A method and apparatus for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs
By identifying target areas in ADAS high-frequency PCBs and combining optical imaging and laser drilling metallization testing, the problem of deviation in interlayer alignment accuracy detection results in existing detection methods is solved, achieving high-precision and reliable interlayer alignment status characterization, which is suitable for quality control of ADAS high-frequency multilayer PCBs.
Patent Information
- Application Number
- CN202610552832.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-24
- Publication Date
- 2026-11-10
AI Technical Summary
Existing PCB interlayer alignment accuracy testing methods are insufficient to fully reflect the actual alignment status of electrical interconnects, resulting in significant deviations in the test results and failing to provide accurate and reliable quantitative basis for subsequent process optimization and reliability assessment.
By identifying the target area of a multilayer PCB, optical offset data and electrical data are obtained through optical observation window perspective imaging and laser drilling combined with metallization continuity resistance testing. The alignment deviation results between each signal layer are obtained through weighted fusion processing.
It achieves high-precision and reliable interlayer alignment accuracy detection, which can truly reflect the microscopic electrical effects introduced by the process, improve the accuracy and reliability of the detection results, and is suitable for precision manufacturing quality control of ADAS high-frequency multilayer PCBs.
Smart Images

Figure CN123065347A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of PCB inspection technology, and more specifically, to a method and apparatus for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs. Background Technology
[0002] With the rapid development of Advanced Driver Assistance Systems (ADAS), extremely high demands are placed on the reliability, signal integrity, and high-frequency performance of automotive electronic devices. High-frequency multilayer PCBs, as the main platform for core ADAS modules (such as millimeter-wave radar, camera processing units, and domain controllers), directly determine the impedance continuity, crosstalk control, and eye diagram quality of high-speed differential signals through their interlayer alignment accuracy. In ADAS applications, high-frequency PCBs typically employ a structure of 10 or more layers, combined with low-loss dielectric materials and fine linewidth design to support signal transmission at GHz levels and above. If interlayer alignment deviations exceed the micrometer level, it will lead to signal reflection, increased attenuation, and even functional failure. Therefore, accurate detection of interlayer alignment accuracy has become a critical quality control step in the high-frequency PCB manufacturing process.
[0003] Currently, X-ray imaging or optical alignment methods are commonly used to assess the alignment accuracy of multilayer PCBs. The former indirectly judges the alignment by observing the deviation of inner layer targets or drill holes using X-rays, while the latter relies on optical measurements before and after exposure. However, the aforementioned methods mainly quantify single-dimensional deviations at the optical or imaging level, which is difficult to fully reflect the actual alignment status of electrical interconnects. Relying solely on optical offset data often ignores microscopic electrical changes caused by process factors such as material thermal expansion, resin flow, and uneven copper plating. The inconsistency between optical data and actual electrical performance results in significant deviations in the final alignment accuracy assessment, failing to provide accurate and reliable quantitative basis for subsequent process optimization and reliability determination.
[0004] Therefore, there is a need to provide a method and device for detecting the alignment accuracy of ADAS high-frequency PCB layers, in order to solve the problem that existing PCB layer alignment accuracy detection methods cannot fully reflect the true alignment status of actual electrical interconnections. Summary of the Invention
[0005] The main objective of this invention is to provide a method and apparatus for detecting the alignment accuracy between layers of ADAS high-frequency PCBs, aiming to solve the technical problems mentioned in the background section.
[0006] The present invention adopts the following technical solution: A method for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs includes: The target area of the multilayer PCB is identified, and the target area is windowed to obtain an optical observation window that exposes the top target group. Each signal layer of the multilayer PCB has a pre-set target group. Perspective imaging is performed on the optical observation window, and the target group of each signal layer is compared with the top target group to obtain offset data; Based on the offset data, laser drilling is performed on the target area, and the resulting micro-hole array is sequentially metallized and subjected to on / off resistance tests to obtain electrical data. The offset data and electrical data are weighted and fused to obtain the alignment deviation results between each signal layer.
[0007] Furthermore, prior to the step of identifying the target region of the multilayer PCB, the method further includes: Layered positioning marks are made on the multilayer PCB design layout, and each signal layer of the multilayer PCB is processed according to the positioning marks to form a target group on each signal layer. The processing includes developing, etching, and film removal processes.
[0008] Furthermore, the step of identifying the target region of the multilayer PCB includes: Perform a global optical scan on a multilayer PCB to obtain a board surface image containing all preset target groups; Based on the image of the plate, template matching and positioning of the target graphic are performed to obtain the region boundary of each preset target group; The target region of the multilayer PCB is extracted by extending a preset safety margin based on the geometric center of the boundary of the region, thereby obtaining the target region of the multilayer PCB.
[0009] Further, the step of performing windowing processing on the target region to obtain an optical observation window exposing the top target group includes: Based on the target area, graphical mask data is generated, and the top solder mask layer of the target area is selectively removed according to the graphical mask data to form a local windowed area. Under coaxial illumination, the optical transmittance of the local windowed area is verified. For the local windowed area whose verification result is lower than a preset threshold, the window size is finely adjusted to obtain the observation window. Using the top target group of the target region as a reference, position offset correction is performed on each signal layer in the observation window to obtain the optical observation window that exposes the top target group.
[0010] Further, the step of performing perspective imaging on the optical observation window and comparing the target group of each signal layer with the top target group to obtain offset data includes: Multispectral perspective images are acquired vertically downwards from the optical observation window in a time-division manner to obtain the original superimposed grayscale image; The original superimposed grayscale image is separated by interlayer projection to obtain a separated grayscale image of each signal layer; The edge points of the separated grayscale image of each signal layer are extracted, and the geometric center of the formed edge point set is fitted within the annular region to obtain the center coordinates of the target group of each signal layer. The offset data is obtained by calculating the offset between the center coordinates of each signal layer target group and the center coordinates of the top layer target group.
