A circuit board testing method and system

Abstract

This invention discloses a circuit board inspection method and system, including a control module, an acquisition module, a display module, a database, and an identification module. The control module is connected to the acquisition module, display module, database, and identification module. The acquisition module is used to acquire image data of copper-clad laminates and soldered circuit boards. The identification module is used to identify the circuit board status information based on the image data acquired by the acquisition module. The database includes a primary database and a final database. The display module is used to read and display the display data stored in the final database. This invention ensures that copper-clad laminate errors can be detected and that boards with copper-clad laminate errors are not used for component soldering, thus improving the overall processing speed. Furthermore, the fixed-format data display allows operators to easily and clearly read the circuit board status information.

Description

Technical Field

[0001] This invention relates to the field of circuit board testing technology, and in particular to a circuit board testing method and system. Background Technology

[0002] With the rapid development of electronic integration technology, the integration level of electronic components on circuit boards is constantly increasing, which also requires increasingly complex manufacturing processes and raises costs. Therefore, it is necessary to inspect the circuit board before and after the components are laid out to detect defects and avoid the inability of electrical appliances based on the circuit board to perform their functions due to defects.

[0003] Application No. 202210084673.4 discloses a batch inspection system and method for hardware-driven circuit boards based on LSTM. The system includes a data acquisition module, an inspection module, a processor, a power supply module, a database, an auxiliary inspection module, an information display module, a prediction module, a communication module, and an inspection monitoring module. The main advantage of this invention is the use of the prediction module, which allows it to construct a long short-term memory (LSTM) model for prediction based on historical inspection data. This LSTM model can then be used to predict the defect thresholds for each batch of circuit boards in different workshops and time periods. When the actual inspection data exceeds the predicted threshold, an alarm is issued and it is displayed that the pass rate of that batch of circuit boards has not met the standard. However, this patent only compares images of the soldered circuit boards and does not compare the copper plating process before soldering. This could lead to situations where the soldered images appear fine, but problems exist in the copper plating process. In such cases, the judgment result might be "good," but in reality, copper plating defects still exist. Furthermore, the manual detection process only involves one step and lacks a follow-up safeguard in case of discrepancies between manual and model detection. Therefore, inaccurate results may occur when manual and model detections are inconsistent. Moreover, the patent does not integrate or process the detected results, failing to demonstrate the rigor of the results. It is impossible to intuitively understand how many procedures were involved in the judgment. The accuracy of the patent's results is quantified by a custom percentage (confidence level), which is highly susceptible to subjective influence and may deviate significantly from objective facts, ultimately leading to inaccurate judgments. Summary of the Invention

[0004] The circuit board testing method and system described in this invention solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a circuit board testing method, comprising the following steps:

[0006] S1: Establish an information storage location in the database and collect image data of the circuit board after copper plating;

[0007] S2: Determine if there is an error in the copper pouring. If yes, update the data carrying the copper pouring error information in the corresponding information storage location of the primary database. If no, proceed to step S3.

[0008] S3: Perform the soldering operation, and collect image data of the soldered circuit board after the soldering is completed;

[0009] S4: Determine whether there is an error after welding. If not, update the data carrying the information that there is no welding error in the corresponding information storage location of the primary database. If yes, update the data carrying the welding error and error type in the corresponding location of the primary database.

[0010] S5: Decide whether to sample the data stored in the primary database. If not, add the information that no sample is being sampled to the data in the primary database and store it in the final database. If yes, proceed to step S6.

[0011] S6: Manually determine whether the judgment results of steps S2 and S4 of the circuit board are incorrect. If the manual judgment result is consistent with the judgment results of steps S2 and S4, the data in the primary database is superimposed with the information that there is sampling and the judgment result is consistent and then put into the final database for storage. If the manual judgment result is inconsistent with the judgment results of steps S2 and S4, then proceed to step S7.

[0012] S7: Organize the joint review team to conduct the final data verification, and then put the data in the primary database into the final database after overlaying the information containing the joint review team's verification and the final verification results;

[0013] S8: After retrieving the data stored in the final database, performing calculations and processing to generate display data, the data is then stored in the final database for the display module to read and display.

[0014] Preferably, the information mentioned in step S1 includes: circuit board number, copper plating condition, soldering condition, error type, whether sampling inspection was conducted, sampling inspection results, whether verification inspection was conducted, and verification inspection results.

[0015] Preferably, the determination of whether the copper pouring is erroneous in step S2 is as follows:

[0016] S201: A convolutional neural network model is trained using a copper-clad laminate, and the trained judgment model is tested using a copper-clad laminate test set.

