KR102157356B1 - Recognizing apparatus, recognizing method, mounting apparatus and mounting method - Google Patents

Abstract

There is provided a recognition apparatus capable of avoiding a decrease in position recognition accuracy that occurs when a correspondence relationship between an arrangement of an object to be recognized and a plurality of imaging pixels satisfies a specific condition.
The recognition apparatus of the present invention captures an image of a zoom lens 33 having a variable imaging magnification, an IC chip 4 having a plurality of imaging pixels, and having bumps arranged at predetermined intervals through the zoom lens 33. The chip camera 2 and a storage unit storing shape information of the IC chip 4 including at least information indicating the arrangement interval of the plurality of bumps, and the length according to the imaging pixel and the arrangement interval stored in the storage unit A signal processing unit 1 is provided that adjusts the imaging magnification of the zoom lens 33 based on the displayed information, and recognizes the position of the bump based on a plurality of pixel values in the image captured by the chip camera 2. .

Description

Recognizing apparatus, recognition method, mounting apparatus, and mounting method TECHNICAL FIELD [Recognizing apparatus, recognizing method, mounting apparatus and mounting method]

The present invention relates to a recognition device, a recognition method, a mounting device, and a mounting method.

As described in Patent Document 1, when a semiconductor chip is mounted on a circuit board, it is performed to recognize the holding posture of the semiconductor chip being conveyed toward the circuit board. In the recognition device described in Patent Document 1, the posture of the package is recognized by imaging the bottom surface of the ball grid array type package and obtaining the inclination of the image of the solder ball. At that time, the position of the semiconductor chip is estimated from the positions of the plurality of centers of gravity of the solder balls in the image by linear fitting by the least squares method.

In addition, Patent Literature 2 shows that in the position recognition of a plurality of solder bumps of a semiconductor chip, the recognition accuracy is improved by increasing the contrast (contrast difference) in a captured image.

In addition, Patent Document 3 shows a configuration in which the lens magnification is finely adjusted for an inspection object having a periodic pattern.

Patent Document 1: Japanese Patent Laid-Open Publication No. Hei 7-320062 Patent Document 2: Japanese Patent Laid-Open No. 2005-93839 Patent Document 3: Japanese Patent Laid-Open No. 2011-75310

In the recognition device described in Patent Document 1, Patent Document 2, and the like, a semiconductor chip is imaged using an image sensor having a plurality of image pickup pixels. Then, a recognition process is performed on a plurality of formations on a semiconductor chip such as solder bumps included in the captured image as targets. The narrowing of connection portions such as solder bumps formed on semiconductor chips is progressing. If the number of pixels of the imaging sensor is the same, the number of pixels allocated per solder bump or each pitch between solder bumps decreases as narrowing proceeds. In this case, it is conceivable that the following problems arise as the resolution decreases.

That is, when the arrangement of the solder bumps and image sampling coincide, there arises a problem in that the position recognition accuracy may be lowered compared to the case where they do not coincide. Here, the case where the arrangement of the solder bumps matches the image sampling means that the number of pixels of the imaging pixels corresponding to each pitch of the arrangement of the solder bumps is always constant. In this case, the following phenomenon can be considered. As a premise, the position at which the pixel value peaks is assumed to correspond to the center position (vertical position, etc.) of the solder bump. In addition, in the vicinity of the center position of the solder bump, it is assumed that two pixels take a peak value (or a value close thereto) on average. In this case, if the arrangement of the solder bumps and image sampling coincide, there is a possibility that two pixels always take a value near the peak value for a plurality of solder bumps (that is, always with the same repetition). On the other hand, when the arrangement of the solder bumps and image sampling are inconsistent, there is a high possibility that variations occur for each solder bump, for example, one to three pixels take a peak value for a plurality of solder bumps. In this case, the error in position recognition is within 2 pixels when the arrangement of the solder bumps and the image sampling coincide, and within 1 to 3 pixels when they do not match. Here, for example, if there are a plurality of solder bumps having an error of less than one pixel, it is possible to accurately estimate the position and direction shift of each solder bump. That is, in this example, when the arrangement of the solder bumps and the image sampling coincide, there is a case where the position recognition accuracy is lower than the case of the mismatch.

The present invention has been made in view of the above circumstances, and a recognition device, a recognition method, a mounting device, and a recognition device capable of avoiding a decrease in position recognition accuracy that occurs when a correspondence relationship between an arrangement of a recognition object and a plurality of image pickup pixels satisfies a specific condition. It aims to provide a mounting method.

