Slide block gasket thickness adjusting method based on line laser 3D camera

Abstract

The invention discloses a slide block gasket thickness adjusting method based on a line laser 3D camera. The slide block gasket thickness adjusting method comprises the following steps: S1, acquiring morphology data of three slide blocks; s2, carrying out preprocessing on the obtained morphology data; s3, analyzing the processed morphology data; s4, transmitting the thickness measurement data to a control system; s5, the control system judges whether the thickness and number of the gaskets need to be adjusted or not according to the thickness size data; s6, the control system controls a vacuum suction cup mechanism on the mechanical arm to suck the gasket with the proper thickness; s7, the gasket is placed on the sliding block seat, and whether the position of the gasket is correct or not is determined through an improved image processing algorithm; s8, the control system dynamically adjusts the thickness of the gasket according to the real-time feedback data; and S9, repeating the steps S1 to S8 until the scanning data of the three sliding blocks meet the set requirements. According to the invention, the precision and efficiency of adjusting the thickness of the slide block gasket are obviously improved, and full-automatic operation and real-time feedback control are realized.

Description

Technical Field

[0001] The invention relates to the technical field of slider gasket thickness adjustment, and in particular to a slider gasket thickness adjustment method based on a line laser 3D camera. Background Art

[0002] In modern industrial production processes, the thickness adjustment and control of slider gaskets is an important part of ensuring product quality and performance. Traditional slider gasket thickness measurement methods mainly rely on contact measurement tools such as calipers and mechanical gauges. These methods are not only complex and time-consuming to operate, but are also easily affected by human operation, resulting in low measurement accuracy. In addition, contact measurement may cause damage to the gasket surface, affecting product quality and service life. Especially on automated production lines, this manual measurement method is difficult to achieve efficient and accurate quality control.

[0003] With the development of industrial automation, non-contact measurement technology has gradually been applied. Among them, line laser 3D camera technology has become an important non-contact measurement method due to its high precision and high efficiency. However, the existing technology still has some technical difficulties in the process of adjusting the thickness of the slider gasket, which limits its application effect and accuracy.

[0004] First, the existing line laser 3D camera technology is easily affected by the scanning angle and surface reflection characteristics when acquiring the three-dimensional topography data of the slider gasket, resulting in incomplete data acquisition or interference. These interferences will affect the subsequent thickness measurement and adjustment process, reducing the measurement accuracy and reliability. Especially in the boundary area, the traditional scanning method is difficult to eliminate the dead angle interference and spike interference, resulting in inaccurate topography data.

[0005] Secondly, in the thickness adjustment process, most existing control systems rely on simple measurement data and preset adjustment parameters, lacking real-time dynamic adjustment capabilities. This method is difficult to adapt to complex industrial environments and changing production needs, resulting in insufficient adjustment accuracy and low efficiency. At the same time, traditional thickness adjustment methods usually require manual intervention, which is prone to human errors and difficult to achieve fully automated operation.

[0006] In addition, existing image processing algorithms also have certain limitations in gasket position detection and confirmation. Traditional image processing algorithms are easily affected by lighting conditions and background noise in complex industrial environments, resulting in inaccurate gasket position detection. Especially in the thickness adjustment process of slider gaskets, the position and thickness of the gasket are closely related, and the error in position detection will directly affect the accuracy and effect of thickness adjustment. Summary of the invention

[0007] One purpose of the present invention is to propose a slider gasket thickness adjustment method based on a line laser 3D camera. The present invention significantly improves the accuracy and efficiency of slider gasket thickness adjustment, realizes fully automated operation and real-time feedback control, and meets the needs of modern industrial production for efficient and precise adjustment.