[0011] Furthermore, the step of laser drilling the target area based on the offset data includes: Based on the offset data, a drilling coordinate sequence for the target area is planned, and laser drilling is performed on the target area in stages based on the drilling coordinate sequence to form preliminary microholes; The preliminary micropores are surface-treated, and the surface-treated preliminary micropores are subjected to depth optical inspection from the top layer of the multilayer PCB downwards to obtain a qualified micropore array.
[0012] Furthermore, the step of sequentially metallizing and testing the continuity resistance of the formed micro-hole array includes: The walls of the micro-hole array are cleaned, and the cleaned micro-hole array is subjected to local chemical copper plating treatment to obtain locally metallized micro-holes. The copper layer thickness of the locally metallized micropores is homogenized using a complexing agent to obtain a locally electrically connected array of metallized micropores. The on / off resistance of the metallized micro-hole array was tested using a flying probe test to obtain electrical data.
[0013] Furthermore, the step of performing continuity resistance testing on the metallized microvia array based on flying probe testing to obtain electrical data further includes: Double-ended flying probes are used to contact the center of the top pore and the center of the bottom pore of each micropore in the metallized micropore array, and a preset low pressure is applied to form the original scan dataset. Based on the original scan dataset, a segmented, incremental test voltage is applied sequentially to each microhole to measure its continuity resistance, thereby obtaining the continuity state and resistance sequence of each microhole. Based on the conduction state and resistance sequence, the micro-hole array is classified into anomalies to form an electrical anomaly distribution map labeled with the anomaly type and the corresponding physical location of the micro-hole. Based on the electrical anomaly distribution map and the geometric arrangement of the micropore array, deviation inversion calculation is performed to obtain electrical data containing the actual alignment deviation between each signal layer.
[0014] Further, the step of performing weighted fusion processing on the offset data and electrical data to obtain the alignment deviation results between each signal layer includes: Feature vectors are constructed from the offset data and electrical data. Based on the formed multilayer fusion dataset, optical values of the micropores in each direction are corrected to obtain optical correction data for each signal layer. The ratio of the modulus of the optical correction data of each signal layer to the degree of anomaly in the electrical anomaly distribution map is calculated and weighted accordingly to obtain the alignment deviation data of each signal layer relative to the top layer. The alignment deviation data and the multi-layer fusion dataset are summarized and organized to obtain the alignment deviation results between each signal layer.
[0015] An ADAS high-frequency PCB interlayer alignment accuracy detection device includes: The identification module is used to identify the target area of the multilayer PCB, and to perform windowing processing on the target area to obtain an optical observation window that exposes the top target group, wherein each signal layer of the multilayer PCB has a pre-set target group. The comparison module is used to perform perspective imaging on the optical observation window, compare the target group of each signal layer with the top target group to obtain offset data; The testing module is used to perform laser drilling on the target area based on the offset data, and to perform metallization and on / off resistance tests on the formed micro-hole array in sequence to obtain electrical data. The data processing module is used to perform weighted fusion processing on the offset data and electrical data to obtain the alignment deviation results between each signal layer.
[0016] Beneficial effects: In this invention, an optical observation window is formed by opening a window in the target area. Perspective imaging is used to acquire the positional information of the target groups of each signal layer in a multilayer PCB relative to the top target group, obtaining high-precision optical offset data. Based on this offset data, laser drilling is performed in the same target area to form a micro-via array, completing metallization and continuity resistance testing to obtain corresponding electrical data. A weighted fusion algorithm is used to comprehensively process the optical offset data and electrical data, outputting the quantitative results of the alignment deviation between each signal layer. Combining non-contact high-resolution observation of optical imaging with the electrical performance verification of actual micro-via interconnects, this invention is suitable for precision manufacturing quality control of high-frequency multilayer PCBs in the ADAS field, enabling comprehensive characterization of interlayer alignment status. It improves the accuracy and reliability of interlayer alignment precision detection, ensuring that the detection results not only reflect geometric positional deviations but also truly reflect the microscopic electrical effects introduced by the manufacturing process. While maintaining detection efficiency, it effectively supports the evolution of high-frequency PCBs towards finer linewidths, lower-loss materials, and higher layer counts, ensuring stable RF performance and signal integrity of ADAS core modules in complex automotive environments. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating the ADAS high-frequency PCB interlayer alignment accuracy detection method of the present invention. Figure 2 This is a schematic block diagram of the structure of an ADAS high-frequency PCB interlayer alignment accuracy detection device according to the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0020] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.
[0021] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.
[0022] Reference Figure 1 This invention proposes a method for detecting the alignment accuracy between layers of an ADAS high-frequency PCB, comprising: S1: Identify the target area of the multilayer PCB, perform windowing processing on the target area to obtain an optical observation window that exposes the top target group, wherein each signal layer of the multilayer PCB is preset with a target group; In step S1, during the preceding ADAS high-frequency multilayer PCB circuit layout design stage, a fundamental physical basis has been established for interlayer alignment accuracy detection. Specifically, in the non-functional area at the board edge or in a specially designated test coupon area, a set of nested concentric circular target groups is pre-designed and arranged for each signal layer. A target group refers to a concentric structure composed of multiple copper rings with progressively increasing diameters. The number of rings equals the total number of layers minus the surface layer. The innermost ring has the smallest diameter and serves as the geometric reference for the entire target group. The diameter difference between adjacent layers is set to a known, precise constant, such as 8 micrometers or 10 micrometers, ensuring that the rings in each layer are clearly distinguishable radially, avoiding overlap and confusion. Simultaneously, to ensure accurate positioning during subsequent optical imaging, a high-precision crosshair is superimposed at the geometric center of each ring, providing a center alignment reference. The target pattern and the main functional circuit pattern are synchronously transferred to each layer of copper foil via LDI laser direct imaging or photoplotting, with the overall pattern transfer accuracy controlled within ±5 micrometers. When designing the targets, the optical transmittance characteristics of commonly used low-loss substrates such as PTFE or LCP composites in high-frequency boards were fully considered. Therefore, appropriate copper thickness and ring linewidth (0.15–0.25 mm) were selected to ensure that the inner target still has sufficient contrast in the visible or near-infrared bands. The target group is positioned to avoid high-density wiring areas and is symmetrically distributed at the four corners or multiple points of the board to comprehensively characterize the interlayer alignment of the entire board. The pre-set target structure requires no additional materials or complex processes and can be completed only through layout fine-tuning, directly determining the detection resolution threshold and optical positioning reliability.