In order to solve the above problems, the recognition apparatus of the present invention includes an image pickup magnification adjustment unit capable of changing an image pickup magnification, a plurality of image pickup pixels, and a recognition object having a plurality of first recognition objects arranged at predetermined intervals. An image pickup unit that captures an image through an adjustment unit; a storage unit storing shape information of the recognition object including at least information indicating an arrangement interval of the plurality of first recognition objects; and a length and the storage unit corresponding to the image pickup pixel A signal for adjusting the imaging magnification of the imaging magnification adjusting unit based on information indicating the arrangement interval stored in the image pickup unit, and recognizing the position of the first recognition target based on a plurality of pixel values in the image captured by the imaging unit It characterized in that it comprises a processing unit.

In addition, another recognition device of the present invention is that the signal processing unit adjusts the imaging magnification of the imaging magnification adjustment unit so that an integer multiple of the arrangement interval and the length according to the imaging pixel among the images captured by the imaging unit becomes non-corresponding. It is characterized.

In addition, in the mounting apparatus of the present invention, the object to be recognized is a semiconductor chip, and based on a result of recognition by the signal processing unit, a predetermined correction control is performed on the position or angle of the object to be recognized, so that the object to be recognized is converted to a predetermined circuit board. It is characterized in that it is mounted on.

In addition, the recognition method of the present invention includes an imaging magnification adjustment unit capable of changing an imaging magnification, a plurality of imaging pixels, and a recognition object having a plurality of first recognition targets arranged at predetermined intervals through the imaging magnification adjustment unit. And a storage unit storing shape information of the object to be recognized, which includes at least information indicating an arrangement interval of the plurality of first recognition objects, and a length corresponding to the image capturing pixel and stored in the storage unit. The imaging magnification of the imaging magnification adjustment unit is adjusted based on the information indicating the arrangement interval, and the position of the first recognition target is recognized based on a plurality of pixel values in the image captured by the imaging unit. .

In addition, another recognition method of the present invention is characterized in that the imaging magnification of the imaging magnification adjustment unit is adjusted so that an integer multiple of the arrangement interval and the length according to the imaging pixel among the images captured by the imaging unit becomes non-corresponding.

Further, in the mounting method of the present invention, the recognition object is a semiconductor chip, and the recognition object is mounted on a predetermined circuit board by performing a predetermined correction control on the position or angle of the recognition object based on the recognition result. It is characterized.

According to the present invention, for a recognition object having a shape in which the arrangement of the recognition object and the image sampling match, the arrangement of the recognition object and the image sampling can be made inconsistent by adjusting the imaging magnification. Therefore, it is possible to easily avoid the deterioration of the position recognition accuracy occurring under specific conditions.

1 is a schematic diagram for explaining a configuration example of a mounting apparatus 100 as an embodiment of the present invention.
FIG. 2 is a perspective view schematically showing the positional relationship of the chip camera 2, the zoom lens 33, and the IC chip 4 shown in FIG. 1.
FIG. 3 is a schematic diagram for explaining an example of a captured image by the chip camera 2 shown in FIG. 1.
4 is a schematic diagram for explaining the operation of the zoom lens 33 shown in FIGS. 1 and 2.
5 is a block diagram illustrating an example of an internal configuration of the mounting device 100 shown in FIG. 1.
6 is a flowchart showing an operation flow of the mounting apparatus 100 shown in FIG. 1.
7 is a diagram for describing an example of a fine adjustment amount of a pixel size described with reference to FIG. 4.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. 1 is a schematic diagram for explaining a configuration example of a mounting device 100 (recognition device) as an embodiment of the present invention. The mounting apparatus 100 shown in FIG. 1 is a device for mounting an IC chip 4 (semiconductor chip) on a circuit board 8. The mounting device 100 includes a signal processing unit 1, a chip camera 2, a zoom lens 33, a moving unit 6, a motion controller 61, a substrate stage 7, a fixing table 7a, and a place camera. 10) and the like.

The signal processing unit 1 performs predetermined recognition processing on each image signal acquired from the chip camera 2 and the place camera 10, and the recognition processing result and design information of the IC chip 4 or circuit board 8 The motion controller 61 and the like are controlled according to the same. This signal processing unit 1 is connected to the motion controller 61 via a cable 41. Further, the signal processing unit 1 is connected to the chip camera 2 via a cable 42. Further, the signal processing unit 1 is connected to the place camera 10 via a cable 43. Then, the signal processing unit 1 is connected to the zoom lens 33 via a cable 44.

The motion controller 61 has a drive mechanism using a motor, etc., and moves the moving part 6 in each axis direction of XYZ indicated by an arrow or circle according to the instruction of the signal processing unit 1, and at the same time, XY Change the direction θ on the plane. In other words, the moving part 6 can move a predetermined length in the depth front and rear direction (X direction), left and right direction (Y direction), and vertical direction (Z direction) toward the drawing, and at the same time, the direction of the suction head 5 ( It is possible to change θ) in a predetermined angular range. In addition, the positive direction of the X-axis is from the inside to the forward direction, and the positive direction of the Y and Z-axis is the arrow direction.