[0008] A method for adjusting the thickness of a slider pad based on a line laser 3D camera according to an embodiment of the present invention comprises the following steps:

[0009] S1. Use a line laser 3D camera to scan the front, middle and rear sliders to obtain the topographic data of the three sliders;

[0010] S2. Preprocess the acquired topographic data to eliminate dead angle interference and spike interference;

[0011] S3, by analyzing the processed morphological data, using an improved multi-scale iterative reconstruction algorithm to generate the required gasket thickness dimension data;

[0012] S4, transmitting thickness measurement data to a control system;

[0013] S5. The control system determines whether the thickness and quantity of the gaskets need to be adjusted according to the thickness dimension data;

[0014] S6, the control system controls the vacuum suction cup mechanism on the robot arm to suck the gasket of appropriate thickness;

[0015] S7, placing the gasket on the slider seat, and confirming whether the gasket position is correct by improving the image processing algorithm;

[0016] S8, the control system dynamically adjusts the gasket thickness according to the real-time feedback data;

[0017] S9, repeat steps S1-S8 until the scan data of the three sliders meet the set requirements.

[0018] Optionally, the S1 specifically includes the following steps:

[0019] S11, start the line laser 3D camera to scan the front, middle and rear sliders respectively, and the scanning path of each slider is from left to right and from top to bottom, forming a grid-like scanning pattern;

[0020] S12, obtaining preliminary morphological data of each slider, the preliminary morphological data including height h 前 (x, y), h 中 (x, y), h 后 (x, y) and surface profile C 前 (x, y), C 中 (x, y), C 后(x, y), where x and y represent the horizontal and vertical coordinates of the slider surface, respectively;

[0021] S13, repeatedly scan the preliminary morphological data of each slider multiple times to eliminate the error that may occur in a single scan and form multiple repeated morphological data sets Where i represents the i-th scan, and n is the number of repeated scans;

[0022] S14. For multiple repeated morphological data sets Perform data fusion and use weighted average algorithm to generate the final morphological data H 前 (x, y), H 中 (x, y), H 后 (x, y).

[0023] Optionally, the S2 specifically includes the following steps:

[0024] S21, the preliminary morphological data H of each slider is obtained 前 (x, y), H 中 (x, y), H 后 (x, y), perform boundary detection, identify and mark the boundary area in the slider topography data;

[0025] S22, using the adaptive multi-angle fusion algorithm to process the identified boundary area, dividing the preliminary morphological data of each slider into several small areas, recorded as Where m is the number of segmented regions;

[0026] For each small area R j Perform multi-angle scanning to obtain topographic data at multiple different angles Where θ represents different scanning angles, and p is the number of scanning angles;

[0027] For each small region R j , apply the adaptive weight function W j (θ) Perform weighted fusion on the morphological data at different angles to generate morphological data H′ without blind spot interference j (x, y):

[0028]

[0029] Among them, W j (θ) is the adaptive weight, which is dynamically adjusted according to the scanning quality at each angle;

[0030] All small areas R j No blind spot interference morphological data H′ j (x, y) splicing to generate the overall morphological data H′ after preliminary correction 前(x, y), H′ 中 (x, y), H′ 后 (x, y);

[0031] S23, the topographic data H′ after preliminary correction 前 (x, y), H′ 中 (x, y), H′ 后 (x, y) performs spike interference detection to identify and mark abnormally high or low value points;

[0032] S24, apply the spike interference filtering algorithm to interpolate the marked abnormal points, use the bilinear interpolation method to perform weighted average of the values ​​of the surrounding normal points, and generate the final preprocessed morphological data

[0033]

[0034] Among them, W(u,v) is the interpolation weight, which indicates the contribution of each neighborhood point to the interpolation result in the interpolation process. The selection of weight is determined according to the distance of the surrounding points and the smoothness of the data. The weight of the closer point is larger and the weight of the farther point is smaller. k is the interpolation window size. H'(x+u,y+v) is the topographic data after preliminary correction, indicating the height value at the position (x+u,y+v). The result of the interpolation process H f (x, y) is the final preprocessed height value at position (x, y);

[0035] S25, the final preprocessed morphological data Stored in the control system.

