Dynamic compensation method for light guide plate film precision of multi-modal perception

Abstract

The application discloses a light guide plate film coating precision dynamic compensation method based on multi-modal sensing, relates to the technical field of light guide plate film coating, and solves the problem of insufficient light guide plate film coating precision, comprising the following steps: S1, acquiring image information and pressure data; S2, extracting the positions of light guide plate marks and film layer marks according to the image information, and analyzing film layer surface defects according to the pressure data to optimize the positions of the film layer marks; S3, calculating the deviation between the adjusted positions of the film layer marks and the positions of the light guide plate marks to obtain a position deviation, performing precision compensation on the light guide plate film coating, acquiring a film coating image at a next time node according to the precision compensation, analyzing the precision of the light guide plate film coating, and obtaining a precision change sequence; and S4, analyzing the precision change of the light guide plate film coating and optimizing the precision compensation of the light guide plate film coating. The application can effectively improve the precision of the light guide plate film coating.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of light guide plate film covering, and particularly relates to a light guide plate film covering precision dynamic compensation method based on multi-modal perception. BACKGROUND

[0002] The existing light guide plate has the following defects when performing film covering:

[0003] 1. When the existing light guide plate is performing film covering, the film covering precision is detected after the film covering is completed, the film covering deviation is calculated through the position deviation of the film layer boundary and the light guide plate boundary, so that the light guide plate film covering precision is uncontrollable, and it is difficult to compensate the light guide plate precision.

[0004] 2. When the existing light guide plate is performing film covering, the film covering precision is judged through single data, the judgment accuracy is not high, when there is film covering abnormality (such as bubble, depression), it is difficult to timely adjust the film covering precision according to the film covering abnormality, so that the film covering precision is reduced.

[0005] 3. When the existing light guide plate is performing film covering, the real-time of precision compensation is insufficient, the light guide plate film covering process has continuity, when the precision has error, if the precision is not adjusted in time, the precision error will affect the subsequent film covering process, so that the precision error continues to increase with the continuous film covering process, the film covering result precision error is large, and the film covering accuracy is low.

[0006] Therefore, the application provides a light guide plate film covering precision dynamic compensation method based on multi-modal perception. SUMMARY

[0007] In view of the defects in the prior art, the application aims to provide a light guide plate film covering precision dynamic compensation method based on multi-modal perception, and aims to improve the light guide plate film covering precision.

[0008] In order to achieve the above purpose, the application adopts the following technical scheme: a light guide plate film covering precision dynamic compensation method based on multi-modal perception, and the specific working process of each step is as follows:

[0009] Step S1: Real-time image acquisition is performed on the light guide plate film covering process, the positions of the light guide plate marks and the film layer marks are recorded, image information is obtained, pressure measurement is performed on the film covered part based on the image information, and pressure data is obtained;

[0010] Step S2: According to the image information, the positions of the light guide plate marks and the film layer marks are extracted; according to the pressure data, the film layer surface defects are analyzed, the film layer is adjusted according to the film layer surface defects, and the position of the film layer mark is optimized;

[0011] Step S3: deviation calculation is performed according to the position of the film layer mark and the position of the light guide plate mark, and the position deviation is obtained; precision compensation is performed on the light guide plate film according to the position deviation; the film image of the next time node is acquired according to the precision compensation, and the film precision of the light guide plate is analyzed according to the film image of the next time node; the film precision of the light guide plate in the film process is counted, and the precision change sequence is obtained;

[0012] Step S4: the precision change of the light guide plate film is analyzed according to the precision change sequence, and the precision compensation of the light guide plate film is optimized in combination with the precision compensation of the light guide plate film.

[0013] Further, the specific steps of step S1 are as follows:

[0014] Step S11: alignment marks are set on the light guide plate to obtain light guide plate marks, the corresponding positions of the film layer are marked according to the light guide plate marks to obtain film layer marks, the number of alignment marks is recorded as is×js; the light guide plate marks are recorded as dgb(i,j); the film layer marks are recorded as mcb(i,j);

[0015] The time required for the film coating of the light guide plate is obtained, and a plurality of time nodes are extracted according to the time required for the film coating of the light guide plate, the number of time nodes is recorded as ts; the film coating image of the light guide plate is collected, the positions of the light guide plate marks and the film layer marks are recorded, and the film coating images of different time nodes are obtained in combination with the number of time nodes; the image information of the light guide plate film is formed by the film coating images of different time nodes;

[0016] Step S12: the film-coated part is synchronously monitored according to the image information of the light guide plate film to obtain the pressure borne by the film layer of the light guide plate, and the pressure data is obtained by counting the pressure borne by the film layer.

[0017] Further, the specific steps of step S2 are as follows:

[0018] Step S21: the film coating image of the current time node is acquired according to the image information, the pressure data is acquired according to the film coating image of the current time node, the film layer surface defect is extracted according to the pressure data, and the adjustment direction of the film layer surface defect is controlled by the pressure data through the adjustment of the film layer surface defect combined with the pressure data;

[0019] Step S22: the position of the film layer mark is adjusted and optimized according to the adjustment direction of the film layer surface defect, and the position of the film layer mark after adjustment and optimization is counted.

