FPC board anti-re-lamination detection method based on target detection

Abstract

The invention relates to the technical field of FPC board lamination, and provides an FPC board anti-re-lamination detection method based on target detection, and the method comprises the following steps: S1, obtaining the lamination process information of an FPC board; s2, performing detection processing according to the lamination process information of the FPC board to obtain lamination data of the FPC board; s3, performing verification processing according to the fitting data of the FPC board to establish a fitting information database of the FPC board; s4, obtaining a detection result of anti-re-lamination of the FPC board based on target detection processing by utilizing the lamination information database of the FPC board; compared with an existing detection method, the speed is increased, meanwhile, the precision is improved, detection is more efficient, and the problems that at present, the detection precision of repeatedly-attached products is low, and the detection efficiency is low are solved.

Description

Technical Field

[0001] The invention relates to the technical field of FPC board lamination, and in particular to an FPC board anti-re-lamination detection method based on target detection. Background Art

[0002] As electronic products become increasingly small and miniaturized, flexible circuit (FPC) boards have emerged. FPC board reinforcement can be classified differently according to different material types, including polyimide (PI) reinforcement and polyester (PET) reinforcement. Metal materials can be divided into aluminum reinforcement and stainless steel reinforcement. In addition, there are FR4 reinforcement and other materials. Different reinforcement materials have different properties. Metal materials can dissipate heat, conduct electricity easily, have high hardness, and PI reinforcement is tough and easy to connect. The thickness of reinforcement is also different. Reinforcement materials such as PI materials are relatively thin, while reinforcement materials such as FR4 and stainless steel are relatively thick. The difference in bonding thickness is large. When the above materials are bonded on the FPC board with a reinforcement machine, there will be repeated bonding, wrong bonding, missing bonding, etc. Usually, the bonding of the product is judged by testing the reinforcement thickness of the material to screen out unqualified products.

[0003] At present, the FPC reinforcement thickness detection mostly relies on manual sampling and visual inspection by workers, which cannot effectively intercept 100% of the repeated bonding products. Usually, a camera is used to detect whether the material is affixed. However, due to different materials, the camera may not be able to recognize or misidentify. In addition, it is impossible to detect whether it is re-bonded only through CCD detection, and an additional detection process is added. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an FPC board anti-re-bonding detection method based on target detection.

[0005] A method for detecting an FPC board against re-lamination based on target detection comprises the following steps: S1. Obtain the bonding process information of the FPC board; S2, performing detection processing according to the bonding process information of the FPC board to obtain the bonding data of the FPC board; S3, performing verification processing according to the bonding data of the FPC board to establish a bonding information database of the FPC board; S4, obtaining the detection result of the FPC board anti-re-lamination based on the target detection processing by utilizing the FPC board lamination information database.

[0006] Furthermore, the obtaining of the bonding process information of the FPC board includes: S1-1, obtain FPC board bonding material information; S1-2, obtaining theoretical bonding thickness data of the FPC board according to the bonding material information of the FPC board and establishing a correlation between the bonding material and the theoretical bonding thickness; S1-3, obtaining bonding material classification information according to the correlation between the bonding material and the theoretical bonding thickness; S1-4, obtaining the bonding process information of the FPC board according to the bonding material classification information; The bonding process information of the FPC board includes a bonding method, a bonding temperature threshold, a pressure threshold and a time threshold.

[0007] Further, performing detection processing according to the bonding process information of the FPC board to obtain the bonding data of the FPC board includes: S2-1, detecting and processing the surface of the FPC board based on the bonding process information of the FPC board using a Kinect depth sensor to obtain corresponding point cloud data; S2-2, performing reverse data processing according to the point cloud data to respectively obtain fitting area data and fitting shape data, and establishing a primary mapping between the fitting shape and the fitting area; S2-3, obtaining fitting size data according to the primary mapping, and establishing a secondary mapping of the fitting size, fitting shape and area; S2-4, obtaining the result of the FPC board's lamination state judgment based on the secondary mapping and grating ruler detection processing; S2-5. Outputting corresponding thickness data as the FPC board bonding data according to the bonding state judgment result of the FPC board.