The moving part 6 is comprised with the suction head 5 at the tip end. The suction head 5 adsorbs an IC (Integrated Circuit) chip 4. In FIG. 1, the IC chip 4 is adsorbed and lifted from the chip cassette 3 of one moving part 6 (left moving part 6), while passing over the chip camera 2 ( The central moving part 6 and the state of moving on the circuit board 8 (right moving part 6) are shown simultaneously. The chip cassette 3, the chip camera 2, the circuit board 8, the place camera 10 and the like are arranged in a positional relationship as shown in FIG. 1.

The adsorption head 5 is configured with an adsorption surface 5a made of, for example, a porous metal. The suction head 5 attaches the IC chip 4 to the suction surface 5a using a negative pressure. In addition, the suction head 5 is heated with the IC chip 4 on the circuit board 8, and a plurality of solder bumps of the IC chip 4 are thermally compressed to a predetermined contact point on the circuit board 8 do. Further, on the suction surface 5a, a mark serving as a reference for recognizing the adhesion state of the IC chip 4 is recorded.

The chip cassette 3 is a container for storing a plurality of IC chips 4. The chip cassette 3 has an open top surface (ie, Z-axis upward).

The IC chip 4 is a semiconductor integrated circuit chip. In this embodiment, the IC chip 4 has a structure called WLCSP (Wafer Level Chip Size (or Scale) Package). A plurality of solder bumps are formed on the back surface of the IC chip 4 (the downward surface of the drawing, that is, the surface opposite to the surface to which the suction head 5 is attached). The IC chip 4 is adsorbed by the suction head 5 from the state accommodated in the chip cassette 3, and is lifted upward in the Z direction. Then, the IC chip 4 is conveyed to the top of the circuit board 8 after passing through the top of the chip camera 2 in that state.

The circuit board 8 is mounted on the substrate stage 7. The substrate stage 7 is supported by a fixing table 7a so that it can be finely adjusted in the XY direction.

The place camera 10 is a camera that photographs the circuit board 8. The signal processing unit 1 performs position recognition processing, etc. around the connection portion with the IC chip 4, based on the image signal acquired from the place camera 10, and the position of the circuit board 8 is shifted from a predetermined reference. Calculate the amount of angular deviation.

The chip camera 2 is attached to the suction head 5 to capture an image of the IC chip 4 in motion, and outputs the captured image signal to the signal processing unit 1. This chip camera 2 is installed at a predetermined position between the installation location of the chip cassette 3 and the installation location of the circuit board 8. The chip camera 2 captures an image of the back surface of the IC chip 4 through the zoom lens 33. The zoom lens 33 has a plurality of lenses and a lens movement mechanism, the focal length of the lens group can be arbitrarily changed within a certain range, and a lens group that does not move the focal plane (that is, does not shift the focus) to be. The zoom lens 33 varies the focal length according to a predetermined control signal input from the outside. The signal processing unit 1 changes the focal length of the zoom lens 33 (that is, the imaging magnification of the chip camera 2) by transmitting a predetermined control signal through the cable 44. 2 is a perspective view schematically showing the positional relationship between the chip camera 2, the zoom lens 33, and the IC chip 4; In addition, the same reference numerals are used for the same configurations in each of the drawings in Figs. 1 and 2. In addition, each arrow direction of XYZ shown in FIG. 2 corresponds to each axis direction of XYZ shown in FIG. 1. In Figs. 1 and 2, the chip camera 2 and the zoom lens 33 are shown in separate form, but the chip camera 2 and the zoom lens 33 may be integrally configured.

The chip camera 2 is configured with a line sensor 21. The line sensor 21 is configured by arranging a plurality of imaging pixels made of solid-state imaging elements such as CMOS (Complementary Metal Oxide Semiconductor) and CCD (Conference of Committee on Disarmament) in a row or in a plurality of rows to ensure exposure time. . The chip camera 2 repeatedly outputs each pixel value, which is a value representing the magnitude of an electric signal obtained by photoelectrically converting incident light to each pixel constituting the line sensor 21, in a predetermined period. In addition, this line sensor 21 outputs an image signal of a monochrome multi-level gradation (ie, gray scale). Accordingly, the pixel value of each pixel output from the line sensor 21 represents a luminance value. In this embodiment, the suction head 5 is in a state in which the IC chip 4 faces the rear surface thereof, that is, the surface formed by arranging a plurality of bumps 4a downward (that is, the chip camera 2 side). It is adsorbed on the adsorption surface 5a. The chip camera 2 outputs each pixel of the line sensor 21 with the IC chip 4 moving at a predetermined speed in the direction indicated by an arrow (i.e., Y direction) as the IC chip movement direction in FIG. Is output a plurality of times by repeating at a predetermined cycle.