[0020] Further, the specific steps of step S21 are as follows:

[0021] Step S211: Obtain the current time node's laminated image, and obtain the pixel number of the laminated image according to the current time node's laminated image, denoted as as t ×bs t ; obtain the pressure data, map the pressure data with the pixels of the laminated image, and obtain the pressure values ylz(a t , b t ) of different pixel positions;

[0022] Step S212: Statistically analyze the pressure values ylz(a t , b t ), cyclically calculate the pressure values of the pixels in the horizontal direction according to the pressure values ylz(a t , b t ), analyze the abnormal values of the pressure values in the horizontal direction, and obtain the horizontal abnormal values spy(a t , b t );

[0023] According to the horizontal abnormal values spy(a t , b t ), perform abnormal analysis on the pressure values at the image pixel (a, b) position:

[0024] When spy(a t , b t ) = 0, the pressure value at the surface image pixel (a, b) position is normal;

[0025] When spy(a t , b t ) > 0, the pressure value at the surface image pixel (a, b) position is too high, and the film layer needs to be marked to be offset in the horizontal direction towards the (a, b) position;

[0026] When spy(a t , b t ) < 0, the pressure value at the surface image pixel (a, b) position is too low, and the film layer needs to be marked to be offset in the horizontal direction away from the (a, b) position;

[0027] According to the pressure values ylz(a t , b t ), cyclically calculate the pressure values of the pixels in the vertical direction, analyze the abnormal values of the pressure values in the vertical direction, and obtain the vertical abnormal values czy(a t , b t );

[0028] According to the vertical abnormal values czy(a t , b t ), perform abnormal analysis on the pressure values at the image pixel (a, b) position:

[0029] Whenczy(a t , b t ) = 0, the pressure value at the surface image pixel (a, b) position is normal;

[0030] Whenczy(a t , b t ) > 0, the pressure value at the surface image pixel (a, b) position is too high, and the film layer mark needs to be offset in the vertical direction towards the (a, b) position;

[0031] Whenczy(a t , b t ) < 0, the pressure value at the surface image pixel (a, b) position is too low, and the film layer mark needs to be offset in the vertical direction away from the (a, b) position;

[0032] The offset amounts of the film layer mark in the horizontal direction and the vertical direction are counted to obtain position offset information, and the film layer is adjusted according to the position offset information.

[0033] Further, the specific steps of the step S22 are as follows:

[0034] Step S221: According to the image information, the position of the light guide plate mark in the image is obtained, the position of the corresponding film layer mark is extracted according to the position of the light guide plate mark in the image, and the position of the film layer mark is recorded as the position spf(i, j) in the horizontal direction and the position czf(i, j) in the vertical direction respectively;

[0035] Step S222: According to the adjustment direction of the film layer surface defect, the position offset information of the film layer mark is obtained, the position offset amount in the horizontal direction is extracted according to the position offset information of the film layer mark, and the position offset amount in the vertical direction is extracted, and the position offset amount in the vertical direction is recorded asczy(i, j);

[0036] Step S223: According to the position spf(i, j) of the film layer mark in the horizontal direction and the position offset amount spy(i, j) in the horizontal direction, the optimized position syh(i, j) of the film layer mark in the horizontal direction is obtained; according to the position czf(i, j) of the film layer mark in the vertical direction and the position offset amountczy(i, j) in the vertical direction, the optimized position cyh(i, j) of the film layer mark in the vertical direction is obtained; the optimized position syh(i, j) of the film layer mark in the horizontal direction and the optimized position cyh(i, j) in the vertical direction are counted to obtain the optimized position of the film layer mark.

[0037] Further, the specific steps of the step S3 are as follows:

[0038] Step S31: Obtain the position of the film layer mark and the position of the light guide plate mark, calculate the position of the film layer mark and the position of the light guide plate mark, obtain the position deviation, analyze the precision compensation of the light guide plate film coating according to the position deviation, and obtain the compensation data;

[0039] Step S32: According to the compensation data, combined with the position deviation, the position deviation of the next time node is predicted to obtain the predicted deviation; obtain the film coating image of the next time node, calculate the actual deviation of the image according to the film coating image of the next time node, analyze the light guide plate film coating precision according to the actual deviation and the predicted deviation, and adjust the subsequent film coating process;

[0040] Step S33: Statistically analyze the light guide plate film coating precision of the film coating process, analyze the light guide plate film coating precision of different time nodes, and construct the precision change sequence according to the light guide plate film coating precision of different time nodes.

[0041] Further, the specific steps of step S31 are as follows:

[0042] Step S311: Obtain the position of the light guide plate mark, extract the horizontal position of the light guide plate mark as dsp(i,j), and extract the vertical position of the light guide plate mark as dcz(i,j); obtain the optimized position of the film layer mark, extract the optimized position of the film layer mark in the horizontal direction as syh(i,j) and in the vertical direction as cyh(i,j) according to the optimized position of the film layer mark;

[0043] Step S312: Calculate the deviation in the horizontal direction according to the horizontal position of the light guide plate mark dsp(i,j) and the optimized position of the film layer mark in the horizontal direction syh(i,j), and obtain the horizontal position deviation ppc.

