A chip pollutant detection device and a detection method

Abstract

The present invention discloses a chip pollutant detection device and a detection method, belonging to the technical field of chip appearance detection. Aiming at the problem in the prior art that it is difficult to distinguish between pollutants on the inner surface and the outer surface of a CMOS chip, the present invention uses two camera components in the chip pollutant detection device to simultaneously capture the CMOS chip to be detected to obtain two CMOS images to be detected at different positions. By using the pixel coordinates of the key point images of the reference plane and the key point images of the pollutants in the CMOS images to be detected, the world coordinates of the key points of the reference plane and the world coordinates of the key points of the pollutants are obtained. The world coordinates of the key points of the reference plane are fitted into a plane, and the distance from the world coordinates of the key points of the pollutants to the plane is calculated to determine whether the pollutants are inner surface pollutants or outer surface pollutants, so as to quickly distinguish between pollutants on the inner surface and the outer surface of the CMOS chip, and improve the detection efficiency and detection accuracy of pollutants on the CMOS chip.

Description

Technical Field

[0001] The present invention relates to the technical field of chip appearance detection, and more specifically, to a chip contaminant detection device and a detection method. Background Art

[0002] In the production process of CMOS chips, one of the steps requires using a glass layer to insulate and isolate devices and the first layer of metal, so as to isolate contaminants and ensure the stable performance of CMOS chip products. However, before and after depositing the glass, contaminants may adhere to the CMOS chips, thus affecting the performance of CMOS chips. The contaminants on the outer surface of the glass can be removed after wiping, while the contaminants on the inner surface of the glass cannot be removed. Therefore, it is very necessary to accurately distinguish the inner surface contaminants and the outer surface contaminants of CMOS chips.

[0003] In the prior art, the methods for distinguishing the inner surface contaminants and the outer surface contaminants of CMOS chips are as follows: (1) Manual detection. Manual detection is to observe whether there are other contaminants under the glass after the production line employees wipe the dirt on the outer surface of the CMOS chip under a high-power microscope. However, this method has low efficiency, high cost, and is also prone to missed and misjudged due to employee fatigue and slack. (2) 2D appearance detection. Through 2D appearance detection, the contaminants on the surface of the CMOS chip can be directly detected, but it cannot distinguish whether the contaminants are on the outer surface or the inner surface of the CMOS chip. (3) Laser three-dimensional projection technology. The laser three-dimensional projection technology requires an additional moving platform to cooperate with laser scanning for detecting the contaminants on the surface of the CMOS chip. This method has a complex structure, high cost, and relatively low efficiency. (4) Structured light three-dimensional scanning technology. The structured light three-dimensional scanning technology requires adding a structured light projector, which has high cost, and needs to continuously capture more than a dozen or even more images and then perform three-dimensional reconstruction, with low efficiency and is not suitable for detecting this type of problem. Summary of the Invention

[0004] 1. Technical Problems to be Solved

[0005] Aiming at the problem in the prior art that it is difficult to distinguish the inner surface contaminants and the outer surface contaminants of CMOS chips, the present invention provides a chip contaminant detection device and a detection method, which can quickly distinguish the inner surface contaminants and the outer surface contaminants of CMOS chips, reduce the misjudgment rate of traditional detection schemes, and improve the detection efficiency and detection accuracy of CMOS chip contaminants.

[0006] 2. Technical Solutions

[0007] The object of the present invention is achieved by the following technical solutions.

[0008] A chip pollutant detection device includes: a first camera assembly, a second camera assembly, a light source, and a lifting stage. The light source is spaced directly above the lifting stage. The first camera assembly includes a first camera and a first telecentric lens, and one end of the first telecentric lens is connected to the first camera. The second camera assembly includes a second camera and a second telecentric lens, and one end of the second telecentric lens is connected to the second camera. The first camera assembly is spaced directly above the light source, and the second camera assembly is arranged on one side of the first camera assembly. A center line included angle is formed between the first camera assembly and the second camera assembly, and the angle of the center line included angle between the first camera assembly and the second camera assembly is set between 10° and 30°.

