High-precision analysis method and system for three-dimensional topography of printed circuit board solder joints

Abstract

The present invention discloses a high-precision analysis method and system for the three-dimensional morphology of solder joints on a printed circuit board (PCB). The method comprises the following steps: constructing a high-precision three-dimensional model to obtain three-dimensional point cloud data; obtaining characteristic data of the solder joint area and determining whether a cold solder joint exists through fusion analysis; obtaining surface roughness and reflectivity loss data of the solder joint area, calculating the solder joint solidification line density, obtaining the cold solder joint direction disorder, and determining the cold solder joint situation; obtaining the minimum spacing between adjacent solder joints and detecting bridging by combining conductive path curvature verification and material diffusion characteristics; obtaining solder joint surface texture gradient and reflectivity attenuation data to perform oxidized solder joint processing; and training the model through a convolutional neural network to construct a solder joint analysis model and obtain solder joint analysis results. The present invention has the advantages of accurately detecting and analyzing cold solder joints, cold solder joints, bridging, and oxidation defects in PCB solder joints through high-precision three-dimensional scanning and deep learning of convolutional neural networks, significantly improving production efficiency and product reliability.

Description

Technical Field

[0001] The present invention relates to solder joint analysis, and in particular to a method and system for high-precision three-dimensional topography analysis of solder joints on a printed circuit board. Background Art

[0002] Printed circuit boards (PCBs) are crucial components of electronic products, and soldering quality directly impacts their performance and reliability. Solder joints are the key connection between electronic components and circuit boards, and their topography directly determines the stability and mechanical strength of the electrical connection. With the advancement of miniaturization and higher performance in electronic devices, 3D topography analysis technology has emerged. Through high-precision 3D reconstruction, it can more comprehensively capture information such as solder joint height and slope, enabling assessment of solder joint quality.

[0003] The three-dimensional morphology analysis methods for printed circuit board solder joints currently available on the market usually rely on traditional two-dimensional imaging technology or limited three-dimensional scanning technology, which often lacks in accuracy and comprehensiveness. Most methods use a rough analysis based on a single sensor or point cloud data, which cannot fully obtain the detailed features of the solder joint area, and can easily lead to misjudgment or omission of defects such as false solder joints and cold solder joints. Traditional methods often ignore the complexity of the solder joint area, such as the morphological characteristics of the non-wetted area, the detailed analysis of surface roughness and reflectivity loss, and the accurate measurement of the tiny spacing between solder joints, resulting in an inability to accurately evaluate the welding quality. Although some methods use three-dimensional point cloud data, they usually use low-precision scanning devices and cannot construct a high-precision three-dimensional model of the solder joint. Therefore, there are obvious deficiencies in the accuracy and reliability of defect detection. Summary of the Invention

[0004] To improve existing methods and systems, a high-precision analysis method and system for the three-dimensional morphology of printed circuit board solder joints is provided. This method uses high-precision three-dimensional scanning and deep learning of convolutional neural networks to accurately detect and analyze defects such as cold solder joints, cold solder joints, bridging, and oxidation in printed circuit board solder joints, providing intelligent and automated solder joint quality assessment and repair solutions, significantly improving production efficiency and product reliability.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] High-precision analysis method for the three-dimensional topography of printed circuit board solder joints, including:

[0007] A high-precision three-dimensional model of the solder joint area is constructed by using a scanning path of a spiral progressive trajectory planning algorithm through a three-dimensional scanning device, and three-dimensional point cloud data of the solder joint area is obtained;

[0008] Based on the 3D point cloud data of the solder joint area, the contact interface height difference, effective contact area ratio and morphological feature data of the non-wetting area of ​​the solder joint area are obtained. Through fusion analysis, it is determined whether there is a cold solder joint, and repair or re-soldering is performed based on the severity.

[0009] Based on the three-dimensional point cloud data of the solder joint area, the surface roughness and reflectivity loss data of the solder joint area are obtained. By calculating the solidification line density of the solder joint, the cold welding direction disorder is obtained and the cold welding condition of the solder joint is judged.