[0017] S202: Input the collected copper-clad image data into the trained judgment model for judgment, obtain the judgment result, and transmit the judgment result back to the convolutional neural network model for data update;

[0018] S203: Output the judgment result.

[0019] Preferably, the determination of whether there is an error after welding in step S4 is as follows:

[0020] S401: Collect circuit boards of different types of soldering errors as training sets to train the convolutional neural network model to obtain a trained judgment model. This model is used to output the soldering error result and the soldering error type.

[0021] S402: Input the collected post-weld image data into the trained judgment model for judgment, obtain the judgment result, and transmit the judgment result back to the convolutional neural network model for data update;

[0022] S403: Output the judgment result.

[0023] Preferably, the data format stored in the ultimate database is as follows: circuit board number (D1) - copper plating status (D2) - soldering status (D3) - error type (D4) - whether to conduct a spot check (D5) - spot check result (D6) - whether to conduct a verification check (D7) - verification check result (D8); the data of D1 is the actual circuit board number, the data of D2 is 1 (copper plating error exists) or 0 (copper plating error does not exist), the data of D3 is 1 (soldering error exists) or 0 (soldering error does not exist), the data of D4 is the text of the soldering error type or 0 (soldering error does not exist), the data of D5 is 1 (spot check exists) or 0 (spot check does not exist), the data of D6 is 1 (spot check result is consistent with the model judgment result) or 0 (spot check result is inconsistent with the model judgment result) or * (spot check does not exist), the data of D7 is 1 (verification check exists) or 0 (verification check does not exist), and the data of D8 is the text of the specific verification check result or 0 (verification check does not exist).

[0024] Preferably, the display data format in step S8 is: D1-circuit board status-accuracy level; the circuit board status specifically refers to the textual display of data from one of D7, D4, D3, or D2, wherein the data priority is D7 > D4 > D3 > D2. When a high-priority data is 0, data with lower priority is considered, and when the lowest-priority data is also 0, the textual information of the status data when the lowest-priority data is 0 is output; the accuracy level is calculated as: 1 + D5 + D7.

[0025] A circuit board inspection system, applied to the circuit board inspection method described above, is characterized by comprising a control module, an acquisition module, a display module, a database, and an identification module; the control module is connected to the acquisition module, the display module, the database, and the identification module respectively; the acquisition module is used to acquire image data of copper-clad laminates and soldered circuit boards; the identification module is used to identify circuit board status information based on the image data acquired by the acquisition module; the database includes a primary database and a final database; the primary database is used to store data obtained by the identification module through model recognition; the final database module is used to store data including data obtained by the identification module through model recognition and manual recognition; and the display module is used to read the display data stored in the final database and display it.

[0026] An electronic device includes a processor and a storage device; the storage device stores a program that is executed by the processor to implement the method described above.

[0027] A computer medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described above.

[0028] The beneficial effects of this invention are as follows:

[0029] Firstly, this invention adds a copper plating judgment step before the conventional inspection of circuit boards with soldered components by setting two steps: copper plating judgment and post-soldering judgment. This ensures that copper plating errors can be detected on the circuit board, and also prevents the board with copper plating errors from being used for component soldering when they occur, and also eliminates the need for soldering error judgment. On the one hand, this can reduce the loss of components and solder, and on the other hand, it can also simplify the computation of the soldering error model and improve the overall computing speed.

[0030] Secondly, this invention displays data in a fixed format (D1-circuit board status-accuracy level), allowing staff to easily and clearly read the circuit board status information. Furthermore, because this patent incorporates the calculation and presentation of accuracy levels, each circuit board status judgment result can correspond to an accuracy level. With each additional level of detection, the accuracy level increases by one level, thus demonstrating the accuracy of the circuit board status data. The standardized data display also facilitates subsequent data retrieval. Attached Figure Description

[0031] Figure 1 This is a schematic flowchart of a circuit board testing method according to the present invention;

[0032] Figure 2 This is a schematic diagram of the circuit board testing system according to the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0035] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0036] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0037] An embodiment of the present invention provides a circuit board testing method, such as... Figure 1 As shown, it includes the following steps:

[0038] S1: Establish an information storage location in the database and collect image data of the circuit board after copper plating;

[0039] The database here includes a primary database and a final database. The primary database stores data obtained through model recognition, while the final database module stores data obtained through both model recognition and manual identification. The data in the final database includes both raw and displayed data. The information mentioned above regarding "establishing information storage locations in the database" includes: circuit board number, copper plating status, soldering status, error type, whether sampling inspection was conducted, sampling inspection results, whether verification inspection was conducted, and verification inspection results. Verification inspection is conducted by the joint review team. Furthermore, the data stored in the primary and final databases does not overlap. Even if the status results (copper plating status, error type, etc.) of a certain circuit board are inconsistent between the primary and final databases, each status result data will be retained for future reference in case of disputes.