[0009] Further, the angle of the center line included angle between the first camera assembly and the second camera assembly is set to 15°.

[0010] A chip pollutant detection method includes the following steps:

[0011] Construct a chip pollutant detection device;

[0012] Calibrate the chip pollutant detection device to obtain calibration parameters, where the calibration parameters include a fundamental matrix, a first camera assembly projection matrix, and a second camera assembly projection matrix;

[0013] Establish an image pixel coordinate system, place the CMOS chip to be detected on the lifting stage, turn on the light source, and the first camera assembly and the second camera assembly simultaneously capture the CMOS chip to be detected to obtain a first camera assembly detection image and a second camera assembly detection image; obtain the image pixel coordinates of the first camera assembly detection image and the image pixel coordinates of the second camera assembly detection image. The image pixel coordinates of the first camera assembly detection image include reference plane key point image pixel coordinates and pollutant key point image pixel coordinates;

[0014] Establish a world coordinate system, and obtain the reference plane key point world coordinates and pollutant key point world coordinates through the relationship between the image pixel coordinates of the first camera assembly detection image and the world coordinates and the relationship between the image pixel coordinates of the second camera assembly detection image and the world coordinates;

[0015] Fit the reference plane key point world coordinates into a plane, calculate the distance from the pollutant key point world coordinates to the plane, and determine whether the pollutant is an inner surface pollutant or an outer surface pollutant.

[0016] Further, select a calibration board, fix the calibration board on the lifting stage, and the first camera assembly and the second camera assembly simultaneously capture the calibration board to obtain a first camera assembly calibration image and a second camera assembly calibration image;

[0017] Establish an image pixel coordinate system to obtain the image pixel coordinates of the calibration image of the first camera component and the image pixel coordinates of the calibration image of the second camera component;

[0018] Calculate the fundamental matrix from the image pixel coordinates of the calibration image of the first camera component and the image pixel coordinates of the calibration image of the second camera component;

[0019] Establish a world coordinate system, set the world coordinates of the calibration board, and calculate the projection matrix of the first camera component from the relationship between the image pixel coordinates of the calibration image of the first camera component and the world coordinates of the calibration board, and calculate the projection matrix of the second camera component from the relationship between the image pixel coordinates of the calibration image of the second camera component and the world coordinates of the calibration board.

[0020] Further, obtain the relationship between the image pixel coordinates and the world coordinates of the detection image of the first camera component through the projection matrix of the first camera component, and obtain the relationship between the image pixel coordinates and the world coordinates of the detection image of the second camera component through the projection matrix of the second camera component.

[0021] Further, obtain the world coordinates of the key points on the reference plane and the world coordinates of the key points of the pollutant by using the relationship between the image pixel coordinates and the world coordinates of the detection image of the first camera component and the relationship between the image pixel coordinates and the world coordinates of the detection image of the second camera component.

[0022] Further, the relationship between the image pixel coordinates and the world coordinates of the detection image of the first camera component is expressed as:

[0023]

[0024] where, P1 represents the projection matrix of the first camera component, u1 represents the horizontal coordinate of the image pixel of the detection image of the first camera component, v1 represents the vertical coordinate of the image pixel of the detection image of the first camera component, X w represents the x-axis coordinate of the key points on the reference plane and the key points of the pollutant in the world coordinate system, Y w represents the y-axis coordinate of the key points on the reference plane and the key points of the pollutant in the world coordinate system, Z w represents the z-axis coordinate of the key points on the reference plane and the key points of the pollutant in the world coordinate system;

[0025] The relationship between the image pixel coordinates and the world coordinates of the detection image of the second camera component is expressed as:

[0026]

[0027] where, P2 represents the projection matrix of the second camera component, u2 represents the horizontal coordinate of the image pixel of the detection image of the second camera component, v2 represents the vertical coordinate of the image pixel of the detection image of the second camera component.

