Automatic imaging method of scribing machine

Abstract

The invention relates to the field of machine vision, in particular to an automatic imaging method of a scribing machine, which comprises the following steps of: constructing a relational expression among camera height, image brightness and light source brightness by adopting a Bessel curved surface; the position with the highest definition of the obtained image is used as a focusing position; the corresponding light source brightness is obtained through a Bessel curved surface relational expression; and adjusting the light source brightness, calculating the image contrast, and setting the corresponding light source brightness as the imaging light source brightness when the contrast is within the standard range. The accurate focusing position is obtained through automatic focusing, a clear image can be automatically obtained, parameters do not need to be manually set, the Bessel curved surface is adopted to construct the relational expression of the height, the image brightness and the light source brightness, the corresponding light source brightness is obtained through the Bessel curved surface relational expression according to the camera height and the set standard brightness, and the image brightness is accurately obtained. And images with appropriate brightness and contrast are obtained through automatic brightness adjustment, and the method has good adaptability to wafer products with different specifications and thicknesses and large brightness and darkness differences.

Description

Technical Field

[0001] The present invention relates to the field of machine vision, and in particular to a method for automatic imaging of a dicing machine. Background Art

[0002] As an important part of the dicing machine equipment, the vision system is equivalent to the eyes of the entire equipment, and the quality of imaging is the basis for the operation of the vision system, directly affecting subsequent alignment, cutting, knife mark detection, etc.

[0003] The three elements of the traditional imaging method are as Figure 1 shown, using a camera at a fixed position and a fixed light intensity. However, when the height of the detection target changes or is inconsistent, situations such as defocusing are likely to occur, especially when detecting precision micro-targets such as wafers. At the same time, it is difficult to simultaneously satisfy the brightness and contrast of the detected target image.

[0004] In view of this, we propose a method for automatic imaging of a dicing machine to solve the existing problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for automatic imaging of a dicing machine to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A method for automatic imaging of a dicing machine, the steps include:

[0007] S1: The camera moves up and down along the Z-axis direction, and a relational expression of height, image brightness, and light source brightness is constructed using a Bezier surface;

[0008] S2: The camera moves up and down along the Z-axis direction through a first stepping, continuously captures images, and obtains the position with the highest image clarity;

[0009] S3: Select a position range according to the position with the highest image clarity obtained by the first stepping, move within the selected position range through a second stepping, and use the position with the highest image clarity obtained by the second stepping as the focusing position;

[0010] S4: According to the height obtained in S3 and the set standard brightness, the corresponding light source brightness is obtained through the Bezier surface relational expression constructed in S1;

[0011] S5: After adjusting the light source brightness to the value obtained in S4 through an instruction sent by the PC side, record the image and calculate the image contrast;

[0012] S6: If the contrast of the image is within the set standard contrast range, set the corresponding light source brightness as the light source brightness for imaging; if the contrast of the image is not within the set standard contrast range, obtain the light source brightness adjustment range according to the Bessel surface relational expression constructed by the set standard brightness range and S1, and adjust the light source brightness in set steps accordingly. Record the image and calculate the contrast of the image each time the light source brightness is adjusted until the contrast of the image is within the set standard contrast range, then set the corresponding light source brightness as the light source brightness for imaging.

[0013] Further, the specific steps of S1 include:

[0014] A1: Select (N + 1) shooting heights and (M + 1) image brightnesses. The shooting heights are H0, H1, H2, ..., H N , and the image brightnesses are I0, I1, I2, ..., I M ;

[0015] A2: Adjust the shooting height of the camera. When the shooting height of the camera is H0, adjust the light source brightness, record the image and calculate the image brightness, and record the corresponding light source brightnesses of the image brightnesses I0, I1, I2, ..., I M . Similarly, record the corresponding light source brightnesses of the image brightnesses I0, I1, I2, ..., I N when the shooting heights of the camera are H1, H2, ..., H M ;

[0016] A3: Obtain the Bessel surface control points according to H0, H1, H2, ..., H N and I0, I1, I2, ..., I M and the corresponding light source brightnesses, and construct the Bessel surface relational expression.