[0044] Calculate the deviation in the vertical direction according to the vertical position of the light guide plate mark dcz(i,j) and the optimized position of the film layer mark in the vertical direction cyh(i,j), and obtain the vertical position deviation cpc.

[0045] Further, the specific steps of step S31 also include:

[0046] Pre-set horizontal precision compensation sys, perform precision compensation on the optimized position of the film layer mark in the horizontal direction syh(i,j) according to the horizontal precision compensation, and calculate the horizontal precision judgment value spd by combining the horizontal position of the light guide plate mark dsp(i,j) and the horizontal position deviation ppc.

[0047] ;

[0048] The preset horizontal precision compensation sys is traversed, the horizontal precision judgment value in the corresponding state is recorded, the minimum horizontal precision judgment value is obtained, the corresponding preset horizontal precision compensation is extracted, and the optimal horizontal precision compensation is obtained.

[0049] The preset vertical precision compensation cys is traversed, the vertical precision judgment value in the corresponding state is recorded, the minimum vertical precision judgment value is obtained, the corresponding preset vertical precision compensation is extracted, and the optimal vertical precision compensation is obtained.

[0050] The preset vertical precision compensation cys is traversed, the vertical precision judgment value in the corresponding state is recorded, the minimum vertical precision judgment value is obtained, the corresponding preset vertical precision compensation is extracted, and the optimal vertical precision compensation is obtained.

[0051] The optimal horizontal precision compensation and the optimal vertical precision compensation are recorded to obtain compensation data.

[0052] Further, the specific steps of the step S32 are as follows:

[0053] Step S321: According to the compensation data, the optimal horizontal precision compensation and the optimal vertical precision compensation are obtained, the optimal horizontal precision compensation is recorded as zys, the optimal vertical precision compensation is recorded as zyc, the horizontal position deviation ppc and the vertical position deviation cpc are obtained, and the optimal horizontal precision compensation and the optimal vertical precision compensation are combined to calculate the horizontal prediction deviation syc and the vertical prediction deviation cpc.

[0054] The film coating image of the next time node is obtained, the actual deviation of the image is obtained according to the film coating image of the next time node, the horizontal actual deviation is extracted from the actual deviation of the image and recorded as ssj(i,j), and the vertical actual deviation is extracted and recorded as csj(i,j).

[0055] Step S322: The prediction deviation and the actual deviation are analyzed, the horizontal precision compensation judgment is performed through the horizontal actual deviation ssj(i,j) and the horizontal prediction deviation syc, the vertical precision compensation judgment is performed through the vertical actual deviation csj(i,j) and the vertical prediction deviation cpc, and the precision compensation data of the subsequent time node is obtained.

[0056] Further, the specific steps of the step S33 are as follows:

[0057] Step S331: precision compensation data of all time nodes of the film coating process is acquired, the light guide plate film coating precision is analyzed according to the precision compensation data, the horizontal direction precision compensation data is acquired from the precision compensation data, and is recorded as sbc(1) to sbc(ts), the horizontal precision change sequence is constructed according to the precision compensation in the horizontal direction, and is recorded as sbh(t); sbh(t)=sbc(t)+sbh(t-1); wherein: sbc(t) represents the precision compensation data of the horizontal direction of the tth time node;

[0058] Step S332: the precision compensation data of the vertical direction is acquired from the precision compensation data, and is recorded as cbc(1) to cbc(ts), the vertical precision change sequence is constructed according to the precision compensation in the vertical direction, and is recorded as cbh(t); cbh(t)=cbc(t)+cbh(t-1), wherein: cbc(t) represents the precision compensation data of the vertical direction of the tth time node.

[0059] According to the above technical scheme, the beneficial effects of the present application are as follows:

[0060] 1. The present application classifies the film coating process by time variable, analyzes the single time node in the film coating process, effectively analyzes the dynamic light guide plate film coating, discovers the film coating error in time, acquires the film coating image of the single time node, extracts the position of the film layer of the film coating image and the light guide plate, compares the positions of multiple nodes, judges the position deviation of the film coating, compensates the film coating precision according to the film coating deviation, and improves the accuracy of the light guide plate film coating.

[0061] 2. The present application measures the pressure of the film coating part, analyzes the film coating abnormity through the pressure data, converts the abnormal phenomenon into the abnormality of the pressure value, accurately judges the abnormal position, timely processes the abnormality according to the pressure data, and improves the film coating precision of the light guide plate.

[0062] 3. The present application judges the film coating error in real time through multiple time nodes, analyzes the overall precision compensation according to the film coating error of different time nodes, enhances the accuracy of the precision compensation, improves the film coating precision of the light guide plate according to the precision compensation. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to facilitate the understanding of those skilled in the art, the present application will be further described below with reference to the accompanying drawings.