[0028] Further, the calculation formula for fitting the world coordinates of the key points on the reference plane into a plane is:

[0029] Ax + By + Cz + D = 0

[0030] Among them, A represents the component of the plane normal vector on the x-axis, B represents the component of the plane normal vector on the y-axis, C represents the component of the plane normal vector on the z-axis, D represents the distance between the plane and the origin, x represents the x-coordinate of any point on the plane, y represents the y-coordinate of any point on the plane, and z represents the z-coordinate of any point on the plane.

[0031] Further, the calculation formula for the distance from the world coordinates of the pollutant key points to the plane is:

[0032]

[0033] Among them, d represents the distance from the world coordinates of the pollutant key points to the plane, x0 represents the x-coordinate of the pollutant key points, y0 represents the y-coordinate of the pollutant key points, and z0 represents the z-coordinate of the pollutant key points.

[0034] Further, set the system error parameter, and judge whether the pollutant is an inner surface pollutant or an outer surface pollutant through the system error parameter.

[0035] 3. Beneficial effects

[0036] Compared with the prior art, the advantages of the present invention are:

[0037] (1) For a chip pollutant detection device and detection method of the present invention, by constructing a CMOS chip pollutant detection device, it is not necessary to perform three-dimensional reconstruction on the entire CMOS chip product. And compared with ordinary lenses, the telecentric lens used in the present invention has less distortion, the imaging of small dirt is clearer, the focal length is fixed, which can effectively reduce the focusing error, the depth of field is large, and the imaging of the tilted camera is still clear. For a tilted camera, the available field of view is larger, which is beneficial to improving the accuracy and detection efficiency.

[0038] (2) For a chip pollutant detection device and detection method of the present invention, only by calculating the three-dimensional distance from the pollutant to the reference plane through the three-dimensional coordinates of the reference plane and the pollutant key points, it is possible to accurately distinguish whether the pollutant is an inner surface pollutant or an outer surface pollutant, effectively improving the detection efficiency of CMOS chip pollutants.

[0039] (3) The chip contaminant detection device and detection method of the present invention can perform all-round detection on CMOS chips at different angles and positions, improve the detection accuracy. By taking CMOS chip images from different perspectives and using image processing technologies such as edge detection, feature extraction, and image registration, the acquired CMOS chip image data is processed and analyzed, and the parallax between two CMOS chip images is calculated to achieve three-dimensional reconstruction of key points, thereby effectively realizing the detection accuracy of CMOS chip contaminants.

[0040] (4) The chip contaminant detection device and detection method of the present invention utilize a high-performance image processing system and fast feedback control technology to achieve real-time processing of a large amount of image data and stain detection, can promptly discover problems, and in combination with an automated control system, achieve automatic detection and processing of contaminants on the production line, improve production efficiency and quality management level, and thus better meet the requirements of modern industrial production for quality control, with strong practicability and wide applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic structural diagram of the chip contaminant detection device according to an embodiment of the present invention;

[0042] Figure 2 It is a schematic flow diagram of the chip contaminant detection method according to an embodiment of the present invention.

[0043] Explanation of the reference numerals in the drawings: 1. First camera assembly; 11. First camera; 12. First telecentric lens; 2. Second camera assembly; 21. Second camera; 22. Second telecentric lens; 3. Light source; 4. Lifting stage. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The present invention will be described in detail below with reference to the accompanying drawings of the specification and specific embodiments.

[0045] Embodiment

[0046] As Figure 1 shown, a chip contaminant detection device provided in this embodiment. The chip contaminant detection device provided in this embodiment includes a first camera assembly 1, a second camera assembly 2, a light source 3, and a lifting stage 4.

[0047] Specifically in this embodiment, the light source 3 is arranged at intervals directly above the lifting stage 4. The first camera assembly 1 includes a first camera 11 and a first telecentric lens 12, and one end of the first telecentric lens 12 is connected to the first camera 11 to form the first camera assembly 1. The second camera assembly 2 includes a second camera 21 and a second telecentric lens 22, and one end of the second telecentric lens 22 is connected to the second camera 21 to form the second camera assembly 2.