[0010] Based on the 3D point cloud data of the solder joint area, the minimum spacing between adjacent solder joints is obtained. The bridging situation is detected by combining the conductive path curvature verification and material diffusion characteristics, and cutting and suction processing is performed based on the severity.

[0011] Based on the three-dimensional point cloud data of the solder joint area, the surface texture gradient and reflectivity attenuation data of the solder joint are obtained, and the oxidation level of the oxide component is verified to perform oxidation treatment on the solder joint;

[0012] The above four analysis methods are trained through convolutional neural networks to build a solder joint analysis model. The three-dimensional point cloud data of each solder joint area is input into the model to obtain the analysis results.

[0013] Preferably, the step of constructing a high-precision three-dimensional model of the solder joint area by using a three-dimensional scanning device and adopting a spiral progressive trajectory planning algorithm to obtain three-dimensional point cloud data of the solder joint area specifically includes:

[0014] Based on the coaxial confocal laser sensor, the geometric center of the pad is used as the spiral starting point to generate the Archimedean spiral equation and continuously acquire high-density point cloud data along the spiral path;

[0015] Perform point cloud registration based on the acquired high-density point cloud data, convert the point cloud data into a continuous surface through a triangulated meshing algorithm, and construct a high-precision three-dimensional model of the solder joint area;

[0016] Key features are extracted based on the high-precision three-dimensional morphology of the solder joint area to obtain three-dimensional point cloud data of the solder joint area.

[0017] Preferably, the step of obtaining the contact interface height difference, effective contact area ratio, and non-wetting area morphological feature data of the solder joint area based on the three-dimensional point cloud data of the solder joint area, determining whether a cold solder joint occurs through fusion analysis, and performing repair or re-soldering based on the severity specifically includes:

[0018] Extract the height profile of the contact area between the solder joint and the pad, calculate the maximum height difference and average height difference of the contact interface, and determine whether there is overhang or uneven contact;

[0019] The actual contact area of ​​the solder joint area is calculated by the curvature and normal angle change method to obtain the effective contact area ratio;

[0020] Detect high curvature areas, concave areas and areas with sharp slope changes in the solder joint area, identify edge discontinuities and island structures in areas where solder has not spread, and extract morphological features of non-wetting areas;

[0021] The severity level is graded based on the three parameters obtained, and repair welding and re-welding are performed based on the grading results.

[0022] Preferably, the step of obtaining surface roughness and reflectivity loss data of the solder joint area based on the three-dimensional point cloud data of the solder joint area, obtaining the cold welding direction disorder by calculating the solidification line density of the solder joint, and determining the cold welding condition of the solder joint specifically includes:

[0023] Based on the three-dimensional point cloud data of the solder joint area, the surface height change of the solder joint area is obtained to obtain the surface roughness;

[0024] Based on the three-dimensional point cloud data of the solder joint area, different reflectivity data of the solder joint area are obtained and the reflectivity loss is calculated;

[0025] Reconstruct the solidification lines of the welding area based on the three-dimensional point cloud data of the welding spot area, calculate the number and distribution of the solidification lines in the welding spot area, and obtain the density;

[0026] Based on the main direction of the solidification line and the deviation in each direction, the cold welding direction disorder is extracted;

[0027] Through the weighted scoring model, the four characteristics of surface roughness, reflectivity loss, solidification line density and cold welding direction disorder are integrated to make a comprehensive judgment on the cold welding situation.

[0028] Preferably, the method of obtaining the minimum spacing between adjacent solder joints based on the three-dimensional point cloud data of the solder joint area, detecting the bridging situation in combination with the conductive path curvature verification and the material diffusion characteristics, and performing cutting and suction processing based on the severity specifically includes:

[0029] Based on the position of the solder joints in the high-precision 3D model of the solder joint area, the solder joints are spatially calibrated to accurately locate each solder joint;

[0030] By calculating the Euclidean distance between solder joints, finding the minimum value and comparing it with the preset minimum spacing standard, the probability distribution of bridging occurrence is obtained;

[0031] Extract the conductive path of the solder in the solder joint area, build a conductive path model based on the solder extension direction and actual soldering method, calculate the curvature of the conductive path, and obtain the probability distribution of bridging;

[0032] Based on the width, depth and diffusion angle data of the solder diffusion zone, the diffusion coefficient and diffusion morphology of the solder are calculated, the diffusion uniformity and the overall quality of the solder joint are evaluated, and the probability distribution of bridging is obtained;

[0033] Through the weighted scoring model, the minimum spacing between adjacent solder joints, the curvature of the conductive path, and the material diffusion characteristics are comprehensively considered to make a comprehensive judgment on the bridging situation.