[0008] Furthermore, the corresponding point cloud data obtained by detecting and processing the surface of the FPC board based on the Kinect depth sensor according to the bonding process information of the FPC board includes: S2-1-1, according to the bonding process information of the FPC board, using the Kinect depth sensor to perform detection processing based on TOF technology to obtain the transmission signal time t0 and the reception signal time t1 respectively; S2-1-2, the calculation formula for calculating the TOF value according to the received signal time t1 and the transmitted signal time t0 is as follows: TOF=t1-t0; S2-1-3. The calculation formula for calculating the distance information according to the TOF value is as follows: d = C × TOF / 2; S2-1-4, using the distance information to obtain corresponding point cloud data based on a Kinect depth sensor; The TOF value is the time it takes for light to fly a certain distance in a medium, d is the distance the light flies in the medium, and C is the speed of light, C=3×10 8m / s.

[0009] Furthermore, the results of the FPC board bonding state judgment obtained based on the secondary mapping and the grating ruler detection processing include: S2-4-1, obtaining the FPC board bonding starting position data based on the secondary mapping and grating ruler detection processing; S2-4-2, obtaining the bonding termination position data of the FPC board based on the corresponding relationship data between the bonding shape and the bonding size based on the grating ruler detection processing; S2-4-3, using the difference between the bonding start position data and the bonding end position data of the FPC board as bonding position moving distance data; S2-4-4, determine whether the bonding position moving distance data is 0, if so, the FPC board bonding state is obtained as missing bonding, and S2-5 is executed, otherwise, the FPC board bonding state is obtained as bonded, and S2-5 is executed.

[0010] Further, performing verification processing according to the bonding data of the FPC board to establish a bonding information database of the FPC board includes: S3-1, analyzing and judging the bonding data of the FPC board to obtain bonding thickness deviation data; S3-2, using the lamination thickness deviation data to verify the lamination data of the FPC board to obtain the lamination state data of the FPC board; S3-2-1, using the lamination thickness deviation data to correct the lamination data of the FPC board to obtain corrected thickness data; S3-2-2, determine whether the corrected thickness data is the same as the theoretical bonding thickness data of the FPC board. If so, obtain the size, shape, and area data corresponding to the corrected thickness data according to the secondary mapping relationship as the corrected bonding data, and execute S3-2-3. Otherwise, directly execute S3-2-4; S3-2-3, judging whether the corrected bonding data is the same as the theoretical bonding data, if so, outputting the corrected thickness data as qualified data of the bonding state of the FPC board and directly executing S3-3, otherwise, outputting the corrected thickness data as wrong bonding data; S3-2-4, judging whether the corrected thickness data exceeds the theoretical bonding thickness data of the FPC board, if so, outputting the corrected thickness data as re-bonding data, otherwise, outputting the corrected thickness data as missed bonding data; S3-3, judging whether the bonding state data of the FPC board is qualified data, if so, obtaining the corresponding position, shape, thickness and area information respectively and marking them as qualified marking data, otherwise, marking processing is performed according to the unqualified data of the bonding state of the FPC board to obtain the corresponding position mark, shape mark, thickness mark and area mark respectively as unqualified marking data; S3-4, using the qualified mark data and the unqualified mark data to establish a FPC board fitting information database; The bonding thickness deviation data is the deviation value caused by the thickness change during the bonding process due to external factors, and the theoretical bonding data is the theoretical bonding size, shape and area data corresponding to the theoretical bonding thickness data of the FPC board.