[0017] Further, in A2, convert the image from the RGB color space to the Lab color space, and use the value corresponding to the L dimension as the evaluation value of the image brightness.

[0018] Further, in S5 or S6, use the standard deviation of the image brightness as the evaluation value of the image contrast.

[0019] Further, in S2, the first step size is 1; in S3, the second step size is 0.2.

[0020] Further, in S2 or S3, use the Laplacian algorithm to evaluate the image sharpness.

[0021] Further, in S3, take the position with the highest image sharpness obtained in the first step as the center point, and set the upper top end and the lower bottom end of the position range.

[0022] Furthermore, the distance between the upper top end and the center point is 2, and the distance between the lower bottom end and the center point is also 2.

[0023] Furthermore, dual camera lenses are adopted for acquisition, a high-precision Z-axis lifting mechanism is used to adjust the position of the cameras, and an adjustable light source controller is used to adjust the light source brightness.

[0024] Furthermore, the adjustable light source controller is connected to the PC end through the RS-232 interface. The PC end gives a dimming command, and the light source is composed of a combination of a ring light source and a point light.

[0025] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0026] The present invention obtains an accurate focusing position through automatic focusing, can automatically obtain a clear image without manual parameter setting, constructs a relational expression between height, image brightness, and light source brightness using a Bessel surface, and obtains the corresponding light source brightness according to the camera height and the set standard brightness through the Bessel surface relational expression. Automatic brightness adjustment is used to obtain an image with appropriate brightness and contrast, and it has good adaptability to wafer products with different specifications, thicknesses, and large brightness differences. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a schematic diagram of a traditional imaging method in the background art;

[0028] Figure 2 is a schematic diagram of the vision mechanism of the present invention;

[0029] Figure 3 is a schematic diagram of the dual camera lenses of the present invention;

[0030] Figure 4 is a schematic diagram of the method for automatic focusing and automatic brightness adjustment of the present invention.

[0031] In the figure: 1. Camera lens; 2. Light source; 3. Fixed bracket DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] The following further describes the technical solutions of the present invention with reference to the accompanying drawings and specific embodiments.

[0033] Embodiment 1

[0034] As Figure 2 shown, the vision mechanism includes a camera lens, a light source, and a fixed bracket.

[0035] As Figure 3 shown, the camera lens adopts dual camera lenses, that is, lenses with different magnifications are used. The advantage of the low-magnification lens is a large field of view; the advantage of the high-magnification lens is a high magnification and the ability to observe more microscopic details.

[0036] The high-precision Z-axis lifting mechanism can accurately move the camera to the corresponding focusing position.

[0037] The adjustable light source controller is connected to the PC through the RS-232 interface. The PC sends dimming commands. The light source consists of a ring light and a point light. The point light focuses on brightness adjustment, while the ring light focuses on making the texture boundaries of objects more obvious. The highly integrated light source controller can control both the ring light and the point light simultaneously to save space.

[0038] A method for automatic imaging of a dicing machine requires the combination of software and hardware. In terms of software, OpenCV is adopted. OpenCV is not only open-source and free but also has a rich visual algorithm library.

[0039] A method for automatic imaging of a dicing machine first performs automatic focusing and then automatic light adjustment. The steps include:

[0040] S1: The camera moves up and down along the Z-axis direction, and a relationship between height, image brightness, and light source brightness is constructed using a Bezier surface.

[0041] S2: The camera moves up and down along the Z-axis direction in large steps of 1, continuously captures images, and uses the Laplacian algorithm to evaluate the image sharpness to obtain the position with the highest image sharpness.

[0042] S3: Taking the position with the highest image sharpness obtained by large steps of 1 as the center point, set the upper and lower ends of the position range. The distance from the upper end to the center point is 2, and the distance from the lower end to the center point is also 2. According to the selected position range, move within the selected position range in small steps of 0.2, and take the position with the highest image sharpness obtained by small steps of 0.2 as the focusing position, which can not only obtain an accurate focusing position but also prevent the moving speed from being too slow.

[0043] S4: According to the height obtained in S3 and the set standard brightness, the corresponding light source brightness is obtained through the Bezier surface relationship constructed in S1.