[0048] In this embodiment, the first camera assembly 1 is arranged at intervals directly above the light source 3, and the second camera assembly 2 is arranged on one side of the first camera assembly 1. It is worth noting that the first telecentric lens 12 and the second telecentric lens 22 used in this embodiment have small distortion, the imaging of small dirt is clearer, the focal length is fixed, which can effectively reduce the focusing error, the depth of field is large, and the imaging of the tilted camera is still clear. For the tilted camera, the available field of view is larger. Therefore, in this embodiment, arranging the second camera assembly 2 on one side of the first camera assembly 1 forms a center line angle between the first camera assembly 1 and the second camera assembly 2, and the angle of the center line angle between the first camera assembly 1 and the second camera assembly 2 is set between 10° and 30°. Furthermore, the available field of view of the first camera assembly 1 and the second camera assembly 2 is larger, which is beneficial to improving the detection accuracy and detection efficiency. Due to the limitation of the depth of field of the first telecentric lens 12 and the second telecentric lens 22, and the larger the connection angle between the first camera assembly 1 and the second camera assembly 2, the smaller the common, effective, and clear area between the first camera 11 and the second camera 21. Therefore, in this embodiment, more preferably, the angle of the center line angle between the first camera assembly 1 and the second camera assembly 2 is set to 15°. In addition, in this embodiment, it is necessary to make the vertex angle of the center line angle formed between the first camera assembly 1 and the second camera assembly 2 directly face directly above the CMOS chip to be detected, thereby forming a 3D system of binocular telecentric lenses.

[0049] As Figure 2 shown, this embodiment also provides a detection method based on the above-mentioned chip pollutant detection device, and its steps include: constructing the above-mentioned chip pollutant detection device, calibrating the chip pollutant detection device to obtain calibration parameters, and the calibration parameters include the fundamental matrix, the projection matrix of the first camera assembly, and the projection matrix of the second camera assembly. Establish an image pixel coordinate system, place the CMOS chip to be detected on the lifting stage 4, turn on the light source 3, and the first camera assembly 1 and the second camera assembly 2 simultaneously photograph the CMOS chip to be detected to obtain the detection image of the first camera assembly and the detection image of the second camera assembly. Obtain the image pixel coordinates of the detection image of the first camera assembly and the image pixel coordinates of the detection image of the second camera assembly. The image pixel coordinates of the detection image of the first camera assembly include the image pixel coordinates of the key points on the reference plane and the image pixel coordinates of the key points of the pollutants. Establish a world coordinate system, and obtain the world coordinates of the key points on the reference plane and the world coordinates of the key points of the pollutants through the relationship between the image pixel coordinates of the detection image of the first camera assembly and the world coordinates and the relationship between the image pixel coordinates of the detection image of the second camera assembly and the world coordinates. Fit the world coordinates of the key points on the reference plane into a plane, calculate the distance from the world coordinates of the key points of the pollutants to the plane, and judge whether the pollutants are inner surface pollutants or outer surface pollutants.

[0050] Specifically in this embodiment, first, a chip contaminant detection device is constructed to form a 3D system of binocular telecentric lenses. It should be noted that in this embodiment, the angle of the central line formed between the first camera component 1 and the second camera component 2 is set to 15°, so as to obtain more common, effective, and clear areas.

[0051] Furthermore, the chip contaminant detection device is calibrated to obtain calibration parameters, which include the fundamental matrix, the projection matrix of the first camera component, and the projection matrix of the second camera component. Specifically, first, an image pixel coordinate system and a world coordinate system are established. Then, a calibration board is selected and fixed on the lifting stage 4 so that the calibration board is located at the center of the field of view of the 3D system of the binocular telecentric lenses. The first camera component 1 and the second camera component 2 simultaneously capture the calibration board to obtain the calibration image of the first camera component and the calibration image of the second camera component. In this embodiment, the lifting stage 4 is moved along the z-axis by a distance d z After that, the calibration board is captured again by the first camera component 1 and the second camera component 2 simultaneously to obtain the calibration images of the first camera component and the calibration images of the second camera component at different positions. And so on, a pair of calibration images of the first camera component and the calibration images of the second camera component at different positions are obtained in total. Through the established image pixel coordinate system, the image pixel coordinates of the calibration image of the first camera component and the image pixel coordinates of the calibration image of the first camera component where i represents the serial number of the calibration board position, i = 0, 1,..., n, n represents the number of groups of calibration images, represents the horizontal image pixel coordinate of the calibration image of the first camera component, represents the vertical image pixel coordinate of the calibration image of the first camera component, represents the horizontal image pixel coordinate of the calibration image of the second camera component, represents the vertical image pixel coordinate of the calibration image of the second camera component. It should be noted that each group of calibration images includes a calibration image of the first camera component and a calibration image of the second camera component. Then, through the image pixel coordinates of the calibration image of the first camera component and the image pixel coordinates of the calibration image of the second camera component the fundamental matrix is calculated, and the calculation formula is:

[0052]

[0053] where F represents the fundamental matrix.

[0054] Set the world coordinates of the calibration board as where, represents the x-axis coordinate of the calibration board in the world coordinate system, represents the y-axis coordinate of the calibration board in the world coordinate system. It represents the z-axis coordinate of the calibration board in the world coordinate system. The projection matrix of the first camera component and the projection matrix of the second camera component are calculated by calculating the relationship between the image pixel coordinates of the calibration image and the world coordinates of the calibration board for the first camera component and the relationship between the image pixel coordinates of the calibration image and the world coordinates of the calibration board for the second camera component. The calculation formula is as follows:

[0055]

[0056] where s represents the scale factor. In this embodiment, s is an empirical constant. u i represents the abscissa of the image pixel of the calibration image, and v i represents the ordinate of the image pixel of the calibration image, and P represents the projection matrix. In this embodiment, the image pixel coordinates of the calibration image of the first camera component and the image pixel coordinates of the calibration image of the second camera component are substituted into this formula to obtain the projection matrix P1 of the first camera component and the projection matrix P2 of the second camera component.

[0057] Thus, the fundamental matrix F of the calibration parameters, the projection matrix P1 of the first camera component, and the projection matrix P2 of the second camera component are obtained through the calibration chip contaminant detection device.

[0058] Further, an image pixel coordinate system is established. The CMOS chip to be detected is placed on the lifting stage 4. The light source 3 is turned on, and the first camera assembly 1 and the second camera assembly 2 simultaneously capture the CMOS chip to be detected, obtaining the detection image of the first camera assembly and the detection image of the second camera assembly. The image pixel coordinates of the detection image of the first camera assembly and the image pixel coordinates of the detection image of the second camera assembly are obtained. The image pixel coordinates of the detection image of the first camera assembly include the image pixel coordinates of the key points on the reference plane and the image pixel coordinates of the key points of the contaminants. In this embodiment, the image pixel coordinates of the key points on the reference plane are extracted by the edge extraction algorithm in the prior art, and the image pixel coordinates of the key points of the contaminants are extracted by the feature extraction algorithm in the prior art. It should be noted that in this embodiment, the extracted image pixel coordinates of the key points of the contaminants are the image pixel coordinates of the central key points of the contaminants. In this embodiment, through the established image pixel coordinate system, the image pixel coordinates p1(u1, v1) of the detection image of the first camera assembly and the image pixel coordinates p2(u2, v2) of the detection image of the second camera assembly are obtained, where u1 represents the horizontal image pixel coordinate of the detection image of the first camera assembly, v1 represents the vertical image pixel coordinate of the detection image of the first camera assembly, u2 represents the horizontal image pixel coordinate of the detection image of the second camera assembly, and v2 represents the vertical image pixel coordinate of the detection image of the second camera assembly. In this embodiment, in order to reduce the matching difficulty of the feature points in the detection image of the first camera assembly and the detection image of the second camera assembly and improve the matching efficiency, the fundamental matrix F is used to achieve feature matching between the detection image of the first camera assembly and the detection image of the second camera assembly. The relationship between the image pixel coordinates p1(u1, v1) of the detection image of the first camera assembly and the image pixel coordinates p2(u2, v2) of the detection image of the second camera assembly is expressed as:

[0059]

[0060] Thus, the key points on the reference plane and the key points of the contaminants on the detection image of the first camera assembly are located on the corresponding epipolar lines of the detection image of the second camera assembly. Similarly, the key points on the reference plane and the key points of the contaminants on the detection image of the second camera assembly are located on the corresponding epipolar lines of the detection image of the first camera assembly, thereby converting the feature point matching from two-dimensional region matching to linear matching, thus greatly improving the image processing efficiency and accuracy.

