A circuit board surface defect detection device and method

Abstract

Embodiments of the present application provide a circuit board surface defect detection device and method. The circuit board surface defect detection device comprises a defect collection unit and a defect analysis unit; the defect collection unit comprises a precision displacement table, a line laser, a ring light source, an industrial lens, a beam splitter, a face array camera one and a face array camera two; the defect analysis unit comprises a calculation analysis device and a result display device; the detection method is realized based on the detection device. The device and method realize two-dimensional defect detection and three-dimensional defect detection at the same time, and adopt common aperture technology to simplify the process of two-dimensional feature and three-dimensional point cloud registration. In the feature extraction stage, the extraction result of the two-dimensional feature is used as the prior condition of three-dimensional point cloud extraction, which can save the computing power requirement of the detection device and improve the detection efficiency. In the defect recognition process, the information contained in the two-dimensional feature is combined with the three-dimensional point cloud to recognize the defect from multiple dimensions, which can accurately and quickly complete the detection of the circuit board surface defect.

Description

Technical Field

[0001] This application relates to the field of circuit board inspection technology, and in particular to a device and method for detecting surface defects on circuit boards. Background Technology

[0002] With the development of modern electronic technology, printed circuit boards (PCBs) have been widely used in various common fields. Currently, China has become the world's largest PCB manufacturing country, and its PCB exports rank among the top in the world. During the production process, it is inevitable that some defective PCBs will be manufactured. Once these PCBs enter the market, they will pose significant risks. Therefore, to prevent defective PCBs from entering the market, screening them out during the inspection stage has become a crucial part of the production process.

[0003] Based on their characteristics, surface defects on circuit boards can be categorized into two-dimensional (2D) and three-dimensional (3D) defects. Two-dimensional defects include short circuits, open circuits, missing or extra traces, excessive or insufficient trace width, missing pads, component misalignment, and incorrect or incomplete soldering. Three-dimensional defects include bending deformation, cold solder joints, and solder bridging. Common methods for detecting surface defects on circuit boards include 2D inspection and 3D inspection methods such as line laser scanning. 2D inspection uses a camera to capture images of the circuit board, identifying various defects quickly and efficiently, and effectively acquiring the color and texture of the target. However, 2D inspection cannot acquire depth information of the circuit board, making it difficult to identify 3D defects such as cold solder joints, and the results are significantly affected by the shooting angle and lighting conditions. Line laser scanning technology uses structured light projection to acquire the 3D morphology of the circuit board surface. The measurement results are less affected by external lighting conditions, but it cannot acquire the color and texture of the circuit board, and it is difficult to identify two-dimensional defects such as component silkscreen printing, short circuits, and open circuits. Summary of the Invention

[0004] To address the limitations of two-dimensional inspection methods in identifying three-dimensional defects such as solder joint defects and the difficulty of three-dimensional inspection methods in identifying two-dimensional defects such as silkscreen markings, short circuits, and open circuits in components, this invention presents a circuit board surface defect detection device and method. This method employs common-aperture technology and multispectral technology to combine two-dimensional and three-dimensional information for a comprehensive multi-dimensional assessment of defect presence. It simultaneously meets the detection requirements for both two-dimensional and three-dimensional defects in circuit board inspection while simplifying the registration process of two-dimensional and three-dimensional information. This enhances both the circuit board inspection capability and inspection efficiency.

[0005] This invention provides a circuit board surface defect detection device, comprising: a defect acquisition unit and a defect analysis unit;

[0006] The defect acquisition unit includes a precision displacement stage, a line laser, a ring light source, an industrial lens, a beam splitter, a first area array camera, and a second area array camera.

[0007] The industrial lens is vertically mounted on the mounting bracket and is used to image defect information on the surface of the circuit board.

[0008] The line laser is installed at an angle, and the optical axis of the line laser is coplanar with the optical axis of the industrial lens and forms an angle. The line laser can emit line lasers, and the spectral range of the emitted lasers is [λ1, λ2]. The three-dimensional information on the circuit board is encoded into light bar signals.

[0009] The ring light source is installed in front of the industrial lens and is coaxial with the industrial lens, and the spectral range of its emitted light is [λ3,λ4].

[0010] The beam splitter is installed behind the industrial lens to split the light beam entering the industrial lens according to the spectral range. The splitting ratio of the beam splitter is 50:50. The exit planes of the two beams are coated with filter films. The spectral bands allowed to pass through the exit plane of one beam are consistent with those of the line laser, which are [λ1,λ2], and the spectral bands allowed to pass through the exit plane of the other beam are consistent with those of the ring light source, which are [λ3,λ4].