[0044] S5: After adjusting the light source brightness to the value obtained in S4 through the instruction sent by the PC, record the image and calculate the image contrast; among them, the standard deviation of the image brightness is used as the evaluation value of the image contrast.

[0045] S6: If the contrast of the image is within the set standard contrast range, the corresponding light source brightness is set to the imaging light source brightness; if the contrast of the image is not within the set standard contrast range, the light source brightness adjustment range is calculated according to the set standard brightness range and the Bezier surface relationship constructed in S1, and the light source brightness is adjusted according to the set step. After each adjustment of the light source brightness, the image is recorded and the image contrast is calculated until the image contrast is within the set standard contrast range, and the corresponding light source brightness is set to the imaging light source brightness.

[0046] Among them, the specific steps of S1 include:

[0047] A1: Select (N+1) shooting heights and (M+1) image brightnesses, where the shooting heights are H0, H1, H2, ..., H N , the image brightness is I0, I1, I2, ..., I M ;

[0048] A2: Adjust the shooting height of the camera. When the shooting height of the camera is H0, adjust the brightness of the light source, record the image and calculate the image brightness. The recorded image brightness is I0, I1, I2, ..., I M The corresponding light source brightness, similarly, the camera shooting height is recorded as H1, H2, ..., H N When the image brightness is I0, I1, I2, ..., I M The corresponding light source brightness; wherein, the image is converted from the RGB color space to the Lab color space, and the value corresponding to the L dimension is used as the evaluation value of the image brightness.

[0049] A3: According to H0, H1, H2, ..., H N and I0, I1, I2, ..., I M And the corresponding light source brightness is used to obtain the Bezier surface control points and construct the Bezier surface relationship.

[0050] The working principle of the automatic imaging method of a dicing machine based on the first embodiment is as follows:

[0051] The Lab color space has a wider color gamut than the RGB color space, so the Lab color space can represent all the color information that the RGB color space can describe. The conversion from RGB color space to Lab color space cannot be done directly. It requires the help of XYZ color space, converting the RGB color space to XYZ color space, and then converting the XYZ color space to Lab color space.

[0052] The conversion relationship from RGB color space to XYZ color space is:

[0053] [X,Y,Z] T= [T]·[R,G,B] T

[0054] where: [T] is a 3×3 matrix:

[0055]

[0056] The conversion relationship formula for converting the XYZ color space to the Lab color space is:

[0057]

[0058] where: X n , Y n , Z n are 0.950456, 1.0, and 1.088754 respectively; the function f is a correction function similar to the Gamma function, when t > 0.008856, when t ≤ 0.008856,

[0059] In the Lab color space, a and b represent chromaticity, and L represents luminance.

[0060] One way to measure contrast is to use the standard deviation of the image luminance. Suppose the image consists of n×m pixels, and the luminance value corresponding to the x×y pixel is I(x,y). The formula for the mean of the image luminance is as follows:

[0061]

[0062] The formula for the standard deviation of the image luminance is as follows:

[0063]

[0064] The larger the standard deviation, the greater the change in the image luminance and the higher the contrast.

[0065] The mathematical model of the N×M-degree Bessel surface is:

[0066]

[0067] where: and are Bessel basis functions; p i,j is the control point of the Bessel surface; u and v are the normalized values of the abscissa and ordinate before the transformation of the Bessel surface respectively; P(u,v) is the value of the Bessel surface after the transformation at the corresponding coordinate point.

[0068] Suppose H S is the set of H0, H1, H2,..., H N and suppose I S is the set of I0, I1, I2,..., IM The set, and the corresponding light source brightness set is L S , then B(H S )·B(I S )·P = L S , where B(H S ) is the Bessel basis function of H S , B(I S ) is the Bessel basis function of I S , and P is the set of Bessel curve control points, and P is obtained in this way.

[0069] Let the height of the focusing position be H, and the set standard brightness be I. Then the corresponding light source brightness is L = B(H)·B(I)·P, and the corresponding image contrast σ is obtained.