[0023] After obtaining the finished multilayer PCB board, an optical observation window is created at the top layer solder mask and surface treatment layer (such as ENIG or OSP) of the target area in the preset target region using precision exposure or laser windowing. The window size is slightly larger than the maximum diameter of the outermost target, ensuring that the copper ring of the top target is fully exposed and serves as a reference for optical imaging. The windowing is positioned based on the cross mark in the target group to avoid damaging any functional circuits. Because the dielectric layer material of the high-frequency board has high transmittance (>70%) for visible and near-infrared light, the superimposed projection image of the multilayer target rings can be observed vertically downwards through the window without thinning or special treatment. The windowing process is a local modification of the standard solder mask pattern, a simple and fast process that is completed after solder masking or before final inspection, without changing the overall PCB structure and electrical performance. The created observation window makes the inner layer targets optically accessible, and the exposed top layer target ring and cross mark together form the zero reference system for calculating all layer deviations, thus ensuring that the test results truly reflect the cumulative alignment error from the inner layer pattern to the finished product, rather than a problem in a single stage.
[0024] S2: Perform perspective imaging on the optical observation window, compare the target group of each signal layer with the top target group to obtain offset data; In step S2, by utilizing the physical properties of the low dielectric constant and low loss factor substrate used in high-frequency PCBs, which have high optical transmittance in the visible or near-infrared bands, the superimposed projection image of the multi-layered nested concentric target structure can be clearly captured vertically downward from the observation window that has been opened on the top layer without any destructive thinning or special modification of the board. An imaging system equipped with a high-resolution industrial camera or stereomicroscope is used, and a suitable near-infrared filter is selected according to the substrate's transmittance characteristics to suppress ambient stray light interference and improve the contrast of the inner copper ring outline. The imaging device vertically acquires multi-layer grayscale images of the target area from directly above the optical observation window. Since each signal layer is pre-set with copper ring structures whose diameters increase from small to large, and the diameter difference between adjacent layers is a fixed precision constant (such as 8μm or 10μm), multiple concentric but gradually increasing ring projection outlines are presented in the same field of view. The innermost smallest diameter ring corresponds to the deepest signal layer, while the outermost largest diameter ring corresponds to the top layer or the signal layer near the surface. The entire superimposed image intuitively reflects the alignment status of each layer's pattern accumulated throughout the entire process. After acquisition, the image is first pre-processed, including grayscale equalization to compensate for light intensity attenuation at different depths, filtering and noise reduction to reduce false edges caused by scattering, and an edge detection algorithm (which can be the Canny operator) is applied to extract the clear boundary points of each ring. Then, the geometric center coordinates of each ring are accurately calculated using circle fitting techniques such as least squares or Hough transform. At the same time, the high-precision crosshair positioning marks pre-set in the top target group are used as absolute reference benchmarks to achieve unified coordinate system alignment of the centers of all layers. On this basis, the coordinate differences between the center of each signal layer ring and the top reference center are compared layer by layer, and the offset vectors in the X and Y directions are calculated respectively. The offset vectors are the preliminary inter-layer alignment deviation data, and their values directly quantify the cumulative misalignment introduced from the inner layer pattern transfer, lamination to the finished product. Due to the special design of increasing ring diameter, even if the deep rings are slightly blurred due to multiple dielectric layer scattering, the point with the strongest edge gradient can still be reliably located, thus ensuring the robustness and repeatability of the offset extraction, which can be controlled within ±3μm. The entire optical transmission imaging and comparison process is completely non-contact and non-destructive, and can be completed using only standard AOI equipment or laboratory-grade microscopic imaging systems, with the advantages of speed, high throughput and low cost.
[0025] The offset data obtained by comparing optical perspective imaging with the center not only intuitively reflects the two-dimensional planar deviation distribution of each signal layer relative to the top reference, but also enables the entire detection method to achieve a progressive verification logic from rapid optical screening to high-precision electrical quantification while maintaining the integrity of the overall PCB structure. This ensures that the final output of interlayer alignment deviation results has both the intuitiveness and efficiency of optical methods, and also takes into account the high-resolution capture capability of electrical testing for deep alignment states.
[0026] S3: Based on the offset data, laser drilling is performed on the target area, and the resulting micro-hole array is sequentially metallized and subjected to on / off resistance tests to obtain electrical data; In step S3, based on the offset data, i.e., the preliminary offset vectors of each layer in the X and Y directions, a UV laser drilling machine is used to precisely process non-critical positions in the target area. The laser operates at a wavelength of 355nm, possessing high precision and high energy density, enabling the formation of fine through-holes with diameters controlled between 0.08 and 0.12mm without damaging surrounding materials. The arrangement of the through-hole array is not fixed but exhibits adaptive characteristics: the hole spacing is set to an integer multiple of the diameter difference between adjacent layers' rings preset during the design phase (e.g., if the diameter difference between adjacent layers is 10μm, the hole spacing can be set as a multiple thereof to match the deviation resolution scale). The array is arranged in a straight line or a small grid pattern, and its coverage length and range are dynamically adjusted according to the magnitude of the optical offset vector. When a large deviation is detected, the array is appropriately extended to ensure that at least all possible deviation threshold sampling points are included, thereby avoiding missing critical transition positions or generating too many invalid holes. Drilling positioning relies on the preset crosshair positioning marks and the center of the optical observation window in the top target group as a reference, achieving automatic centering through a machine vision system with a positioning accuracy better than ±5μm. The laser parameters are modulated to ensure that each via depth penetrates all dielectric layers until it reaches the underlying copper foil, but without piercing the underlying copper surface to maintain the overall structural integrity of the PCB. The resulting microvia array traverses each layer of copper rings in a multi-layered nested concentric target group, establishing a physical channel for electrical continuity testing. Furthermore, due to the extremely small aperture and the fact that each target group contains only 10 to 30 vias, the impact on the integrity of high-frequency signals, impedance continuity, and electromagnetic compatibility is negligible, fully meeting the stringent electrical performance requirements of ADAS high-frequency PCBs.