[0036] Optionally, S3 specifically includes the following steps:

[0037] S31 preprocessed morphological data Perform multi-angle scanning to generate topographic data sets at multiple angles Where θ represents different scanning angles and p is the number of scanning angles:

[0038]

[0039] Among them, H θ (x, y) represents the topographic data at the scanning angle θ, x and y are the horizontal and vertical coordinates of the slider surface, respectively, I(x′, y′, z′) is the reflection intensity of the scanned surface at the spatial position (x′, y′, z′), which represents the reflection of the slider surface to the laser, G θ(xx′, yy′, zz′) is the scanning response function, which describes the response characteristics of laser scanning at different angles. The integral areas [a, b] and [c, d] are the boundaries of the scanning area, indicating the range covered by the scanning;

[0040] S32 normalizes the topographic data at each angle to eliminate the influence of angle differences. The normalized topographic data is H′ θ (x, y);

[0041] S33 uses an adaptive weight fusion algorithm to perform weighted fusion on the topography data from multiple angles to generate comprehensive topography data H c (x, y):

[0042]

[0043] Among them, H c (x, y) is the fused comprehensive morphological data, W(θ) is the adaptive weight, which is dynamically adjusted according to the quality of each angle data, μ θ and σ θ are the mean and standard deviation of the topographic data at angle θ, respectively, which are used to measure the central tendency and dispersion of the data;

[0044] S34 for comprehensive morphological data H c (x, y) is processed at multiple scales to generate a multi-scale hierarchical model

[0045]

[0046] Among them, L k (x, y) is the hierarchical model at the kth level, n k is the window size, indicating the scale range of hierarchical processing, W k (i, j) is the weight function of the kth level, which is used to perform weighted processing on the comprehensive topographic data at different positions;

[0047] At each scale level, S35 uses an adaptive residual correction algorithm based on a neural network to calculate the morphological residual R of the current level. k (x, y):

[0048]

[0049] Among them, R k (x, y) is the shape residual of the kth level, w i is the weight, which is dynamically adjusted according to the output of the neural network;

[0050] S36 for the residual error R k(x, y) is subjected to adaptive smoothing based on wavelet transform, and the residual data after filtering is R k，f (x, y):

[0051] R k，f (x, y) = W-1{H w (u,v)·W{R k (x, y)}};

[0052] Among them, W and W -1 denote wavelet transform and inverse wavelet transform respectively, H w (u, v) is an adaptive wavelet filter function, which is used to smooth the residual data and eliminate noise;

[0053] S37 smooths the residual data R k，f (x, y) is added to the hierarchical model L of the current level k (x, y), update the hierarchical model:

[0054] L k+1 (x, y) = L k (x,y)+R k，f (x, y);

[0055] Among them, L k+1 (x, y) is the updated hierarchical model;

[0056] S38 repeats steps S35 to S37 until all hierarchical models at all levels have completed iterative reconstruction, generating the final multi-scale iterative reconstruction morphological data L q (x, y);

[0057] S39 is used to reconstruct the final multi-scale iterative morphological data L q (x, y) for thickness dimension analysis, using thickness calculation formula T(x, y):

[0058] T(x, y) = L q (x, y)-B(x, y);

[0059] Among them, T(x, y) is the thickness dimension data of the required gasket generated, and B(x, y) is the reference surface data of the bottom of the slider, which represents the reference surface height of the slider at different positions.

[0060] Optionally, the S5 specifically includes the following steps:

[0061] S51, the calculated gasket thickness dimension data T 前 (x, y), T 中 (x, y), T 后 (x, y) is transmitted to the control system;

[0062] S52, the control system performs statistical analysis on the gasket thickness dimension data at each position and calculates the average thickness at each position

[0063]

[0064] Where N is the number of measurement points, T(x i ,y i ) is the position (x i ,y i ) at the thickness dimension data;

[0065] S53, the control system compares the calculated average thickness with the preset standard thickness T 标准 Compare and calculate thickness deviation ΔT 前 , ΔT 中 , ΔT 后 :

[0066]

[0067] Among them, ΔT is the thickness deviation, which indicates the difference between the current thickness and the standard thickness;

[0068] S54. Based on the thickness deviation ΔT, the control system determines whether the thickness and quantity of the gaskets need to be adjusted, as follows:

[0069] If |ΔT|≤∈, the current thickness is considered to be within the allowable error range and no adjustment is required;

[0070] If |ΔT|>∈, the thickness or quantity of the gasket needs to be adjusted, where ∈ is the allowable thickness error range;

[0071] S55, the control system determines the adjustment scheme according to the positive and negative thickness deviation:

[0072] If ΔT>0, it means the current thickness is too large and the thickness or quantity of the gasket needs to be reduced;

[0073] If ΔT<0, it means the current thickness is insufficient and the thickness or quantity of the gasket needs to be increased;

[0074] S56, the control system generates an adjustment instruction, including the thickness or quantity of the gasket to be adjusted, and the adjustment instruction includes a specific adjustment parameter ΔT 调整 :

[0075]

[0076] Where, ΔT 调整 is the thickness difference that needs to be adjusted.