[0064] Figure 1 The present application is a method schematic diagram;

[0065] Figure 2 The present application is a marker adjustment schematic diagram;

[0066] Figure 3 A film deviation calculation schematic diagram of the present application; DETAILED DESCRIPTION

[0067] The technical solutions of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0068] Embodiment one

[0069] Please refer to Figure 1 The multi-modal perception light guide plate film precision dynamic compensation method of the present application comprises:

[0070] Step S1: Real-time image acquisition is performed on the light guide plate film coating process, the positions of the light guide plate marks and the film layer marks are recorded, image information is obtained, pressure measurement is performed on the coated part based on the image information, and pressure data is obtained;

[0071] Step S11: Alignment marks are set on the light guide plate to obtain light guide plate marks, the corresponding positions of the film layer are marked according to the light guide plate marks to obtain film layer marks, the number of alignment marks is denoted as is×js; the light guide plate marks are denoted as dgb(i,j) according to the number of alignment marks; the film layer marks are denoted as mcb(i,j);

[0072] It should be noted that dgb(i,j) represents the light guide plate mark in the ith row and the jth column, and mcb(i,j) represents the film layer mark corresponding to dgb(i,j).

[0073] The time required for coating the light guide plate is obtained, a plurality of time nodes are extracted according to the time required for coating the light guide plate, the number of time nodes is denoted as ts; the light guide plate film image is collected by a high-precision vision camera, the positions of the light guide plate marks and the film layer marks are recorded, the image information of the light guide plate film is obtained in combination with the number of time nodes; and the image information of the light guide plate film is obtained from the film images at different time nodes.

[0074] Step S12: The coated part is synchronously monitored according to the image information of the light guide plate film, the pressure on the film layer of the light guide plate is tested by a pressure sensor, the pressure on the film layer is counted, and pressure data is obtained;

[0075] Please refer to Figure 2; Step S2: according to the image information, the position of the light guide plate mark and the position of the film layer mark are extracted; according to the pressure data, the film layer surface defect is analyzed, the film layer is adjusted according to the film layer surface defect, and the position of the film layer mark is optimized; the position of the film layer mark after adjustment is recorded;

[0076] Step S21: according to the image information, the film coating image of the current time node is obtained, the pressure data is obtained according to the film coating image of the current time node, the film layer surface defect is extracted according to the pressure data, and the film layer is adjusted through the film layer surface defect combined with the pressure data; the adjustment direction of the film layer surface defect is controlled by the pressure data;

[0077] Step S211: obtain the film coating image of the current time node, obtain the pixel number of the film coating image according to the film coating image of the current time node, and record as t ×bs t ; obtain the pressure data, map the pressure data with the pixels of the film coating image, and obtain the pressure values ylz(a t , b t ) of different pixel positions;

[0078] It should be noted that ylz(a t , b t ) represents the pressure value at the image pixel (a, b) position at the tth time node; (a, b) represents the ath in horizontal direction and the bth in vertical direction.

[0079] Step S212: statistics of the pressure values ylz(a t , b t ) of different pixel positions, according to the pressure values ylz(a t , b t ), the pressure values of the pixels in the horizontal direction are calculated circularly, the abnormal values of the pressure values in the horizontal direction are analyzed, and the horizontal abnormal values spy(a t , b t ) are obtained;

[0080] ;

[0081] Wherein, ylc((a-1) t , b t ) represents the pressure value at the image pixel (a-1, b) position at the tth time node; ylc((a+1) t , b t ) represents the pressure value at the image pixel (a+1, b) position at the tth time node;

[0082] According to the horizontal abnormal value spy(a t , bt Anomaly analysis was performed on the pressure values ​​at pixel (a, b) in the image:

[0083] When spy(a t b t When ) = 0, the pressure value at the pixel (a, b) position of the surface image is normal;

[0084] When spy(a t b t If the pressure value at pixel (a, b) on the surface image is too high, the film marking needs to be shifted horizontally towards the position (a, b).

[0085] When spy(a t b t If the pressure value at pixel (a, b) on the surface image is too low, the film marking needs to be shifted horizontally away from the position (a, b).

[0086] It should be noted that: when shifting towards the position (a, b), if there is a film layer marker at the position (a+1, b), then the position closer to the film layer marker is (a+0.5, b). This example shows that the horizontal shift is 0.5 towards the position (a, b); similarly, the shift is made away from the position (a, b).

[0087] Based on the pressure value ylz (a t b t The pressure values ​​of pixels in the vertical direction are calculated iteratively, and outliers in the vertical pressure values ​​are analyzed to obtain the vertical outlier czy(a). t b t );

[0088] ;

[0089] Among them, ylc(a t (b-1) t ) represents the pressure value at the image pixel (a, b-1) at the t-th time node; ylc(a t (b+1) t ) represents the pressure value at the image pixel (a, b+1) at the t-th time node;

[0090] Based on the vertical outlier czy(a) t b t Anomaly analysis was performed on the pressure values ​​at pixel (a, b) in the image:

[0091] When czy(a t b tWhen ) = 0, the pressure value at the pixel (a, b) position of the surface image is normal;

[0092] When czy(a t b t If the pressure value at pixel (a, b) on the surface image is too high, the film marking needs to be shifted vertically towards the position (a, b).

[0093] When czy(a t b t If the pressure value at pixel (a, b) on the surface image is too low, the film marking needs to be shifted vertically away from the position (a, b).

[0094] The offset of the film layer markings in the horizontal and vertical directions is statistically analyzed to obtain position offset information;

[0095] It should be noted that by calculating the forces in the horizontal and vertical directions independently, the accuracy of pressure anomaly detection is improved. At the same time, based on the offset of the membrane markings in the horizontal and vertical directions, the offset values ​​are made clearer, thus improving the accuracy of the offset.

