Composite optical film slip monitoring system and method based on multi-source data analysis

Abstract

The application discloses a composite optical film slip monitoring system and method based on multi-source data analysis, and relates to the technical field of optical detection. The method comprises the following steps: acquiring an optical path distribution field of a composite optical film in a measurement field of view; calculating first-order partial derivatives of the optical path distribution field along orthogonal first and second directions to obtain first and second partial derivative fields; calculating the partial derivative of the first partial derivative field along the second direction to obtain a first mixed partial derivative field, and calculating the partial derivative of the second partial derivative field along the first direction to obtain a second mixed partial derivative field; calculating the difference between the first and second mixed partial derivative fields to construct a second-order mixed partial derivative difference field; comparing each field value in the difference field with a preset threshold to generate a binary label matrix; identifying slip candidate points according to the binary label matrix and converting the slip candidate points into physical coordinates to mark the position area of interlayer slip in the measurement field of view.

Description

Technical Field

[0001] This invention relates to the field of optical detection technology, specifically to a composite optical film slip monitoring system and method based on multi-source data analysis. Background Technology

[0002] Composite optical films are functional optical components made up of multiple layers of light-transmitting films of different materials that are precisely bonded together. During the manufacturing process and service, each film layer may experience micron-level interlayer relative displacement due to thermal and moisture expansion, stress relaxation, or interface aging, which is called "slippage". Slippage directly leads to uneven brightness and darkness of the image and optical distortion, which is a key defect affecting display quality.

[0003] The common paradigm of current slip monitoring technology is to directly or indirectly measure the position difference vector δ between layer A and layer B. Regardless of whether optical imaging, digital image correlation, capacitive grid or embedded sensing is used, they all rely on the same logical premise, that is, there exists an observation medium that can simultaneously, with the same reference, and without disturbance acquire the spatial coordinates of corresponding material points on the two layers, and use this to calculate the relative displacement.

[0004] However, this premise does not hold true in composite optical film structures.

[0005] After the composite optical film is bonded, the inner interface between layer A and layer B is encapsulated in an impenetrable stack, and each layer has a different refractive index and elastic modulus. Any non-destructive detection medium must pass through the heterogeneous interface to act on the measured point, and the slippage itself will change the geometry of the interface (misalignment, tilt or local gap), thereby changing the propagation path of the detection medium.

[0006] Taking optical detection as an example, slippage causes changes in the interface refraction conditions, distorting the geometric mapping relationship of the imaging system. The image point displacement ΔP is composed of the superposition of the actual slippage ΔP1 and the refractive distortion ΔP2. Both are determined by the slippage state δ, yet they are mutually causal. Contact detection also faces the dilemma of probe stress rewriting the local slippage state.

[0007] The resulting problem is that the measurement signal is a coupled function of the slip state and the distortion of the observation channel, and the two are inseparable; the system outputs only a single observation value ΔP, but is simultaneously affected by two independent variables, the actual displacement and the interface morphology, resulting in fewer equations than unknowns and no unique solution for the inversion.

[0008] This predicament does not stem from insufficient sensor accuracy or environmental noise interference, but rather from a cognitive misalignment between the measurement paradigm itself and the physical structure of the object being measured. That is, at the moment the bonding is completed, the physical quantity of "absolute position difference between layers" has lost the possibility of independent and undisturbed observation; any technical path that attempts to approximate this quantity cannot escape the cycle of the observation behavior rewriting the target quantity itself.

[0009] To address the aforementioned problems, this invention proposes a composite optical film slip monitoring system and method based on multi-source data analysis. Summary of the Invention

[0010] The purpose of this invention is to provide a composite optical film slip monitoring system and method based on multi-source data analysis to solve the problems raised in the prior art.

[0011] To achieve the above objectives, the present invention provides the following technical solution:

[0012] A method for monitoring slippage of composite optical films based on multi-source data analysis includes the following steps:

[0013] S1. Obtain the optical path distribution field of the composite optical film within the measurement field of view;

[0014] S2. Calculate the first partial derivative of the optical path distribution field along the first direction to obtain the first partial derivative field, and calculate the first partial derivative of the optical path distribution field along the second direction to obtain the second partial derivative field;

[0015] S3. Calculate the first partial derivative of the first partial derivative field along the second direction to obtain the first mixed partial derivative field, and calculate the first partial derivative of the second partial derivative field along the first direction to obtain the second mixed partial derivative field.

[0016] S4. Calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field to obtain the second-order mixed partial derivative difference field;

[0017] S5. Compare each field value in the second-order mixed partial derivative difference field with a preset threshold.

[0018] S6. Based on the comparison results, mark the location area where the composite optical film undergoes interlayer slippage in the measurement field of view.

[0019] S1 further includes the following:

[0020] The optical path distribution field refers to the two-dimensional scalar distribution of the accumulated optical path length at various spatial positions within the measurement field of view after the light beam passes through the composite optical film.

[0021] The optical path length is the integral of the product of the geometric path length and the refractive index of the medium over the beam propagation path.

[0022] The steps for obtaining the optical path distribution field are as follows:

[0023] A collimated beam is incident perpendicularly onto the first surface of the composite optical film;

[0024] The phase distribution of the transmitted wavefront emitted from the second surface of the composite optical film is acquired by a wavefront sensor.

[0025] Based on the conversion relationship between the phase distribution of the transmitted wavefront and the detection wavelength, the optical path length at each position within the measurement field of view is calculated point by point.

[0026] The optical path lengths at all positions within the measurement field of view are constructed as the optical path distribution field.

[0027] S2 further includes the following:

[0028] The optical path distribution field is stored in discretized matrix form, and this matrix is ​​denoted as the optical path distribution field matrix L;

[0029] The row direction of the matrix is ​​defined as the first direction, and the column direction is defined as the second direction;

[0030] Let L(i,j) be the element in the matrix located at row i and column j;

[0031] Where i is the row number, and its value ranges from 1 to i to M, and M is the total number of rows in the matrix;

[0032] j is the column index, and its value range is 1≤j≤N, where N is the total number of columns in the matrix;

[0033] Let Δx be the physical spatial interval between adjacent elements of the matrix in the first direction, and Δy be the physical spatial interval between adjacent elements of the matrix in the second direction;

[0034] Calculate the first-order partial derivative of the matrix along the first direction to obtain the first partial derivative field matrix Px(i,j);

[0035] Since discrete matrices cannot be directly differentiated, numerical difference is used to approximate partial derivatives, and different difference schemes are selected according to the position of the elements.