[0077] Optionally, the S7 specifically includes the following steps:

[0078] S71, moving the gasket sucked by S6 to above the slider seat, and adjusting the posture of the robot arm so that the gasket is aligned with the predetermined position of the slider seat;

[0079] S72, using an improved image processing algorithm to perform positioning detection on the gasket, and obtaining the initial position data P of the gasket on the slider seat 初 (x, y), where x and y are the horizontal and vertical coordinates of the spacer on the slider seat;

[0080] S73, initial position data P of gasket 初 (x, y) performs edge detection to identify the edge contour E(x, y) of the gasket:

[0081]

[0082] in, Respectively represent the gradients of the pad position data in the x and y directions;

[0083] S74, compare the detected gasket edge profile E(x, y) with the standard edge profile E(x, y) of the slider seat. 标准 (x, y) is matched and the position deviation ΔP(x, y) is calculated:

[0084] ΔP(x, y) = E 标准 (x, y)-E(x, y);

[0085] Among them, ΔP(x, y) is the deviation of the shim position from the standard position, which represents the difference between the actual position of the shim on the slider seat and the expected position;

[0086] S75. Based on the position deviation ΔP (x, y), the control system determines whether the position of the gasket is correct, as follows:

[0087] If |ΔP(x, y)|≤∈1, the gasket position is considered correct;

[0088] If |ΔP(x, y)|>∈1, the position of the gasket needs to be adjusted, where ∈1 is the allowable deviation range;

[0089] S76, the control system generates a position adjustment instruction according to the position deviation ΔP (x, y), and the adjustment instruction includes a specific adjustment parameter ΔP 调整 (x, y):

[0090] ΔP 调整 (x, y) = ΔP(x, y);

[0091] Among them, ΔP 调整 (x, y) is the position deviation value that needs to be adjusted;

[0092] S77, sending the position adjustment instruction to the actuator, the actuator performs the actual position adjustment operation according to the instruction, ensuring that the gasket position matches the standard position, and finally confirming whether the gasket position is correct.

[0093] The beneficial effects of the present invention are:

[0094] The present invention uses a line laser 3D camera to perform multi-angle scanning to obtain complete three-dimensional morphological data of the slider gasket, and eliminates dead angle interference and spike-shaped interference through an adaptive multi-angle fusion algorithm to generate accurate morphological data. The morphological data is analyzed and processed through an improved multi-scale iterative reconstruction algorithm to generate the required gasket thickness dimension data to ensure the accuracy and reliability of thickness adjustment. In addition, the gasket position is accurately detected and confirmed through an improved image processing algorithm to ensure the correct position of the gasket on the slider seat, providing an accurate reference for thickness adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0096] Figure 1 A flowchart of a slider gasket thickness adjustment method based on a line laser 3D camera proposed by the present invention;

[0097] Figure 2 This is a schematic diagram of an improved multi-scale iterative reconstruction algorithm for a slider gasket thickness adjustment method based on a line laser 3D camera proposed in the present invention. DETAILED DESCRIPTION

[0098] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0099] refer to Figure 1-2 , a slider gasket thickness adjustment method based on a line laser 3D camera, comprising the following steps:

[0100] S1. Use a line laser 3D camera to scan the front, middle and rear sliders to obtain the topographic data of the three sliders;

[0101] S2. Preprocess the acquired topographic data to eliminate dead angle interference and spike interference;

[0102] S3, by analyzing the processed morphological data, using an improved multi-scale iterative reconstruction algorithm to generate the required gasket thickness dimension data;

[0103] S4, transmitting thickness measurement data to a control system;

[0104] S5. The control system determines whether the thickness and quantity of the gaskets need to be adjusted according to the thickness dimension data;

[0105] S6, the control system controls the vacuum suction cup mechanism on the robot arm to suck the gasket of appropriate thickness;

[0106] S7, placing the gasket on the slider seat, and confirming whether the gasket position is correct by improving the image processing algorithm;

[0107] S8, the control system dynamically adjusts the gasket thickness according to the real-time feedback data;

[0108] S9, repeat steps S1-S8 until the scan data of the three sliders meet the set requirements.