[0096] Step S22: Adjust and optimize the position of the film markings according to the adjustment direction of the film surface defects, and statistically analyze the position of the film markings after adjustment and optimization;

[0097] Step S221: Based on the image information, obtain the position of the light guide plate mark in the image, extract the position of the corresponding film mark based on the position of the light guide plate mark in the image, and record the position of the film mark as the horizontal position spf(i,j) and the vertical position czf(i,j) respectively.

[0098] Step S222: Obtain the position offset information of the film mark according to the adjustment direction of the film surface defect. Based on the position offset information of the film mark, extract the position offset in the horizontal direction and denot it as spy(i,j). Extract the position offset in the vertical direction and denot it as czy(i,j).

[0099] Step S223: Obtain the optimized position syh(i,j) of the film layer mark in the horizontal direction according to the position spf(i,j) of the film layer mark in the horizontal direction combined with the position offset spy(i,j) in the horizontal direction, obtain the optimized position cyh(i,j) of the film layer mark in the vertical direction according to the position czf(i,j) of the film layer mark in the vertical direction combined with the position offset czy(i,j) in the vertical direction, and perform statistics on the optimized position syh(i,j) of the film layer mark in the horizontal direction and the optimized position cyh(i,j) of the film layer mark in the vertical direction to obtain the optimized position of the film layer mark.

[0100] Step S3: Calculate the position deviation according to the position of the adjusted film layer mark and the position of the light guide plate mark, obtain the position deviation, perform precision compensation on the light guide plate film coating according to the position deviation, obtain the film coating image at the next time node according to the precision compensation, analyze the precision of the light guide plate film coating according to the film coating image at the next time node, and perform statistics on the precision of the light guide plate film coating during the film coating process to obtain the precision change sequence.

[0101] Please refer to Figure 3 Step S31: Obtain the position of the film layer mark and the position of the light guide plate mark, calculate the position of the film layer mark and the position of the light guide plate mark, obtain the position deviation, analyze the precision compensation of the light guide plate film coating according to the position deviation, and obtain the compensation data.

[0102] Step S311: Obtain the position of the light guide plate mark, extract the horizontal position dsp(i,j) of the light guide plate mark according to the position of the light guide plate mark, and extract the vertical position dcz(i,j) of the light guide plate mark; obtain the optimized position of the film layer mark, and extract the optimized position syh(i,j) of the film layer mark in the horizontal direction and the optimized position cyh(i,j) of the film layer mark in the vertical direction according to the optimized position of the film layer mark.

[0103] Step S312: Calculate the deviation in the horizontal direction according to the horizontal position dsp(i,j) of the light guide plate mark and the optimized position syh(i,j) of the film layer mark in the horizontal direction to obtain the horizontal position deviation ppc.

[0104] ;

[0105] Calculate the deviation in the vertical direction according to the vertical position dcz(i,j) of the light guide plate mark and the optimized position cyh(i,j) of the film layer mark in the vertical direction to obtain the vertical position deviation cpc.

[0106] ;

[0107] It should be noted that: by calculating the mean of the horizontal position deviation and the vertical position deviation of all mark positions, the accuracy optimization of the horizontal position deviation and the vertical position deviation is limited, and the calculation efficiency of the accuracy optimization is improved.

[0108] Step S313: preset the horizontal accuracy compensation sys, and perform accuracy compensation on the optimized position syh(i,j) of the film layer mark in the horizontal direction according to the horizontal accuracy compensation, calculate the horizontal position dsp(i,j) of the light guide plate mark and the horizontal position deviation ppc to obtain the horizontal accuracy judgment value spd;

[0109] ;

[0110] It should be noted that: the value range of sys is [-ppc, ppc]; when the value of sys exceeds ppc, the overall deviation of the film layer is large, so it is limited, and when calculating the deviation, the offset direction is eliminated by absolute value, so the accuracy compensation is expanded by -ppc to enhance the calculation accuracy.

[0111] Iterate the preset horizontal accuracy compensation sys, record the horizontal accuracy judgment value in the corresponding state, obtain the minimum horizontal accuracy judgment value, extract the corresponding preset horizontal accuracy compensation, and obtain the optimal horizontal accuracy compensation;

[0112] Preset the vertical accuracy compensation cys, and perform accuracy compensation on the optimized position cyh(i,j) of the film layer mark in the vertical direction according to the vertical accuracy compensation, calculate the vertical position dcz(i,j) of the light guide plate mark and the vertical position deviation cpc to obtain the vertical accuracy judgment value cpd;

[0113] ;

[0114] Iterate the preset vertical accuracy compensation cys, record the vertical accuracy judgment value in the corresponding state, obtain the minimum vertical accuracy judgment value, extract the corresponding preset vertical accuracy compensation, and obtain the optimal vertical accuracy compensation;

[0115] Record the optimal horizontal accuracy compensation and the optimal vertical accuracy compensation to obtain the compensation data.