[0011] The area array camera one and the area array camera two have the same parameters and are both mounted on the imaging plane of the industrial lens, respectively receiving the two beams from the beam splitter;

[0012] The precision displacement stage is installed below the industrial lens and ensures that the table surface of the precision displacement stage is within the depth of field of the industrial lens. The precision displacement stage is used to drive the circuit board to move in a single axis.

[0013] The defect analysis unit includes a calculation and analysis device and a result display device;

[0014] The computational analysis device is responsible for extracting and identifying two-dimensional and three-dimensional defects of the circuit board from the multispectral information of the circuit board collected in the defect acquisition unit.

[0015] The result display device is used to display the multispectral information results of the circuit board and the defect identification results.

[0016] In some embodiments, the defect acquisition unit employs multispectral encoding and decoding technology and common aperture technology to encode two-dimensional and three-dimensional defects on the circuit board surface into different spectral bands, and acquires information in the two spectral bands through the common aperture area array camera one and the area array camera two.

[0017] In some embodiments, the defect analysis unit uses the multispectral information of the circuit board acquired by the defect acquisition unit to decode the line laser information in the [λ1,λ2] band to obtain the three-dimensional shape of the circuit board surface. Each point in the three-dimensional point cloud is represented by coordinates (x,y,z). The image of the [λ3,λ4] band is processed to obtain the two-dimensional defects on the circuit surface. The color of each point on the image is represented by (R,G,B).

[0018] This invention provides a method for detecting surface defects on a circuit board, based on the circuit board surface defect detection device described in any of the above embodiments, the method comprising:

[0019] Step 1: Mount and fix the circuit board to be tested on the precision displacement stage, and turn on the precision displacement stage to move the circuit board.

[0020] Step 2: Turn on the line laser and the ring light source. The line laser and the ring light source illuminate the circuit board with a beam of light. The beam of light is reflected by the surface of the circuit board and enters the industrial lens for imaging.

[0021] Step 3: The beam splitter separates the light beam entering the industrial lens according to the spectral range, and then the beam is captured by two area array cameras to obtain two circuit board images with different spectral bands.

[0022] Step 4: Extract the center of the light stripe from the online laser image using a line laser light stripe center extraction algorithm, and calculate the three-dimensional point cloud coordinates of the extracted light stripe center;

[0023] Step 5: Index the corresponding texture color information based on the center position of the extracted light stripe, thus completing the extraction and registration of two-dimensional and three-dimensional information;

[0024] Step 6: Using image segmentation and recognition algorithms, the features such as components and solder joints in the two-dimensional image are segmented and identified. The segmentation results of the two-dimensional image are directly used for the pre-segmentation of the three-dimensional point cloud.

[0025] Step 7: Use a circuit board defect recognition algorithm to identify the segmented two-dimensional texture image and three-dimensional point cloud, and display the defect recognition results through a display device.

[0026] In some embodiments, step four, extracting the light stripe center from the online laser image using a line laser light stripe center extraction algorithm, and calculating the three-dimensional point cloud coordinates, includes:

[0027] The center of the light stripe is extracted from the online laser image using a line laser light stripe center extraction algorithm, resulting in a set of n light stripe center coordinates {(X1,Y1),……,(X n ,Y n )}, calculate the 3D point cloud coordinates (x, y, z) according to the formula;

[0028] For laser stripe center extraction, the centroid method can be used, and the calculation formula is as follows:

[0029]

[0030] Where I(X,Y) is the gray value of a pixel in the line laser image, X is the horizontal coordinate of a pixel in the image, Y is the vertical coordinate of a pixel in the image, X1 is the horizontal coordinate of the center of the laser stripe extracted by the centroid method, and Y1 is the vertical coordinate of the center of the laser stripe extracted by the centroid method.

[0031] Since the coordinates of the light stripe center calculated using the centroid method are in the image coordinate system and the unit is the number of pixels, it is necessary to transform the light stripe center coordinates to the world coordinate system and the unit to common units such as mm. The transformation formula is as follows:

[0032]

[0033] Where x', y', z' are the coordinates of the light stripe center in the world coordinate system, s is the pixel size of the camera, X0 and Y0 are the coordinates of the image center, and f is the focal length of the industrial lens;

[0034] By combining the coordinate calculation results from Formula 2 with the equation Ax + By + Cz + D = 0 for the light plane, we obtain the three-dimensional point cloud coordinates (x, y, z).

