 408, and if the mark detection has failed, the process proceeds to the abnormal termination processing routine 436. In routines 408 to 412, the lower right end mark 6 is
Detect the position of b. If the position resolution for the lower right end mark 6b has been successfully completed, the process advances to routine 414;
Calculate the amount of rotational deviation from the difference in the X coordinates of the left and right marks0
If the deviation amount in the rotational direction is within the allowable value, the process proceeds from 1 judgment routine 416 to 422; otherwise, the process proceeds to judgment routine 418, in which the number of rotational deviation corrections that have already been performed is checked. If the predetermined number of corrections have been completed, it is assumed that automatic correction is impossible and the process ends abnormally. If not, the substrate moving table 20 is caused to perform correction operations around two axes, and the process returns to routine 402. In routines 422 to 432, the upper right end mark 6e and the upper left end mark 6d are detected in the same manner as in steps 402 to 414. 1 judgment routine 432 to 43 when position detection is completed normally for all marks.
Proceeding to step 4, the amount of deviation in the X and Y directions for aligning the printed circuit board with the reference position is calculated, and the board moving table 20 is driven in a direction to correct this. Through this program processing, the printed circuit board 1 is located at the origin of the standard xy coordinate system, and by numerically controlling the board movement table 20 using this as the starting point, the mounting position of each component is placed directly below the suction head 25. Can be aligned sequentially.

FIG. 16(B) shows the routines 404, 414, 424 and 43 for mark position detection in the above program.
2 is a detailed flowchart of a subprogram corresponding to 0. In this program, after initializing various parameters in a start routine 440, the number of the designated parameter of the standard pattern is set to "0" in a routine 442.

When this number is "0", the first search area (X-work to X2) is specified in routine 444, and fine pattern matching using standard pattern 201 is executed in routine 446. In other words, the signals 44a and 44 in FIG.
is output. When the pattern detection circuit outputs the end signal 60f, the routine 448 is executed and the signal %44n is output.
and 44i are output and the contents of registers 65 and 66 are read into the memory 41. In the 0th order routine 450, the pattern number is incremented by 1. If the pattern number is 1 or less, the judgment routine 452 returns to the routine 444 and the standard Second search area (y-yz) using pattern 202
The same pattern matching as above is performed. Second
If pattern matching is completed in the search area,
Proceeding to routine 454, it is determined whether or not the two pieces of matching degree data read into the memory 41 are greater than or equal to a predetermined value.
After performing the calculation, the process proceeds to the termination routine 464. If the softness is less than the specified value, a routine 458 checks the number of times the binarization threshold has been changed. The binarization threshold has an influence on the matching degree of pattern matching, and if the change has already been made a predetermined number of times, the routine advances to routine 460 and abnormal termination processing is performed. Otherwise routine 46
2 changes the binarization threshold slightly from the current value.

Return to routine 442. To change the binarization threshold, use signal 4.
This is achieved by changing the value of the data output as 4f.

FIG. 17 shows the routine 312 (322°33
2) is a detailed program flowchart. In this program, first, from the supply device 22 to the work pallet 2,
3, the substrate moving table 20 is moved to the next component mounting position according to pre-stored position data (routine 474), and the work pallet 23 is moved forward by a predetermined distance (routine 476).
.

The work pallet stops when the component 10 is directly under the suction head 25.Next, the suction head 25 is lowered a predetermined distance from the initial position (routine 478), and the suction valve is opened to suck the component (routine 480)
. When this is completed, the suction head is raised to a position where the lower ends of the five parts are completely separated from the top surface of the work pallet (routine 482), and the work pallet is returned to the initial position, that is, the position where the next part is received (routine 484). ).

When the work pallet is retracted, the suction head 25 is lowered a predetermined distance (routine 486), and the reflection plate 27 is inserted directly below the component (routine 488). Stops at the position that matches the focal length.