[0116] Step S32: according to the compensation data, combine the position deviation to predict the position deviation of the next time node, obtain the predicted deviation; obtain the film coating image of the next time node, calculate the actual deviation of the image according to the film coating image of the next time node, analyze the light guide plate film coating accuracy according to the actual deviation and the predicted deviation, and adjust the subsequent film coating process;

[0117] Step S321: According to the compensation data, the optimal horizontal precision compensation and the optimal vertical precision compensation are obtained, and the optimal horizontal precision compensation is denoted as zys, and the optimal vertical precision compensation is denoted as zyc. The horizontal position deviation ppc and the vertical position deviation cpc are obtained, and the optimal horizontal precision compensation and the optimal vertical precision compensation are combined to calculate the horizontal prediction deviation syc and the vertical prediction deviation cpc, syc = zys + ppc, and cpc = zyc + cpc.

[0118] The film covering image of the next time node is obtained, and the actual deviation of the image is obtained according to the film covering image of the next time node. The horizontal actual deviation is extracted from the actual deviation of the image, denoted as ssj(i, j), and the vertical actual deviation is extracted, denoted as csj(i, j).

[0119] Step S322: The prediction deviation and the actual deviation are analyzed, the horizontal precision compensation is judged by the horizontal actual deviation ssj(i, j) and the horizontal prediction deviation syc, the vertical precision compensation is judged by the vertical actual deviation csj(i, j) and the vertical prediction deviation cpc, and the precision compensation data of the subsequent time node is obtained.

[0120] Step S33: The light guide plate film covering precision in the film covering process is counted, the light guide plate film covering precision of different time nodes is analyzed, and the precision change sequence is constructed according to the light guide plate film covering precision of different time nodes.

[0121] Step S331: The precision compensation data of all time nodes in the film covering process is obtained, the light guide plate film covering precision is analyzed according to the precision compensation data, the horizontal direction precision compensation data is obtained from the precision compensation data, denoted as sbc(1) to sbc(ts), the horizontal precision change sequence is constructed according to the precision compensation in the horizontal direction, denoted as sbh(t); sbh(t) = sbc(t) + sbh(t-1); It should be noted that t∈[1, ts], when t = 1, sbh(t-1) = 0, and sbc(t) represents the precision compensation data of the horizontal direction of the tth time node.

[0122] Step S332: The vertical direction precision compensation data is obtained from the precision compensation data, denoted as cbc(1) to cbc(ts), the vertical precision change sequence is constructed according to the precision compensation in the vertical direction, denoted as cbh(t); cbh(t) = cbc(t) + cbh(t-1); It should be noted that when t = 1, cbh(t-1) = 0, and cbc(t) represents the precision compensation data of the vertical direction of the tth time node.

[0123] It should be noted that the accuracy change is the change value obtained after the accuracy compensation is continued on the existing change, so when calculating the accuracy change of the current time node, all accuracy compensations before the current time node need to be counted; the accuracy change directly reflects the geometric position change of the film layer in the film coating process.

[0124] Step S4: analyzing the accuracy change of the light guide plate film coating according to the accuracy change sequence, and optimizing the accuracy compensation of the light guide plate film coating in combination with the accuracy compensation of the light guide plate film coating.

[0125] According to the accuracy change sequence, the horizontal accuracy change sequence and the vertical accuracy change sequence are analyzed respectively; the horizontal accuracy change sequence sbh(t) is obtained, the horizontal change sequence sbh(t) is accumulated to obtain the horizontal change sum, and the light guide plate film coating is shifted in the horizontal direction according to the horizontal change sum; the vertical accuracy change sequence cbh(t) is obtained, the vertical change sequence cbh(t) is accumulated to obtain the vertical change sum, and the light guide plate film coating is shifted in the vertical direction according to the vertical change sum; the accuracy compensation of the light guide plate film coating is optimized according to the results of the overall shift of the light guide plate film coating in the horizontal direction and the overall shift of the light guide plate film coating in the vertical direction.

[0126] The preferred embodiments disclosed above are only used to help explain the present application. The preferred embodiments do not describe all the details, nor limit the present application to only the specific embodiments. Obviously, according to the content of the present application, many modifications and changes can be made. The present application selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can well understand and utilize the present application. The present application is limited only by the claims and their full scope and equivalents.

Claims

1. A method for dynamic compensation of precision of a light guide plate film for multi-modal sensing, characterized in that, The application comprises the following steps: Step S1: Real-time image acquisition is performed on the light guide plate film coating process, the positions of the light guide plate marks and the film layer marks are recorded, image information is obtained, pressure measurement is performed on the coated part based on the image information, and pressure data is obtained; Step S2: According to the image information, the positions of the light guide plate marks and the film layer marks are extracted; the film layer surface defects are analyzed according to the pressure data, the film layer is adjusted according to the film layer surface defects, and the position of the film layer mark is optimized; Step S3: Deviation calculation is performed according to the position of the film layer mark and the position of the light guide plate mark, the position deviation is obtained, precision compensation is performed on the light guide plate film coating according to the position deviation; the next time node coating image is obtained according to the precision compensation, the light guide plate film coating precision is analyzed according to the next time node coating image; the light guide plate film coating precision in the coating process is counted, and a precision change sequence is obtained; The specific steps of step S3 are as follows: Step S31: The positions of the film layer marks and the light guide plate marks are obtained, the positions of the film layer marks and the light guide plate marks are calculated, the position deviation is obtained, the precision compensation of the light guide plate film coating is analyzed according to the position deviation, and compensation data is obtained; Step S32: According to the compensation data, the position deviation of the next time node is predicted in combination with the position deviation, and a predicted deviation is obtained; the next time node coating image is obtained, the actual deviation of the image is calculated according to the next time node coating image, the light guide plate film coating precision is analyzed according to the actual deviation and the predicted deviation, and the subsequent coating process is adjusted; Step S33: The light guide plate film coating precision in the coating process is counted, the light guide plate film coating precision at different time nodes is analyzed, and a precision change sequence is constructed according to the light guide plate film coating precision at different time nodes; Step S4: The precision change of the light guide plate film coating is analyzed according to the precision change sequence, and the precision compensation of the light guide plate film coating is optimized in combination with the precision compensation of the light guide plate film coating.