[0035] The beneficial effects of the above embodiments include:

[0036] This device and method, while simultaneously achieving two-dimensional and three-dimensional defect detection, simplifies the registration process between two-dimensional features and three-dimensional point clouds by employing a common-aperture technique. In the feature extraction stage, the extracted two-dimensional features are used as a priori conditions for three-dimensional point cloud extraction, saving computational resources and improving detection efficiency. Furthermore, during defect identification, the information contained in the two-dimensional features is combined with the three-dimensional point cloud to identify defects from multiple dimensions, enabling accurate and rapid detection of surface defects on circuit boards. Attached Figure Description

[0037] The accompanying drawings illustrate, by way of example and not limitation, the various embodiments discussed herein.

[0038] Figure 1 This is a schematic diagram of the components of a circuit board surface defect detection device;

[0039] Figure 2 This is a flowchart illustrating the method for detecting tilt and rotation errors using this device.

[0040] Symbol explanation:

[0041] 1-Precision displacement stage; 2-Line laser; 3-Ring light source; 4-Industrial lens; 5-Area scan camera one; 6-Beam splitter; 7-Area scan camera two; 8-Result display device; 9-Computational analysis device. Detailed Implementation

[0042] In order to gain a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for reference and illustration only and are not intended to limit the embodiments of this application.

[0043] In the embodiments described in this application, it should be noted that, unless otherwise stated and limited, the term "connection" should be interpreted broadly. For example, it can be an electrical connection, or a connection between two internal components. It can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above term according to the specific circumstances.

[0044] It should be noted that the terms "first," "second," and "third" used in the embodiments of this application are merely used to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first," "second," and "third" can be interchanged in a specific order or sequence where permitted. It should be understood that the objects distinguished by "first," "second," and "third" can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in an order other than those illustrated or described herein.

[0045] This invention proposes a circuit board surface defect detection device, which consists of a defect acquisition unit and a defect analysis unit, as detailed below. Figure 1 The defect acquisition unit consists of a precision displacement stage 1, a line laser 2, a ring light source 3, an industrial lens 4, a first area array camera 5, a beam splitter 6, and a second area array camera 7. The defect analysis unit includes a calculation and analysis device 9 and a result display device 8.

[0046] The defect acquisition unit employs multispectral encoding and decoding technology and common-aperture technology to encode two-dimensional and three-dimensional defects on the circuit board surface into different spectral bands, and acquires information from both spectral bands using a common-aperture area array camera. This simplifies the registration process between two-dimensional and three-dimensional information, effectively acquiring both defects on the circuit board surface. The industrial lens 4 is vertically mounted on the mounting bracket to image the defect information on the circuit board surface. The line laser 2 is mounted at an angle, with its optical axis coplanar with and at a certain angle to the optical axis of the industrial lens 4. The line laser 2 emits line laser light with a spectral range of [λ1, λ2], encoding the three-dimensional information on the circuit board as light stripe signals. The ring light source 3 is mounted in front of the industrial lens 4 and coaxial with the lens, with an output light spectral range of [λ3, λ4]. The beam splitter 6 is mounted behind the industrial lens 4 to split the incoming light beam according to its spectral range. The beam splitter 6 has a splitting ratio of 50:50. The exit planes of both beams are coated with filters. The spectral bands allowed to pass through the exit plane of one beam are consistent with those of the line laser 2, namely [λ1, λ2]; the spectral bands allowed to pass through the exit plane of the other beam are consistent with those of the ring light source 3, namely [λ3, λ4]. Two area scan cameras, area scan camera 5 and area scan camera 7, have identical parameters and are both mounted on the imaging plane of the industrial lens 4, receiving the two beams from the beam splitter 6 respectively. A precision displacement stage 1 is mounted below the industrial lens 4, ensuring that the stage surface of the precision displacement stage 1 is within the depth of field of the industrial lens 4. The precision displacement stage 1 is used to drive the circuit board to perform single-axis movement.

[0047] The defect analysis unit utilizes the multispectral information of the circuit board acquired by the defect acquisition unit to decode the line laser information in the [λ1, λ2] bands, obtaining the three-dimensional topography of the circuit board surface. Each point in the three-dimensional point cloud is represented by coordinates (x, y, z). The image in the [λ3, λ4] band is processed to obtain two-dimensional defects on the circuit surface, with the color of each point in the image represented by (R, G, B). Furthermore, image segmentation and image recognition techniques are used to segment and identify features such as components and solder joints on the circuit board based on the two-dimensional information, including color and texture. Since the two area-array cameras share the same aperture, the two-dimensional and three-dimensional information has been registered, allowing the segmentation and recognition results to be directly used for the extraction and identification of three-dimensional defects. This effectively reduces the computational power required for three-dimensional defect identification and improves the identification rate. The computational analysis device 9 is responsible for extracting and identifying two-dimensional and three-dimensional defects of the circuit board from the multispectral information acquired by the defect acquisition unit. The result display device 8 displays the multispectral information results of the circuit board and the defect identification results.