[0011] Furthermore, the thickness deviation data of the FPC board is obtained by analyzing and processing the bonding data, including: S3-1-1, respectively obtaining the bonding material, bonding temperature and bonding pressure data of the FPC board according to the bonding data of the FPC board; S3-1-2, determine whether the bonding material of the FPC board corresponds to the theoretical bonding thickness data of the FPC board, if so, execute S3-1-3, otherwise, obtain the bonding thickness data of the FPC board that has a one-to-one correspondence with the bonding material of the FPC board, and output the difference between the bonding thickness data of the FPC board and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2; S3-1-3, judging whether the bonding temperature of the FPC board exceeds the bonding temperature threshold in the bonding process information of the FPC board, if so, executing S3-1-4, otherwise, directly executing S3-1-5; S3-1-4, determine whether the bonding temperature of the FPC board is the first thickness interference, if so, obtain the first interference thickness data generated by the interference, and output the difference between the first interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2, otherwise, execute S3-1-5; S3-1-5, judging whether the bonding pressure of the FPC board exceeds the bonding pressure threshold in the bonding process information of the FPC board, if so, executing S3-1-6, otherwise, outputting the bonding thickness deviation data as 0, and directly executing S3-2; S3-1-6, determine whether the bonding pressure of the FPC board that exceeds the bonding pressure threshold is the second thickness interference, if so, obtain the second interference thickness data generated by the interference, and output the difference between the second interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and execute S3-2, otherwise, output the bonding thickness deviation data as 0, and execute S3-2; Wherein, the thickness interference is an external factor that affects the bonding thickness value; Among them, different materials have different corresponding relationships with different thicknesses, different materials correspond to different bonding processes, and different bonding processes correspond to different bonding temperatures and pressure thresholds.

[0012] Furthermore, the detection results of the FPC board anti-re-lamination obtained by using the FPC board lamination information database based on the target detection process include: S4-1, using the FPC board fitting information database to establish an initial training model based on target detection processing; S4-1-1, using the historical data and real-time data in the FPC board fitting information database as a training set, and using unqualified marked data as a verification set; S4-1-2, using the training set as input and the unqualified labeled data as output, establishing an initial training model based on the Faster R-CNN algorithm; S4-2, using the initial training model to obtain a target detection model; S4-3, obtaining the detection result of the FPC board anti-re-bonding according to the target detection model.

[0013] Further, obtaining a target detection model using the initial training model includes: S4-2-1, using the verification set to input the initial training model to obtain an initial output result; S4-2-2, determine whether the initial output result corresponds to the subset of the verification set. If so, execute S4-2-3. Otherwise, use the non-corresponding subset of the verification set as the updated training set and return to S4-1-2; S4-2-3, respectively obtaining abnormal thickness data corresponding to the abnormal fitting parameters as abnormal data, using the abnormal data to input the initial training model to obtain a model output result, and determining whether the model output result corresponds to the verification set, if so, outputting the initial training model as the target detection model, otherwise, using the non-corresponding verification set as an updated training set, and returning to S4-1-2; The abnormal bonding parameters include abnormal temperature and pressure parameters exceeding bonding temperature and bonding pressure thresholds; and the abnormal thickness data include first interference thickness data and second interference thickness data.

[0014] Furthermore, the detection results of the FPC board anti-re-bonding obtained according to the target detection model include: S4-3-1, obtaining unqualified data information of the bonding state of the corresponding FPC board according to the target detection model; S4-3-2, obtaining corresponding re-attachment data information according to the unqualified data information of the bonding state of the FPC board; S4-3-3, obtaining FPC board anti-re-lamination alarm information and alarm position information according to the re-lamination data information; S4-3-4, performing traceability processing according to the FPC board anti-re-lamination alarm information and the alarm location information to obtain the FPC board anti-re-lamination detection result; S4-3-4-1, tracing back to the exact bonding location of the FPC board according to the anti-rebonding alarm information and alarm location information of the FPC board to obtain the bonding status of the corresponding FPC board; S4-3-4-2, judging whether the bonding state of the FPC board is already bonded, if so, executing S4-3-4-3, otherwise, outputting the detection result of the FPC board anti-rebonding as inaccurate; S4-3-4-3, determine whether the bonding state of the FPC board is repeated bonding, if so, output the FPC board anti-re-bonding detection result as accurate, otherwise, output the FPC board anti-re-bonding detection result as inaccurate.