3 schematically shows an example of the captured image 200 by the chip camera 2. In FIG. 3, each axial direction of XYZ shown in FIG. 1 corresponds to the direction indicated by each arrow of XY and a circle chart of Z. In addition, in FIG. 3, as for the positive direction of each axis, XY is an arrow direction, and Z is a direction from front to inside. In the example shown in Fig. 3, the picked-up image 200 includes an image of the front of the rear surface of the IC chip 4 and an image of almost the entire surface of the suction surface 5a. In addition, a pair of symbols 5b and a line 5c indicating a reference region are marks actually drawn on the suction surface 5a. These marks are marks that serve as a reference for recognizing the adhered state of the IC chip 4. However, it is not necessary to provide all of the symbol 5b and the reference line 5c. In addition, in the example shown in Fig. 3, the IC chips 4 have a plurality of bumps 4a arranged at equal intervals in the XY direction.

Next, the operation of the zoom lens 33 and the like will be described with reference to FIG. 4. FIG. 4 is a schematic diagram for explaining an image captured by the line camera 2 of partial regions 201 and 202 on the back surface of the IC chip 4 shown in FIG. 3. 4A to 4C are schematic diagrams of an image captured at a certain imaging magnification A, and FIGS. 4D to 4F are images taken at a different imaging magnification B (however, B<A). It is a schematic diagram of. 4A shows the four bumps 4a and each pixel Px in the area 201 when the area 201 shown by the dotted line in FIG. 3 is imaged in the XY direction with a total of 36 pixels of 6 pixels. ) Is a diagram schematically showing the correspondence relationship. FIG. 4B is a diagram schematically showing the pixel values of each pixel Px shown in FIG. 4A. It indicates that the pixel values of each of the four pixels Px at the four corners indicated by hatching are larger than the pixel values of the other pixels Px (pixels shown in white). 4C is a diagram schematically showing the pixel values G of the six pixels Px in the second row from the bottom of FIG. 4B.

On the other hand, FIG. 4D shows four bumps 4a and each pixel in the area 202 when the area 202 shown by the chain line in FIG. 3 is imaged in the XY direction with a total of 36 pixels of 6 pixels. It is a diagram schematically showing the correspondence relationship of (Px). Further, in (d) of FIG. 4, the size of each pixel is large in correspondence with the difference in the area 201 and the area 202 (here, the area 201 <the area 202 area) ( That is, the area of the imaging area for each pixel is wider in Fig. 4(d) than in Fig. 4(a)). FIG. 4E is a diagram schematically showing the pixel values of each pixel Px shown in FIG. 4D. It is indicated that the pixel values of the nine pixels Px, which are hatched by a diagonal line rising to the right, are larger than the pixel values of the other pixels Px. In addition, the pixel values of the 11 white pixels Px in total are the lowest, and the pixel values of the total of 16 pixels Px, which are hatched by a diagonal line downward to the right, have pixel values between them. 4(f) is a diagram schematically showing the pixel values G of the 6 pixels Px in the second row from the bottom of FIG. 4E.

4A to 4C are cases in which the arrangement of the solder bumps 4a and image sampling coincide. That is, the pitch (PitchX) between the solder bumps 4a in the X direction is equal to four of the size (PixSizeX) of the imaging pixel (Px). Similarly, the pitch PitchY between the solder bumps 4a in the Y direction corresponds to the size of the image pickup pixel Px (PixSizeY). In this case, the pixel value of each pixel Px has the same value repeatedly for other bumps 4a, for example, as shown in Fig. 4C. That is, in this case, the pixel values G of the two pixels corresponding to the right bump 4a and the pixel values G of the two pixels corresponding to the left bump 4a are the same.

On the other hand, the case shown in Figs. 4D to 4F is a case where the arrangement of the solder bumps 4a and image sampling are not consistent. That is, the pitch PitchX between the solder bumps 4a in the X direction does not coincide with the length obtained by an integer multiple of the size of the imaging pixel Px (PixSizeXa). Similarly, the pitch PitchY between the solder bumps 4a in the Y direction does not coincide with the length obtained by an integer multiple of the size (PixSizeYa) of the imaging pixel Px. In this case, the pixel value of each pixel Px represents a value having a different change with respect to the different bump 4a, for example, as shown in Fig. 4F. In this case, the maximum value of the pixel value G corresponding to the right bump 4a is greater than the maximum value of the pixel value G corresponding to the left bump 4a. In addition, as compared with FIG. 4C, the variation in pixel values corresponding to each bump 4a is larger in FIG. 4F. In this case, the same phenomenon as in the case described for the subject of the present invention has been confirmed. That is, there is a possibility that the change in the pixel value of Fig. 4(f) improves the position recognition accuracy compared to Fig. 4(c).