[0048] The steps for using the aforementioned circuit board surface defect detection device to detect surface defects on a circuit board are as follows:

[0049] Step 1: Mount and fix the circuit board to be tested on the precision displacement stage, and turn on the precision displacement stage to move the circuit board at a constant speed.

[0050] Step two: Turn on the line laser and the ring light source. The line laser and the ring light source illuminate the circuit board with a beam of light. The beam of light is reflected by the surface of the circuit board and enters the industrial lens for imaging.

[0051] Step 3: The beam splitter separates the light beam entering the industrial lens according to the spectral range, and then the beam is captured by two area array cameras to obtain two circuit board images with different spectral bands.

[0052] Step 4: Extract the center of the light stripe from the online laser image using a line laser light stripe center extraction algorithm, and calculate the three-dimensional point cloud coordinates.

[0053] Specifically, the center of the light stripe is extracted from the online laser image using a line laser light stripe center extraction algorithm, resulting in a set of n light stripe center coordinates {(X1,Y1),……,(X n ,Y n The three-dimensional point cloud coordinates (x, y, z) are calculated according to the formula.

[0054] For laser stripe center extraction, the centroid method can be used, and the calculation formula is as follows:

[0055]

[0056] Where I(X,Y) is the gray value of a pixel in the line laser image, X is the x-coordinate of the pixel in the image, and Y is the y-coordinate of the pixel in the image. X1 is the x-coordinate of the center of the laser beam extracted by the centroid method, and Y1 is the y-coordinate of the center of the laser beam extracted by the centroid method.

[0057] Since the coordinates of the light stripe center calculated using the centroid method are in the image coordinate system and the unit is pixels, it is necessary to transform the light stripe center coordinates to the world coordinate system (generally, the camera coordinate system is set to the world coordinate system), and convert the units to common units such as mm. The transformation formula is as follows:

[0058]

[0059] Where x', y', z' are the coordinates of the light stripe center in the world coordinate system, s is the pixel size of the camera, X0 and Y0 are the coordinates of the image center, and f is the focal length of the industrial lens.

[0060] By combining the coordinate calculation results from Formula 2 with the equation Ax + By + Cz + D = 0 for the light plane, we obtain the three-dimensional point cloud coordinates (x, y, z).

[0061] Step 5: Based on the extracted light stripe center position {(X1,Y1),……,(X n ,Yn )}, indexing to the corresponding texture color information {(R1,G1,B1),……,(R n G n B n This means that the extraction and registration of two-dimensional and three-dimensional information are completed. The registered point cloud can be represented as (x,y,z,R,G,B).

[0062] Step six involves using image segmentation and recognition algorithms to segment and identify features such as components and solder joints in the two-dimensional image. The segmentation results of the two-dimensional image are then directly used for the pre-segmentation of the three-dimensional point cloud.

[0063] Step 7: Use a circuit board defect recognition algorithm to identify the segmented two-dimensional texture image and three-dimensional point cloud, and display the defect recognition results through a display device.

[0064] This device and method, while simultaneously achieving two-dimensional and three-dimensional defect detection, simplifies the registration process between two-dimensional features and three-dimensional point clouds by employing a common-aperture technique. In the feature extraction stage, the extracted two-dimensional features are used as a priori conditions for three-dimensional point cloud extraction, saving computational resources and improving detection efficiency. Furthermore, during defect identification, the information contained in the two-dimensional features is combined with the three-dimensional point cloud to identify defects from multiple dimensions, enabling accurate and rapid detection of surface defects on circuit boards.

[0065] The technical solutions described in the embodiments of this application can be combined arbitrarily without conflict.

[0066] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A circuit board surface defect detection device, characterized in that, include: Defect acquisition unit and defect analysis unit; The defect acquisition unit includes a precision displacement stage, a line laser, a ring light source, an industrial lens, a beam splitter, a first area array camera, and a second area array camera. The industrial lens is vertically mounted on the mounting bracket and is used to image defect information on the surface of the circuit board. The line laser is installed at an angle, and the optical axis of the line laser is coplanar with the optical axis of the industrial lens and forms an angle. The line laser can emit line laser light, and the spectral range of the emitted laser light is [λ1, λ2]. The three-dimensional information on the circuit board is encoded into light bar signals. The ring light source is installed in front of the industrial lens and is coaxial with the industrial lens, and the spectral range of its emitted light is [λ3, λ4]. The beam splitter is installed behind the industrial lens to split the light beam entering the industrial lens according to the spectral range. The splitting ratio of the beam splitter is 50:

50. The exit planes of the two beams are coated with filter films. The spectral bands allowed to pass through the exit plane of one beam are consistent with those of the line laser, which are [λ1, λ2], and the spectral bands allowed to pass through the exit plane of the other beam are consistent with those of the ring light source, which are [λ3, λ4]. The area array camera one and the area array camera two have the same parameters and are both mounted on the imaging plane of the industrial lens, respectively receiving the two beams from the beam splitter; The precision displacement stage is installed below the industrial lens and ensures that the table surface of the precision displacement stage is within the depth of field of the industrial lens. The precision displacement stage is used to drive the circuit board to move in a single axis. The defect analysis unit includes a calculation and analysis device and a result display device; The computational analysis device is responsible for extracting and identifying two-dimensional and three-dimensional defects of the circuit board from the multispectral information of the circuit board collected in the defect acquisition unit. The result display device is used to display the multispectral information results and defect identification results of the circuit board; The defect acquisition unit uses multispectral encoding and decoding technology and common aperture technology to encode two-dimensional and three-dimensional defects on the circuit board surface into different spectral bands, and acquires information in the two spectral bands through the common aperture area array camera one and the area array camera two.

2. The circuit board surface defect detection device according to claim 1, characterized in that, The defect analysis unit uses the multispectral information of the circuit board collected by the defect acquisition unit to decode the line laser information in the [λ1, λ2] band to obtain the three-dimensional shape of the circuit board surface. Each point in the three-dimensional point cloud is represented by coordinates (x, y, z). The image of the [λ3, λ4] band is processed to obtain the two-dimensional defects on the circuit surface. The color of each point on the image is represented by (R, G, B).

3. A method for detecting surface defects on a circuit board, based on the circuit board surface defect detection device according to claim 1 or 2, characterized in that, The method includes: Step 1: Mount and fix the circuit board to be tested on the precision displacement stage, and turn on the precision displacement stage to move the circuit board. Step 2: Turn on the line laser and the ring light source. The line laser and the ring light source illuminate the circuit board with a beam of light. The beam of light is reflected by the surface of the circuit board and enters the industrial lens for imaging. Step 3: The beam splitter separates the light beam entering the industrial lens according to the spectral range, and then the beam is captured by two area array cameras to obtain two circuit board images with different spectral bands. Step 4: Extract the center of the light stripe from the online laser image using a line laser light stripe center extraction algorithm, and calculate the three-dimensional point cloud coordinates of the extracted light stripe center; Step 5: Index the corresponding texture color information based on the center position of the extracted light stripe, thus completing the extraction and registration of two-dimensional and three-dimensional information; Step 6: Using image segmentation and recognition algorithms, the features of components and solder joints in the two-dimensional image are segmented and identified. The segmentation results of the two-dimensional image are directly used for the pre-segmentation of the three-dimensional point cloud. Step 7: Use a circuit board defect recognition algorithm to identify the segmented two-dimensional texture image and three-dimensional point cloud, and display the defect recognition results through a display device.

4. The method for detecting surface defects on a circuit board according to claim 3, characterized in that, Step four involves extracting the center of the light stripe from the online laser image using a line laser light stripe center extraction algorithm, and then calculating the three-dimensional point cloud coordinates of the extracted light stripe center. This includes: The center of the light stripe is extracted from the online laser image using a line laser light stripe center extraction algorithm. n The set of center coordinates of the light stripes {( X 1, Y 1),……,( X n , Y n )}, calculate the coordinates of the three-dimensional point cloud according to the formula ( x , y , z ); The centroid method is used to extract the center of the laser stripe, and the calculation formula is as follows: in, I ( X , Y ) represents the grayscale value of a pixel in a line laser image. X Let x be the x-coordinate of a pixel in the image. Y y is the ordinate of a pixel in the image. X 1 represents the x-coordinate of the center of the laser stripe extracted by the centroid method. Y 1 represents the ordinate of the center of the laser stripe extracted by the centroid method; Since the coordinates of the light stripe center calculated using the centroid method are in the image coordinate system and the unit is the number of pixels, it is necessary to transform the light stripe center coordinates to the world coordinate system and the unit to the commonly used mm unit. The conversion formula is as follows: in, x ', y ', z 'These are the coordinates of the center of the light stripe in the world coordinate system.' s The pixel size of the camera. X 0 and Y 0 represents the coordinates of the image center. f For industrial lens focal length; Combine the coordinate calculation results in the transformation formula with the equation of the light plane A. x +B y +C z +D=0, obtain the three-dimensional point cloud coordinates. x , y , z ).

Images