[0034] Preferably, the step of obtaining the surface texture gradient and reflectivity attenuation data of the solder joint based on the three-dimensional point cloud data of the solder joint area and performing the oxidized solder joint treatment by verifying the oxidation level of the oxide component specifically includes:

[0035] Based on the high-precision 3D model of the solder joint area, the normal vector of each point is calculated, and the change between the normal vectors is calculated to obtain the texture gradient;

[0036] Based on different angles and lighting conditions, the reflectivity of the solder joint area is measured, the reflectivity loss of the solder joint area is calculated, and the thickness and quality of the surface oxide layer are evaluated;

[0037] By testing the oxide composition, the degree of oxidation of the solder joints can be determined, and oxidized solder joints of different oxidation levels can be treated.

[0038] Preferably, the above four analysis methods are trained by convolutional neural network to construct a solder joint analysis model, and the three-dimensional point cloud data of each solder joint area is input into the model to obtain the analysis results, which specifically includes:

[0039] Based on the above four solder joint anomaly detection and judgment processes, the detection and judgment processes are trained through a convolutional neural network to obtain a trained solder joint analysis model;

[0040] Input the real-time collected 3D point cloud data into the solder joint analysis model to obtain the existing problems of the solder joints and the analysis process results;

[0041] By continuously collecting and annotating new solder joint data, the optimization model is updated to improve its prediction accuracy.

[0042] Furthermore, a high-precision three-dimensional topography analysis system for printed circuit board solder joints is proposed, including:

[0043] Data acquisition module: The data acquisition module collects high-precision three-dimensional point cloud data of the welding spot area through a three-dimensional scanning device and a spiral progressive trajectory planning algorithm;

[0044] Cold solder joint detection module: The cold solder joint detection module analyzes whether there is a cold solder joint through the solder joint feature data and determines the necessity of repair soldering or re-soldering;

[0045] Cold welding detection module: The cold welding detection module determines the cold welding condition of the weld point based on surface roughness, reflectivity loss, solidification line density and cold welding direction disorder and performs corresponding processing;

[0046] Bridging detection module: The bridging detection module detects solder joint bridging through the minimum spacing between solder joints, conductive path curvature and material diffusion characteristics, and performs cutting and suction processing according to the severity;

[0047] Oxidation Detection Module: The oxidation detection module analyzes and processes oxidized solder joints based on the solder joint surface texture gradient, reflectivity attenuation data, and oxidation level of the oxide components;

[0048] Neural training module: This module uses a convolutional neural network to train the analysis model based on four solder joint anomaly detection methods and continuously optimizes the model to improve prediction accuracy;

[0049] Processor: The processor is used to process the calculation process of each formula and the construction calculation process of each model.

[0050] Compared with the prior art, the advantages of the present invention are:

[0051] By adopting a spiral progressive trajectory planning algorithm, combined with a high-precision 3D scanning device, it is possible to comprehensively collect 3D point cloud data of the solder joint area and construct an accurate 3D model through an advanced triangulated meshing algorithm, ensuring high-precision solder joint detection. Secondly, based on rich 3D point cloud data and combining multiple parameters such as contact interface height difference, effective contact area ratio, surface roughness, and reflectivity loss, this method can comprehensively evaluate solder joint quality and accurately determine problems such as false solder joints, cold solder joints, bridging, and oxidation, providing a scientific basis for repair, re-soldering, and repair. At the same time, deep learning training using convolutional neural networks is used to continuously optimize the model and enhance the accuracy and predictive capabilities of the analysis. This method can achieve automated and efficient solder joint detection, reduce human error, improve production efficiency, and adopt customized repair strategies for different types of solder joint defects, significantly improving the reliability of the solder joints and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic diagram of the method proposed in the present invention;