[0036] For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows:

[0037] Px(i,j)=[L(i+1,j)−L(i−1,j)] / (2·Δx);

[0038] For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows:

[0039] Px(1,j)=[L(2,j)−L(1,j)] / Δx;

[0040] For the lower boundary row element with row number i=M, the backward difference formula is used for calculation, as follows:

[0041] Px(M,j)=[L(M,j)−L(M−1,j)] / Δx;

[0042] Calculate the first-order partial derivative of the matrix along the second direction to obtain the second partial derivative field matrix Py(i,j):

[0043] For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows:

[0044] Py(i,j)=[L(i,j+1)−L(i,j−1)] / (2·Δy);

[0045] For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows:

[0046] Py(i,1)=[L(i,2)−L(i,1)] / Δy;

[0047] For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows:

[0048] Py(i,N)=[L(i,N)−L(i,N−1)] / Δy;

[0049] The calculated Px(i,j) matrix is ​​used as the first partial derivative field, and the Py(i,j) matrix is ​​used as the second partial derivative field.

[0050] S3 further includes the following:

[0051] Based on the first partial derivative field matrix Px(i,j), the second partial derivative field matrix Py(i,j), and the number of rows M, the number of columns N, the first direction physical space interval Δx, and the second direction physical space interval Δy of the optical path distribution field matrix L(i,j);

[0052] Calculate the first-order partial derivative of the first partial derivative field matrix Px(i,j) along the second direction to obtain the first mixed partial derivative field matrix Pxy(i,j);

[0053] For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows:

[0054] Pxy(i,j)=[Px(i,j+1)−Px(i,j−1)] / (2·Δy);

[0055] For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows:

[0056] Pxy(i,1)=[Px(i,2)−Px(i,1)] / Δy;

[0057] For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows:

[0058] Pxy(i,N)=[Px(i,N)−Px(i,N−1)] / Δy;

[0059] Calculate the first-order partial derivative of the second partial derivative field matrix Py(i,j) along the first direction to obtain the second mixed partial derivative field matrix Pyx(i,j);

[0060] For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows:

[0061] Pyx(i,j)=[Py(i+1,j)−Py(i−1,j)] / (2·Δx);

[0062] For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows:

[0063] Pyx(1,j)=[Py(2,j)−Py(1,j)] / Δx;

[0064] For the lower boundary row element with row number i=M, the backward difference formula is used for calculation:

[0065] Pyx(M,j)=[Py(M,j)−Py(M−1,j)] / Δx;

[0066] The calculated Pxy(i,j) matrix is ​​used as the first mixed partial field, and the Pyx(i,j) matrix is ​​used as the second mixed partial field.

[0067] Wherein, the values ​​of each element in the first mixed partial derivative field Pxy(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the first direction and then along the second direction.

[0068] The values ​​of each element in the second mixed partial derivative field Pyx(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the second direction and then along the first direction.

[0069] S4 further includes the following:

[0070] Obtain the first mixed partial-guided field matrix Pxy(i,j) and the second mixed partial-guided field matrix Pyx(i,j);

[0071] For each row number i, the range of values ​​for i is 1 ≤ i ≤ M;

[0072] For each column index j, the range of j is 1≤j≤N;

[0073] The element difference between the first mixed partial derivative field matrix Pxy(i,j) and the second mixed partial derivative field matrix Pyx(i,j) at the same position (i,j) is calculated using the following formula:

[0074] D(i,j)=Pxy(i,j)−Pyx(i,j);

[0075] Where D(i,j) is the difference between the first mixed partial field and the second mixed partial field at the corresponding spatial position;

[0076] Traverse all row indices i and column indices j, and arrange all calculated D(i,j) according to the original row and column order to construct a second-order mixed partial derivative difference field matrix D.

[0077] S5 further includes the following:

[0078] Obtain the second-order mixed partial derivative difference field matrix D. The elements of the second-order mixed partial derivative difference field matrix D are denoted as D(i,j), where the row number i, column number j, number of rows M, and number of columns N are defined in the same way as in S4.

[0079] A preset threshold T is set, which is a critical value used to distinguish the effective signal caused by slip from the background fluctuation;

[0080] The absolute value of each element D(i,j) is compared with the preset threshold T.

[0081] A binary label matrix B is generated based on the comparison results. The binary label matrix B has the same number of rows M and columns N as the second-order mixed partial derivative difference field matrix D.

[0082] The element B(i,j) at the position corresponding to D(i,j) in the binary label matrix B takes values ​​as follows:

[0083] When |D(i,j)|>T, B(i,j) is assigned the value 1, where 1 indicates that the corresponding position is a candidate point for sliding.

[0084] When |D(i,j)|≤T, B(i,j) is assigned the value 0, where 0 indicates that the corresponding position is a non-slip point.

[0085] S6 further includes the following:

[0086] Obtain the binary label matrix B, wherein the number of rows M and the number of columns N of the binary label matrix B are consistent with the optical path distribution field matrix L;

[0087] Iterate through all elements B(i,j) of the binary label matrix B;

[0088] The value range of row number i is 1≤i≤M, and the value range of column number j is 1≤j≤N;

[0089] The spatial position (i,j) corresponding to the element B(i,j) that takes the first label value is identified as a sliding candidate point;

[0090] Based on the row number i and column number j of the matrix, and combined with the first direction physical space interval Δx and the second direction physical space interval Δy, the matrix index coordinates of the sliding candidate points are converted into physical coordinates in the measurement field of view;

[0091] The physical coordinates are calculated as follows:

[0092] x = x0 + (j−1)·Δx;

[0093] y = y0 + (i−1)·Δy;

[0094] Where x is the physical coordinate value along the first direction in the measurement field of view, and y is the physical coordinate value along the second direction in the measurement field of view;

[0095] x0 and y0 are the initial physical coordinates of the origin of the measurement field of view in the first direction and the second direction, respectively;

[0096] The physical coordinates of all candidate sliding points are collected to form a set of sliding positions;

[0097] A visual marker is generated at the corresponding physical coordinates of the measurement field of view. The visual marker is used to indicate the specific location area where interlayer slippage occurs in the composite optical film.

[0098] The composite optical film slip monitoring system based on multi-source data analysis includes an optical path distribution field acquisition module, a first-order partial derivative field calculation module, a mixed partial derivative field calculation module, a second-order mixed partial derivative difference field calculation module, a threshold comparison module, and a slip position marking module.