[0061] Further, a world coordinate system is established. Through the relationship between the image pixel coordinates p1(u1, v1) of the detection image of the first camera assembly and the world coordinates, and the relationship between the image pixel coordinates p2(u2, v2) of the detection image of the second camera assembly and the world coordinates, the world coordinates of the key points on the reference plane and the world coordinates of the key points of the contaminants are obtained. Specifically, in this embodiment, the relationship between the image pixel coordinates of the detection image of the first camera assembly and the world coordinates is represented by the projection matrix P1 of the first camera assembly, and the relationship is expressed as:

[0062]

[0063] The relationship between the image pixel coordinates and the world coordinates of the image detected by the second camera component is represented by the second camera component projection matrix P2, and the relationship is expressed as:

[0064]

[0065] Thus, the x-axis coordinate X of the reference plane key point and the pollutant key point in the world coordinate system, the y-axis coordinate Y of the reference plane key point and the pollutant key point in the world coordinate system, and the z-axis coordinate Z of the reference plane key point and the pollutant key point in the world coordinate system are obtained from the relationship between the image pixel coordinates and the world coordinates of the image detected by the first camera component and the relationship between the image pixel coordinates and the world coordinates of the image detected by the second camera component. w , the y-axis coordinate Y of the reference plane key point and the pollutant key point in the world coordinate system w , and the z-axis coordinate Z of the reference plane key point and the pollutant key point in the world coordinate system w The linear equation of. According to the principle of triangulation reconstruction in the prior art, the world coordinates (X w , Y w , Z w ) of the reference plane key point and the pollutant key point are calculated.

[0066] Furthermore, the world coordinates of the reference plane key point are fitted to a plane, the distance from the world coordinates of the pollutant key point to the plane is calculated, and it is determined whether the pollutant is an inner surface pollutant or an outer surface pollutant. Specifically, the calculation formula for fitting the world coordinates of the reference plane key point to a plane is:

[0067] Ax + By + Cz + D = 0

[0068] where A represents the component of the plane normal vector on the x-axis, B represents the component of the plane normal vector on the y-axis, C represents the component of the plane normal vector on the z-axis, D represents the distance between the plane and the origin, x represents the x-coordinate of any point on the plane, y represents the y-coordinate of any point on the plane, and z represents the z-coordinate of any point on the plane.

[0069] Furthermore, the calculation formula for the distance from the world coordinates of the pollutant key point to the plane is:

[0070]

[0071] where d represents the distance from the world coordinates of the pollutant key point to the plane, x0 represents the x-coordinate of the pollutant key point, y0 represents the y-coordinate of the pollutant key point, and z0 represents the z-coordinate of the pollutant key point

[0072] Furthermore, in this embodiment, a system error parameter δ is set. When the distance d from the key point of the contaminant to the plane is less than δ, the contaminant in the CMOS chip image to be detected is an inner surface contaminant. When the distance d from the key point of the contaminant to the plane is greater than or equal to δ, the contaminant in the CMOS chip image to be detected is an outer surface contaminant. Thus, in this embodiment, only by calculating the three-dimensional distance from the contaminant to the reference plane through the three-dimensional coordinates of the reference plane and the key point of the contaminant can the contaminant be accurately distinguished as an inner surface contaminant or an outer surface contaminant, effectively improving the detection efficiency of contaminants in the CMOS chip.