[0053] Figure 2 This is a schematic diagram of point cloud data acquisition proposed by the present invention;

[0054] Figure 3 This is a schematic diagram of the cold solder joint detection proposed by the present invention;

[0055] Figure 4 This is a schematic diagram of the cold welding detection proposed by the present invention;

[0056] Figure 5 This is a schematic diagram of the bridge detection proposed by the present invention;

[0057] Figure 6 This is a schematic diagram of the oxidation detection proposed by the present invention;

[0058] Figure 7 A schematic diagram of the solder joint analysis model proposed in the present invention;

[0059] Figure 8 This is a diagram of the architecture of the electronic equipment in this solution;

[0060] Figure 9 This is a schematic diagram of the computer-readable storage medium structure in this solution. DETAILED DESCRIPTION

[0061] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0062] High-precision 3D topography analysis system for printed circuit board solder joints, including:

[0063] Data acquisition module: The data acquisition module collects high-precision three-dimensional point cloud data of the welding spot area through a three-dimensional scanning device and a spiral progressive trajectory planning algorithm;

[0064] Cold solder joint detection module: The cold solder joint detection module analyzes whether there is a cold solder joint in the solder joint through the solder joint feature data, and determines the necessity of repair soldering or re-soldering;

[0065] Cold welding detection module: The cold welding detection module determines the cold welding condition of the weld point based on surface roughness, reflectivity loss, solidification line density and cold welding direction disorder and performs corresponding processing;

[0066] Bridging detection module: The bridging detection module detects solder joint bridging through the minimum spacing between solder joints, conductive path curvature and material diffusion characteristics, and performs cutting and suction processing according to the severity;

[0067] Oxidation Detection Module: The oxidation detection module analyzes and processes oxidized solder joints based on the solder joint surface texture gradient, reflectivity attenuation data, and oxidation level of the oxide components;

[0068] Neural training module: This module uses a convolutional neural network to train the analysis model based on four solder joint anomaly detection methods and continuously optimizes the model to improve prediction accuracy;

[0069] Processor: The processor is used to process the calculation process of each formula and the construction calculation process of each model.

[0070] See Figure 1 As shown, the high-precision analysis method for the three-dimensional morphology of solder joints on a printed circuit board includes:

[0071] Step 1: Using a 3D scanning device and a spiral progressive trajectory planning algorithm to scan the path, a high-precision 3D model of the solder joint area is constructed to obtain 3D point cloud data of the solder joint area;

[0072] Step 2: Based on the 3D point cloud data of the solder joint area, the contact interface height difference, effective contact area ratio, and morphological feature data of the non-wetting area are obtained. Through fusion analysis, it is determined whether a cold solder joint exists, and repair or re-soldering is performed based on the severity.

[0073] Step 3: Based on the 3D point cloud data of the solder joint area, obtain the surface roughness and reflectivity loss data of the solder joint area. By calculating the solidification line density of the solder joint, obtain the cold welding direction disorder and determine the cold welding condition of the solder joint.

[0074] Step 4: Based on the 3D point cloud data of the solder joint area, the minimum spacing between adjacent solder joints is obtained. The bridging situation is detected by combining the conductive path curvature verification and material diffusion characteristics, and cutting and suctioning are performed based on the severity.

[0075] Step 5: Based on the 3D point cloud data of the solder joint area, obtain the solder joint surface texture gradient and reflectivity attenuation data, and perform oxidation treatment on the solder joint by verifying the oxidation level of the oxide component;

[0076] Step 6: Use convolutional neural networks to train the above four analysis methods, build a solder joint analysis model, input the three-dimensional point cloud data of each solder joint area into the model, and obtain the analysis results.