[0027] After laser drilling, the resulting microvia array undergoes metallization and continuity / resistance testing to obtain reliable electrical data. Specifically, standard PCB post-processing is employed, including plasma cleaning to remove via wall residue, chemical copper plating to form the initial conductive layer, and full-board copper plating. This results in the deposition of a uniform, continuous copper layer approximately 15 to 25 μm thick on the wall of each microvia, electrically connecting all the target rings traversing the layers. Because the target rings of each signal layer employ a nested concentric circle design, with diameters increasing progressively from small to large and the diameter difference between adjacent layers being a known, precise constant, when there is an actual alignment deviation between layers, the microvia may fall precisely inside the copper region of a certain ring, resulting in low-resistance conduction, or it may fall outside the ring in the dielectric region, maintaining a high-resistance open circuit. Subsequently, using a flying probe tester or a dedicated micro-probe fixture, each microvia is contacted from both the top and bottom layers, and a low-voltage test signal is applied to each microvia in the array, recording the continuity (conductivity or open circuit) and corresponding resistance value of each via. The measured on / off modes are precisely compared with the theoretical conduction threshold pre-calculated based on the optical offset vector. For example, if the k-th micro-hole in a certain direction unexpectedly conducts, it indicates that the cumulative alignment deviation in that direction has exceeded the sum of the diameter differences of the previous k layers. Based on this, the actual deviation value of the signal layer can be deduced. The resolution of electrical verification can reach the ±8 to 12 μm range. By converting the information of deep targets that are difficult to locate precisely in optical imaging into clear binary electrical signals and quantized resistance data, not only is the depth and accuracy of detection improved, but the non-destructive and automated characteristics of the entire process are also maintained.
[0028] S4: The offset data and electrical data are weighted and fused to obtain the alignment deviation results between each signal layer.
[0029] In step S4, the preliminary optical offset vector and the electrical continuity test data are weighted and fused to combine the advantages of the two complementary detection methods, thereby obtaining the true alignment deviation between each signal layer relative to the top layer (or reference layer). This enables a reliable quantitative determination of the interlayer alignment accuracy of the ADAS high-frequency multilayer PCB finished board. In the aforementioned steps, the preliminary X and Y offset vectors of each signal layer target group relative to the top layer target group have been obtained through optical perspective imaging. The offset vectors are calculated based on image edge detection and circle center fitting, which are intuitive, fast, and non-contact. However, due to factors such as light scattering from the high-frequency substrate, cumulative blurring of the dielectric layer thickness, and contrast attenuation of deep targets, its accuracy often has certain uncertainties in deeper layers, fluctuating within the range of ±15 to 25 μm. Meanwhile, the micro-hole array formed by adaptive laser drilling establishes a physical and electrical channel that runs through all signal layers after metallization. Whether each micro-hole falls within a certain copper ring determines the on / off state at that location. Since the diameter of the target ring increases strictly and the diameter difference between adjacent layers is a known precise constant, when the micro-hole array is arranged at a fixed interval along a specific direction, the turning point of conduction / open circuit precisely corresponds to the critical position where the actual cumulative deviation exceeds a certain threshold. This binary electrical signal (conduction or open circuit) plus the quantized information of the resistance value has extremely high resolution (up to ±8~12μm), which is particularly suitable for verifying the reliability of optical data at deep targets. Therefore, a preliminary deviation coordinate system for each signal layer is constructed based on the optical offset vector. The X / Y offset value of each layer is mapped to a unified reference plane. The on / off mode obtained from the electrical test is used as a key verification threshold point for comparison. For example, when the optical prediction indicates that the kth micro-hole in a certain direction should be in an open circuit state, but the actual test shows that it is conductive, it indicates that the actual cumulative deviation in that direction has exceeded the sum of the diameter differences of the previous k layers. The optical calculation value is corrected a second time using the electrical conduction point as the anchor point. The two are fused by a weighted algorithm (the optical data can have a higher weight for continuous quantities, and the electrical threshold can have a higher weight for discrete confirmation) to eliminate the systematic error caused by optical blurring. Finally, the corrected alignment deviation result of each layer is output.
[0030] Weighted fusion processing not only improves the overall accuracy and confidence of the detection, but also fully demonstrates the complementary characteristics of optical and electrical methods in high-frequency PCB interlayer alignment detection: the optical part provides rapid, full-field preliminary deviation distribution and vector direction information, while the electrical part provides precise threshold confirmation of deep targets through the on / off resistance test after microvia metallization. The combination of the two can generate a comprehensive interlayer alignment deviation report, including quantitative indicators such as the maximum X / Y deviation value of each signal layer relative to the top layer, the magnitude of vector synthesis deviation, the root mean square error of RMS, and the interlayer deviation matrix. At the same time, it can also draw a deviation distribution heatmap. The system generates multi-layer target perspective overlay images and electrical on / off state matrices. When the optical and electrical results are highly consistent, the fused value can be directly adopted as the final conclusion. If slight inconsistencies occur (mostly due to deep optical blurring), the electrical threshold point is prioritized for correction to ensure accurate and reliable deviation judgment, which is most sensitive to the integrity of high-frequency signals. The final determination of whether it is qualified is based on the typical interlayer alignment tolerance of ADAS high-frequency PCB (e.g., ±40~50μm), with process optimization suggestions, thereby providing efficient and reliable technical support for ADAS system to control the interlayer alignment quality of high-frequency multi-layer PCBs.
[0031] In one embodiment, prior to the step of identifying the target region of a multilayer PCB, the method further includes: Layered positioning marks are made on the multilayer PCB design layout, and each signal layer of the multilayer PCB is processed according to the positioning marks to form a target group on each signal layer. The processing includes developing, etching, and film removal processes.