[0040] The data format stored in the primary database is: circuit board number (D1) - copper coverage (D2) - soldering status (D3) - error type (D4); the data format of the raw data stored in the final database is: circuit board number (D1) - copper coverage (D2) - soldering status (D3) - error type (D4) - whether to conduct sampling inspection (D5) - sampling inspection result (D6) - whether to conduct verification inspection (D7) - verification inspection result (D8), and the displayed data format is: D1 - circuit board status - accuracy level.

[0041] The data for D1 is the actual circuit board number; the data for D2 is 1 (copper plating error exists) or 0 (copper plating error does not exist); the data for D3 is 1 (soldering error exists) or 0 (soldering error does not exist); the data for D4 is text indicating the soldering error type or 0 (soldering error does not exist); the data for D5 is 1 (sampling inspection exists) or 0 (sampling inspection does not exist); the data for D6 is 1 (sampling inspection result matches the model judgment result) or 0 (sampling inspection result does not match the model judgment result) or * (sampling inspection does not exist); the data for D7 is 1 (verification inspection exists) or 0 (verification inspection does not exist); the data for D8 is text indicating the specific verification inspection result or 0 (verification inspection does not exist). The circuit board status and accuracy level will be explained in detail later.

[0042] S2: Determine if there is an error in the copper pouring. If yes, update the data carrying the copper pouring error information in the corresponding information storage location of the primary database. If no, proceed to step S3. The specific steps for determining if there is an error in the copper pouring are as follows:

[0043] S201: A convolutional neural network model is trained using copper-clad laminate images, and the trained judgment model is tested using a copper-clad laminate test set.

[0044] S202: Input the collected copper-clad image data into the trained judgment model for judgment, obtain the judgment result, and transmit the judgment result back to the convolutional neural network model for data update;

[0045] S203: Output the judgment result.

[0046] S3: Perform the soldering operation, and collect image data of the soldered circuit board after the soldering is completed;

[0047] S4: Determine if there was an error after welding. If not, update the data containing the information that welding was not problematic in the corresponding information storage location of the primary database. If yes, update the data containing the information that welding was problematic and the type of error in the corresponding location of the primary database. The specific steps for determining if there was an error after welding are as follows:

[0048] S401: Collect circuit boards of different types of soldering errors as training sets to train the convolutional neural network model to obtain a trained judgment model. This model is used to output the soldering error result and the soldering error type.

[0049] The types of soldering errors here include missing components, misaligned components, and incorrect soldering.

[0050] S402: Input the collected post-weld image data into the trained judgment model for judgment, obtain the judgment result, and transmit the judgment result back to the convolutional neural network model for data update;

[0051] S403: Output the judgment result.

[0052] S5: Decide whether to sample the data stored in the primary database. If not, add the information that no sample is being sampled to the data in the primary database and store it in the final database. If yes, proceed to step S6.

[0053] S6: Manually determine whether the judgment results of steps S2 and S4 of the circuit board are incorrect. If the manual judgment result is consistent with the judgment results of steps S2 and S4, the data in the primary database is superimposed with the information that there is sampling and the judgment result is consistent and then put into the final database for storage. If the manual judgment result is inconsistent with the judgment results of steps S2 and S4, then proceed to step S7.

[0054] S7: Organize the joint review team to conduct the final data verification, and then put the data in the primary database into the final database after overlaying the information containing the joint review team's verification and the final verification results;

[0055] S8: After retrieving the data stored in the final database, performing calculations and processing to generate display data, the data is then stored in the final database for the display module to read and display.