[0077] See Figure 2 As shown, a high-precision three-dimensional model of the solder joint area is constructed by using a scanning path of a spiral progressive trajectory planning algorithm through a three-dimensional scanning device, and the three-dimensional point cloud data of the solder joint area is obtained, specifically including:

[0078] Based on the coaxial confocal laser sensor, the geometric center of the pad is used as the spiral starting point to generate the Archimedean spiral equation and continuously acquire high-density point cloud data along the spiral path;

[0079] Perform point cloud registration based on the acquired high-density point cloud data, convert the point cloud data into a continuous surface through a triangulated meshing algorithm, and construct a high-precision three-dimensional model of the solder joint area;

[0080] Key features are extracted based on the high-precision three-dimensional morphology of the solder joint area to obtain three-dimensional point cloud data of the solder joint area.

[0081] Specifically, the Archimedean spiral can be expressed by the following equation:

[0082] ;

[0083] in, is the distance from the geometric center of the pad, is the helix angle, 、 are constants, controlling the starting radius of the spiral and the radial increment per unit angle respectively;

[0084] Based on this spiral equation, the laser sensor collects point cloud data along a spiral path. The sampling interval and scanning rate of the sensor determine the density of the point cloud data.

[0085] After obtaining high-density point cloud data, the source point cloud and the target point cloud are aligned through the transformation matrix, and the point clouds are registered so that the data collected from different perspectives can be correctly aligned.

[0086] See Figure 3 As shown in the figure, based on the three-dimensional point cloud data of the solder joint area, the contact interface height difference, effective contact area ratio and morphological feature data of the non-wetting area of ​​the solder joint area are obtained. Through fusion analysis, it is determined whether there is a cold solder joint, and repair welding or re-soldering is performed based on the severity. The specific processing includes:

[0087] Extract the height profile of the contact area between the solder joint and the pad, calculate the maximum height difference and average height difference of the contact interface, and determine whether there is overhang or uneven contact;

[0088] The actual contact area of ​​the solder joint area is calculated by the curvature and normal angle change method to obtain the effective contact area ratio;

[0089] Detect high curvature areas, concave areas and areas with sharp slope changes in the solder joint area, identify edge discontinuities and island structures in areas where solder has not spread, and extract morphological features of non-wetting areas;

[0090] The severity level is graded based on the three parameters obtained, and repair welding and re-welding are performed based on the grading results.

[0091] Specifically, the height profile of the contact area between the solder joint and the pad is extracted from the constructed 3D model of the solder joint area. By calculating the curvature and normal angle changes of the solder joint area, the geometric characteristics of the actual contact area can be further analyzed. The curvature represents the degree of surface bending and can be calculated by the change of the normal vector of each contact point.

[0092] The flatness of the surface morphology is determined by the change in the direction of the normal line, and the change in the angle between the normal lines of adjacent triangles is measured by the formula:

[0093] ;

[0094] in, is the angle, is the normal vector of two adjacent patches;

[0095] By integrating the surface of the contact region, the area of ​​the contact region can be obtained;

[0096] Through curvature analysis, normal angle change and surface slope calculation, high curvature areas, concave areas and areas with sharp slope changes in the solder joint area can be detected, and areas where the solder is not evenly spread or not wetted can be obtained;

[0097] The non-wetting area usually presents edge discontinuity or island-like structure. By detecting the continuity of the edge of the solder joint area, it is determined whether there is a poor contact area;

[0098] Severity classification: For mild defect areas, local repair welding is used to repair them, while for severe defect areas, complete re-welding may be required to ensure the quality of the solder joints.

[0099] See Figure 4 As shown, based on the three-dimensional point cloud data of the solder joint area, the surface roughness and reflectivity loss data of the solder joint area are obtained. By calculating the solidification line density of the solder joint, the cold welding direction disorder is obtained, and the cold welding situation of the solder joint is judged. Specifically, the following are performed:

[0100] Based on the three-dimensional point cloud data of the solder joint area, the surface height change of the solder joint area is obtained to obtain the surface roughness;

[0101] Based on the three-dimensional point cloud data of the solder joint area, different reflectivity data of the solder joint area are obtained and the reflectivity loss is calculated;

[0102] Reconstruct the solidification lines of the welding area based on the three-dimensional point cloud data of the welding spot area, calculate the number and distribution of the solidification lines in the welding spot area, and obtain the density;

[0103] Based on the main direction of the solidification line and the deviation in each direction, the cold welding direction disorder is extracted;

[0104] Through the weighted scoring model, the four characteristics of surface roughness, reflectivity loss, solidification line density and cold welding direction disorder are integrated to make a comprehensive judgment on the cold welding situation.