[0032] In the above embodiments, during the manufacturing process of ADAS high-frequency multilayer PCBs, layer positioning marking is planned for each signal layer during the circuit layout design stage. Specifically, in the non-functional area on the board edge or in a specially set test coupon area, a set of nested concentric circle target groups is designed for each layer. Each target group consists of multiple copper rings with progressively increasing diameters. The number of rings corresponds to the total number of layers minus the surface layer, with the innermost ring having the smallest diameter as a benchmark. The diameter difference between adjacent layers is fixed as a precise constant, such as 8 micrometers or 10 micrometers, to ensure that the diameter sequence strictly corresponds one-to-one with the layer sequence. At the same time, a high-precision cross-shaped positioning mark is superimposed at the geometric center of each ring for subsequent imaging alignment. The target pattern, along with the main functional circuit, is transferred to each copper foil layer via LDI laser direct imaging or photoplotting. The pattern transfer accuracy is controlled within ±5 micrometers, fully considering the low loss of the high-frequency board. The optical transmittance characteristics of materials such as PTFE or LCP are considered. Appropriate copper thickness and ring linewidth are selected, preferably between 0.15 and 0.25 mm, to ensure sufficient contrast of the inner layer targets in the visible or near-infrared bands. The target group is positioned to avoid high-density wiring areas and is symmetrically distributed at the four corners or multiple points of the board to comprehensively characterize the alignment of the entire board. After the layout design is completed, the processing stage begins. For each signal layer of the multilayer PCB, a development process is performed sequentially to reveal the photoresist pattern, an etching process to remove the copper foil in non-patterned areas, and a stripping process to remove the remaining photoresist. This precisely forms the corresponding target group on each signal layer. Standard wet processing ensures that the target copper ring is completely preserved and highly consistent with the design layout. Through such layered positioning marks and pre-processed target groups, clear and identifiable interlayer feature marks are provided for identifying the target area. The entire process is seamlessly integrated into the conventional multilayer board pre-lamination inner layer processing flow without changing any electrical functions.
[0033] In one example, the step of identifying the target region of a multilayer PCB includes: Perform a global optical scan on a multilayer PCB to obtain a board surface image containing all preset target groups; Based on the plate image, template matching and positioning of the target graphic are performed to obtain the region boundary of each preset target group; The target region of the multilayer PCB is extracted by extending a preset safety margin based on the geometric center of the boundary of the region, thereby obtaining the target region of the multilayer PCB.
[0034] In the above embodiments, a global optical scan is performed on the entire completed multilayer PCB laminate. A high-resolution linear scan camera or area scan camera, together with a uniform illumination system, is used to acquire a complete board image containing all preset target groups at once. Preferably, this is done on an automated optical inspection (AOI) device. The scanning resolution is set to 10 micrometers / pixel or higher to ensure that the target details are clearly visible. The acquired board image is in RGB or grayscale format and includes target information at the top solder mask opening and the inner layer projection outline partially transmitted through the solder mask. Based on the image, template matching is performed to locate the target graphic. Normalized cross-correlation (NCC) or shape-based matching algorithms are used to correlate the pre-extracted standard target template from the design layout with the actual image region by region, finding the region with the highest matching degree. This determines the precise region boundary of each pre-set target group, represented by a rectangular or polygonal outline that surrounds all visible rings and crosshairs. After obtaining the region boundary coordinates of each target group, a preset safety margin is uniformly expanded outwards from the geometric center as a reference point. The margin is set to 1.2 to 1.5 times the diameter of the outermost ring or a fixed value such as 1.5 to 2 mm, ensuring that the expanded area completely covers all possible target graphics while providing sufficient buffer to prevent boundary misjudgment or windowing damage to functional circuits. The resulting polygonal or rectangular region is the final extracted multi-layer PCB target area. The entire recognition process is completed using non-contact optical methods, fully utilizing the efficiency of global scanning and the robustness of template matching. It can accurately locate multiple target positions against complex board backgrounds, providing a reliable positioning basis for precise local processing.
[0035] In one example, the step of windowing the target region to obtain an optical observation window exposing the top target group includes: Based on the target area, graphical mask data is generated, and the top solder mask layer of the target area is selectively removed according to the graphical mask data to form a local windowed area. Under coaxial illumination, the optical transmittance of the local windowed area is verified. For the local windowed area whose verification result is lower than a preset threshold, the window size is finely adjusted to obtain the observation window. Using the top target group of the target region as a reference, position offset correction is performed on each signal layer in the observation window to obtain the optical observation window that exposes the top target group.
[0036] In the above embodiments, for the target area, graphical mask data is generated based on the coordinates of the area. Specifically, in the CAM software, the target area is converted into a negative image of the solder mask layer, that is, only the target area is generated with a windowed image while the rest of the solder mask layer remains intact. Then, the mask data is input into a solder mask exposure machine or a laser direct imaging device to selectively expose and develop the top solder mask layer of a multilayer PCB, thereby achieving local windowing of the target area, removing the solder mask material at the corresponding location and exposing the top copper target group. After windowing, the optical transmittance of these locally windowed areas is verified under coaxial illumination. The light intensity transmission ratio of the central area of the window is measured using an integrating sphere or a dedicated transmittance meter. For transmittance lower than a preset threshold (such as the design expectation value), the transmittance is verified. The window boundary contour (85%) is finely adjusted. The window boundary is automatically or manually extended outward by several micrometers until the transmittance meets the standard. This avoids the impact of residual adhesive or surface treatment layer on the subsequent imaging clarity. The finely adjusted window becomes the formal optical observation window. Using the cross mark or circular geometric center of the top target group in the target area as a reference, the position offset of each signal layer target in the window field of view is corrected. Specifically, by acquiring multiple frames of images with different focal depths or using the known interlayer thickness to estimate the focal plane position, the coordinate system of the imaging system is adjusted layer by layer to align the center of each layer target with the top reference. This corrects and compensates for the slight overall offset and lens distortion that may exist after pressing, ensuring that the projection of all layer targets is centered on the top layer during perspective imaging.