[0099] The optical path distribution field acquisition module is used to acquire the optical path distribution field of the composite optical film within the measurement field of view;

[0100] The first-order partial derivative field calculation module is used to calculate the first-order partial derivative of the optical path distribution field along the first direction to generate a first partial derivative field, and to calculate the first-order partial derivative of the optical path distribution field along the second direction to generate a second partial derivative field.

[0101] The hybrid partial derivative field calculation module is used to calculate the first partial derivative of the first partial derivative field along the second direction to generate the first hybrid partial derivative field, and to calculate the first partial derivative of the second partial derivative field along the first direction to generate the second hybrid partial derivative field.

[0102] The second-order mixed partial derivative difference field calculation module is used to calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field, and construct a second-order mixed partial derivative difference field.

[0103] The threshold comparison module is used to compare each field value in the second-order mixed partial derivative difference field with a preset threshold, and generate a binary label matrix based on the comparison result;

[0104] The slip position marking module is used to identify slip candidate points according to the binary marking matrix, convert the matrix index coordinates of the slip candidate points into physical coordinates in the measurement field of view, and generate a visual mark at the corresponding position in the measurement field of view to indicate the location area where the composite optical film undergoes interlayer slip.

[0105] Compared with the prior art, the beneficial effects of the present invention are:

[0106] 1. This invention fundamentally eliminates the contamination of the measurement signal by the observation channel distortion by reconstructing the slip monitoring object as the second-order mixed partial derivative asymmetry of the optical path distribution field. This asymmetry index is naturally insensitive to scalar distortion fields caused by refractive index fluctuations, interface tilt, and thermal drift, but only responds to the non-conservative optical path component caused by interlayer slip. Therefore, this invention does not require any distortion compensation or clock synchronization at the hardware level, nor does it rely on an external spatial reference point. It directly filters out all observation disturbances from the original measurement data, obtaining a pure slip characteristic signal.

[0107] 2. This invention breaks through the inherent paradigm of existing technologies that "use positional difference as the measurand," and no longer attempts to solve for the interlayer absolute displacement vector, which loses the possibility of independent observation due to bonding and encapsulation. Instead, it utilizes the mathematical characteristic that the optical path distribution field loses its scalar potential property after slip occurs, and directly characterizes the spatial distribution of the slip shear gradient through the difference of mixed partial derivatives. This method completely eliminates the reliance of traditional technologies on the memory of initial coincidence points, artificial markers, or historical benchmarks, and for the first time realizes the autonomous positioning of interlayer slip of composite optical films without any prior reference conditions. Attached Figure Description

[0108] Figure 1 This is a schematic diagram of the composite optical film slip monitoring system and method based on multi-source data analysis of the present invention. Detailed Implementation

[0109] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0110] Example: Figure 1 As shown, the present invention provides a technical solution.

[0111] A method for monitoring slippage of composite optical films based on multi-source data analysis includes the following steps:

[0112] S1. Obtain the optical path distribution field of the composite optical film within the measurement field of view;

[0113] S2. Calculate the first partial derivative of the optical path distribution field along the first direction to obtain the first partial derivative field, and calculate the first partial derivative of the optical path distribution field along the second direction to obtain the second partial derivative field;

[0114] S3. Calculate the first partial derivative of the first partial derivative field along the second direction to obtain the first mixed partial derivative field, and calculate the first partial derivative of the second partial derivative field along the first direction to obtain the second mixed partial derivative field.

[0115] S4. Calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field to obtain the second-order mixed partial derivative difference field;

[0116] S5. Compare each field value in the second-order mixed partial derivative difference field with a preset threshold.

[0117] S6. Based on the comparison results, mark the location area where the composite optical film undergoes interlayer slippage in the measurement field of view.

[0118] S1 further includes the following:

[0119] The optical path distribution field refers to the two-dimensional scalar distribution of the accumulated optical path length at various spatial positions within the measurement field of view after the light beam passes through the composite optical film.

[0120] The optical path length is the integral of the product of the geometric path length and the refractive index of the medium over the beam propagation path.

[0121] The steps for obtaining the optical path distribution field are as follows:

[0122] A collimated beam is incident perpendicularly onto the first surface of the composite optical film;

[0123] The phase distribution of the transmitted wavefront emitted from the second surface of the composite optical film is acquired by a wavefront sensor.

[0124] Based on the conversion relationship between the phase distribution of the transmitted wavefront and the detection wavelength, the optical path length at each position within the measurement field of view is calculated point by point.

[0125] The optical path lengths at all positions within the measurement field of view are constructed as the optical path distribution field.

[0126] S2 further includes the following:

[0127] The optical path distribution field is stored in discretized matrix form, and this matrix is ​​denoted as the optical path distribution field matrix L;

[0128] The row direction of the matrix is ​​defined as the first direction, and the column direction is defined as the second direction;

[0129] Let L(i,j) be the element in the matrix located at row i and column j;

[0130] Where i is the row number, and its value ranges from 1 to i to M, and M is the total number of rows in the matrix;

[0131] j is the column index, and its value range is 1≤j≤N, where N is the total number of columns in the matrix;

[0132] Let Δx be the physical spatial interval between adjacent elements of the matrix in the first direction, and Δy be the physical spatial interval between adjacent elements of the matrix in the second direction;

[0133] Calculate the first-order partial derivative of the matrix along the first direction to obtain the first partial derivative field matrix Px(i,j);

[0134] Since discrete matrices cannot be directly differentiated, numerical difference is used to approximate partial derivatives, and different difference schemes are selected according to the position of the elements.

[0135] For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows:

[0136] Px(i,j)=[L(i+1,j)−L(i−1,j)] / (2·Δx);

[0137] For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows:

[0138] Px(1,j)=[L(2,j)−L(1,j)] / Δx;

[0139] For the lower boundary row element with row number i=M, the backward difference formula is used for calculation, as follows:

[0140] Px(M,j)=[L(M,j)−L(M−1,j)] / Δx;

[0141] Calculate the first-order partial derivative of the matrix along the second direction to obtain the second partial derivative field matrix Py(i,j):

[0142] For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows:

[0143] Py(i,j)=[L(i,j+1)−L(i,j−1)] / (2·Δy);

[0144] For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows:

[0145] Py(i,1)=[L(i,2)−L(i,1)] / Δy;

[0146] For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows:

[0147] Py(i,N)=[L(i,N)−L(i,N−1)] / Δy;

[0148] The calculated Px(i,j) matrix is ​​used as the first partial derivative field, and the Py(i,j) matrix is ​​used as the second partial derivative field.