[0105] Specifically, for the three-dimensional point cloud data of the solder joint area, the surface height change is calculated, the average value of the absolute difference between the height of each point and its average height is calculated, that is, the arithmetic mean roughness, and the average value of the square of the difference between the height of each point and the average height is calculated, that is, the root mean square roughness, to obtain the surface roughness;

[0106] Obtain different reflectivity data of the solder joint area as the reflectivity of each point. After calculating the reflectivity loss of each point, calculate the average reflectivity loss, that is, the average value of the reflectivity loss of all points;

[0107] The spatial position of the solidification line is obtained by fitting the temperature distribution data of the solder joint area. The solidification line is usually closely related to the change of the surface temperature. The solidification line is extracted by the temperature threshold. The number and distribution density of the solidification lines are calculated by analyzing the extracted solidification lines.

[0108] Cold welding usually causes the solidification line direction of the welding area to be disordered and the main direction to deviate. By calculating the main direction of the solidification line and the deviation degree of each direction, the disorder degree of the cold welding direction can be extracted.

[0109] Based on the four features extracted above (surface roughness, reflectivity loss, solidification line density, and cold weld direction disorder), a weighted scoring model is used to conduct a comprehensive evaluation of cold welds.

[0110] Take appropriate measures based on the evaluation results, including mild cold welding: repair welding can be considered, moderate cold welding: re-welding can be considered; severe cold welding: re-welding or re-re-welding is required.

[0111] See Figure 5 As shown in the figure, based on the 3D point cloud data of the solder joint area, the minimum spacing between adjacent solder joints is obtained. The bridging situation is detected by combining the conductive path curvature verification and material diffusion characteristics, and cutting and suction processing is performed based on the severity. Specifically, it includes:

[0112] Based on the position of the solder joints in the high-precision 3D model of the solder joint area, the solder joints are spatially calibrated to accurately locate each solder joint;

[0113] By calculating the Euclidean distance between solder joints, finding the minimum value and comparing it with the preset minimum spacing standard, the probability distribution of bridging occurrence is obtained;

[0114] Extract the conductive path of the solder in the solder joint area, build a conductive path model based on the solder extension direction and actual soldering method, calculate the curvature of the conductive path, and obtain the probability distribution of bridging;

[0115] Based on the width, depth and diffusion angle data of the solder diffusion zone, the diffusion coefficient and diffusion morphology of the solder are calculated, the diffusion uniformity and the overall quality of the solder joint are evaluated, and the probability distribution of bridging is obtained;

[0116] Through the weighted scoring model, the minimum spacing between adjacent solder joints, the curvature of the conductive path, and the material diffusion characteristics are comprehensively considered to make a comprehensive judgment on the bridging situation.

[0117] Specifically, the positions of the solder joints are extracted from the 3D point cloud data of the solder joint area. By calculating the Euclidean distance between the solder joints, the spacing between adjacent solder joints can be evaluated and compared with the preset minimum spacing standard to obtain the probability distribution of bridging occurrence.

[0118] By analyzing the conductive path of the solder in the solder joint area, we can establish a model of the solder expansion direction and the actual soldering method, evaluate the probability of bridging, and calculate the curvature of the conductive path. The conductive path is a curve composed of a series of discrete points. The curvature can be calculated by the change in the normal between two points. The formula is:

[0119] ;

[0120] in, is the curvature, is the tangent vector of the curve, is the derivative of the tangent vector;

[0121] Calculate the relationship between the curvature change of the conductive path and the bridging probability. The area with greater curvature may indicate a high bridging probability.

[0122] Obtain the width, depth and diffusion angle data of the solder diffusion zone, calculate the diffusion coefficient of the solder, and evaluate the uniformity of the solder diffusion. If the diffusion coefficient is too large or the diffusion is uneven, bridging may occur.