Therefore, in the present embodiment, the signal processing unit 1 includes a length (PixSizeX or PixSizeXa or PixSizeY or PixSizeYa) according to the imaging pixel Px, and a bump 4a stored in a predetermined storage unit inside the signal processing unit 1. ), the focal length of the zoom lens 33, that is, the imaging magnification of the chip camera 2, is adjusted so that the arrangement of the bumps 4a and the image sampling are inconsistent based on the information indicating the arrangement interval PitchX or PitchY of ).

Next, with reference to FIG. 5, an example of the internal configuration of the control operation of the mounting apparatus 100 described with reference to FIG. 1 will be described. The signal processing unit 1 has a host computer 11, an input unit 12, an output unit 13, and a storage unit 14. The host computer 11 includes a main memory 11a, and the main memory 11a includes an image memory 11b. The host computer 11 has a CPU (Central Processing Unit) inside, and controls each unit by executing a predetermined program stored in the storage unit 14 or the like. The input unit 12 is an interface for inputting signals from the place camera 10 and the chip camera 2 to the host computer 11. Image data representing images captured from the place camera 10 and the chip camera 2 are input to the input unit 12.

The output section 13 is an interface for outputting signals from the host computer 11 to the motion controller 61 and the zoom lens 33. From the output unit 13, a correction amount corresponding to information (ΔX, ΔY, Δθ) indicating a correction amount for correcting and controlling the positional shift and angle shift of the moving unit 6 with respect to the motion controller 61 (or A signal indicating this)) is output. Also, a signal for changing the focal length of the zoom lens 33 is output from the output unit 13.

Here, with reference to FIG. 3, the information (ΔX, ΔY, Δθ) indicating the amount of correction for correcting the positional displacement and the angular displacement will be described. In Fig. 3, the reference point P0 is a point corresponding to the rotation center of the suction head 5 in the θ direction. The coordinate value of the reference point P0 can be calculated based on the reference symbol 5b or the reference line 5c on the suction surface 5a. The center point Pc is a point corresponding to the center point (or the center of gravity) of the IC chip 4. The coordinates of the center point Pc are the coordinate values of the plurality of bumps 4a recognized from the captured image of the chip camera 2 and the design values of the IC chip 4 (or other IC chips 4 as a reference). It can be calculated based on the measured value). ?X represents the amount of shift between the reference point P0 and the center point Pc in the X direction. ?Y represents the amount of deviation between the reference point P0 and the center point Pc in the Y direction. ?X and ?Y are calculated as deviations of the respective coordinate values of the reference point P0 and the center point Pc. Then, the good is calculated as the difference between the direction as the reference of the IC chip 4 and the actual direction. In particular, the coordinate values of the plurality of bumps 4a recognized from the captured image of the chip camera 2 and the design values of the coordinates and arrangement relationship of each bump 4a (or other IC chips 4 as a reference) It can be calculated based on the measured value).

In Fig. 5, the storage unit 14 is, for example, a non-volatile memory, and the design information 14a of the chip cassette 3, the IC chip 4, the circuit board 8, etc., and the zoom described with reference to Fig. 4 Focal length control information 14c, which is information used when indicating the focal length of the lens 33, is stored. This focal length control information 14c is configured as, for example, a table storing a correspondence relationship between the identification number of the IC chip 4 and the information indicating the specifications of the chip camera 2 and the focal length to be set.

In addition, although not shown in Fig. 1, in addition to the above configuration, the signal processing unit 1 is, for example, an interface for transmitting a predetermined control signal to the place camera 10 or the chip camera 2, or from the motion controller 61. It includes an interface for receiving a predetermined control signal. Further, the storage unit 14 also stores information necessary for controlling the motion controller 61 and the like. In addition, the design information 14a includes the external appearance of the chip cassette 3, IC chip 4, and circuit board 8, reference marks, information indicating the shape and position of each connection terminal, etc., and set values at the time of thermal compression bonding. And the like. In addition, the design information is not limited to the design value in the drawing, and may be, for example, an actual measured value of the actual IC chip 4 as a reference.