[0149] S3 further includes the following:

[0150] Based on the first partial derivative field matrix Px(i,j), the second partial derivative field matrix Py(i,j), and the number of rows M, the number of columns N, the first direction physical space interval Δx, and the second direction physical space interval Δy of the optical path distribution field matrix L(i,j);

[0151] Calculate the first-order partial derivative of the first partial derivative field matrix Px(i,j) along the second direction to obtain the first mixed partial derivative field matrix Pxy(i,j);

[0152] For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows:

[0153] Pxy(i,j)=[Px(i,j+1)−Px(i,j−1)] / (2·Δy);

[0154] For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows:

[0155] Pxy(i,1)=[Px(i,2)−Px(i,1)] / Δy;

[0156] For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows:

[0157] Pxy(i,N)=[Px(i,N)−Px(i,N−1)] / Δy;

[0158] Calculate the first-order partial derivative of the second partial derivative field matrix Py(i,j) along the first direction to obtain the second mixed partial derivative field matrix Pyx(i,j);

[0159] For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows:

[0160] Pyx(i,j)=[Py(i+1,j)−Py(i−1,j)] / (2·Δx);

[0161] For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows:

[0162] Pyx(1,j)=[Py(2,j)−Py(1,j)] / Δx;

[0163] For the lower boundary row element with row number i=M, the backward difference formula is used for calculation:

[0164] Pyx(M,j)=[Py(M,j)−Py(M−1,j)] / Δx;

[0165] The calculated Pxy(i,j) matrix is ​​used as the first mixed partial field, and the Pyx(i,j) matrix is ​​used as the second mixed partial field.

[0166] Wherein, the values ​​of each element in the first mixed partial derivative field Pxy(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the first direction and then along the second direction.

[0167] The values ​​of each element in the second mixed partial derivative field Pyx(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the second direction and then along the first direction.

[0168] S4 further includes the following:

[0169] Obtain the first mixed partial-guided field matrix Pxy(i,j) and the second mixed partial-guided field matrix Pyx(i,j);

[0170] For each row number i, the range of values ​​for i is 1 ≤ i ≤ M;

[0171] For each column index j, the range of j is 1≤j≤N;

[0172] The element difference between the first mixed partial derivative field matrix Pxy(i,j) and the second mixed partial derivative field matrix Pyx(i,j) at the same position (i,j) is calculated using the following formula:

[0173] D(i,j)=Pxy(i,j)−Pyx(i,j);

[0174] Where D(i,j) is the difference between the first mixed partial field and the second mixed partial field at the corresponding spatial position;

[0175] Traverse all row indices i and column indices j, and arrange all calculated D(i,j) according to the original row and column order to construct a second-order mixed partial derivative difference field matrix D.

[0176] S5 further includes the following:

[0177] Obtain the second-order mixed partial derivative difference field matrix D. The elements of the second-order mixed partial derivative difference field matrix D are denoted as D(i,j), where the row number i, column number j, number of rows M, and number of columns N are defined in the same way as in S4.

[0178] A preset threshold T is set, which is a critical value used to distinguish the effective signal caused by slip from the background fluctuation;

[0179] The absolute value of each element D(i,j) is compared with the preset threshold T.

[0180] A binary label matrix B is generated based on the comparison results. The binary label matrix B has the same number of rows M and columns N as the second-order mixed partial derivative difference field matrix D.

[0181] The element B(i,j) at the position corresponding to D(i,j) in the binary label matrix B takes values ​​as follows:

[0182] When |D(i,j)|>T, B(i,j) is assigned the value 1, where 1 indicates that the corresponding position is a candidate point for sliding.

[0183] When |D(i,j)|≤T, B(i,j) is assigned the value 0, where 0 indicates that the corresponding position is a non-slip point.

[0184] S6 further includes the following:

[0185] Obtain the binary label matrix B, wherein the number of rows M and the number of columns N of the binary label matrix B are consistent with the optical path distribution field matrix L;

[0186] Iterate through all elements B(i,j) of the binary label matrix B;

[0187] The value range of row number i is 1≤i≤M, and the value range of column number j is 1≤j≤N;

[0188] The spatial position (i,j) corresponding to the element B(i,j) that takes the first label value is identified as a sliding candidate point;

[0189] Based on the row number i and column number j of the matrix, and combined with the first direction physical space interval Δx and the second direction physical space interval Δy, the matrix index coordinates of the sliding candidate points are converted into physical coordinates in the measurement field of view;

[0190] The physical coordinates are calculated as follows:

[0191] x = x0 + (j−1)·Δx;

[0192] y = y0 + (i−1)·Δy;

[0193] Where x is the physical coordinate value along the first direction in the measurement field of view, and y is the physical coordinate value along the second direction in the measurement field of view;

[0194] x0 and y0 are the initial physical coordinates of the origin of the measurement field of view in the first direction and the second direction, respectively;

[0195] The physical coordinates of all candidate sliding points are collected to form a set of sliding positions;

[0196] A visual marker is generated at the corresponding physical coordinates of the measurement field of view. The visual marker is used to indicate the specific location area where interlayer slippage occurs in the composite optical film.

[0197] The composite optical film slip monitoring system based on multi-source data analysis includes an optical path distribution field acquisition module, a first-order partial derivative field calculation module, a mixed partial derivative field calculation module, a second-order mixed partial derivative difference field calculation module, a threshold comparison module, and a slip position marking module.

[0198] The optical path distribution field acquisition module is used to acquire the optical path distribution field of the composite optical film within the measurement field of view;

[0199] The first-order partial derivative field calculation module is used to calculate the first-order partial derivative of the optical path distribution field along the first direction to generate a first partial derivative field, and to calculate the first-order partial derivative of the optical path distribution field along the second direction to generate a second partial derivative field.

[0200] The hybrid partial derivative field calculation module is used to calculate the first partial derivative of the first partial derivative field along the second direction to generate the first hybrid partial derivative field, and to calculate the first partial derivative of the second partial derivative field along the first direction to generate the second hybrid partial derivative field.

[0201] The second-order mixed partial derivative difference field calculation module is used to calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field, and construct a second-order mixed partial derivative difference field.

[0202] The threshold comparison module is used to compare each field value in the second-order mixed partial derivative difference field with a preset threshold, and generate a binary label matrix based on the comparison result;

[0203] The slip position marking module is used to identify slip candidate points according to the binary marking matrix, convert the matrix index coordinates of the slip candidate points into physical coordinates in the measurement field of view, and generate a visual mark at the corresponding position in the measurement field of view to indicate the location area where the composite optical film undergoes interlayer slip.