[0123] See Figure 6 As shown, based on the three-dimensional point cloud data of the solder joint area, the surface texture gradient and reflectivity attenuation data of the solder joint are obtained, and the oxidation solder joint treatment is performed by verifying the oxidation level of the oxide component. Specifically, the following steps are involved:

[0124] Based on the high-precision 3D model of the solder joint area, the normal vector of each point is calculated, and the change between the normal vectors is calculated to obtain the texture gradient;

[0125] Based on different angles and lighting conditions, the reflectivity of the solder joint area is measured, the reflectivity loss of the solder joint area is calculated, and the thickness and quality of the surface oxide layer are evaluated;

[0126] By testing the oxide composition, the degree of oxidation of the solder joints can be determined, and oxidized solder joints of different oxidation levels can be treated.

[0127] Specifically, the texture gradient reflects the change between normal vectors and describes the change of surface texture. The normal vector of each point is calculated based on the 3D point cloud data, and the angle between adjacent normal vectors is calculated. For each point, the amplitude of the normal vector change is calculated as the local texture change.

[0128] Reflectivity is the ratio of the intensity of light reflected from a surface to the intensity of the incident light. Reflectivity loss is evaluated by measuring the reflectivity at different angles. The reference reflectivity is the reflectivity of an unoxidized surface. Reflectivity loss is the difference between the reflectivity of the current solder joint and the reference reflectivity. The formula is:

[0129] ;

[0130] in, is the reflectivity difference, is the reference reflectivity, is the reflectivity at each point, is the angle of incidence, is the observation angle;

[0131] By extracting the oxide components of the solder joint area, the degree of oxidation is determined based on the oxide composition. The degree of oxidation can be measured by the ratio of the oxide components. Based on the calculated results of the oxidation degree, the oxidation level can be divided into multiple levels, and mild oxidation can be lightly cleaned or simply treated, moderate oxidation can be surface cleaned or repaired, and severe oxidation can be re-soldered or surface deoxidized.

[0132] See Figure 7 As shown in the figure, the above four analysis methods are trained by convolutional neural network to build a solder joint analysis model. The three-dimensional point cloud data of each solder joint area is input into the model to obtain the analysis results, including:

[0133] Based on the above four solder joint anomaly detection and judgment processes, the detection and judgment processes are trained through a convolutional neural network to obtain a trained solder joint analysis model;

[0134] Input the real-time collected 3D point cloud data into the solder joint analysis model to obtain the existing problems of the solder joints and the analysis process results;

[0135] By continuously collecting and annotating new solder joint data, the optimization model is updated to improve its prediction accuracy.

[0136] Furthermore, the method according to the embodiment of the present application can also be used with the aid of Figure 8 The electronic device architecture shown in FIG. Figure 8 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the method and system for high-precision analysis of the three-dimensional topography of solder joints on a printed circuit board provided in this application. The electronic device 500 may also include a terminal interface 508. Of course, Figure 8The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 8 One or more components of an electronic device are shown.

[0137] Figure 9 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 9 , a computer-readable storage medium 600 according to one embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by a processor, the method and system for high-precision analysis of the three-dimensional topography of printed circuit board solder joints according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc.

[0138] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0139] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0140] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A high-precision analysis method for the three-dimensional morphology of solder joints on a printed circuit board, characterized in that: include: A high-precision three-dimensional model of the solder joint area is constructed by using a scanning path of a spiral progressive trajectory planning algorithm through a three-dimensional scanning device. Key features are extracted based on the high-precision three-dimensional model of the solder joint area to obtain three-dimensional point cloud data of the solder joint area. Based on the three-dimensional point cloud data of the solder joint area, the height profile of the contact area between the solder joint and the pad is extracted, and the maximum height difference and average height difference of the contact interface are calculated. The actual contact area of ​​the solder joint area is calculated by the curvature and normal angle change method to obtain the effective contact area ratio. The high curvature area, concave area and area with sharp slope change in the solder joint area are detected, and the edge discontinuity and island-like structure of the unspread solder area are identified to extract the morphological characteristics of the unwetted area. Determine whether a cold weld occurs through fusion analysis, and perform repair or re-welding based on the severity; Based on the three-dimensional point cloud data of the solder joint area, the surface roughness and reflectivity loss data of the solder joint area are obtained, the density of the solder joint solidification line is calculated, and the cold welding direction disorder is obtained based on the main direction of the solidification line and the deviation in each direction, and the cold welding condition of the solder joint is judged; Based on the location of solder joints in the high-precision 3D model of the solder joint area, the minimum spacing between adjacent solder joints is obtained. The curvature of the conductive path of the solder in the solder joint area and the diffusion coefficient of the solder are combined to detect bridging conditions, and cutting and suctioning are performed based on the severity. The diffusion coefficient of the solder is calculated by obtaining the width, depth and diffusion angle data of the solder joint diffusion zone; Based on the high-precision 3D model of the solder joint area, the solder joint surface texture gradient and reflectivity attenuation data are obtained, and the oxidized solder joint treatment is performed by verifying the oxidation level of the oxide component; The above four analysis methods are trained through convolutional neural networks to build a solder joint analysis model. The three-dimensional point cloud data of each solder joint area is input into the solder joint analysis model to obtain the analysis results.