[0070] The set standard contrast range is [σ min , σ max , and the set standard brightness range is [I min , I max . If σ ∈ [σ min , σ max , then L is set as the light source brightness for imaging; if then the light source brightness is adjusted within the range of [L min , L max until the corresponding σ ∈ [σ min , σ max , and then the corresponding L is set as the light source brightness for imaging; where, L min = B(H)·B(I min )·P, L max = B(H)·B(I max )·P.

[0071] The above specific embodiments are only several preferred embodiments of the present invention. Based on the technical solution of the present invention and the relevant revelations of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A method for automatic imaging of a dicing machine, characterized by the steps Including: S1: The camera moves up and down along the Z-axis direction, and uses a Bezier surface to construct the relationship between height, image brightness, and light source brightness. S2: The camera moves up and down along the Z-axis direction through a first step, continuously captures images, and obtains the position with the highest image clarity. S3: Select a position range based on the position with the highest image clarity obtained by the first step, move within the selected position range with a second step, and use the position with the highest image clarity obtained by the second step as the focusing position. S4: According to the height obtained in S3 and the set standard brightness, obtain the corresponding light source brightness through the Bezier surface relationship constructed in S1. S5: After adjusting the light source brightness to the value obtained in S4 through the instruction sent by the PC side, record the image and calculate the image contrast. S6: If the contrast of the image is within the set standard contrast range, set the corresponding light source brightness as the light source brightness for imaging; if the contrast of the image is not within the set standard contrast range, obtain the light source brightness adjustment range according to the set standard brightness range and the Bezier surface relationship constructed in S1, and adjust the light source brightness according to the set step. After each adjustment of the light source brightness, record the image and calculate the image contrast until the contrast of the image is within the set standard contrast range, and then set the corresponding light source brightness as the light source brightness for imaging.

2. The method for automatic imaging of a dicing machine according to claim 1, wherein The specific steps of S1 include: A1: Select (N + 1) shooting heights and (M + 1) image brightness levels. The shooting heights are H0, H1, H2, ..., H N , and the image brightness levels are I0, I1, I2, ..., I M ; A2: Adjust the shooting height of the camera. When the shooting height of the camera is H0, adjust the light source brightness, record the image and calculate the image brightness, and record the image brightness as I0, I1, I2, ..., I M The corresponding light source brightness. Similarly, record the image brightness as I0, I1, I2, ..., I when the shooting heights of the camera are H1, H2, ..., H N ; M The corresponding light source brightness; A3: Obtain the control points of the Bezier surface based on H0, H1, H2, ..., H N and I0, I1, I2, ..., I M and construct the Bezier surface relation according to the corresponding light source brightness.

3. The method for automatic imaging of a dicing machine according to claim 2, characterized in that: In A2, the image is converted from the RGB color space to the Lab color space, and the value corresponding to the L dimension is used as the evaluation value of the image brightness.

4. The method for automatic imaging of a dicing machine according to claim 1, characterized in that: In S5 or S6, the standard deviation of the image brightness is used as the evaluation value of the image contrast.

5. The method for automatic imaging of a dicing machine according to claim 1, characterized in that: In S2, the first step is 1; in S3, the second step is 0.

2.

6. The method for automatic imaging of a dicing machine according to claim 1, characterized in that: In S2 or S3, the Laplacian algorithm is used to evaluate the image clarity.

7. The method for automatic imaging of a dicing machine according to claim 1, characterized in that: In S3, taking the position with the highest image clarity obtained by the first step as the center point, set the upper top end and the lower bottom end of the position range.

8. The method for automatic imaging of a dicing machine according to claim 7, characterized in that: The distance between the upper top end and the center point is 2, and the distance between the lower bottom end and the center point is also 2.

9. The method for automatic imaging of a dicing machine according to claim 1, characterized in that: A dual-camera lens is used for acquisition, a high-precision Z-axis lifting mechanism is used to adjust the position of the camera, and an adjustable light source controller is used to adjust the light source brightness.

10. The method for automatic imaging of a dicing machine according to claim 9, characterized in that: The adjustable light source controller is connected to the PC through the RS-232 interface. The PC gives the dimming command, and the light source is composed of a ring light source combined with a point light.

Images