[0109] In this implementation, S1 specifically includes the following steps:

[0110] S11, start the line laser 3D camera to scan the front, middle and rear sliders respectively, and the scanning path of each slider is from left to right and from top to bottom, forming a grid-like scanning pattern;

[0111] S12, obtaining preliminary morphological data of each slider, the preliminary morphological data including height h 前 (x, y), h 中 (x, y), h 后 (x, y) and surface profile C 前 (x, y), C 中 (x, y), C 后 (x, y), where x and y represent the horizontal and vertical coordinates of the slider surface, respectively;

[0112] S13, repeatedly scan the preliminary morphological data of each slider multiple times to eliminate the error that may occur in a single scan and form multiple repeated morphological data sets Where i represents the i-th scan, and n is the number of repeated scans;

[0113] S14. For multiple repeated morphological data sets Perform data fusion and use weighted average algorithm to generate the final morphological data H 前 (x, y), H 中 (x, y), H 后 (x, y).

[0114] In this implementation, S2 specifically includes the following steps:

[0115] S21, the preliminary morphological data H of each slider is obtained 前 (x, y), H 中 (x, y), H后 (x, y) performs boundary detection to identify and mark the boundary area in the slider topography data;

[0116] S22, using the adaptive multi-angle fusion algorithm to process the identified boundary area, dividing the preliminary morphological data of each slider into several small areas, recorded as Where m is the number of segmented regions;

[0117] For each small area R j Perform multi-angle scanning to obtain topographic data at multiple different angles Where θ represents different scanning angles, and p is the number of scanning angles;

[0118] For each small region R j , apply the adaptive weight function W j (θ) Perform weighted fusion on the morphological data at different angles to generate morphological data H′ without blind spot interference j (x, y):

[0119]

[0120] Among them, W j (θ) is the adaptive weight, which is dynamically adjusted according to the scanning quality at each angle;

[0121] All small areas R j No blind spot interference morphological data H′ j (x, y) splicing to generate the overall morphological data H′ after preliminary correction 前 (x, y), H′ 中 (x, y), H′ 后 (x, y);

[0122] S23, the topographic data H′ after preliminary correction 前 (x, y), H′ 中 (x, y), H′ 后 (x, y) performs spike interference detection to identify and mark abnormal high or low value points;

[0123] S24, apply the spike interference filtering algorithm to interpolate the marked abnormal points, use the bilinear interpolation method to perform weighted average of the values ​​of the surrounding normal points, and generate the final preprocessed morphological data

[0124]

[0125] Among them, W(u,v) is the interpolation weight, which indicates the contribution of each neighborhood point to the interpolation result in the interpolation process. The selection of weight is determined according to the distance of the surrounding points and the smoothness of the data. The weight of the closer point is larger and the weight of the farther point is smaller. k is the interpolation window size. H'(x+u,y+v) is the topographic data after preliminary correction, indicating the height value at the position (x+u,y+v). The result of the interpolation process H f (x, y) is the final preprocessed height value at position (x, y);

[0126] S25, the final preprocessed morphological data Stored in the control system.