[0204] This embodiment addresses the micron-level interlayer slip detection scenario of composite optical films used in display panels. It employs a composite optical film slip monitoring system and method based on multi-source data analysis to achieve precise positioning and visual marking of interlayer slip in the composite optical film. In this embodiment, the measurement field of view is set as a rectangular optical detection area. A helium-neon laser is used as the detection light source, with a detection wavelength λ=632.8nm. The origin of the measurement field of view has initial physical coordinates of 0μm in both the first and second directions, i.e., x0=0μm, y0=0μm. The physical spatial interval of the optical path distribution field discrete matrix is ​​Δx=5μm in the first direction and Δy=5μm in the second direction. The matrix has a total of M=100 rows and N=80 columns. A preset threshold T=0.02μm⁻² is used to distinguish the effective slip signal from background fluctuations. The unit of all optical path-related calculations is unified in μm, and the unit of partial derivative-related calculations is derived from difference operations as μm⁻¹ or μm⁻².

[0205] In this embodiment, the composite optical film is a functional optical element consisting of three layers of transparent thin films precisely bonded together. During the manufacturing process, micron-level interlayer slippage occurs due to thermal and hygroscopic expansion, requiring monitoring of the slippage position using this method. The entire monitoring process relies on a dedicated slippage monitoring system, which consists of an optical path distribution field acquisition module, a first-order partial derivative field calculation module, a mixed partial derivative field calculation module, a second-order mixed partial derivative difference field calculation module, a threshold comparison module, and a slippage position marking module. These modules work together to complete the entire process from optical path data acquisition to slippage position marking. The specific monitoring implementation steps are as follows.

[0206] First, the optical path distribution field of the composite optical film within the measurement field of view is acquired using the optical path distribution field acquisition module. The optical path distribution field is a two-dimensional scalar distribution of the accumulated optical path lengths at various spatial positions within the measurement field of view after the light beam passes through the composite optical film. The optical path length is the integral of the product of the geometric path length and the refractive index of the medium over the beam propagation path. Specifically, the acquisition process involves: a collimated helium-neon laser is perpendicularly incident on the first surface of the composite optical film. After the laser penetrates the film, a wavefront sensor collects the phase distribution of the transmitted wavefront exiting from the second surface of the composite optical film. Then, based on the conversion relationship between the phase distribution of the transmitted wavefront and the detection wavelength of 632.8 nm, the optical path lengths at 100×80 spatial positions within the measurement field of view are calculated point by point. Finally, the optical path lengths at all positions are constructed as a discretized optical path distribution in row and column order. The process distribution field matrix L is denoted by L(i,j) (1≤i≤100, 1≤j≤80), where i is the row number and j is the column number. For example, L(1,1)=25.0μm, L(1,2)=25.1μm, L(2,1)=25.05μm, L(50,40)=30.0μm, and L(100,80)=35.0μm. The values ​​of each element in the matrix fluctuate regularly with the film layer distribution of the composite optical film and the change of the refractive index of the medium.

[0207] After the optical path distribution field matrix L is constructed, the first-order partial derivative field calculation module calculates the first-order partial derivatives of the matrix along the orthogonal first and second directions, generating the first partial derivative field matrix Px(i,j) and the second partial derivative field matrix Py(i,j), where the row direction of the matrix is ​​defined as the first direction and the column direction is defined as the second direction. Since discrete matrices cannot be directly differentiated, this embodiment uses numerical differencing to approximate partial derivatives. The central difference, forward difference, or backward difference scheme is selected based on the element's position in the matrix. The calculation rules and specific implementations for different difference schemes are as follows: For the first partial derivative field matrix Px(i,j), which is the first-order partial derivative of the optical path distribution field matrix along the first direction, for non-boundary row elements whose row indices satisfy 2≤i≤99, the central difference formula Px(i,j)=[L(i+1,j)−L(i−1,j)] / (2・Δx) is used for calculation. For example, when calculating Px(50,40), given L(51,40)=30.2μm and L(49,40)=29.8μm, substituting these values ​​yields Px(50,40)=(30.2-29.8) / (2×5)=0.04 μm⁻¹; When the upper boundary row element of row number i=1 is used, the forward difference formula Px(1,j)=[L(2,j)−L(1,j)] / Δx is used for calculation. For example, Px(1,40), given L(2,40)=28.5μm and L(1,40)=28.0μm, substituting them, we get Px(1,40)=(28.5-28.0) / 5=0.1μm⁻¹; When the row number is... The lower boundary row element with i=100 is calculated using the backward difference formula Px(M,j)=[L(M,j)−L(M−1,j)] / Δx. For example, for Px(100,40), given L(100,40)=32.0μm and L(99,40)=31.7μm, substituting them, we get Px(100,40)=(32.0-31.7) / 5=0.06μm⁻¹.

[0208] For the second partial derivative field matrix Py(i,j), which is the first-order partial derivative of the optical path distribution field matrix along the second direction, when the column index satisfies 2≤j≤79 for non-boundary column elements, the central difference formula Py(i,j)=[L(i,j+1)−L(i,j−1)] / (2・Δy) is used for calculation. For example, when calculating Py(50,40), given L(50,41)=30.1μm and L(50,39)=29.9μm, substituting them gives Py(50,40)=(30.1-29.9) / (2×5)=0.02μm⁻¹; when the column index j=1 for the left boundary column elements, the forward difference formula Py(i,1)=[L(i,j+1)−L(i,j−1)] / (2・Δy) is used for calculation. 2) Calculate using [−L(i,1)] / Δy. For example, for Py(50,1), given L(50,2)=27.3μm and L(50,1)=27.0μm, substituting, we get Py(50,1)=(27.3-27.0) / 5=0.06μm⁻¹; When the right boundary column element of column number j=80, use the backward difference formula Py(i,N)=[L(i,N)−L(i,N−1)] / Δy to calculate. For example, for Py(50,80), given L(50,80)=33.0μm and L(50,79)=32.8μm, substituting, we get Py(50,80)=(33.0-32.8) / 5=0.04μm⁻¹. After completing the calculation of 100×80 elements according to the above difference formula, the complete first partial derivative field matrix Px and second partial derivative field matrix Py are obtained, which serve as the basis for subsequent mixed partial derivative field calculations.