Further, the chip camera 2 has a line sensor 21, an A/D converter (analog digital converter) 22, and an output unit 23. The A/D converter 22 converts the analog pixel value of each pixel output from the line sensor 21 into a digital signal. The output unit 23 converts the digital signal sequence output from the A/D converter 22 into a digital image signal of a predetermined format and outputs it. The image signal output from the output unit 23 is directly stored in the image memory 11b via the input unit 12, for example, by a DMA (Direct Memory Access) method.

Next, an operation example of the mounting apparatus 100 described with reference to FIGS. 1 and 5 will be described with reference to FIG. 6 and the like. 6 is a flowchart showing an operation flow of the mounting device 100.

A predetermined input (not shown) of information specifying the IC chip 4 or the circuit board 8 while preparing it at a predetermined position on the chip cassette 3 or circuit board 8 accommodating the IC chip 4 After setting in the signal processing unit 1 via the apparatus, when the operator performs a predetermined instruction operation, the mounting apparatus 100 starts operation accordingly. Upon starting the operation, the signal processing unit 1 uses the design information 14a of the IC chip 4 stored in the storage unit 14 and the focal length control information ( With reference to 14c), by setting the focal length of the zoom lens 33 to a predetermined value, the imaging magnification of the chip camera 2 is finely adjusted (step S100).

Fine adjustment of the imaging magnification of the chip camera in step S100 only needs to be performed once for the first IC chip 4 when the same IC chip 4 is mounted multiple times in succession. In addition, the fine adjustment amount of the imaging magnification can be set based on the following calculation result, for example.

That is, referring to FIG. 4, the size of the pixel Px in the X direction (PixSizeXa) and the size in the Y direction (PixSizeYa) of the pixel Px after fine adjustment are the pitch (PitchX) and Y in the X direction between the solder bumps 4a. The size in the X direction (PixSizeX) and the size in the Y direction (PixSizeY) of the pixel Px at a predetermined reference magnification, which is a magnification when the pitch in the direction (PitchY) and the pitch of the bumps 4a are integer multiples of the pixel size. It can be calculated using the following equation. However, since the imaging magnification has the same value in the X and Y directions, when the arrangement of the solder bumps and the image sampling coincide in either X or Y, a fine adjustment amount is calculated for the coinciding direction.

PixSizeXa=PitchX/ax (however, ax=PitchX/PixSizeX-kx)

PixSizeYa=PitchY/ay (however, ay=PitchY/PixSizeY-ky)

Here, the value of PitchX is a relative value based on the size of PixSizeX. Further, the value of PitchY is a relative value based on the size of PixSizeY. kx and ky are constants for setting the amount to be finely adjusted, and can be set based on, for example, the number of arrangements N of the bumps 4a. That is, for example, kx=1/Nx (Nx is the number of bumps 4a in the X direction) and ky=1/Ny (Ny is the number of bumps 4a in the Y direction).

In addition, the setting values of the imaging magnification are the size in the X direction (PixSizeXa) and the size in the Y direction (PixSizeYa) of the pixel Px after fine adjustment, and the size in the X direction of the pixel (Px) at the reference magnification (PixSizeX) and Y It can be determined from the size of the direction (PixSizeY) as PixSizeXa/PixSizeX (magnification in the X direction) and PixSizeYa/PixSizeY (magnification in the Y direction). The setting value of the imaging magnification is adjusted using the calculated value of either the X direction or the Y direction when the calculated value of the magnification in the X direction and the magnification in the Y direction is different. Or, when using a line sensor, set the calculated value in the X direction as the imaging magnification, and finely adjust the PixSizeY in the Y direction by adjusting the moving speed at the time of imaging in the Y direction, and the calculated value of the magnification in the X and Y directions will be It is possible to respond in other cases.

7 shows the calculation result of the adjustment amount using the above equation. The horizontal axis is PitchX/PixSizeX (or PitchY/PixSizeY), and the vertical axis is the value of PixSizeXa (or PixSizeYa). The unit is an arbitrary unit when the pixel size PixSizeX=PixSizeY=1. In addition, the solid line shows the calculated value when the number of bumps 4a is N=2, the dotted line is the number N=4, and the chain line is the number N=6. It can be seen from FIG. 7 that if the bump spacing is 10 times or more of the pixel size, the fine adjustment amount is 5% or less.

When the processing of step S100 is completed, next, the signal processing unit 1 outputs a predetermined control signal to the motion controller 61, thereby adsorbing the moving unit 6 on the chip cassette 3 next to the IC chip ( It moves upward of 4) (step S101). Next, in accordance with the instruction from the signal processing unit 1, the motion controller 61 picks the IC chip 4 from the chip cassette 3 from the suction head 5 (step S102). Next, in accordance with an instruction from the signal processing unit 1, the motion controller 61 starts moving the moving unit 6 toward the substrate stage 7 (step S103). Here, in accordance with the instruction from the signal processing unit 1, the chip camera 2 captures an image of the IC chip 4 (step S104). Next, the signal processing unit 1 calculates the amount of positional displacement of the IC chip 4 and the like based on the image captured by the chip camera 2 (step S105).