[0056] The displayed data format is: D1-Circuit Board Status-Accuracy Level. The circuit board status specifically refers to the textual display of data from one of D7, D4, D3, or D2, with data priority in the order D7 > D4 > D3 > D2. When a higher priority data point is 0, lower priority data is considered, and when the lowest priority data point is also 0, the textual information of the status data when the lowest priority data point is 0 is output. The accuracy level is calculated as: 1 + D5 + D7. That is, when the circuit board has not undergone a sampling inspection step, D5 is 0, and D7 is also 0, resulting in an accuracy level of 1. Level 1 indicates that the displayed result has passed model verification. When the circuit board has undergone a sampling inspection step and the results of the manual sampling inspection are consistent with the model sampling inspection, D5 is 1, and D7 is 0, resulting in an accuracy level of 2. When the circuit board is inconsistent after the sampling inspection step and requires review by a joint audit team, D5 is 1, and D7 is 1, resulting in an accuracy level of 3. In addition, each level corresponds to a color identifier, which can appear in the form of font color, background color, and pre-mark color. These font colors, background colors, and pre-mark colors can be used to identify part or all of the displayed data. Each accuracy level corresponds to a color identifier to help staff quickly and clearly distinguish the level corresponding to each color identifier. Specifically, level one corresponds to yellow, level two corresponds to green, and level three corresponds to red. Since the human eye and human thinking patterns will default to red as a warning color, red is chosen to mark the circuit boards that are inconsistent during the sampling inspection and need to be checked by the joint review team. This serves as a warning to the staff and also indicates that there is a judgment error on the circuit board (whether it is a model error or a manual inspection error, this level must represent at least one type of judgment error). The reason for setting the color of the secondary level as green is that the human eye and human thinking patterns tend to associate green with safety and reasonableness and passability. Therefore, green is given to circuit board data that has been sampled and whose sampling results are consistent with the model's judgment results. This means that the model's judgment and the results of manual inspection are consistent. Relatively speaking, the level of "safe and reasonable passability" (corresponding to "trustworthy" in this patent) is relatively high and the probability of the model's judgment being correct is also relatively high. Therefore, green is used to mark it.

[0057] The circuit board status is displayed in several ways: "Copper Plating Error", "Missing Component", "Component Misalignment", "Soldering Error", and "No Soldering Error". This display result also has a priority level: Joint Review Group Conclusion > Manual Sampling Inspection Conclusion > Model Judgment Conclusion. When a higher priority data point is 0, lower priority data is considered. If the lowest priority data point is also 0, then the text information of the status data when the lowest priority data point is 0 is output.

[0058] For example, if a circuit board of model ABC has no errors in copper plating during model inspection, but has soldering errors (component misalignment), and is not subject to random inspection, then the original data of this circuit board in the final database is: abc-0-1-No errors in copper plating, component errors-0-*-0-0, and the displayed data of this circuit board in the final database is: abc-No errors in copper plating, component errors-1;

[0059] For example, a circuit board of model 'def' shows no errors in copper plating during model inspection, but has soldering errors (type 'soldering error'). Random sampling results are inconsistent, and the joint review team's assessment indicates a missing component. In this case, the original data for this circuit board in the final database would be: def-0-1-No copper plating error, soldering error-1-0-1-No copper plating error, missing component. The displayed data for this circuit board in the final database would be: def-No copper plating error, missing component-3.

[0060] For example, a circuit board of model 'def' shows no errors in copper plating during model inspection, but has soldering errors (type 'soldering error'). Random sampling yields consistent results. In this case, the original data for this circuit board in the final database would be: def-0-1-No copper plating error, soldering error-1-1-0-0, and the displayed data in the final database would be: def-No copper plating error, soldering error-2.

[0061] This embodiment also provides a circuit board testing system, such as Figure 2 As shown, the system includes a control module, a data acquisition module, a display module, a database, and a recognition module. The control module is connected to the data acquisition module, the display module, the database, and the recognition module. The data acquisition module is used to acquire image data of the copper-clad laminate and the soldered circuit board. The recognition module is used to recognize the circuit board status information based on the image data acquired by the data acquisition module. The database includes a primary database and a final database. The primary database is used to store data obtained by the recognition module through model recognition. The final database module is used to store data obtained by the recognition module through model recognition and manual recognition. The display module is used to read the display data stored in the final database and display it.

[0062] In summary, this invention, through its two-step process of copper plating judgment and post-soldering judgment, adds a copper plating judgment step before the conventional inspection of soldered circuit boards. This ensures that copper plating errors can be detected on the circuit board and prevents the board with the copper plating error from being used for component soldering, thus avoiding further soldering error judgment. This reduces component and solder waste and simplifies the computational workload of the soldering error model, improving the overall computational speed. Furthermore, this invention uses a fixed-format data display (D1-circuit board status-accuracy level), allowing operators to easily and clearly read the circuit board status information. Because this patent incorporates accuracy level calculation and presentation, each circuit board status judgment result corresponds to an accuracy level. Each additional level of inspection increases the accuracy level, demonstrating the accuracy of the circuit board status data. The standardized data display also facilitates subsequent data retrieval, solving the problems raised in the background art.

