A method for calibrating the position between a die-bonding camera and a die-bonding head

Abstract

The present invention relates to the field of die bonding equipment and provides a position calibration method between a die bonding camera and a die bonding head, comprising placing a reflector in a die bonding area, moving the nozzle center of the die bonding head to the field of view center of the die bonding camera, and performing mark recognition; the die bonding head moves in a coordinate system lattice planned in the calibration area of ​​the die bonding, recording the coordinate position of the die bonding head and the camera coordinates, performing affine transformation on the coordinate position of the die bonding head and the camera coordinates, and solving the affine transformation matrix to obtain the virtual coordinates of the die bonding head, and then obtaining an error table Grid between the coordinate position of the die bonding head and the virtual coordinates. E ; A target position pixel coordinate P is given by the solid crystal camera W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The bilinear interpolation calculation error v is used to obtain the actual target position P of the die bonding head. B The present invention utilizes affine transformation and bilinear interpolation to calibrate the position, thereby effectively improving the precision of the die bonding.

Description

Technical Field

[0001] The present invention relates to the field of die bonding equipment, and in particular to a method for calibrating the position between a die bonding camera and a die bonding head. Background Art

[0002] In die bonding equipment, in order to improve the production efficiency of the equipment, the XY movement of the die bonding head and the Y movement of the camera on the die bonding machine (the die bonding machine only moves in the Y direction) are independent of each other, so it is often necessary to calibrate the positional relationship between the two coordinate systems in the die bonding area to convert between the coordinates. The existing method is to first calibrate the pixels of the camera on the die bonding machine, and then align the center of the camera's field of view with the center of the die bonding machine head at three or more locations of the die bonding machine position. The base coordinates of the pixel coordinates of the camera's field of view and the XY coordinates of the die bonding machine head are transformed by linear interpolation. Although this method is fast, it does not take into account the mechanical errors of the XY motion platform of the die bonding machine head and the Y motion platform of the camera on the die bonding machine and the positional relationship between the two coordinate systems, and is not suitable for high-precision die bonding operations. Summary of the Invention

[0003] The problem solved by the present invention is how to provide a position calibration method that utilizes the XY matrix of the crystal bonding head to perform point movement in the crystal bonding area, captures the position through the crystal bonding camera, and calibrates the coordinates of the crystal bonding head and the pixel coordinates of the crystal bonding camera through affine transformation and bilinear interpolation, thereby effectively improving the accuracy of crystal bonding.

[0004] To solve the above problems, the present invention provides a position calibration method between a die bonding camera and a die bonding head, comprising: placing a reflector in the die bonding area, adjusting the intensity of the camera light source, moving the nozzle center of the die bonding head to the center of the field of view of the die bonding camera, and performing mark recognition on the nozzle center;

[0005] A coordinate system matrix consisting of N rows and M columns is planned in the calibration area of ​​the die bonding. The die bonding head moves in the coordinate system matrix and the coordinate position of the die bonding head is recorded for each movement. And the camera coordinates of the mark at the center of the nozzle captured by the solid crystal camera

[0006] The coordinate position of the die bonding head generated by each movement of the die bonding head and camera coordinates Perform affine transformation to solve the affine transformation matrix Get the transformation relationship between coordinates;

[0007] According to the camera coordinates of the Mark at the center of the nozzle and the affine transformation matrix Obtain the virtual coordinates of the die bonding head Then find the coordinate position of the solid crystal head With virtual coordinates Error table Grid E ;

[0008] The solid crystal camera guides a target position pixel coordinate P W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The error v is calculated by bilinear interpolation, and then the actual target position P of the die bonding head is obtained. B .

[0009] Furthermore, the row and column spacing of the coordinate system dot matrix consisting of N rows and M columns is 0.5 mm.

[0010] Furthermore, the coordinate position of the bonding head generated by each movement of the bonding head and camera coordinates Perform affine transformation to solve the affine transformation matrix include:

[0011] XY coordinates of the die bonding head at each point Perform calculations and records to form a grid of the solid crystal head coordinate system B (N*M), the position of the light spot of the die bonding head for each movement Identify and record to form a world coordinate grid guided by the die-bonding camera W (N*M);

[0012] According to the coordinate system of the die bonding head Grid B (N*M) and world coordinate grid W (N*M), calculate the affine transformation matrix

[0013] Furthermore, the lattice grid of the solid crystal head coordinate system B (N*M) and world coordinate grid W (N*M), calculate the affine transformation matrix include:

[0014] Grid of the solid crystal head coordinate system B Each set of bonding head coordinate positions in (N*M) And the corresponding world coordinate lattice Grid W Camera coordinates in (N*M) Establish the following relationship:

[0015]

[0016] The integration is expanded to get:

[0017]

[0018] The typical thread least squares form is formed, that is, B = Ax; by x = (A T A) -1 A T B, find the Grid B (N*M) and Grid W Affine transformation matrix between (N*M) in:

[0019]

[0020] Furthermore, the camera coordinates of the Mark at the center of the nozzle and the affine transformation matrix Obtain the virtual coordinates of the die bonding head In the virtual coordinate The expression is:

[0021]

[0022] Furthermore, the coordinate position of the bonding head is obtained With virtual coordinates Error table Grid E In the error table Grid E The expression is:

[0023]

[0024] Furthermore, the solid crystal camera guides a target position pixel coordinate P W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The error v is calculated by bilinear interpolation, and then the actual target position P of the die bonding head is obtained. B ,include:

[0025] Given the target position pixel coordinate P W =(ξ,η);

[0026] Obtain the target virtual coordinates P through reflection transformation A , the expression is

[0027] According to the error table Grid E Bilinear interpolation calculation error v, where error v = (Δx, Δy);

[0028] Obtain the actual target position P B , the expression is:

[0029] Furthermore, the error table Grid E , get the target position pixel coordinate P W The four adjacent points and the corresponding error v are:

[0030]

[0031] According to bilinear interpolation, we get:

[0032]

[0033] in:

[0034]

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The position calibration method of the present invention fully considers the mechanical errors of the verticality and straightness of the crystal bonding head movement and the positional relationship (rotation, scale) between the two coordinate systems. On the basis of using affine transformation to establish linear transformations of verticality, rotation and scale, an error table is introduced to perform bilinear interpolation fitting on nonlinear errors such as the straightness of the axis itself, thereby further improving the crystal bonding accuracy of the crystal bonding machine. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the overall process of an embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of the position structure of the die bonding camera and the die bonding head according to an embodiment of the present invention;

[0039] Figure 3 Schematic diagram of the principle structure of the die bonding head coordinate system lattice according to an embodiment of the present invention;

[0040] Figure 4 Schematic diagram of the principle structure of the world coordinate lattice guided by the die bonding camera according to an embodiment of the present invention;

[0041] Figure 5 Schematic diagram of the linear mechanical error between the two axis systems of the die bonding head and the die bonding camera according to an embodiment of the present invention;

[0042] Figure 6 Schematic diagram of nonlinear mechanical error between a die bonding head and a die bonding camera according to an embodiment of the present invention;

[0043] Figure 7 Schematic diagram of the principle structure of bilinear interpolation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0045] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood broadly. For example, they may refer to fixed, detachable, or integral connections; mechanical connections; direct connections or indirect connections via an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0046] Throughout this specification, references to the terms "embodiment," "one embodiment," and "an implementation" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or implementation are included in at least one embodiment or implementation of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or implementation. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or implementations.

[0047] like Figure 1 As shown, the present invention provides a position calibration method between a die bonding camera and a die bonding head, comprising: placing a reflector in a die bonding area, adjusting the intensity of the camera light source, moving the nozzle center of the die bonding head to the center of the field of view of the die bonding camera, and performing mark recognition on the nozzle center;

[0048] A coordinate system matrix consisting of N rows and M columns is planned in the calibration area of ​​the die bonding. The die bonding head moves in the coordinate system matrix and the coordinate position of the die bonding head is recorded for each movement. And the camera coordinates of the mark at the center of the nozzle captured by the solid crystal camera

[0049] The coordinate position of the die bonding head generated by each movement of the die bonding head and camera coordinates Perform affine transformation to solve the affine transformation matrix Get the transformation relationship between coordinates;

[0050] According to the camera coordinates of the Mark at the center of the nozzle and the affine transformation matrix Obtain the virtual coordinates of the die bonding head Then find the coordinate position of the solid crystal head With virtual coordinates Error table Grid E ;

[0051] The solid crystal camera guides a target position pixel coordinate P W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The error v is calculated by bilinear interpolation, and then the actual target position P of the die bonding head is obtained. B .

[0052] Specifically:

[0053] S1: As attached Figure 2 As shown, a flat reflective lens is placed in the die bonding area. This allows light from the die bonding camera's light source to be reflected by the reflective lens through a long through-hole in the die bonding head, perpendicular to the reflective mirror surface, forming a circular light spot with sharp edges within the die bonding camera's field of view. The die bonding head's XY coordinates are then moved to align the light spot with the center of the die bonding camera's field of view, and mark learning and recognition of the light spot is performed.

[0054] S2: As attached Figure 3 As shown in the figure, the bonding area is planned to consist of N rows and M columns with a spacing of 0.5mm for the XY motion of the bonding head. The XY coordinates of the bonding head at each point are Perform calculations and records to form a grid of the solid crystal head coordinate system B (N*M), as attached Figure 4 As shown. The position of the light spot of the die bonding head for each movement Identify and record to form a world coordinate grid guided by the die-bonding camera W (N*M).

[0055] S3: For mechanical errors such as perpendicularity, rotation between axis systems and scale ratio differences, see the attached Figure 5 As shown, it can be converted through a linear relationship, that is, we can use affine transformation to express the relationship between the verticality, rotation and scale ratio between the two axis systems. The dot matrix Grid obtained by step S2 B (N*M) and Grid W (N*M), we can calculate the transformation matrix R of the affine transformation.