[0073] Thus, a chip contaminant detection device and a detection method provided in this embodiment can perform a full-range detection on CMOS chips at different angles and positions, improving the detection accuracy. By capturing CMOS chip images from different perspectives and using image processing technologies such as edge detection, feature extraction, and image registration, the obtained CMOS chip image data is processed and analyzed, and the parallax between two CMOS chip images is calculated to achieve the three-dimensional reconstruction of key points, thereby effectively realizing the detection accuracy of contaminants in the CMOS chip. In addition, in this embodiment, by using a high-performance image processing system and a fast feedback control technology, the real-time processing of a large amount of image data and stain detection are realized, problems can be discovered in a timely manner, and combined with an automated control system, the automatic detection and processing of stains on the production line are realized, improving the production efficiency and quality management level, and thus better meeting the requirements of modern industrial production for quality control, with strong practicability and wide applicability.

[0074] The above schematically describes the present invention and its implementation manners. This description is not restrictive. Without departing from the spirit or basic features of the present invention, the present invention can be implemented in other specific forms. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. Any reference numeral in the claims should not limit the claimed claim. Therefore, if those of ordinary skill in the art are inspired by it and, without creative design, design a structural manner and an embodiment similar to the technical solution without departing from the purpose of this creation, they shall fall within the protection scope of the present invention. In addition, the word "comprising" does not exclude other elements or steps, and the word "a" before an element does not exclude including "a plurality of" such elements. The multiple elements stated in the product claims can also be implemented by one element through software or hardware. The words such as "first" and "second" are used to represent names and do not represent any specific order.

Claims

1. A chip contaminant detection method, characterized in that: include: Calibrate the chip contaminant detection device to obtain calibration parameters, the calibration parameters including a basic matrix, a first camera assembly projection matrix, and a second camera assembly projection matrix; An image pixel coordinate system is established, a CMOS chip to be inspected is placed on a lifting stage (4), a light source (3) is connected, and the first camera component (1) and the second camera component (2) simultaneously photograph the CMOS chip to be inspected, thereby obtaining an inspection image of the first camera component and an inspection image of the second camera component; Acquire the image pixel coordinates of the image detected by the first camera assembly and the image pixel coordinates of the image detected by the second camera assembly, wherein the image pixel coordinates of the image detected by the first camera assembly include the image pixel coordinates of the key points of the reference surface and the image pixel coordinates of the key points of the pollutant; Establish a world coordinate system, obtain the relationship between the image pixel coordinates of the image detected by the first camera component and the world coordinates through the projection matrix of the first camera component, and obtain the relationship between the image pixel coordinates of the image detected by the second camera component and the world coordinates through the projection matrix of the second camera component; Then, the world coordinates of the key points of the reference surface and the world coordinates of the key points of the pollutants are obtained by using the relationship between the image pixel coordinates of the image detected by the first camera component and the world coordinates and the relationship between the image pixel coordinates of the image detected by the second camera component and the world coordinates; Fit the world coordinates of the key points of the reference plane into a plane, calculate the distance from the world coordinates of the key points of the pollutant to the plane, and determine whether the pollutant is an inner surface pollutant or an outer surface pollutant; The following chip contaminant detection device is adopted: a first camera assembly (1), a second camera assembly (2), a light source (3) and a lifting platform (4), wherein the light source (3) is arranged at a distance just above the lifting platform (4), the first camera assembly (1) comprises a first camera (11) and a first telecentric lens (12), one end of the first telecentric lens (12) is connected to the first camera (11), the second camera assembly (2) comprises a second camera (21) and a second telecentric lens (22), one end of the second telecentric lens (22) is connected to the second camera (21), the first camera assembly (1) is arranged at a distance just above the light source (3), the second camera assembly (2) is arranged on one side of the first camera assembly (1), a centerline angle is formed between the first camera assembly (1) and the second camera assembly (2), and the angle of the centerline angle between the first camera assembly (1) and the second camera assembly (2) is set between 10° and 30°.

2. A chip contaminant detection method according to claim 1, characterized in that: The included angle between the center lines of the first camera assembly (1) and the second camera assembly (2) is set to 15°.