FIG. 18 shows a detailed program flowchart of the component mounting routine 315 of FIG. 13. In this program, after returning the reflector to its initial position (routine 502)
, the suction head 25 is lowered to the surface position of the printed circuit board (routine 504), and further lowered by a certain distance to apply back pressure for temporary bonding (routine 506);
When this is finished, return to the initial position (routine 508)
.

Next, the component position correction routine 313 (323
, 333) is shown in the first part.
This will be explained with reference to Figure 9.

As already explained in FIGS. 2 to 4, in this embodiment, three different imaging regions are set for the LSI component and the SSI component, and two imaging regions are set for the R component.
Based on the relationship between the positions of the lead wires detected in each imaging area, the attitude and positional deviation of the picked-up electronic component is determined. Images of these areas are taken by the second television camera 30, and if there is a positional shift in the electronic components.

Positional deviations in the X and Y directions can be detected using the moving table.
The correction is made by moving the JIs plate, and only the positional deviation in the rotational direction is corrected by the rotational operation of the suction head.

The program of the component position correction routine shown in FIG.
The objective lens 31 is moved to a position, and in routine 524 the position of the lead wire Lla imaged at the end is detected.8 Next, in routine 526, the objective lens 31 is moved to a second detection position, for example, the imaging area 13b. Then, in routine 528, the position of the lead wire 11b imaged at the end is detected. When the positions of the two lead wires are detected in this manner, the amount of rotational deviation of the electronic component around the Z axis can be determined from the X and Y coordinates of each lead wire in routine 530. Also, Figures 2 and 3.

As can be understood from FIG. 5, the lead wire is bent into a crank shape, so the lead wire is
Although the lengths of lla and llb vary in the y-axis direction, they have high positional accuracy in the axial direction orthogonal to this. Therefore, the first. Two lead wires 11a detected at the second detection position.

The X coordinate of 11b can be used to accurately calculate the position of the electronic component in the X-axis direction, and this calculation is performed in routine 530. In routine 532, the suction The head driving device is operated in the positional deviation correction direction. In the determination routine 534, the correction amount is compared with a predetermined value, and if the correction amount is large, the process returns to the routine 522 and repeats the above-described operation, and if not, the process proceeds to the routine 536. Routine 5
In step 36, it is determined whether the type of part currently being processed is an R part. If it is an R part, the routine advances to routine 538, where the Y coordinate of the component is calculated from the Y coordinates of the two lead wires that have already been detected, and the routine advances to end routine 546. LSI parts,
In the case of SSI parts, determination routines 536 to 540
, 542, the lead wire 11c is detected at the third detection position 7213c, and the routine 5 is executed based on the Y coordinate.
44, calculate the exact Y coordinate of the part.

In the above component position correction program, the routines 524, 528 and 542 for detecting the position of the lead wire detect a specific lead wire from among a large number of lead wires having the same shape included in the imaging field of view. It is necessary to devise ways to select the

Therefore, in the present invention, as shown in FIG.
52) To detect multiple lead lines, a method is adopted in which the set search area is divided into multiple sections and the maximum change between the standard pattern and the part pattern cut out from the image is determined for each divided area. . In this way, the target endmost lead wire can be specified from the relative positional relationship of the plurality of detected lead wires. In this coarse position detection routine 554, it is determined whether or not the position detection has been successful. If unsuccessful, the process proceeds to an abnormal termination routine 560, and if successful, the process proceeds to a routine 556. In routine 556, a narrow search area is set from the coarse position data of the lead wires that have already been identified, similar to that used for cross mark detection, and pattern matching is performed using a precise standard pattern without sampling. do.

This allows accurate X and Y coordinates of the target lead line to be determined. Routine 558 determines whether or not the precise position detection is successful, and if not, proceeds to an abnormal termination routine, otherwise proceeds to termination routine 562.

Details of the rough position detection routine 552 will be explained with reference to the subprogram flowchart of FIG. 21 and FIG. 22.