[0037] In one example, the step of performing perspective imaging on the optical observation window and comparing the target group of each signal layer with the top target group to obtain offset data includes: Multispectral perspective images are acquired vertically downwards from the optical observation window in a time-division manner to obtain the original superimposed grayscale image; The original superimposed grayscale image is separated by interlayer projection to obtain a separated grayscale image of each signal layer; The edge points of the separated grayscale image of each signal layer are extracted, and the geometric center of the formed edge point set is fitted within the annular region to obtain the center coordinates of the target group of each signal layer. The offset data is obtained by calculating the offset between the center coordinates of each signal layer target group and the center coordinates of the top layer target group.
[0038] In the above embodiments, at the formed optical observation window, an imaging system equipped with a multispectral light source and a high-resolution camera is used to acquire multiple sets of perspective images vertically downwards from the window in a time-division manner. Specifically, multiple wavelengths such as visible light, near-infrared 850nm, or 950nm are used for sequential illumination and synchronous acquisition. Different wavelengths have different transmission and reflection characteristics for the dielectric layer and copper ring. The original image sequence acquired in the time-division manner is fused by channels to form an original superimposed grayscale image, which simultaneously superimposes the projection contours of all target rings from the top layer to the inner layer. Interlayer projection separation processing is performed on this superimposed grayscale image. This can be achieved using deconvolution algorithms based on known interlayer thickness and spectral response, U-Net-style separation networks trained by deep learning, or reconstruction techniques of multi-focal plane image stacks, to decompose the mixed projection into the unique signal layers of each signal layer. The grayscale images are separated, and each separated image highlights the edge of the corresponding layer's ring while suppressing interference from other layers. Edge extraction is performed on each separated grayscale image, using Canny edge detection or Sobel operator combined with subpixel refinement to obtain the edge point set of the target ring for each layer. Then, geometric center fitting is performed on these edge points within a preset effective area of the ring. Least square circle fitting, minimum bounding circle, or Hough circle transform combined with RANSAC can be used to remove outliers. Finally, the precise center coordinates of the target group for each signal layer are calculated. Using the center coordinates of the top target group as the zero reference point, the center coordinates of the other layers are subtracted from it to calculate the offset data in the X and Y directions. The offset data quantifies the inter-layer alignment deviation accumulated from the top layer to each inner layer throughout the entire processing process.
[0039] In one example, the step of laser drilling the target area based on the offset data includes: Based on the offset data, a drilling coordinate sequence for the target area is planned, and laser drilling is performed on the target area in stages based on the drilling coordinate sequence to form preliminary microholes; The preliminary micropores are surface-treated, and the surface-treated preliminary micropores are subjected to depth optical inspection from the top layer of the multilayer PCB downwards to obtain a qualified micropore array.
[0040] In the above embodiment, the offset vector is directly mapped to the physical coordinate system of the target area. Using the top-level crosshair as the absolute origin, the precise position of each expected sampling point is calculated based on the magnitude and direction of the offset, forming one or more straight line sequences, grid sequences, or coordinate sequences adaptively adjusted according to the main direction of the deviation. This ensures that the drilling points can effectively cover the possible deviation range and avoid the critical edge transition area of the target ring. After planning, the coordinate sequence is loaded using an ultraviolet laser drilling device, and a staged drilling operation is performed. The first stage rapidly drills a guide hole in the top layer with lower energy and higher speed. The second stage gradually increases the pulse energy and controls the focal depth, gradually expanding the hole diameter to the range of 0.08–0.12 mm. Simultaneously, it penetrates the dielectric material layer by layer until it approaches the bottom copper foil but does not completely penetrate the bottom copper surface, thereby forming a preliminary micro-hole array within the target area. The micro-holes physically traverse multiple layers of dielectric and copper ring structures from the top layer to the inner layer. After laser drilling, the microvias undergo preliminary surface treatment, including plasma cleaning to remove resin residue and carbide residue generated during drilling, and chemical micro-etching or immersion in drilling descaling solution, to achieve a clean and sufficiently rough surface on the hole walls. Using an optical microscope or confocal microscopy system equipped with a long working distance objective lens, the microvias are inspected vertically downwards through the top optical window of the multilayer PCB. Layer-by-layer focusing scanning or optical coherence tomography is used to check the integrity of each microvia wall, the presence of obvious taper anomalies, whether the bottom is clean and exposes the underlying copper surface, and the presence of blockages or residues. Simultaneously, the accuracy of the hole positions relative to the pre-planned coordinates and their reasonable relative position to the target ring is verified. Finally, microvia arrays with acceptable morphology and no obvious processing damage are selected as valid samples for subsequent metallization and electrical testing.
[0041] In one embodiment, the step of sequentially metallizing and conducting continuity and resistance tests on the formed microporous array includes: The walls of the micro-hole array are cleaned, and the cleaned micro-hole array is subjected to local chemical copper plating treatment to obtain locally metallized micro-holes. The copper layer thickness of the locally metallized micropores is homogenized using a complexing agent to obtain a locally electrically connected array of metallized micropores. The on / off resistance of the metallized micro-hole array was tested using a flying probe test to obtain electrical data.
[0042] In the above embodiments, plasma cleaning equipment or chemical cleaning solutions are used to thoroughly treat the inner walls of all micropores to remove residual organic debris, oxides, and trace contaminants from laser drilling, resulting in a uniformly activated surface on the pore walls. The cleaned micropore array is then fed into the chemical copper plating process. However, unlike conventional full-board copper plating, a localized selective chemical copper plating process is used here. Through mask protection or spray deposition, a thin layer of chemical copper (0.5–1.5 μm thick) is deposited only on the pore walls and near the pore openings, thus obtaining locally metallized micropores. Other board surfaces outside the target area remain free of additional copper deposition, avoiding unnecessary plating interference. After localized chemical copper deposition, a complexing agent system is introduced to homogenize the deposited copper layer. The complexing agent forms a stable complex with copper ions. By controlling the soaking time and solution pH, the copper layer thickness is made uniform across the entire depth of the pore wall, avoiding excessive thickness at the top or bottom. This ultimately forms a locally electrically connected metallized micropore array. At this point, the copper layer on the wall of each qualified micropore electrically connects the copper rings of the target cells that pass through it. Further, the continuity resistance of this metallized micropore array is tested using flying probe microscopy to obtain electrical data.