[0056]

[0057] Grid B (N*M) and Grid W Each group in (N*M) The following relationships can be established:

[0058]

[0059] Integration and expansion can be obtained:

[0060]

[0061] Become the form of the typical thread least squares, B = Ax, which can be obtained by x = (A T A) -1 A T B, find the Grid B (N*M) and Grid W Affine transformation matrix between (N*M)

[0062] S4: Since the axis itself has mechanical error in straightness, and the straightness error is nonlinear, as shown in the attached Figure 6 Therefore, we use the affine transformation matrix Calculated

[0063] The actual movement position of the die bonding head There are errors, such as Figure 6 As shown. The straightness error table is Grid E , record each original data and the error v after affine transformation ij ,Right now

[0064]

[0065] S5: Straightness error can be solved by bilinear interpolation method, as shown in the attached Figure 7 As shown, given a target position P W =(ξ,η), find the actual P B =(x,y),

[0066]

[0067] The error v = (Δx, Δy), which is represented by the error table Grid E , get the 4 adjacent points and the corresponding error v:

[0068]

[0069] Bilinear interpolation yields:

[0070]

[0071] in:

[0072]

[0073] Finally, the actual position is obtained.

[0074] Although the present disclosure is disclosed as above, the protection scope of the present disclosure is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present disclosure, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A method for calibrating the position between a die bonding camera and a die bonding head, characterized in that: include: Place the reflector in the die bonding area, adjust the camera light intensity, move the nozzle center of the die bonding head to the center of the die bonding camera's field of view, and mark the nozzle center. A coordinate system matrix consisting of N rows and M columns is planned in the calibration area of ​​the die bonding. The die bonding head moves in the coordinate system matrix and the coordinate position of the die bonding head is recorded for each movement. And the camera coordinates of the mark at the center of the nozzle captured by the solid crystal camera The coordinate position of the die bonding head generated by each movement of the die bonding head and camera coordinates Perform affine transformation to solve the affine transformation matrix Get the transformation relationship between coordinates; According to the camera coordinates of the Mark at the center of the nozzle and the affine transformation matrix Obtain the virtual coordinates of the die bonding head Then find the coordinate position of the solid crystal head With virtual coordinates Error table Grid E ; The solid crystal camera guides a target position pixel coordinate P W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The error v is calculated by bilinear interpolation, and then the actual target position P of the die bonding head is obtained. B .

2. The position calibration method between the die bonding camera and the die bonding head according to claim 1, characterized in that: The row and column spacing of the coordinate system dot matrix consisting of N rows and M columns is 0.5 mm.

3. The position calibration method between the die bonding camera and the die bonding head according to claim 1, characterized in that: The coordinate position of the bonding head generated by each movement of the bonding head and camera coordinates Perform affine transformation to solve the affine transformation matrix include: The coordinate position of the die bonding head at each point Perform calculations and records to form a grid of the solid crystal head coordinate system B (N*M), the position of the light spot of the die bonding head for each movement Identify and record to form a world coordinate grid guided by the die-bonding camera W (N*M); According to the coordinate system of the die bonding head Grid B (N*M) and world coordinate grid W (N*M), calculate the affine transformation matrix 4. The position calibration method between the die bonding camera and the die bonding head according to claim 3, characterized in that: According to the solid crystal head coordinate system lattice Grid B (N*M) and world coordinate grid W (N*M), calculate the affine transformation matrix include: Grid of the solid crystal head coordinate system B Each set of bonding head coordinate positions in (N*M) and the corresponding world coordinate lattice Grid W Camera coordinates in (N*M) Establish the following relationship: The integration is expanded to get: The typical thread least squares form is formed, that is, B = Ax; by x = (A T A) -1 A T B, find the Grid B (N*M) and Grid W Affine transformation matrix between (N*M) in:

5. The position calibration method between the die bonding camera and the die bonding head according to claim 1, characterized in that: The camera coordinates of the Mark based on the center of the nozzle and the affine transformation matrix Obtain the virtual coordinates of the die bonding head In the virtual coordinate The expression is:

6. The position calibration method between the die bonding camera and the die bonding head according to claim 5, characterized in that: Then the coordinate position of the solid crystal head is obtained With virtual coordinates Error table Grid E In the error table Grid E The expression is:

7. The position calibration method between the die bonding camera and the die bonding head according to claim 6, characterized in that: The solid crystal camera guides a target position pixel coordinate P W , obtain the target virtual coordinate P through reflection transformation A , and then according to the error table Grid E The error v is calculated by bilinear interpolation, and then the actual target position P of the die bonding head is obtained. B ,include: Given the target position pixel coordinate P W =(ξ,η); Obtain the target virtual coordinates P through reflection transformation A , the expression is According to the error table Grid E The error v is calculated by bilinear interpolation, where error v = (Δx, Δy); Obtain the actual target position P B , the expression is:

8. The position calibration method between the die bonding camera and the die bonding head according to claim 7, characterized in that: According to the error table Grid E The bilinear interpolation calculation error v includes: From the error table Grid E , get the target position pixel coordinate P W The coordinates of the four adjacent points and the corresponding error v are: According to bilinear interpolation, we get: in:

Images