[0127] In this implementation, S3 specifically includes the following steps:

[0128] S31 preprocessed morphological data Perform multi-angle scanning to generate topographic data sets at multiple angles Where θ represents different scanning angles and p is the number of scanning angles:

[0129]

[0130] Among them, H θ (x, y) represents the topographic data at the scanning angle θ, x and y are the horizontal and vertical coordinates of the slider surface, respectively, I(x′, y′, z′) is the reflection intensity of the scanned surface at the spatial position (x′, y′, z′), which represents the reflection of the slider surface to the laser, G θ (xx′, yy′, zz′) is the scanning response function, which describes the response characteristics of laser scanning at different angles. The integral areas [a, b] and [c, d] are the boundaries of the scanning area, indicating the range covered by the scanning;

[0131] S32 normalizes the topographic data at each angle to eliminate the influence of angle differences. The normalized topographic data is H′ θ (x, y);

[0132] S33 uses an adaptive weight fusion algorithm to perform weighted fusion on the topography data from multiple angles to generate comprehensive topography data H c (x, y):

[0133]

[0134] Among them, H c (x, y) is the fused comprehensive morphological data, W(θ) is the adaptive weight, which is dynamically adjusted according to the quality of each angle data, μ θ and σ θare the mean and standard deviation of the topographic data at angle θ, respectively, which are used to measure the central tendency and dispersion of the data;

[0135] S34 for comprehensive morphological data H c (x, y) is processed at multiple scales to generate a multi-scale hierarchical model

[0136]

[0137] Among them, L k (x, y) is the hierarchical model at the kth level, n k is the window size, indicating the scale range of hierarchical processing, W k (i, j) is the weight function of the kth level, which is used to perform weighted processing on the comprehensive topographic data at different positions;

[0138] At each scale level, S35 uses an adaptive residual correction algorithm based on a neural network to calculate the morphological residual R of the current level. k (x, y):

[0139]

[0140] Among them, R k (x, y) is the shape residual of the kth level, w i is the weight, which is dynamically adjusted according to the output of the neural network;

[0141] S36 for the residual error R k (x, y) is subjected to adaptive smoothing based on wavelet transform, and the residual data after filtering is R k，f (x, y):

[0142] R k，f (x, y) = W-1{H w (u,v)·W{R k (x, y)}};

[0143] Among them, W and W -1 denote wavelet transform and inverse wavelet transform respectively, H w (u, v) is an adaptive wavelet filter function, which is used to smooth the residual data and eliminate noise;

[0144] S37 smooths the residual data R k，f (x, y) is added to the hierarchical model L of the current level k (x, y), update the hierarchical model:

[0145] L k+1 (x, y) = L k (x,y)+R k，f(x, y);

[0146] Among them, L k+1 (x, y) is the updated hierarchical model;

[0147] S38 repeats steps S35 to S37 until all hierarchical models at all levels have completed iterative reconstruction, generating the final multi-scale iterative reconstruction morphological data L q (x, y);

[0148] S39 is used to reconstruct the final multi-scale iterative morphological data L q (x, y) for thickness dimension analysis, using thickness calculation formula T(x, y):

[0149] T(x, y) = L q (x, y)-B(x, y);

[0150] Among them, T(x, y) is the thickness dimension data of the required gasket generated, and B(x, y) is the reference surface data of the bottom of the slider, which represents the reference surface height of the slider at different positions.

[0151] In this implementation, S5 specifically includes the following steps:

[0152] S51, the calculated gasket thickness dimension data T 前 (x, y), T 中 (x, y), T 后 (x, y) is transmitted to the control system;

[0153] S52, the control system performs statistical analysis on the gasket thickness dimension data at each position and calculates the average thickness at each position

[0154]

[0155] Where N is the number of measurement points, T(x i ,y i ) is the position (x i ,y i ) at the thickness dimension data;

[0156] S53, the control system compares the calculated average thickness with the preset standard thickness T 标准 Compare and calculate thickness deviation ΔT 前 , ΔT 中 , ΔT 后 :

[0157]

[0158] Among them, ΔT is the thickness deviation, which indicates the difference between the current thickness and the standard thickness;

[0159] S54. Based on the thickness deviation ΔT, the control system determines whether the thickness and quantity of the gaskets need to be adjusted, as follows:

[0160] If |ΔT|≤∈, the current thickness is considered to be within the allowable error range and no adjustment is required;

[0161] If |ΔT|≤∈, the thickness or quantity of the gasket needs to be adjusted, where ∈ is the allowable thickness error range;

[0162] S55, the control system determines the adjustment scheme according to the positive and negative thickness deviation:

[0163] If ΔT>0, it means the current thickness is too large and the thickness or quantity of the gasket needs to be reduced;

[0164] If ΔT<0, it means the current thickness is insufficient and the thickness or quantity of the gasket needs to be increased;

[0165] S56, the control system generates an adjustment instruction, including the thickness or quantity of the gasket to be adjusted, and the adjustment instruction includes a specific adjustment parameter ΔT 调整 :

[0166]

[0167] Where, ΔT 调整 is the thickness difference that needs to be adjusted.