In this program, first, in routine 602, a coarse standard pattern used for pattern matching is determined by a signal 44.
It is set in the register 64 as .

use! ! The A quasi-pattern differs depending on the type of electronic component and the imaging area, that is, which of the first to third detection positions is selected, and is specified by parameter setting before branching to this subprogram.

For example, at the second detection position of the LSI component, the imaging area 13b
Lead wires are projected as shown in FIG. 22 (A) or (B), and the pattern shown in FIG. 23 is used as a standard pattern for detecting the position of the tip of each lead wire. This standard pattern 220 consists of 12×12 pixels, and the 22nd
As shown by a broken line frame 221 in Figure (A), it includes a characteristic pattern that matches the tip of the lead wire. Similar standard patterns are prepared for each type of lead wire and each detection position.

In the routine 604, a parameter specifying the sampling interval and a parameter for dividing the imaging area into a plurality of sections are output as signals 44a and 44c to the timing control circuit 60 and the area dividing circuit 70, respectively.

As a result, a search area divided into sections 210 to 214 is set, for example, as shown in FIGS. 22(A) and 22(B). Determined by the size and spacing of the lead wires included in the section. 608 is a routine that detects the position of the partial pattern that most matches the standard pattern for each of the divided sections, whereby the data processing device 40 outputs a start instruction signal 44d to the area dividing circuit, and the pattern The detection circuit 5o operates. When the region dividing circuit outputs the end signal 70e, a routine 610 reads data for each section from the coordinate memory 73 and the coincidence degree memory 72. In routine 612, only those data whose degree of coincidence exceeds a predetermined value are selected as lead line candidate point data.

614 is a routine that determines whether or not there is selected candidate point data; if there is no data that corresponds to the candidate, the process proceeds to routine 624, and checks whether or not the binarization threshold has already been changed a specified number of times. judge. If this judgment indicates that the threshold value change has been completed the specified number of times, the process proceeds to routine 626 and is treated as an abnormal termination process; otherwise, the process proceeds to routine 626.
8, the threshold value is changed and the processing from routine 608 is performed again. If there is data that is a candidate point for a lead line, the routine proceeds from routine 614 to 616, and the positional relationship of the candidate points is examined. Here, two
If more than one candidate point exists, one candidate point with the maximum degree of coincidence is left among them, and the others are deleted. This is because data registration may occur in two or more sections for one lead wire depending on the positional relationship between the sections and the lead wire. In routine 620, the presence or absence of a target lead line is determined from among the candidate points arranged in this manner based on the positional relationship of the X and Y coordinates. If the target lead line located at the end is found, the routine proceeds to routine 622, where the position coordinates of a specific point on this lead line, for example, the coordinates of point Q in FIG. 22, are calculated, and the process proceeds to end routine 630. Otherwise, the process proceeds to routine 624, where it is determined whether or not to change the binarization threshold and perform pattern detection again.

Next, details of the precise position detection routine 556 will be explained with reference to the program flowchart of FIG. 24 and FIGS. 25 and 26. Since the coarse position detection described above uses a sampled coarse standard pattern, the determined X and Y coordinates of the specific point Q include an error corresponding to the sampling interval. Therefore, as shown in FIG. Position detection is performed using precise standard patterns shown in Figures 26 (A), (B), and (C).These standard patterns also consist of 12 x 12 pixels.
1st. The accurate X coordinate of point Q is determined from the center of the X coordinate determined in the second search area, and the accurate Y coordinate of point Q is determined from the Ysi mark determined in the third search area.

In the program shown in FIG. 24, first, in routine 642, a number "O" is set in a parameter indicating the standard pattern to be used. When this number is "0", routine 644
In steps 648 to 648, for example, using the standard pattern 231 in FIG. 26(A), pattern matching is performed with the search area specified as X1 to x2, and the detected The data processing device operates to register in a predetermined memory area. The above parameters are incremented by 1 in routine 650. Determination routine 652 determines whether this value is greater than 1, and if greater than 1, proceeds to routine 654, otherwise returns to 644. As a result, the standard pattern 232 is used corresponding to the number "1", and X, to X.