[0043] In one embodiment, the step of performing continuity resistance testing on the metallized microvia array based on flying probe testing to obtain electrical data further includes: Double-ended flying probes are used to contact the center of the top pore and the center of the bottom pore of each micropore in the metallized micropore array, and a preset low pressure is applied to form the original scan dataset. Based on the original scan dataset, a segmented, incremental test voltage is applied sequentially to each microhole to measure its continuity resistance, thereby obtaining the continuity state and resistance sequence of each microhole. Based on the conduction state and resistance sequence, the micro-hole array is classified into anomalies to form an electrical anomaly distribution map labeled with the anomaly type and the corresponding physical location of the micro-hole. Based on the electrical anomaly distribution map and the geometric arrangement of the micropore array, deviation inversion calculation is performed to obtain electrical data containing the actual alignment deviation between each signal layer.
[0044] In the above embodiments, the specific process of obtaining electrical data by flying probe testing of the metallized microvia array includes: applying double-ended flying probe contact to the center of the top and bottom apertures of each microvia in the metallized microvia array; positioning the test probe on the copper surface of the aperture with extremely low contact pressure; applying a preset low voltage (from a few millivolts to tens of millivolts) for initial scanning; and collecting and forming an original scan dataset, which contains the initial on / off state, initial contact resistance value, and position coordinate information of each microvia. Based on the original scan dataset, a segmented increasing test voltage is applied to each microvia, for example, gradually increasing from 10mV to 500mV or higher; each voltage segment is maintained for several seconds and the corresponding resistance change is recorded, thereby obtaining the complete conduction state and resistance sequence of each microvia, which can clearly distinguish between truly conductive low-resistance vias, edge-contact high-resistance vias, and completely open high-resistance vias. Based on the conduction state and resistance sequence, the entire microvia array is classified for anomalies. Vias with resistance values exceeding the normal range are marked as anomalies, and the anomaly types are further subdivided according to the resistance range and trend, such as complete open circuit, intermittent contact, excessively high contact resistance, and copper layer discontinuity. Each anomaly type is mapped one-to-one with the physical coordinates of the corresponding microvia, forming an electrical anomaly distribution map labeled with the anomaly type and specific microvia location. Using the geometric coordinates of the target area as the base map, the spatial distribution pattern of the anomaly points is visually displayed. Finally, based on the electrical anomaly distribution map and the geometric arrangement of the microvia array itself, deviation inversion calculation is performed. Utilizing the known microvia spacing and the increasing diameter of the target ring, the actual cumulative alignment deviation of each signal layer relative to the top layer is inferred from the location of the anomaly points. For example, if a certain microvia changes from open circuit to conduction in a certain direction, it indicates that the deviation in that direction has reached the sum of the diameter differences of the previous layers, thus calculating precise electrical data containing the actual alignment deviation between each signal layer.
[0045] In one embodiment, the step of performing weighted fusion processing on the offset data and electrical data to obtain the alignment deviation result between each signal layer includes: Feature vectors are constructed from the offset data and electrical data. Based on the formed multilayer fusion dataset, optical values of the micropores in each direction are corrected to obtain optical correction data for each signal layer. The ratio of the modulus of the optical correction data of each signal layer to the degree of anomaly in the electrical anomaly distribution map is calculated and weighted accordingly to obtain the alignment deviation data of each signal layer relative to the top layer. The alignment deviation data and the multi-layer fusion dataset are summarized and organized to obtain the alignment deviation results between each signal layer.
[0046] In the above embodiments, feature vectors are constructed from the offset data obtained by optical testing and the abnormal data obtained by electrical testing, respectively. The optical offset vectors are organized into multi-dimensional coordinate vectors according to the layer order. At the same time, the abnormal positions, abnormal types, and resistance sequences in the electrical anomaly distribution map are quantized and encoded into feature vectors of the corresponding layers. Then, the two are merged to form a multi-layer fusion dataset. In this fusion dataset, optical values are corrected for micro-apertures in each direction. Using the electrical conduction / open circuit threshold points as anchor points, secondary correction is performed on the center positioning deviation caused by edge blurring or light scattering that may exist in the optical calculation, resulting in more accurate optical correction data for each signal layer. The ratio between the magnitude of the optical correction data of each signal layer and the degree of anomaly in the corresponding electrical anomaly distribution map is calculated. For example, the density of abnormal points or the average level of abnormal resistance is used as the quantification index of the degree of anomaly. The fusion weight of each signal layer is dynamically allocated through this ratio, so that layers with clear optical data and fewer anomalies are given higher weights, while layers with blurred optical data but clear electrical thresholds rely more on electrical results. This yields the alignment deviation data of each signal layer relative to the top layer. These deviation data simultaneously include X-axis, Y-axis components and synthetic vector values. The alignment deviation data of each signal layer is summarized and organized with the multilayer fusion dataset. The deviation values of all layers are statistically analyzed, including the calculation of the maximum deviation, average deviation, standard deviation and deviation direction distribution trend between layers. After removing individual extreme outliers, the final alignment deviation results between each signal layer are generated and output in the form of numerical reports and deviation vector matrices. This can be directly used to determine whether the interlayer alignment tolerance requirements of ADAS high-frequency PCBs are met.
[0047] refer to Figure 2 An ADAS high-frequency PCB interlayer alignment accuracy detection device, comprising: The identification module 100 is used to identify the target area of the multilayer PCB, and to perform windowing processing on the target area to obtain an optical observation window that exposes the top target group, wherein each signal layer of the multilayer PCB is preset with a target group. The comparison module 200 is used to perform perspective imaging on the optical observation window, compare the target group of each signal layer with the top target group to obtain offset data; The test module 300 is used to perform laser drilling on the target area based on the offset data, and to perform metallization and on / off resistance tests on the formed micro-hole array in sequence to obtain electrical data. The data processing module 400 is used to perform weighted fusion processing on the offset data and electrical data to obtain the alignment deviation results between each signal layer.