3. A chip contaminant detection method according to claim 1 or 2, comprising the following steps: Constructing chip contaminant detection devices; The calibration parameters are obtained by selecting a calibration plate, fixing the calibration plate on a lifting platform (4), and simultaneously photographing the calibration plate with the first camera component (1) and the second camera component (2) to obtain a calibration image of the first camera component and a calibration image of the second camera component; establishing an image pixel coordinate system to obtain image pixel coordinates of the calibration image of the first camera component and image pixel coordinates of the calibration image of the second camera component; the image pixel coordinates of the calibration image of the first camera component are expressed as , the image pixel coordinates of the calibration image of the second camera assembly are expressed as , where i represents the serial number of the calibration plate position, , n represents the number of calibration image groups, represents the image pixel abscissa of the calibration image of the first camera assembly, represents the image pixel ordinate of the first camera assembly calibration image, represents the image pixel abscissa of the calibration image of the second camera assembly, represents the image pixel ordinate of the second camera assembly calibration image; the basic matrix is ​​calculated by the image pixel coordinates of the first camera assembly calibration image and the image pixel coordinates of the second camera assembly calibration image; the calculation formula of the basic matrix is: Wherein, F represents the basic matrix; establish a world coordinate system, set the world coordinates of the calibration plate, and the world coordinates of the calibration plate are expressed as ,in, Indicates the x-axis coordinate of the calibration plate in the world coordinate system. Indicates the y-axis coordinate of the calibration plate in the world coordinate system. represents the z-axis coordinate of the calibration plate in the world coordinate system; the projection matrix of the first camera component is calculated by the relationship between the image pixel coordinates of the calibration image of the first camera component and the world coordinates of the calibration plate, and the projection matrix of the second camera component is calculated by the relationship between the image pixel coordinates of the calibration image of the second camera component and the world coordinates of the calibration plate. The calculation formula of the projection matrix is: Where s represents the scale factor, represents the image pixel horizontal coordinate of the calibration image, represents the image pixel ordinate of the calibration image, P represents the projection matrix, and the image pixel coordinates of the calibration image of the first camera assembly and the image pixel coordinates of the calibration image of the second camera assembly are substituted into the formula to obtain the first camera assembly projection matrix P1 and the second camera assembly projection matrix P2.

4. A chip contaminant detection method according to claim 1, characterized in that: The relationship between the image pixel coordinates and the world coordinates of the image detected by the first camera component is expressed as: Wherein, P1 represents the projection matrix of the first camera assembly, u1 represents the horizontal coordinate of the image pixel of the image detected by the first camera assembly, v1 represents the vertical coordinate of the image pixel of the image detected by the first camera assembly, and X w Indicates the x-axis coordinates of the datum surface key points and the pollutant key points in the world coordinate system, w Indicates the Y-axis coordinates of the datum surface key points and the pollutant key points in the world coordinate system, Z w Indicates the z-axis coordinates of the datum surface key points and the pollutant key points in the world coordinate system; The relationship between the image pixel coordinates and the world coordinates of the image detected by the second camera component is expressed as: Among them, P2 represents the projection matrix of the second camera assembly, u2 represents the horizontal coordinate of the image pixel of the image detected by the second camera assembly, and v2 represents the vertical coordinate of the image pixel of the image detected by the second camera assembly.

5. A chip contaminant detection method according to claim 4, characterized in that: The calculation formula for fitting the world coordinates of the key points of the reference surface into a plane is: Among them, A represents the component of the plane normal vector on the x-axis, B represents the component of the plane normal vector on the y-axis, C represents the component of the plane normal vector on the z-axis, D represents the distance between the plane and the origin, x represents the x-coordinate of any point on the plane, y represents the y-coordinate of any point on the plane, and z represents the z-coordinate of any point on the plane.

6. A chip contaminant detection method according to claim 5, characterized in that: The distance calculation formula from the world coordinates of the pollutant key point to the plane is: Among them, d represents the distance from the world coordinates of the pollutant key point to the plane, x0 represents the x coordinate of the pollutant key point, y0 represents the y coordinate of the pollutant key point, and z0 represents the z coordinate of the pollutant key point.

7. A chip contaminant detection method according to claim 6, characterized in that: The system error parameters are set, and the contaminants are judged as inner surface contaminants or outer surface contaminants according to the system error parameters.

Images