Pattern matching is performed using the search area as the search area.

Routine 654 determines whether the degree of matching obtained by pattern matching twice is greater than or equal to a predetermined value. - If the softness is low, proceed to routines 660 and 662 to change the threshold for video binarization, and repeat the processing from routine 642. If a high match is obtained, routines 654 to 656
, and detect the width of the lead line from the two X coordinates. This lead width is determined in a routine 658, and it is determined whether or not it falls within an allowable range of standard values predetermined for each component. If the error is too large, routine 660 is entered to change the video binarization threshold.

If it is within the allowable range, the process proceeds to routines 664 and 666, and pattern matching using the standard pattern 233 is executed in the third search range @y-y2.

This matching result is determined in routine 668.

If a high degree of coincidence is obtained, the routine proceeds to routine 670, where the X and Y coordinates of the center point Q of the lead line are calculated. Next, in routine 672, the calculation result is stored in a predetermined memory area corresponding to the imaging field of view, and the process proceeds to end routine 676.

In addition, although the above description describes the position detection of the lead wire extending in the y-axis direction of the imaging area 13b, the first lead wire that is in the opposite direction or orthogonal relationship thereto is detected. Third imaging area 13a, 1
3c, the standard pattern 23 shown in FIG. 26(D)
4 is required.

Also second. In the third imaging area, the X coordinate and Yf'! are used to calculate the lead width and its center position. The i mark has an inverse relationship. Regarding these matters, in each routine of the program shown in Fig. 24, the processing contents can be determined by referring to the parameters indicating the type of part and the imaging area at that time, so this program can be used in multiple types. can be commonly applied to multiple imaging fields of view of parts. Also, 1lil$ in Figures 26 and 15! ! Patterns can be shared.

FIG. 27 is a time chart showing the operations of each part from picking up one part to mounting it on the board under the program control described above.

As explained above, the automatic assembly apparatus of the present invention uses two imaging devices, the first imaging device detects the position of the board, and the second imaging device detects the position of the component. . By using two imaging devices in this way for detection targets located at different heights,
Focusing work is no longer necessary, reducing processing time, improving detection accuracy, and simplifying the imaging system positioning mechanism. Furthermore, according to the present invention, the position of a component is determined from the results of feature detection performed in a plurality of imaging fields, so the position of components of different sizes can be detected using the same recognition method. If a method of switching the imaging field of view by movement is adopted, these recognition processes can be realized in a short poem with a simple mechanism. In this case, by illuminating the part from the back and photographing its silhouette image as shown in Figure 6, it is possible to simply prevent the part shape from being distorted due to light reflection on the curved surface of the part. Since the pattern on the surface of the board on the back of the component can be blocked and only the component can be imaged, it has the advantage of preventing misrecognition of the pattern.

In addition, in the example, a light reflecting plate is inserted between the component and the board,
The horizontally projected light from the projector is reflected upward by the light reflecting plate to capture a silhouette image of the part, but this reflecting plate is inserted when detecting the position of the part, and is inserted from directly below the suction head when loading the part. As long as it can be evacuated, the work pallet 24 shown in FIG. 6 is used also as a light reflecting plate, and the element 27
．． It is also possible to have a structure in which 28 is omitted.

In this case, the part 24 of the work pallet parts receiver is made of a transparent member, and a light source is built inside the part 24, or a reflector is attached to the inside of the part 24, so that the light from the projector 29 is reflected upward. do it.

The embodiment describes an assembly operation in which an electronic component such as an LSI is stacked on a terminal forming surface of a printed circuit board, but the characteristic configuration of the present invention is an assembly operation in which an electronic component such as an LSI is stacked on a terminal forming surface of a printed circuit board. It is clear that this method can also be applied to the automation of other assembly operations.