[0209] Subsequently, the hybrid partial derivative field calculation module calculates the partial derivatives of the first partial derivative field along the second direction and the partial derivatives of the second partial derivative field along the first direction, respectively, to obtain the first hybrid partial derivative field matrix Pxy(i,j) and the second hybrid partial derivative field matrix Pyx(i,j). Pxy(i,j) is a discrete approximation of the second-order hybrid partial derivative of the optical path distribution field after differentiating along the first direction and then along the second direction. Pyx(i,j) is a discrete approximation of the second-order hybrid partial derivative of the optical path distribution field after differentiating along the second direction and then along the first direction. The calculation of both is also based on the numerical difference method, and uses the same physical space intervals Δx and Δy and matrix row and column numbers M and N as the first-order partial derivatives.

[0210] The first mixed partial derivative field matrix Pxy(i,j) is calculated as the first-order partial derivative of Px(i,j) along the second direction. For non-boundary column elements where the column index satisfies 2≤j≤79, the central difference formula Pxy(i,j)=[Px(i,j+1)−Px(i,j−1)] / (2・Δy) is used. For example, when calculating Pxy(50,40), given Px(50,41)=0.045μm⁻¹ and Px(50,39)=0.035μm⁻¹, substituting them gives Pxy(50,40)=(0.045-0.035) / (2×5)=0.001μm⁻². For left boundary column elements where column index j=1, the forward difference formula Pxy(i,1)=[Px(i,2)] / [Px(i,j+1)−Px(i,j−1)] / (2・Δy) is used. The calculation is performed using -Px(i,1)] / Δy. For example, given Pxy(50,1), where Px(50,2)=0.05μm⁻¹ and Px(50,1)=0.03μm⁻¹, substituting these values ​​gives Pxy(50,1)=(0.05-0.03) / 5=0.004μm⁻². When the right boundary column element has column number j=80, the backward difference formula is used. The formula Pxy(i,N)=[Px(i,N)−Px(i,N−1)] / Δy is used for calculation. For example, for Pxy(50,80), given Px(50,80)=0.07μm⁻¹ and Px(50,79)=0.06μm⁻¹, substituting them into the formula gives Pxy(50,80)=(0.07-0.06) / 5=0.002μm⁻².

[0211] The second mixed partial derivative field matrix Pyx(i,j) is calculated as the first-order partial derivative of Py(i,j) along the first direction. For non-boundary row elements with row indices satisfying 2≤i≤99, the central difference formula Pyx(i,j)=[Py(i+1,j)−Py(i−1,j)] / (2・Δx) is used. For example, when calculating Pyx(50,40), given Py(51,40)=0.022μm⁻¹ and Py(49,40)=0.018μm⁻¹, substituting them gives Pyx(50,40)=(0.022-0.018) / (2×5)=0.0004μm⁻². For upper boundary row elements with row indices i=1, the forward difference formula Pyx(1,j)=[Py(2,j)−Py( Calculate 1,j)] / Δx, such as Pyx(1,40), given Py(2,40)=0.025μm⁻¹ and Py(1,40)=0.02μm⁻¹, substituting gives Pyx(1,40)=(0.025-0.02) / 5=0.001μm⁻²; when the lower boundary row element of row number i=100, the backward difference formula Pyx The calculation is as follows: (M,j) = [Py(M,j)−Py(M−1,j)] / Δx. For example, given Pyx(100,40), where Py(100,40) = 0.03μm⁻¹ and Py(99,40) = 0.026μm⁻¹, substituting these values ​​gives Pyx(100,40) = (0.03-0.026) / 5 = 0.0008μm⁻². After calculating all elements, the first mixed partial-guided field matrix Pxy and the second mixed partial-guided field matrix Pyx, both with a value of 100×80, are obtained.

[0212] Next, the second-order mixed partial derivative difference field calculation module calculates the difference between the first and second mixed partial derivative fields, constructing the second-order mixed partial derivative difference field matrix D. The specific calculation rules are as follows: For each row index i (1≤i≤100) and column index j (1≤j≤80) in the matrix, calculate the element difference between Pxy(i,j) and Pyx(i,j) at the same position (i,j). The calculation formula is D(i,j)=Pxy(i,j)−Pyx(i,j). After traversing all row and column indices to complete the difference calculation for all elements, arrange the results according to the original row and column order to obtain the second-order mixed partial derivative difference field matrix D, which has the same dimension as the original optical path distribution field matrix. In this embodiment, the difference field element values ​​in the non-slip region exhibit small fluctuations, such as D(50,40)=0.001-0.0004=0.0006μm⁻² and D(1,40)=0.0015-0.001=0.0005μm⁻²; while in the region where interlayer slip occurs in the composite optical film, the difference field element values ​​show significant abrupt changes, such as D(60,50)=0.0304-0.0004=0.03μm⁻² and D(70,60)=0.0258-0.0008=0.025μm⁻². These abrupt changes are the core characteristic signals for slip monitoring.

[0213] After the difference field matrix D is constructed, the threshold comparison module compares each field value in the matrix with a preset threshold T, and generates a binary label matrix B based on the comparison result. In this embodiment, the preset threshold T = 0.02 μm⁻², and the binary label matrix B and the difference field matrix D are 100×80 matrices of the same dimension. The rule for the value of its element B(i,j) is as follows: when |D(i,j)|>T, it indicates that the partial derivative difference at this position is an effective signal caused by slip, and B(i,j) is assigned a value of 1, representing that this position is a slip candidate point; when |D(i,j)|≤T, it indicates that the partial derivative difference at this position is background fluctuation, and B(i,j) is assigned a value of 0, representing that this position is a non-slip point. The difference field matrix is ​​evaluated element by element according to this rule. For example, if |D(50,40)|=0.0006μm⁻²≤0.02μm⁻², then B(50,40)=0; if |D(60,50)|=0.03μm⁻²>0.02μm⁻², then B(60,50)=1; if |D(70,60)|=0.025μm⁻²>0.02μm⁻², then B(70,60)=1. Finally, a binary label matrix B consisting only of 0 and 1 is obtained, which clearly distinguishes the slip candidate points and non-slip points in the measurement field of view.