2. The multi-modal sensing dynamic compensation method of light guide plate film precision according to claim 1, characterized in that, The specific steps of step S1 are as follows: Step S11: Alignment marks are set on the light guide plate to obtain light guide plate marks, the corresponding positions of the film layer are marked according to the light guide plate marks to obtain film layer marks, the number of alignment marks is is×js, the light guide plate marks are denoted as dgb(i,j), and the film layer marks are denoted as mcb(i,j); The time required for the light guide plate film coating is obtained, a plurality of time nodes are extracted according to the time required for the light guide plate film coating, the number of time nodes is ts, the light guide plate film coating image is acquired, the positions of the light guide plate marks and the film layer marks are recorded, and the coating images at different time nodes are obtained in combination with the number of time nodes; the image information of the light guide plate film coating is composed of the coating images at different time nodes; Step S12: The coated part is synchronously monitored according to the image information of the light guide plate film coating, the pressure borne by the film layer of the light guide plate is obtained, the pressure borne by the film layer is counted, and pressure data is obtained.

3. The multi-modal sensing dynamic compensation method of light guide plate film precision according to claim 1, characterized in that, The specific steps of step S2 are as follows: Step S21: according to the image information, obtaining the film image of the current time node, obtaining the pressure data according to the film image of the current time node, extracting the film layer surface defect according to the pressure data, adjusting the film layer through the film layer surface defect combined with the pressure data, and controlling the adjustment direction of the film layer surface defect through the pressure data; Step S22: adjusting and optimizing the position of the film layer mark according to the adjustment direction of the film layer surface defect, and counting the position of the film layer mark after adjustment and optimization.

4. The multi-modal sensing dynamic compensation method of light guide plate film precision according to claim 3, characterized in that, The specific steps of the step S21 are as follows: Step S211: Obtain the current time node's laminated image, and obtain the pixel number of the laminated image according to the current time node's laminated image, denoted as as t ×bs t ; obtain the pressure data, map the pressure data with the pixels of the laminated image, and obtain the pressure values ylz(a t , b t ) of different pixel positions; Step S212: statistics of pressure values ylz(a t , b t ) of different pixel positions, loop calculation of pressure values of pixels in horizontal direction according to pressure values ylz(a t , b t ), analysis of abnormal values of pressure values in horizontal direction, and obtaining horizontal abnormal values spy(a t , b t ); Anomaly analysis of pressure values at image pixel (a, b) locations according to horizontal anomaly values spy(a t , b t ): When spy(a t , b t ) = 0, it indicates that the pressure value at the image pixel (a, b) position is normal; When spy(a t , b t ) > 0, it indicates that the pressure value at the image pixel (a, b) position is too high, and the film layer is marked to be offset in the horizontal direction to the (a, b) position. When spy(a t , b t ) < 0, it indicates that the pressure value at the image pixel (a, b) position is too low, and the film layer is marked to be offset in the horizontal direction away from the (a, b) position; According to the pressure value ylz(a t , b t ), the pressure value of the pixel in the vertical direction is calculated in a loop, the abnormal value of the pressure value in the vertical direction is analyzed, and the vertical abnormal value czy(a t , b t ) is obtained. The pressure value at the image pixel (a, b) location is analyzed for outliers according to the vertical outlier value czy(a t , b t ): Whenczy(a t , b t ) = 0, it indicates that the pressure value at the image pixel (a, b) position is normal; When czv(a t , b t ) > 0, it indicates that the pressure value at the image pixel (a, b) position is too high, and the film layer is marked to be offset in the vertical direction to the (a, b) position. When czv(a t , b t ) < 0, it indicates that the pressure value at the image pixel (a, b) position is too low, and the film layer is marked to be offset in the vertical direction away from the (a, b) position; The offset amount of the film layer mark in the horizontal direction and the vertical direction is counted to obtain the position offset information, and the film layer is adjusted according to the position offset information.