[0015] The advantages of the present invention are: The target detection-based FPC board anti-re-bonding detection method provided by the present invention reduces redundant CCD detection processes, makes the detection process simpler, and greatly improves the detection speed. Through the FPC board anti-re-bonding alarm information, the bonding site of the corresponding FPC board can be accurately traced to achieve 100% effective interception. Compared with the existing detection method, the speed is improved while the accuracy is improved, making the detection more efficient, and solving the current problems of low detection accuracy and low detection efficiency of repeated bonding products. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention provides a flow chart of a method for monitoring the height of an FPC board against re-sticking. DETAILED DESCRIPTION

[0017] Example 1 Figure 1 As shown, this embodiment provides an FPC board anti-re-bonding detection method based on target detection, comprising the following steps: S1. Obtain the bonding process information of the FPC board; S2, performing detection processing according to the bonding process information of the FPC board to obtain the bonding data of the FPC board; S3, performing verification processing according to the bonding data of the FPC board to establish a bonding information database of the FPC board; S4, obtaining the detection result of the FPC board anti-re-lamination based on the target detection processing by utilizing the FPC board lamination information database.

[0018] S1 specifically includes: S1-1, obtain FPC board bonding material information; S1-2, obtaining theoretical bonding thickness data of the FPC board according to the bonding material information of the FPC board and establishing a correlation between the bonding material and the theoretical bonding thickness; S1-3, obtaining bonding material classification information according to the correlation between the bonding material and the theoretical bonding thickness; S1-4, obtaining the bonding process information of the FPC board according to the bonding material classification information; The bonding process information of the FPC board includes a bonding method, a bonding temperature threshold, a pressure threshold and a time threshold.

[0019] In this embodiment, a method for detecting anti-re-bonding of an FPC board based on target detection is provided, wherein the reinforcement material information includes: PI material, PET material, PU material, FR4 material, and stainless steel material; The FPC board reinforcement material information acquisition corresponding to the bonding thickness information includes: PI material reinforcement thickness is 12.5-50μm, PET material reinforcement thickness is 12.5-50μm, PU material reinforcement thickness is 10-30μm, FR4 material reinforcement thickness is 0.1-2mm, stainless steel material reinforcement thickness is 0.5-3mm; The relationship between the reinforcing material and the bonding thickness is as follows: the reinforcing material with a thickness less than 0.1 mm includes PI material, PET material and PU material, and the reinforcing material with a thickness greater than 0.1 mm includes FR4 material and stainless steel material; The reinforcement material classification information is: thin material and thick material.

[0020] S2 specifically includes: S2-1, detecting and processing the surface of the FPC board based on the bonding process information of the FPC board using a Kinect depth sensor to obtain corresponding point cloud data; S2-2, performing reverse data processing according to the point cloud data to respectively obtain fitting area data and fitting shape data, and establishing a primary mapping between the fitting shape and the fitting area; S2-3, obtaining fitting size data according to the primary mapping, and establishing a secondary mapping of the fitting size, fitting shape and area; S2-4, obtaining the result of the FPC board's lamination state judgment based on the secondary mapping and grating ruler detection processing; S2-5. Outputting corresponding thickness data as the FPC board bonding data according to the bonding state judgment result of the FPC board.

[0021] In this embodiment, a method for detecting FPC board re-bonding based on target detection is provided, wherein the bonding shape includes a regular rectangle and an irregular polygon, the bonding area is formed with the outer contour line of the bonding shape as the boundary, and the bonding size includes 2mm*2mm and 300mm*30mm.

[0022] S2-1 specifically includes: S2-1-1, according to the bonding process information of the FPC board, using the Kinect depth sensor to perform detection processing based on TOF technology to obtain the transmission signal time t0 and the reception signal time t1 respectively; S2-1-2, the calculation formula for calculating the TOF value according to the received signal time t1 and the transmitted signal time t0 is as follows: TOF=t1-t0; S2-1-3. The calculation formula for calculating the distance information according to the TOF value is as follows: d = C × TOF / 2; S2-1-4, using the distance information to obtain corresponding point cloud data based on a Kinect depth sensor; The TOF value is the time it takes for light to fly a certain distance in a medium, d is the distance the light flies in the medium, and C is the speed of light, C=3×10 8 m / s.