In step S105, the signal processing unit 1 calculates, for example, the amount of positional shift or angle shift of the IC chip 4 as follows. (1) First, the signal processing unit 1 corrects the grayscale luminance or lens aberration of a predetermined pixel of the captured image captured by the chip camera 2 by referring to information for correcting the already confirmed luminance deviation or lens aberration. The image processing to be performed is executed. However, this correction process can be omitted depending on the configuration. (2) Next, the signal processing unit 1 sets a recognition target area corresponding to each bump 4a based on the design value of the IC chip 4 read out from the storage unit 14. (3) Next, the signal processing unit 1 obtains the position of the center of gravity of the luminance value from the luminance value (= pixel value) of each pixel for each recognition target region set in step S202. (4) Next, the signal processing unit 1 determines the amount of deviation in the X direction (ΔX) and the amount of deviation in the Y direction (ΔY) of the IC chip 4 based on the X and Y coordinates of each bump 4a. ) And the angle shift amount (△θ). Here, the signal processing unit 1 estimates the arrangement direction of each bump 4a using the least squares method or the like, based on the X and Y coordinate values of the plurality of bumps 4a, for example. Next, the coordinates of the center point Pc of the IC chip 4 are obtained based on the estimated arrangement direction of each bump 4a. In addition, the signal processing unit 1 recognizes the reference point 5b or the reference line 5c described with reference to FIG. 3, and becomes a reference when calculating the coordinates and angle deviation (Δθ) of the reference point P0 based on this. Find the direction. Further, the signal processing unit 1 determines the amount of displacement in the X direction (ΔX), Y based on the obtained arrangement direction of the bumps 4a, the coordinates of the center point Pc, the coordinates of the reference point P0, and the reference direction. The amount of directional deviation (ΔY) and the amount of angular deviation (Δθ) are obtained.

Next, the signal processing unit 1 outputs a predetermined instruction to the motion controller 61, and the motion controller 61 transfers the moving unit 6 to the IC chip 4 of the circuit board 8 on the substrate stage 7 ) Moves upward to the mounting position (step S106). Next, in accordance with the instruction of the signal processing unit 1, the place camera 10 captures an image of the mounting area of the IC chip 4 on the circuit board 8 (step S107). The signal processing unit 1 recognizes a shift amount or an angular shift in the XY direction of the mounting position with respect to the mounting area of the IC chip 4 with respect to the image captured by the place camera 10 (step S108).

Next, the signal processing unit 1 determines each of the moving units 6 based on the amount of positional and angular deviation of the IC chip 4 calculated in step S105 and the amount of positional displacement of the mounting area of the circuit board 8. The correction amount is calculated (step S109). The signal processing unit 1 considers the result of recognition of the circuit board 8, and then determines the position and direction based on the amount of deviation (ΔX, ΔY, ΔZ, Δθ) of the IC chip 4. A correction amount is calculated when adjusting the position and angle θ of the moving unit 6 in the XYZ direction so as to match as much as possible.

Next, the signal processing unit 1 transmits a signal indicating each correction amount determined in step S109 to the motion controller 61 (step S110).

And, after the motion controller 61 corrects the positional shift or angular shift of the moving part 6, the moving part 6 lowers the height, and the IC chip 4 is mounted on the connection part of the circuit board 8 for thermal compression bonding. (Step S111).

As described above, in the present embodiment, as the zoom lens 33 as an imaging magnification adjustment unit in which the mounting device 100 can change the imaging magnification, and an imaging unit having a plurality of imaging pixels, a plurality of bumps ( 4a) A chip camera 2 (imaging unit) for capturing an image of an IC chip 4 (object to be recognized) having (first recognition target) through a zoom lens 33 is provided. In addition, the mounting device 100 stores design information 14a indicating shape information of the IC chip 4 including at least information indicating the arrangement interval of the plurality of bumps 4a (such as PitchX in Fig. 4). A storage unit 14 is provided. In addition, the mounting device 100 is based on the information indicating the length (PixSizeX in Fig. 4, etc.) according to the imaging pixel and the arrangement interval of the bumps 4a stored in the storage unit 14 to the zoom lens 33. A signal processing unit 1 is provided that adjusts the image pickup magnification obtained by the chip camera 2 and recognizes the position of the bump 4a based on a plurality of pixel values in the image captured by the chip camera 2. Here, the signal processing unit 1 adjusts the imaging magnification so that an integer multiple of the array interval and the length according to the imaging pixel among the images captured by the chip camera 2 becomes non-corresponding. Accordingly, according to the present embodiment, by adjusting the imaging magnification for a recognition object having a shape in which the arrangement of the recognition object and the image sampling match, the arrangement of the recognition object and the image sampling can be made inconsistent. Accordingly, even under such conditions, it is possible to easily avoid a decrease in position recognition accuracy.