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method of detecting a circuit board, characterized by, It comprises the following steps: S1: establishing information storage position in database, collecting copper-clad circuit board picture data; the information comprises: circuit board label, copper-clad condition, soldering condition, error type, whether to be sampled, sampling result, whether to be checked and checking result; S2: judging whether the copper-clad is wrong, if yes, updating the data carrying copper-clad error information in the corresponding information storage position of the primary database, if no, then step S3 is performed; S3: performing soldering operation, collecting the soldered circuit board picture data after soldering; S4: judging whether the soldering is wrong, if no, then updating the data carrying no soldering error information in the corresponding information storage position of the primary database, if yes, then storing the data carrying soldering error and error type in the corresponding position of the primary database, the data format stored in the primary database is: circuit board label D1-copper-clad condition D2-soldering condition D3-error type D4; wherein D1 is the actual circuit board label; D2 is 1 or 0, wherein 1 means copper-clad error exists, and 0 means copper-clad error does not exist; D3 is 1 or 0, wherein 1 means soldering error exists, and 0 means soldering error does not exist; D4 is the text of soldering error type or 0, wherein 0 means soldering error does not exist; S5: deciding whether to sample the data stored in the primary database, if no, then storing the data in the primary database in the ultimate database after adding the information carrying no sampling, if yes, then step S6 is performed; S6: artificially judging whether the results of steps S2 and S4 are wrong, if the artificial judgment result is consistent with the results of steps S2 and S4, then storing the data in the primary database in the ultimate database after adding the information carrying sampling and consistent results, if the artificial judgment result is not consistent with the results of steps S2 and S4, then step S7 is performed; S7: the organization of the joint review group for the final data check, and the data in the primary database superimposed on the presence of joint review group check and final check results information into the ultimate database storage, the ultimate database storage of the original data format is: circuit board label D1 - copper case D2 - welding case D3 - error type D4 - whether the sampling D5 - sampling results D6 - whether the nuclear D7 - nuclear results D8; wherein D5 is 1 or 0, wherein 1 is the presence of sampling, 0 is not the presence of sampling; D6 is 1 or 0 or , wherein 1 is the sampling result and the model judgment result is consistent, 0 is the sampling result and the model judgment result is not consistent, , 0 is not the presence of sampling; D7 is 1 or 0, wherein 1 is the presence of nuclear, 0 is not the presence of nuclear; D8 is the specific nuclear results text or 0, wherein 0 is not the presence of nuclear; S8: retrieving the data stored in the ultimate database, generating display data after processing, and storing in the ultimate database for the display module to read and display, the display data format is: D1-circuit board state-accuracy level; wherein the circuit board state is the text display of one data in D7, D4, D3 or D2, the data priority is D7>D4>D3>D2, when the data with high priority is 0, the data with low priority is considered, and when the data with the lowest priority is also 0, the text information of the state data with the lowest priority 0 is output; the calculation method of the accuracy level is: 1+D5+D7.

2. The method of claim 1, wherein The judgment of whether the copper-clad is wrong in step S2 is as follows: S201: training a convolutional neural network model using copper-clad board picture data, testing the convolutional neural network model using a copper-clad board test set to obtain a trained judgment model; S202: inputting the collected copper-clad picture data into the trained judgment model for judgment, obtaining the judgment result, and transmitting the judgment result back to the convolutional neural network model for data updating; S203: outputting the judgment result.

3. The method of claim 1, wherein The judgment of whether the soldering is wrong in step S4 is as follows: S401: Collect different kinds of circuit boards with welding error types as training sets, train the convolutional neural network model to obtain a trained judgment model, and output the welding error result and the welding error type; S402: Input the collected picture data of the welded circuit board into the trained judgment model for judgment, obtain the judgment result, and reversely transmit the judgment result to the convolutional neural network model for data updating; S403: Output the judgment result.

4. A circuit board inspection system for use in a circuit board inspection method as claimed in any one of the claims 1-3, characterized in that The control module is connected with the acquisition module, the display module, the database and the identification module; the acquisition module is used for acquiring picture data of the copper-clad plate and the welded circuit board; the identification module is used for identifying the state information of the circuit board according to the picture data acquired by the acquisition module; the database includes a primary database and a final database; the primary database is used for storing data identified by the model in the identification module; the final database module is used for storing data identified by the model and manually identified in the identification module; and the display module is used for reading and displaying the display data stored in the final database.

5. An electronic device comprising: A processor and a storage device; The storage device stores a program, and the program is executed by the processor to realize the method of any one of claims 1-3.

6. A computer medium having stored thereon a computer program, wherein, The program is executed by the processor to realize the method of any one of claims 1-3.

Images