[0023] S2-4 specifically includes: S2-4-1, obtaining the FPC board bonding starting position data based on the secondary mapping and grating ruler detection processing; S2-4-2, obtaining the bonding termination position data of the FPC board based on the corresponding relationship data between the bonding shape and the bonding size based on the grating ruler detection processing; S2-4-3, using the difference between the bonding start position data and the bonding end position data of the FPC board as bonding position moving distance data; S2-4-4, determine whether the bonding position moving distance data is 0, if so, the FPC board bonding state is obtained as missing bonding, and S2-5 is executed, otherwise, the FPC board bonding state is obtained as bonded, and S2-5 is executed.

[0024] S3 specifically includes: S3-1, analyzing and judging the bonding data of the FPC board to obtain bonding thickness deviation data; S3-2, using the lamination thickness deviation data to verify the lamination data of the FPC board to obtain the lamination state data of the FPC board; S3-3, judging whether the bonding state data of the FPC board is qualified data, if so, obtaining the corresponding position, shape, thickness and area information respectively and marking them as qualified marking data, otherwise, marking processing is performed according to the unqualified data of the bonding state of the FPC board to obtain the corresponding position mark, shape mark, thickness mark and area mark respectively as unqualified marking data; S3-4, using the qualified mark data and the unqualified mark data to establish a FPC board fitting information database; The bonding thickness deviation data is the deviation value caused by the thickness change during the bonding process due to external factors, and the theoretical bonding data is the theoretical bonding size, shape and area data corresponding to the theoretical bonding thickness data of the FPC board.

[0025] S3-1 specifically includes: S3-1-1, respectively obtaining the bonding material, bonding temperature and bonding pressure data of the FPC board according to the bonding data of the FPC board; S3-1-2, determine whether the bonding material of the FPC board corresponds to the theoretical bonding thickness data of the FPC board, if so, execute S3-1-3, otherwise, obtain the bonding thickness data of the FPC board that has a one-to-one correspondence with the bonding material of the FPC board, and output the difference between the bonding thickness data of the FPC board and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2; S3-1-3, judging whether the bonding temperature of the FPC board exceeds the bonding temperature threshold in the bonding process information of the FPC board, if so, executing S3-1-4, otherwise, directly executing S3-1-5; S3-1-4, determine whether the bonding temperature of the FPC board is the first thickness interference, if so, obtain the first interference thickness data generated by the interference, and output the difference between the first interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2, otherwise, execute S3-1-5; S3-1-5, judging whether the bonding pressure of the FPC board exceeds the bonding pressure threshold in the bonding process information of the FPC board, if so, executing S3-1-6, otherwise, outputting the bonding thickness deviation data as 0, and directly executing S3-2; S3-1-6, determine whether the bonding pressure of the FPC board that exceeds the bonding pressure threshold is the second thickness interference, if so, obtain the second interference thickness data generated by the interference, and output the difference between the second interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and execute S3-2, otherwise, output the bonding thickness deviation data as 0, and execute S3-2; Wherein, the thickness interference is an external factor that affects the bonding thickness value; Among them, different materials have different corresponding relationships with different thicknesses, different materials correspond to different bonding processes, and different bonding processes correspond to different bonding temperatures and pressure thresholds.

[0026] S3-2 specifically includes: S3-2-1, using the lamination thickness deviation data to correct the lamination data of the FPC board to obtain corrected thickness data; S3-2-2, determine whether the corrected thickness data is the same as the theoretical bonding thickness data of the FPC board. If so, obtain the size, shape, and area data corresponding to the corrected thickness data according to the secondary mapping relationship as the corrected bonding data, and execute S3-2-3. Otherwise, directly execute S3-2-4; S3-2-3, judging whether the corrected bonding data is the same as the theoretical bonding data, if so, outputting the corrected thickness data as qualified data of the bonding state of the FPC board and directly executing S3-3, otherwise, outputting the corrected thickness data as wrong bonding data; S3-2-4, judging whether the corrected thickness data exceeds the theoretical bonding thickness data of the FPC board, if so, outputting the corrected thickness data as re-bonding data, otherwise, outputting the corrected thickness data as missed bonding data; S4 specifically includes: S4-1, using the FPC board fitting information database to establish an initial training model based on target detection processing; S4-2, using the initial training model to obtain a target detection model; S4-3, obtaining the detection result of the FPC board anti-re-bonding according to the target detection model.