[Brief explanation of drawings]

FIG. 1 shows an example of a printed circuit board 1 on which electronic components are mounted, and FIGS. 2, 3, and 4 each show an example of the shape of an electronic component 10 mounted on the printed circuit board. Figure 5 is a diagram showing the connection state between the terminal on the printed circuit board and the lead wire end of the electronic component, and Figure 6 is a schematic diagram of the mechanical part of the device that automatically mounts the electronic component on the printed circuit board. 7 is a diagram for explaining the optical system of the first imaging device of the automatic loading device, FIG. 8 is a block diagram showing the overall configuration of the control system of the automatic loading device, and FIG. 9 is a perspective view shown in FIG. 10 is a block diagram showing an example of a specific configuration of the pattern detection circuit 50 in FIG. 8, FIG. 12 is a diagram showing a specific configuration of the timing control circuit 60 in FIG. 9, and FIG. 12 is a diagram showing a specific configuration of the area dividing circuit 70 in FIG. 9. FIG. 13 is a program flowchart showing a schematic control procedure of the automatic mounting device, FIG. 14 is a diagram showing the relationship between the shape of the positioning cross mark formed on the printed circuit board 1 and the imaging field of view, and FIG. 15(A) and 15(B) are diagrams each showing a standard pattern applied to detecting the position of the cross mark, and FIG. 16(A) is a program showing details of the printed circuit board position correction routine 302 in FIG. 13. - Flowchart: FIG. 16(B) is a program flowchart showing a more detailed procedure of the mark position detection routine 404 in FIG. 16(A), and FIG. 17 is a program flowchart showing a more detailed procedure of the mark position detection routine 404 in FIG. 13. 18 is a program flowchart showing details of the component mounting routine 315 in FIG. 13, and FIG. 19 is a program flowchart showing details of the component mounting routine 315 in FIG.
Program flowchart showing details of No. 13, No. 20
The figure shows the lead wire position detection routine 524 (5
28, 542), FIG. 21 is a program flowchart showing more detailed steps of the lead wire rough position detection routine 552 of FIG. 20, and FIGS. 22(A) and (B) are imaging Diagram 23 for explaining the relationship between lead lines included in the field of view and area division
The figure shows an example of a standard pattern applied for rough position detection, FIG. 24 is a program flowchart showing a more detailed procedure of the lead wire fine position detection routine 556 of FIG. 20, and FIG. An explanatory diagram of the pattern search area set for the above-mentioned precise position detection, FIGS. 26(A) to 26(D)
) are diagrams each showing an example of a standard pattern applied for precise position detection, and FIG. is a printed circuit board moving table, 2
2 is a parts supply device, 23 is a work pallet, 25 is a vacuum suction head for holding parts, 27 is a light reflection plate, 3
0 is a television camera for detecting component position, 34 is a television camera for detecting board position l, 31.35 is an objective lens, 32.
36 is an eyepiece lens, and 29.37 is a projector. To the 7th eye / tρC ) 72 Kokou 1st evening (A) (B) 21 / 17th rod / 3rd eye

Claims (1)

[Claims] 1. a component moving means for moving a component having a plurality of lead wires to a predetermined position on a board to be mounted; an irradiation means for irradiating the lead wires of the moved component with light from above the board; an imaging means for imaging at least two component regions of the lead wire of the component by transmitted light of the light irradiated by the irradiation means; and standards for at least two specific regions to be detected including the tips of the lead wires of the component; According to a storage means for storing position information and reference position information of a specific area to be detected read from the storage means,
means for moving the imaging region of the imaging means, and dividing the imaging region imaged by the imaging means after the movement based on the size or interval of the lead wires of the component, and the lead wires existing in the divided regions; A device for detecting the position of a component, comprising means for detecting the position of a partial pattern of the lead wire located at the tip thereof as seen from the imaging means.