Further, in the present embodiment, since the lens magnification adjustment mechanism is configured to be automatically adjusted such as, for example, electric zoom, even if the layout of the target IC chip 4 changes, it is possible to automatically cope with it. However, the focal length of the zoom lens 33 may be manually set, for example.

In addition, as a method of obtaining the same effect as in this embodiment, a method of acquiring an image in a state intentionally rotated with respect to the chip camera 2 can be considered. However, it is necessary to rotate the rotation shaft before and after mounting the IC chip 4 There is an effect on the mounting time of the IC chip 4. On the other hand, if the lens magnification is finely adjusted as in the present embodiment, fine adjustment only needs to be made once before the processing is executed, and there is no influence on the mounting time of the IC chip 4.

In addition, by including lenses with different focal lengths (for example, cylindrical lenses) along the axial direction in a plurality of lens groups constituting the zoom lens 33, the magnification can be adjusted independently in the two-dimensional direction. It is possible to do.

Incidentally, in the above description, bumps arranged at the same interval are shown, but they do not necessarily have to be the same interval. In that case, for example, if the pixel size after adjustment becomes a real number rather than an integer in calculation, the deviation of the value below the decimal point is used as an objective function, and the pixel size at which the deviation becomes the maximum when the pixel size is finely adjusted. Just select

Further, in the above embodiment, the recognition target is the solder bump 4a on the IC chip 4, but the position recognition by the chip camera 2 is, for example, a semiconductor chip or other circuit mounted by packaging other than CSP. It can also be applied to parts and passive parts. In addition, it is not limited to the mounting device, and for example, it can be configured as a device used in the inspection step of the bump 4a in the step before being accommodated in the chip cassette 3. In addition, the chip camera 2 may be configured using an area sensor instead of the line sensor 21.

1 Signal processing unit 2 Chip camera (imaging unit)
3 Chip cassette 4 IC chip (object to be recognized)
4a bump (first recognition target) 5 suction head
6 Moving part 7 Board stage
8 circuit board 11 host computer
33 Zoom Lens 61 Motion Controller
100 mounting device

Claims (6)

An imaging magnification adjusting unit capable of changing the imaging magnification;
An imaging unit that has a plurality of imaging pixels and captures an image of a recognition object having a plurality of first recognition targets arranged at predetermined intervals via the imaging magnification adjustment unit;
A storage unit for storing shape information of the recognition object including at least information indicating an arrangement interval of the plurality of first recognition objects;
The imaging magnification of the imaging magnification adjustment unit is adjusted based on information indicating a length according to the imaging pixel and the arrangement interval stored in the storage unit, and based on a plurality of pixel values in the image captured by the imaging unit, the It includes; a signal processing unit for recognizing the position of the first recognition object,
The recognition apparatus, wherein the signal processing unit adjusts the imaging magnification of the imaging magnification adjusting unit so that an integer multiple of the array interval and the length corresponding to the imaging pixel among the images captured by the imaging unit becomes non-corresponding. delete The object to be recognized is a semiconductor chip,
A mounting apparatus, characterized in that the recognition object is mounted on a predetermined circuit board by performing a predetermined correction control on the position or angle of the object to be recognized based on a recognition result by the signal processing unit according to claim 1. An imaging magnification adjusting unit capable of changing the imaging magnification;
An imaging unit that has a plurality of imaging pixels and captures an image of a recognition object having a plurality of first recognition targets arranged at predetermined intervals via the imaging magnification adjustment unit;
Using a storage unit for storing shape information of the object to be recognized, including at least information indicating an arrangement interval of the plurality of first recognition objects
The imaging magnification of the imaging magnification adjustment unit is adjusted based on information indicating a length according to the imaging pixel and the arrangement interval stored in the storage unit, and based on a plurality of pixel values in the image captured by the imaging unit, the Recognize the location of the first recognition object,
The recognition method, characterized in that the imaging magnification of the imaging magnification adjustment unit is adjusted so that an integer multiple of the arrangement interval and the length corresponding to the imaging pixel among the images captured by the imaging unit becomes non-corresponding. delete The object to be recognized is a semiconductor chip,
A mounting method, characterized in that the recognition object is mounted on a predetermined circuit board by performing a predetermined correction control on the position or angle of the object to be recognized based on the recognition result according to claim 4.

Images