2. The high-precision three-dimensional topography analysis method for printed circuit board solder joints according to claim 1, characterized in that: The method of using a spiral progressive trajectory planning algorithm to construct a high-precision three-dimensional model of the solder joint area by using a three-dimensional scanning device, extracting key features based on the high-precision three-dimensional model of the solder joint area, and obtaining three-dimensional point cloud data of the solder joint area specifically includes: Based on the coaxial confocal laser sensor, the geometric center of the pad is used as the spiral starting point to generate the Archimedean spiral equation and continuously acquire high-density point cloud data along the spiral path; Perform point cloud registration based on the acquired high-density point cloud data, convert the point cloud data into a continuous surface through a triangulated meshing algorithm, and construct a high-precision three-dimensional model of the solder joint area; Key features are extracted based on the high-precision three-dimensional model of the solder joint area to obtain three-dimensional point cloud data of the solder joint area.

3. The high-precision three-dimensional topography analysis method of printed circuit board solder joints according to claim 1, characterized in that: The method of obtaining surface roughness and reflectivity loss data of the solder joint area based on the three-dimensional point cloud data of the solder joint area, calculating the density of the solder joint solidification line, obtaining the cold welding direction disorder based on the main direction of the solidification line and the deviation in each direction, and determining the cold welding condition of the solder joint specifically includes: Based on the three-dimensional point cloud data of the solder joint area, the surface height change of the solder joint area is obtained to obtain the surface roughness; Based on the three-dimensional point cloud data of the solder joint area, different reflectivity data of the solder joint area are obtained and the reflectivity loss is calculated; Reconstruct the solidification lines of the welding area based on the three-dimensional point cloud data of the welding spot area, calculate the number and distribution of the solidification lines in the welding spot area, and obtain the solidification line density; Based on the main direction of the solidification line and the deviation in each direction, the cold welding direction disorder is extracted; Through the weighted scoring model, the four characteristics of surface roughness, reflectivity loss, solidification line density and cold welding direction disorder are integrated to make a comprehensive judgment on the cold welding situation.

4. The high-precision three-dimensional topography analysis method for printed circuit board solder joints according to claim 1, characterized in that: The method of obtaining the surface texture gradient of the solder joint based on the high-precision three-dimensional model of the solder joint area, obtaining reflectivity attenuation data, and performing the oxidized solder joint treatment by verifying the oxidation level of the oxide component specifically includes: Normal vector calculation is performed based on the high-precision 3D model of the solder joint area. The normal vector of each point is extracted and the variation between the normal vectors is calculated to obtain the texture gradient. Based on different angles and lighting conditions, the reflectivity of the solder joint area is measured, the reflectivity loss of the solder joint area is calculated, and the thickness and quality of the surface oxide layer are evaluated; By testing the oxide composition, the degree of oxidation of the solder joints can be determined, and oxidized solder joints of different oxidation levels can be treated.

5. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the high-precision analysis method for the three-dimensional morphology of printed circuit board solder joints as described in any one of claims 1-4.

6. A computer-readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, the method for high-precision analysis of the three-dimensional topography of solder joints on a printed circuit board according to any one of claims 1 to 4 is implemented.

Images