[0214] Finally, the slip position marking module identifies slip candidate points based on the binary marking matrix B and converts them into physical coordinates in the measurement field of view, completing the marking of the interlayer slip position region. First, it iterates through all elements in the binary marking matrix B, identifying the matrix index coordinates (i,j) corresponding to elements with a value of 1 as slip candidate points. In this embodiment, the identified slip candidate points are (60,50), (70,60), and the surrounding continuous matrix coordinates with a value of 1, forming a set of slip candidate points. Then, according to the conversion formula between matrix index coordinates and physical coordinates, the matrix coordinates of the slip candidate points are converted into physical coordinates in the measurement field of view. The conversion formulas are x=x0+(j−1)・Δx and y=y0+(i−1)・Δy, where x is the physical coordinate value along the first direction in the measurement field of view, and y is the physical coordinate value along the second direction in the measurement field of view. In this embodiment, x0=0μm, y0=0μm, Δx=5μm, Δy=5μm. The sliding candidate point (60,50) is transformed to obtain x=(50-1)×5=245μm and y=(60-1)×5=295μm; the sliding candidate point (70,60) is transformed to obtain x=(60-1)×5=295μm and y=(70-1)×5=345μm.

[0215] After the physical coordinates of all slip candidate points are collected into a slip position set, a visual marker is generated at the corresponding physical coordinates in the measurement field of view. For example, a red highlight mark is marked in the (245μm,295μm), (295μm,345μm) and surrounding slip position set areas on the optical detection interface. This clearly indicates the specific location area where interlayer slip of the composite optical film occurs, thus completing the entire slip monitoring process.

[0216] This embodiment achieves high-precision monitoring of interlayer slip in composite optical films by combining the above-described system and method. In practical applications, this method reconstructs the slip monitoring object as the second-order mixed partial derivative asymmetry of the optical path distribution field, fundamentally eliminating the contamination of the measurement signal by the observation channel distortion. This asymmetry index is naturally insensitive to the scalar distortion field caused by refractive index fluctuations, interface tilt, and thermal drift, and only responds to the non-conservative optical path component caused by interlayer slip. Therefore, no distortion compensation or clock synchronization is required at the hardware level, and no external spatial reference point is needed. All observation disturbances are directly filtered out from the original measurement data, and a pure slip characteristic signal is obtained.

[0217] Meanwhile, this embodiment breaks through the inherent paradigm of traditional slip monitoring technology that "takes position difference as the measurand". Instead of trying to solve the interlayer absolute displacement vector that loses the possibility of independent observation due to bonding and encapsulation, it utilizes the mathematical characteristics of the optical path distribution field losing scalar potential properties after slip occurs. By directly characterizing the spatial distribution of slip shear gradient through the difference of mixed partial derivatives, it completely gets rid of the memory dependence of traditional technology on initial coincidence point, artificial mark or historical benchmark. It realizes the autonomous positioning of interlayer slip of composite optical film under the condition of no prior reference. The slip position error obtained by monitoring is controlled at the micrometer level, which fully meets the high-precision detection requirements of composite optical film for display panel.

[0218] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for monitoring slippage of composite optical films based on multi-source data analysis, characterized in that: Includes the following steps: S1. Obtain the optical path distribution field of the composite optical film within the measurement field of view; S2. Calculate the first partial derivative of the optical path distribution field along the first direction to obtain the first partial derivative field, and calculate the first partial derivative of the optical path distribution field along the second direction to obtain the second partial derivative field; S3. Calculate the first partial derivative of the first partial derivative field along the second direction to obtain the first mixed partial derivative field, and calculate the first partial derivative of the second partial derivative field along the first direction to obtain the second mixed partial derivative field. S4. Calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field to obtain the second-order mixed partial derivative difference field; S5. Compare each field value in the second-order mixed partial derivative difference field with a preset threshold. S6. Based on the comparison results, mark the location region in the measurement field of view where the composite optical film undergoes interlayer slippage; S1 further includes the following: The optical path distribution field refers to the two-dimensional scalar distribution of the accumulated optical path length at various spatial positions within the measurement field of view after the light beam passes through the composite optical film. The optical path length is the integral of the product of the geometric path length and the refractive index of the medium over the beam propagation path. The steps for obtaining the optical path distribution field are as follows: A collimated beam is incident perpendicularly onto the first surface of the composite optical film; The phase distribution of the transmitted wavefront emitted from the second surface of the composite optical film is acquired by a wavefront sensor. Based on the conversion relationship between the phase distribution of the transmitted wavefront and the detection wavelength, the optical path length at each position within the measurement field of view is calculated point by point. The optical path lengths at all positions within the measurement field of view are constructed as the optical path distribution field; S3 further includes the following: Based on the first partial derivative field matrix Px(i,j), the second partial derivative field matrix Py(i,j), and the number of rows M, the number of columns N, the first direction physical space interval Δx, and the second direction physical space interval Δy of the optical path distribution field matrix L(i,j); Calculate the first-order partial derivative of the first partial derivative field matrix Px(i,j) along the second direction to obtain the first mixed partial derivative field matrix Pxy(i,j); For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows: Pxy(i,j)=[Px(i,j+1)−Px(i,j−1)] / (2·Δy); For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows: Pxy(i,1)=[Px(i,2)−Px(i,1)] / Δy; For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows: Pxy(i,N)=[Px(i,N)−Px(i,N−1)] / Δy; Calculate the first-order partial derivative of the second partial derivative field matrix Py(i,j) along the first direction to obtain the second mixed partial derivative field matrix Pyx(i,j); For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows: Pyx(i,j)=[Py(i+1,j)−Py(i−1,j)] / (2·Δx); For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows: Pyx(1,j)=[Py(2,j)−Py(1,j)] / Δx; For the lower boundary row element with row number i=M, the backward difference formula is used for calculation: Pyx(M,j)=[Py(M,j)−Py(M−1,j)] / Δx; The calculated Pxy(i,j) matrix is ​​used as the first mixed partial field, and the Pyx(i,j) matrix is ​​used as the second mixed partial field. Wherein, the values ​​of each element in the first mixed partial derivative field Pxy(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the first direction and then along the second direction. The element values ​​in the second mixed partial derivative field Pyx(i,j) are discrete approximations of the second-order mixed partial derivative values ​​obtained by first differentiating the optical path distribution field along the second direction and then along the first direction. S4 further includes the following: Obtain the first mixed partial-guided field matrix Pxy(i,j) and the second mixed partial-guided field matrix Pyx(i,j); For each row number i, the range of values ​​for i is 1 ≤ i ≤ M; For each column index j, the range of j is 1≤j≤N; The element difference between the first mixed partial derivative field matrix Pxy(i,j) and the second mixed partial derivative field matrix Pyx(i,j) at the same position (i,j) is calculated using the following formula: D(i,j)=Pxy(i,j)−Pyx(i,j); Where D(i,j) is the difference between the first mixed partial field and the second mixed partial field at the corresponding spatial position; Traverse all row indices i and column indices j, and arrange all calculated D(i,j) according to the original row and column order to construct a second-order mixed partial derivative difference field matrix D.