5. The multi-modal sensing dynamic compensation method of light guide plate film precision according to claim 3, characterized in that, The specific steps of the step S22 are as follows: Step S221: according to the image information, obtaining the position of the light guide plate mark in the image, extracting the position of the corresponding film layer mark according to the position of the light guide plate mark in the image, and recording the position of the film layer mark as the position spf(i,j) in the horizontal direction and the position czf(i,j) in the vertical direction respectively; Step S222: according to the adjustment direction of the film layer surface defect, obtaining the position offset information of the film layer mark, extracting the horizontal position offset amount according to the position offset information of the film layer mark, recording the horizontal position offset amount as spy(i,j), extracting the vertical position offset amount, recording the vertical position offset amount as czy(i,j); Step S223: according to the position spf(i,j) of the film layer mark in the horizontal direction and the horizontal position offset amount spy(i,j), the optimized position syh(i,j) of the film layer mark in the horizontal direction is obtained; according to the position czf(i,j) of the film layer mark in the vertical direction and the vertical position offset amount czy(i,j), the optimized position cyh(i,j) of the film layer mark in the vertical direction is obtained; the optimized position of the film layer mark in the horizontal direction syh(i,j) and in the vertical direction cyh(i,j) is counted to obtain the optimized position of the film layer mark.

6. The multi-modal sensing dynamic compensation method of light guide plate film precision according to claim 1, characterized in that, The specific steps of the step S31 are as follows: Step S311: obtaining the position of the light guide plate mark, extracting the horizontal position dsp(i,j) of the light guide plate mark and the vertical position dcz(i,j) of the light guide plate mark according to the position of the light guide plate mark, obtaining the optimized position of the film layer mark, and extracting the optimized position syh(i,j) of the film layer mark in the horizontal direction and the optimized position cyh(i,j) of the film layer mark in the vertical direction according to the optimized position of the film layer mark; Step S312: calculating the horizontal deviation according to the horizontal position dsp(i,j) of the light guide plate mark and the optimized position syh(i,j) of the film layer mark in the horizontal direction to obtain the horizontal position deviation ppc; According to the vertical position dcz(i,j) of the light guide plate mark and the optimized position cyh(i,j) of the film layer mark in the vertical direction, the vertical deviation is calculated to obtain the vertical position deviation cpc. 7.The multi-modal sensing dynamic compensation method of LPTFE film precision according to claim 1, characterized in that, The specific steps of the step S31 further include: A preset horizontal precision compensation sys is used to compensate the optimized position syh(i, j) of the film layer mark in the horizontal direction, and a horizontal precision judgment value spd is obtained by combining the horizontal position of the light guide plate mark dsp(i, j) and the horizontal position deviation ppc; ； Wherein, is represents that the film layer mark exists in is row, and js represents that the film layer mark exists in js column; The preset horizontal precision compensation sys is traversed, the horizontal precision judgment value in the corresponding state is recorded, the minimum horizontal precision judgment value is obtained, the corresponding preset horizontal precision compensation is extracted, and the optimal horizontal precision compensation is obtained; A preset vertical precision compensation cys is used to compensate the optimized position cyh(i, j) of the film layer mark in the vertical direction, and a vertical precision judgment value cpd is obtained by combining the vertical position of the light guide plate mark dcz(i, j) and the vertical position deviation cpc; The preset vertical precision compensation cys is traversed, the vertical precision judgment value in the corresponding state is recorded, the minimum vertical precision judgment value is obtained, the corresponding preset vertical precision compensation is extracted, and the optimal vertical precision compensation is obtained; The optimal horizontal precision compensation and the optimal vertical precision compensation are recorded, and compensation data is obtained. 8.The multi-modal sensing dynamic compensation method of LPTFE film precision according to claim 1, characterized in that, The specific steps of the step S32 are as follows: Step S321: According to the compensation data, the optimal horizontal precision compensation and the optimal vertical precision compensation are obtained, the optimal horizontal precision compensation is recorded as zys, the optimal vertical precision compensation is recorded as zyc, the horizontal position deviation ppc and the vertical position deviation cpc are obtained, and the horizontal prediction deviation syc and the vertical prediction deviation cyc are obtained by combining the optimal horizontal precision compensation and the optimal vertical precision compensation; The film coating image of the next time node is obtained, the actual deviation of the image is obtained according to the film coating image of the next time node, the horizontal actual deviation is extracted from the actual deviation of the image, and is recorded as ssj(i, j); The vertical actual deviation is extracted and is recorded as csj(i, j); Step S322: The prediction deviation and the actual deviation are analyzed, the horizontal precision compensation is judged by the horizontal actual deviation ssj(i, j) and the horizontal prediction deviation syc, the vertical precision compensation is judged by the vertical actual deviation csj(i, j) and the vertical prediction deviation cyc, and the precision compensation data of the subsequent time node is obtained. 9.The multi-modal sensing dynamic compensation method of LPTFE film precision according to claim 1, characterized in that, The specific steps of the step S33 are as follows: Step S331: The precision compensation data of all time nodes of the film coating process is obtained, the precision compensation data is used to analyze the light guide plate film coating precision, the horizontal direction precision compensation data is obtained from the precision compensation data, and is recorded as sbc(1) to sbc(ts), the horizontal precision change sequence sbh(t) is constructed according to the horizontal direction precision compensation, and sbh(t) = sbc(t) + sbh(t-1); wherein: sbc(t) represents the precision compensation data of the horizontal direction of the tth time node; Step S332: Obtain the precision compensation data in the vertical direction from the precision compensation data, denoted as cbc(1) to cbc(ts), and construct a vertical precision change sequence denoted as cbh(t) according to the precision compensation in the vertical direction; cbh(t) = cbc(t) + cbh(t-1), wherein cbc(t) represents the precision compensation data in the vertical direction at the tth time node.

Images