[0168] In this implementation, S7 specifically includes the following steps:

[0169] S71, moving the gasket sucked by S6 to above the slider seat, and adjusting the posture of the robot arm so that the gasket is aligned with the predetermined position of the slider seat;

[0170] S72, using an improved image processing algorithm to perform positioning detection on the gasket, and obtaining the initial position data P of the gasket on the slider seat 初 (x, y), where x and y are the horizontal and vertical coordinates of the pad on the slider seat;

[0171] S73, initial position data P of gasket 初 (x, y) performs edge detection to identify the edge contour E(x, y) of the gasket:

[0172]

[0173] in, Respectively represent the gradients of the pad position data in the x and y directions;

[0174] S74, compare the detected gasket edge profile E(x, y) with the standard edge profile E(x, y) of the slider seat. 标准(x, y) is matched and the position deviation ΔP(x, y) is calculated:

[0175] |ΔP(x, y)=E 标准 (x, y)-E(x, y);

[0176] Among them, ΔP(x, y) is the deviation of the shim position from the standard position, which represents the difference between the actual position of the shim on the slider seat and the expected position;

[0177] S75. Based on the position deviation ΔP (x, y), the control system determines whether the position of the gasket is correct, as follows:

[0178] If |ΔP(x, y)|≤∈1, the gasket position is considered correct;

[0179] If |ΔP(x, y)|>∈1, the position of the gasket needs to be adjusted, where ∈1 is the allowable deviation range;

[0180] S76, the control system generates a position adjustment instruction according to the position deviation ΔP (x, y), and the adjustment instruction includes a specific adjustment parameter ΔP 调整 (x, y):

[0181] |ΔP 调整 (x, y) = ΔP(x, y);

[0182] Among them, ΔP 调整 (x, y) is the position deviation value that needs to be adjusted;

[0183] S77, sending the position adjustment instruction to the actuator, the actuator performs the actual position adjustment operation according to the instruction, ensuring that the gasket position matches the standard position, and finally confirming whether the gasket position is correct.

[0184] Embodiment 1:

[0185] In the production workshop of a high-precision machinery manufacturing company, the thickness adjustment of the slider gasket has always been a crucial part of the production process. The company needs to produce a batch of slider gaskets for high-precision instruments. The thickness requirements of these gaskets are very strict, and the allowable error range is only ±0.01mm. However, the traditional measurement and adjustment methods mainly rely on manual operation and contact measuring tools, resulting in low production efficiency, insufficient accuracy, and easy human operation errors, affecting product quality. In order to solve these problems, the company introduced the present invention for testing.

[0186] In actual production, this embodiment selects a batch of 100 slider gaskets for experiment, and uses the traditional method and the method of the present invention to measure and adjust the thickness, so as to verify the feasibility and effectiveness of the present invention.

[0187] At the beginning of the experiment, each slider gasket was first scanned using a line laser 3D camera. The camera was mounted on a flexibly adjustable robotic arm to ensure that the scan could cover the entire gasket surface. During the scanning process, the camera acquired the gasket's topographic data from different angles. Through the adaptive multi-angle fusion algorithm, the dead angle interference and spike-shaped interference that may occur during the scanning process were eliminated, and high-precision 3D topographic data was generated.

[0188] The acquired topographic data is transmitted to the control system, and the system pre-processes the data. The improved multi-scale iterative reconstruction algorithm is used to perform layered processing and residual correction on the topographic data to generate accurate gasket thickness dimension data. These data are compared with the preset standard thickness to calculate the thickness deviation. If the thickness deviation exceeds the allowable error range, the system will generate an adjustment instruction to control the vacuum suction cup mechanism on the robot arm to perform thickness adjustment operations.