2. The composite optical film slip monitoring method based on multi-source data analysis according to claim 1, characterized in that: S2 further includes the following: The optical path distribution field is stored in discretized matrix form, and this matrix is ​​denoted as the optical path distribution field matrix L; The row direction of the matrix is ​​defined as the first direction, and the column direction is defined as the second direction; Let L(i,j) be the element in the matrix located at row i and column j; Where i is the row number, and its value ranges from 1 to i to M, and M is the total number of rows in the matrix; j is the column index, and its value range is 1≤j≤N, where N is the total number of columns in the matrix; Let Δx be the physical spatial interval between adjacent elements of the matrix in the first direction, and Δy be the physical spatial interval between adjacent elements of the matrix in the second direction; Calculate the first-order partial derivative of the matrix along the first direction to obtain the first partial derivative field matrix Px(i,j); Since discrete matrices cannot be directly differentiated, numerical difference is used to approximate partial derivatives, and different difference schemes are selected according to the position of the elements. For non-boundary row elements with row number i satisfying 2 ≤ i ≤ M−1, the central difference formula is used for calculation, as follows: Px(i,j)=[L(i+1,j)−L(i−1,j)] / (2·Δx); For the elements in the upper boundary row with row number i=1, the forward difference formula is used for calculation, as follows: Px(1,j)=[L(2,j)−L(1,j)] / Δx; For the lower boundary row element with row number i=M, the backward difference formula is used for calculation, as follows: Px(M,j)=[L(M,j)−L(M−1,j)] / Δx; Calculate the first-order partial derivative of the matrix along the second direction to obtain the second partial derivative field matrix Py(i,j): For non-boundary column elements with column index j satisfying 2≤j≤N−1, the central difference formula is used for calculation, as follows: Py(i,j)=[L(i,j+1)−L(i,j−1)] / (2·Δy); For the left boundary column element with column index j=1, the forward difference formula is used for calculation, as follows: Py(i,1)=[L(i,2)−L(i,1)] / Δy; For the right boundary column elements where column index j=N, the backward difference formula is used for calculation, as follows: Py(i,N)=[L(i,N)−L(i,N−1)] / Δy; The calculated Px(i,j) matrix is ​​used as the first partial derivative field, and the Py(i,j) matrix is ​​used as the second partial derivative field.

3. The composite optical film slip monitoring method based on multi-source data analysis according to claim 1, characterized in that: S5 further includes the following: Obtain the second-order mixed partial derivative difference field matrix D. The elements of the second-order mixed partial derivative difference field matrix D are denoted as D(i,j), where the row number i, column number j, number of rows M, and number of columns N are defined in the same way as in S4. A preset threshold T is set, which is a critical value used to distinguish the effective signal caused by slip from the background fluctuation; The absolute value of each element D(i,j) is compared with the preset threshold T. A binary label matrix B is generated based on the comparison results. The binary label matrix B has the same number of rows M and columns N as the second-order mixed partial derivative difference field matrix D. The element B(i,j) at the position corresponding to D(i,j) in the binary label matrix B takes values ​​as follows: When |D(i,j)|>T, B(i,j) is assigned the value 1, where 1 indicates that the corresponding position is a candidate point for sliding. When |D(i,j)|≤T, B(i,j) is assigned the value 0, where 0 indicates that the corresponding position is a non-slip point.

4. The composite optical film slip monitoring method based on multi-source data analysis according to claim 3, characterized in that: S6 further includes the following: Obtain the binary label matrix B, wherein the number of rows M and the number of columns N of the binary label matrix B are consistent with the optical path distribution field matrix L; Iterate through all elements B(i,j) of the binary label matrix B; The range of row number i is 1≤i≤M, and the range of column number j is 1≤j≤N; The spatial position (i,j) corresponding to the element B(i,j) that takes the first label value is identified as a sliding candidate point; Based on the row number i and column number j of the matrix, and combined with the first direction physical space interval Δx and the second direction physical space interval Δy, the matrix index coordinates of the sliding candidate points are converted into physical coordinates in the measurement field of view; The physical coordinates are calculated as follows: x = x0 + (j−1)·Δx; y = y0 + (i−1)·Δy; Where x is the physical coordinate value along the first direction in the measurement field of view, and y is the physical coordinate value along the second direction in the measurement field of view; x0 and y0 are the initial physical coordinates of the origin of the measurement field of view in the first direction and the second direction, respectively; The physical coordinates of all candidate sliding points are collected to form a set of sliding positions; A visual marker is generated at the corresponding physical coordinates of the measurement field of view. The visual marker is used to indicate the specific location area where interlayer slippage occurs in the composite optical film.

5. A composite optical film slip monitoring system based on multi-source data analysis, applied to the composite optical film slip monitoring method based on multi-source data analysis as described in any one of claims 1-4, characterized in that: It includes an optical path distribution field acquisition module, a first-order partial derivative field calculation module, a mixed partial derivative field calculation module, a second-order mixed partial derivative difference field calculation module, a threshold comparison module, and a slip position marking module; The optical path distribution field acquisition module is used to acquire the optical path distribution field of the composite optical film within the measurement field of view; The first-order partial derivative field calculation module is used to calculate the first-order partial derivative of the optical path distribution field along the first direction to generate a first partial derivative field, and to calculate the first-order partial derivative of the optical path distribution field along the second direction to generate a second partial derivative field. The hybrid partial derivative field calculation module is used to calculate the first partial derivative of the first partial derivative field along the second direction to generate the first hybrid partial derivative field, and to calculate the first partial derivative of the second partial derivative field along the first direction to generate the second hybrid partial derivative field. The second-order mixed partial derivative difference field calculation module is used to calculate the difference between the first mixed partial derivative field and the second mixed partial derivative field, and construct a second-order mixed partial derivative difference field. The threshold comparison module is used to compare each field value in the second-order mixed partial derivative difference field with a preset threshold, and generate a binary label matrix based on the comparison result; The slip position marking module is used to identify slip candidate points according to the binary marking matrix, convert the matrix index coordinates of the slip candidate points into physical coordinates in the measurement field of view, and generate a visual mark at the corresponding position in the measurement field of view to indicate the location area where the composite optical film undergoes interlayer slip.

Images