[0027] S4-1 specifically includes: S4-1-1, using the historical data and real-time data in the FPC board fitting information database as a training set, and using unqualified marked data as a verification set; S4-1-2. Using the training set as input and the unqualified labeled data as output, an initial training model is established based on the Faster R-CNN algorithm.

[0028] S4-2 specifically includes: S4-2-1, using the verification set to input the initial training model to obtain an initial output result; S4-2-2, determine whether the initial output result corresponds to the subset of the verification set. If so, execute S4-2-3. Otherwise, use the non-corresponding subset of the verification set as the updated training set and return to S4-1-2; S4-2-3, respectively obtaining abnormal thickness data corresponding to the abnormal fitting parameters as abnormal data, using the abnormal data to input the initial training model to obtain a model output result, and determining whether the model output result corresponds to the verification set, if so, outputting the initial training model as the target detection model, otherwise, using the non-corresponding verification set as an updated training set, and returning to S4-1-2; The abnormal bonding parameters include abnormal temperature and pressure parameters exceeding bonding temperature and bonding pressure thresholds; and the abnormal thickness data include first interference thickness data and second interference thickness data.

[0029] S4-3 specifically includes: S4-3-1, obtaining unqualified data information of the bonding state of the corresponding FPC board according to the target detection model; S4-3-2, obtaining corresponding re-attachment data information according to the unqualified data information of the bonding state of the FPC board; S4-3-3, obtaining FPC board anti-re-lamination alarm information and alarm position information according to the re-lamination data information; S4-3-4, performing traceability processing according to the FPC board anti-re-lamination alarm information and the alarm location information to obtain the FPC board anti-re-lamination detection result.

[0030] S4-3-4-1, tracing back to the exact bonding location of the FPC board according to the anti-rebonding alarm information and alarm location information of the FPC board to obtain the bonding status of the corresponding FPC board; S4-3-4-2, judging whether the bonding state of the FPC board is already bonded, if so, executing S4-3-4-3, otherwise, outputting the detection result of the FPC board anti-rebonding as inaccurate; S4-3-4-3, determine whether the bonding state of the FPC board is repeated bonding, if so, output the FPC board anti-re-bonding detection result as accurate, otherwise, output the FPC board anti-re-bonding detection result as inaccurate.

[0031] For those skilled in the art, the invention is not limited to the details of the above-mentioned exemplary embodiments, and the invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the invention; therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the invention is limited by the appended claims rather than the above description, and it is intended that all changes within the meaning and scope of the equivalent elements of the claims are included in the invention.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any minor modifications, equivalent substitutions and improvements made to the above embodiments based on the technical essence of the present invention should be included in the protection scope of the technical solution of the present invention.

Claims

1. A method for detecting FPC board re-bonding based on target detection, characterized in that: The following steps are involved: S1. Obtain the bonding process information of the FPC board; S2, performing detection processing according to the bonding process information of the FPC board to obtain the bonding data of the FPC board; S3, performing verification processing according to the bonding data of the FPC board to establish a bonding information database of the FPC board; S4, obtaining the detection result of the FPC board anti-re-lamination based on the target detection processing by utilizing the FPC board lamination information database.

2. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 1, characterized in that: The obtaining of the bonding process information of the FPC board includes: S1-1, obtain FPC board bonding material information; S1-2, obtaining theoretical bonding thickness data of the FPC board according to the bonding material information of the FPC board and establishing a correlation between the bonding material and the theoretical bonding thickness; S1-3, obtaining bonding material classification information according to the correlation between the bonding material and the theoretical bonding thickness; S1-4, obtaining the bonding process information of the FPC board according to the bonding material classification information; The bonding process information of the FPC board includes a bonding method, a bonding temperature threshold, a pressure threshold and a time threshold.

3. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 2, characterized in that: Performing detection and processing according to the bonding process information of the FPC board to obtain the bonding data of the FPC board includes: S2-1, detecting and processing the surface of the FPC board based on the bonding process information of the FPC board using a Kinect depth sensor to obtain corresponding point cloud data; S2-2, performing reverse data processing according to the point cloud data to respectively obtain fitting area data and fitting shape data, and establishing a primary mapping between the fitting shape and the fitting area; S2-3, obtaining fitting size data according to the primary mapping, and establishing a secondary mapping of the fitting size, fitting shape and area; S2-4, obtaining the result of the FPC board's lamination state judgment based on the secondary mapping and grating ruler detection processing; S2-5. Outputting corresponding thickness data as the FPC board bonding data according to the bonding state judgment result of the FPC board.

4. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 3, characterized in that: According to the bonding process information of the FPC board, the surface of the FPC board is detected and processed based on the Kinect depth sensor to obtain corresponding point cloud data including: S2-1-1, according to the bonding process information of the FPC board, using the Kinect depth sensor to perform detection processing based on TOF technology to obtain the transmission signal time t0 and the reception signal time t1 respectively; S2-1-2, the calculation formula for calculating the TOF value according to the received signal time t1 and the transmitted signal time t0 is as follows: TOF=t1-t0; S2-1-3. The calculation formula for calculating the distance information according to the TOF value is as follows: d = C × TOF / 2; S2-1-4, using the distance information to obtain corresponding point cloud data based on a Kinect depth sensor; The TOF value is the time it takes for light to fly a certain distance in a medium, d is the distance the light flies in the medium, and C is the speed of light, C=3×10 8 m / s.

5. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 3, characterized in that: The results of the FPC board bonding state judgment obtained based on the secondary mapping and the grating ruler detection processing include: S2-4-1, obtaining the FPC board bonding starting position data based on the secondary mapping and grating ruler detection processing; S2-4-2, obtaining the bonding termination position data of the FPC board based on the corresponding relationship data between the bonding shape and the bonding size based on the grating ruler detection processing; S2-4-3, using the difference between the bonding start position data and the bonding end position data of the FPC board as bonding position moving distance data; S2-4-4, determine whether the bonding position moving distance data is 0, if so, the FPC board bonding state is obtained as missing bonding, and S2-5 is executed, otherwise, the FPC board bonding state is obtained as bonded, and S2-5 is executed.

6. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 3, characterized in that: Performing verification processing according to the bonding data of the FPC board to establish the bonding information database of the FPC board includes: S3-1, analyzing and judging the bonding data of the FPC board to obtain bonding thickness deviation data; S3-2, using the lamination thickness deviation data to verify the lamination data of the FPC board to obtain the lamination state data of the FPC board; S3-2-1, using the lamination thickness deviation data to correct the lamination data of the FPC board to obtain corrected thickness data; S3-2-2, determine whether the corrected thickness data is the same as the theoretical bonding thickness data of the FPC board. If so, obtain the size, shape, and area data corresponding to the corrected thickness data according to the secondary mapping relationship as the corrected bonding data, and execute S3-2-3. Otherwise, directly execute S3-2-4; S3-2-3, judging whether the corrected bonding data is the same as the theoretical bonding data, if so, outputting the corrected thickness data as qualified data of the bonding state of the FPC board and directly executing S3-3, otherwise, outputting the corrected thickness data as wrong bonding data; S3-2-4, judging whether the corrected thickness data exceeds the theoretical bonding thickness data of the FPC board, if so, outputting the corrected thickness data as re-bonding data, otherwise, outputting the corrected thickness data as missed bonding data; S3-3, judging whether the bonding state data of the FPC board is qualified data, if so, obtaining the corresponding position, shape, thickness and area information respectively and marking them as qualified marking data, otherwise, marking processing is performed according to the unqualified data of the bonding state of the FPC board to obtain the corresponding position mark, shape mark, thickness mark and area mark respectively as unqualified marking data; S3-4, using the qualified mark data and the unqualified mark data to establish a FPC board fitting information database; The bonding thickness deviation data is the deviation value caused by the thickness change during the bonding process due to external factors, and the theoretical bonding data is the theoretical bonding size, shape and area data corresponding to the theoretical bonding thickness data of the FPC board.

7. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 6, characterized in that: The thickness deviation data of the FPC board obtained by analyzing and processing the bonding data includes: S3-1-1, respectively obtaining the bonding material, bonding temperature and bonding pressure data of the FPC board according to the bonding data of the FPC board; S3-1-2, determine whether the bonding material of the FPC board corresponds to the theoretical bonding thickness data of the FPC board, if so, execute S3-1-3, otherwise, obtain the bonding thickness data of the FPC board that has a one-to-one correspondence with the bonding material of the FPC board, and output the difference between the bonding thickness data of the FPC board and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2; S3-1-3, judging whether the bonding temperature of the FPC board exceeds the bonding temperature threshold in the bonding process information of the FPC board, if so, executing S3-1-4, otherwise, directly executing S3-1-5; S3-1-4, determine whether the bonding temperature of the FPC board is the first thickness interference, if so, obtain the first interference thickness data generated by the interference, and output the difference between the first interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and directly execute S3-2, otherwise, execute S3-1-5; S3-1-5, judging whether the bonding pressure of the FPC board exceeds the bonding pressure threshold in the bonding process information of the FPC board, if so, executing S3-1-6, otherwise, outputting the bonding thickness deviation data as 0, and directly executing S3-2; S3-1-6, determine whether the bonding pressure of the FPC board that exceeds the bonding pressure threshold is the second thickness interference, if so, obtain the second interference thickness data generated by the interference, and output the difference between the second interference thickness data and the theoretical bonding thickness data of the FPC board as the bonding thickness deviation data, and execute S3-2, otherwise, output the bonding thickness deviation data as 0, and execute S3-2; The thickness interference is an external factor that affects the value of the bonding thickness.

8. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 6, characterized in that: The detection results of the FPC board anti-re-lamination obtained by using the FPC board lamination information database based on the target detection processing include: S4-1, using the FPC board fitting information database to establish an initial training model based on target detection processing; S4-1-1, using the historical data and real-time data in the FPC board fitting information database as a training set, and using unqualified marked data as a verification set; S4-1-2, using the training set as input and the unqualified labeled data as output, establishing an initial training model based on the Faster R-CNN algorithm; S4-2, using the initial training model to obtain a target detection model; S4-3, obtaining the detection result of the FPC board anti-re-bonding according to the target detection model.

9. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 8, characterized in that: Obtaining a target detection model using the initial training model includes: S4-2-1, using the verification set to input the initial training model to obtain an initial output result; S4-2-2, determine whether the initial output result corresponds to the subset of the verification set. If so, execute S4-2-3. Otherwise, use the non-corresponding subset of the verification set as the updated training set and return to S4-1-2; S4-2-3, respectively obtaining abnormal thickness data corresponding to the abnormal fitting parameters as abnormal data, using the abnormal data to input the initial training model to obtain a model output result, and determining whether the model output result corresponds to the verification set, if so, outputting the initial training model as the target detection model, otherwise, using the non-corresponding verification set as an updated training set, and returning to S4-1-2; The abnormal bonding parameters include abnormal temperature and pressure parameters exceeding bonding temperature and bonding pressure thresholds; and the abnormal thickness data include first interference thickness data and second interference thickness data.

10. The FPC board anti-re-bonding detection method based on target detection as claimed in claim 8, characterized in that: The detection results of the FPC board anti-re-bonding obtained according to the target detection model include: S4-3-1, obtaining unqualified data information of the bonding state of the corresponding FPC board according to the target detection model; S4-3-2, obtaining corresponding re-attachment data information according to the unqualified data information of the bonding state of the FPC board; S4-3-3, obtaining FPC board anti-re-lamination alarm information and alarm position information according to the re-lamination data information; S4-3-4, performing traceability processing according to the FPC board anti-re-lamination alarm information and the alarm location information to obtain the FPC board anti-re-lamination detection result; S4-3-4-1, tracing back to the exact bonding location of the FPC board according to the anti-rebonding alarm information and alarm location information of the FPC board to obtain the bonding status of the corresponding FPC board; S4-3-4-2, judging whether the bonding state of the FPC board is already bonded, if so, executing S4-3-4-3, otherwise, outputting the detection result of the FPC board anti-rebonding as inaccurate; S4-3-4-3, determine whether the bonding state of the FPC board is repeated bonding, if so, output the FPC board anti-re-bonding detection result as accurate, otherwise, output the FPC board anti-re-bonding detection result as inaccurate.

Images