[0048] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A method for detecting the alignment accuracy between layers of an ADAS high-frequency PCB, characterized in that, include: The target area of the multilayer PCB is identified, and the target area is windowed to obtain an optical observation window that exposes the top target group. Each signal layer of the multilayer PCB has a pre-set target group. Perspective imaging is performed on the optical observation window, and the target group of each signal layer is compared with the top target group to obtain offset data; Based on the offset data, laser drilling is performed on the target area, and the resulting micro-hole array is sequentially metallized and subjected to on / off resistance tests to obtain electrical data. The offset data and electrical data are weighted and fused to obtain the alignment deviation results between each signal layer.
2. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs according to claim 1, characterized in that, Before the step of identifying the target region of a multilayer PCB, the method further includes: Layered positioning marks are made on the multilayer PCB design layout, and each signal layer of the multilayer PCB is processed according to the positioning marks to form a target group on each signal layer. The processing includes developing, etching, and film removal processes.
3. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs according to claim 1, characterized in that, The step of identifying the target region of a multilayer PCB includes: Perform a global optical scan on a multilayer PCB to obtain a board surface image containing all preset target groups; Based on the image of the plate, template matching and positioning of the target graphic are performed to obtain the region boundary of each preset target group; The target region of the multilayer PCB is extracted by extending a preset safety margin based on the geometric center of the boundary of the region, thereby obtaining the target region of the multilayer PCB.
4. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCB according to claim 1, characterized in that, The step of performing windowing processing on the target area to obtain an optical observation window exposing the top target group includes: Based on the target area, graphical mask data is generated, and the top solder mask layer of the target area is selectively removed according to the graphical mask data to form a local windowed area. Under coaxial illumination, the optical transmittance of the local windowed area is verified. For the local windowed area whose verification result is lower than a preset threshold, the window size is finely adjusted to obtain the observation window. Using the top target group of the target region as a reference, position offset correction is performed on each signal layer in the observation window to obtain the optical observation window that exposes the top target group.
5. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs according to claim 1, characterized in that, The step of performing perspective imaging on the optical observation window and comparing the target group of each signal layer with the top target group to obtain offset data includes: Multispectral perspective images are acquired vertically downwards from the optical observation window in a time-division manner to obtain the original superimposed grayscale image; The original superimposed grayscale image is separated by interlayer projection to obtain a separated grayscale image of each signal layer; The edge points of the separated grayscale image of each signal layer are extracted, and the geometric center of the formed edge point set is fitted within the annular region to obtain the center coordinates of the target group of each signal layer. The offset data is obtained by calculating the offset between the center coordinates of each signal layer target group and the center coordinates of the top layer target group.
6. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCB according to claim 1, characterized in that, The step of performing laser drilling on the target area based on the offset data includes: Based on the offset data, a drilling coordinate sequence for the target area is planned, and laser drilling is performed on the target area in stages based on the drilling coordinate sequence to form preliminary microholes; The preliminary micropores are surface-treated, and the surface-treated preliminary micropores are subjected to depth optical inspection from the top layer of the multilayer PCB downwards to obtain a qualified micropore array.
7. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCBs according to claim 1, characterized in that, The steps of sequentially metallizing and conducting continuity and resistance tests on the formed micro-hole array include: The walls of the micro-hole array are cleaned, and the cleaned micro-hole array is subjected to local chemical copper plating treatment to obtain locally metallized micro-holes. The copper layer thickness of the locally metallized micropores is homogenized using a complexing agent to obtain a locally electrically connected array of metallized micropores. The on / off resistance of the metallized micro-hole array was tested using a flying probe test to obtain electrical data.
8. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCB according to claim 7, characterized in that, The step of performing continuity resistance testing on the metallized microvia array based on flying probe testing to obtain electrical data further includes: Double-ended flying probes are used to contact the center of the top pore and the center of the bottom pore of each micropore in the metallized micropore array, and a preset low pressure is applied to form the original scan dataset. Based on the original scan dataset, a segmented, incremental test voltage is applied sequentially to each microhole to measure its continuity resistance, thereby obtaining the continuity state and resistance sequence of each microhole. Based on the conduction state and resistance sequence, the micro-hole array is classified into anomalies to form an electrical anomaly distribution map labeled with the anomaly type and the corresponding physical location of the micro-hole. Based on the electrical anomaly distribution map and the geometric arrangement of the micropore array, deviation inversion calculation is performed to obtain electrical data containing the actual alignment deviation between each signal layer.
9. The method for detecting the interlayer alignment accuracy of ADAS high-frequency PCB according to claim 8, characterized in that, The step of weighted fusing the offset data and electrical data to obtain the alignment deviation results between each signal layer includes: Feature vectors are constructed from the offset data and electrical data. Based on the formed multilayer fusion dataset, optical values of the micropores in each direction are corrected to obtain optical correction data for each signal layer. The ratio of the modulus of the optical correction data of each signal layer to the degree of anomaly in the electrical anomaly distribution map is calculated and weighted accordingly to obtain the alignment deviation data of each signal layer relative to the top layer. The alignment deviation data and the multi-layer fusion dataset are summarized and organized to obtain the alignment deviation results between each signal layer.
10. A device for detecting the alignment accuracy between layers of an ADAS high-frequency PCB, characterized in that, include: The identification module is used to identify the target area of the multilayer PCB, and to perform windowing processing on the target area to obtain an optical observation window that exposes the top target group, wherein each signal layer of the multilayer PCB has a preset target group. The comparison module is used to perform perspective imaging on the optical observation window, compare the target group of each signal layer with the top target group to obtain offset data; The testing module is used to perform laser drilling on the target area based on the offset data, and to perform metallization and on / off resistance tests on the formed micro-hole array in sequence to obtain electrical data. The data processing module is used to perform weighted fusion processing on the offset data and electrical data to obtain the alignment deviation results between each signal layer.