[0189] During the thickness adjustment process, the system first determines whether the thickness of the shim needs to be increased or decreased. According to the instructions, the robot arm picks up the shim of appropriate thickness and places it accurately on the slider seat. The improved image processing algorithm detects the position of the shim in real time to ensure that the shim is placed correctly. The whole process is completed in a few minutes, greatly improving production efficiency.

[0190] In order to verify the effect of the present invention, this embodiment compares the data of the traditional method and the method of the present invention in the process of measuring and adjusting the thickness of the slider pad.

[0191] Among 100 samples, the traditional method takes an average of about 20 minutes to measure and adjust each gasket, while the method of the present invention takes an average of only 4 minutes per gasket. The measurement accuracy of the traditional method is ±0.05mm, while the measurement accuracy of the method of the present invention reaches ±0.01mm, which significantly improves the measurement accuracy and consistency. The following Table 1 is the specific comparison data:

[0192] Table 1 Data comparison between the present invention and the traditional method

[0193]

[0194] During the adjustment process, the traditional method requires manual adjustments for many times, the adjustment accuracy is not high, the error range is large, and the adjustment success rate is only 80.68%. However, the method of the present invention uses real-time dynamic adjustment and feedback control, with high adjustment accuracy, small error range, and an adjustment success rate of up to 97.54%.

[0195] During the experiment, this embodiment records specific data of some slider pads to demonstrate the effectiveness of the method of the present invention.

[0196] Sample 1:

[0197] Traditional method: The measured thickness is 1.08mm, the adjusted thickness is 1.00mm, it takes 20 minutes, and the error is ±0.05mm;

[0198] The method of the present invention: the measured thickness is 1.01mm, the adjusted thickness is 1.00mm, it takes 4 minutes, and the error is ±0.01mm;

[0199] Sample 2:

[0200] Traditional method: The measured thickness is 0.97mm, the adjusted thickness is 1.00mm, it takes 19 minutes, and the error is ±0.05mm;

[0201] The method of the present invention: the measured thickness is 1.00 mm, the thickness after adjustment is 1.00 mm, it takes 4 minutes, and the error is ±0.01 mm;

[0202] Sample 3:

[0203] Traditional method: The measured thickness is 1.10mm, the adjusted thickness is 1.00mm, it takes 21 minutes, and the error is ±0.05mm;

[0204] The method of the present invention: the measured thickness is 1.00 mm, the adjusted thickness is 1.00 mm, the time consumption is 3.5 minutes, and the error is ±0.01 mm;

[0205] It can be seen from the above data that the method of the present invention is superior to the traditional method in terms of measurement accuracy, adjustment time and adjustment accuracy, and effectively solves the problems of inaccurate measurement, imprecise adjustment and low efficiency existing in the traditional method.

[0206] The method of the present invention not only improves production efficiency and reduces manual operation time, but also significantly improves product quality and consistency. Through automated thickness adjustment and real-time feedback control, human errors are reduced, ensuring that the thickness of the slider pad is always within the allowable error range, thereby improving the operating efficiency and product quality of the entire production line.

[0207] In summary, the present invention significantly improves the accuracy and efficiency of slider pad thickness adjustment through high-precision three-dimensional morphology data acquisition, improved multi-scale iterative reconstruction algorithm, real-time dynamic adjustment and feedback control, and improved image processing algorithm, meets the needs of modern industrial production for efficient and precise adjustment, and realizes full automation and intelligence of the thickness adjustment process. These innovations not only improve the performance and reliability of the system, but also provide strong technical support for the intelligent transformation of industrial automation production lines.

[0208] The present invention utilizes a line laser 3D camera to perform multi-angle scanning on the slider gasket, and combines it with an adaptive multi-angle fusion algorithm to eliminate dead angle interference and spike interference in traditional methods, and generates high-precision three-dimensional morphology data. Through multi-angle scanning and data fusion, the integrity and accuracy of the morphology data are ensured, the accuracy of the data is significantly improved, and a reliable basis is provided for subsequent thickness measurement and adjustment.

[0209] In the process of generating thickness dimension data, the present invention introduces an improved multi-scale iterative reconstruction algorithm, and realizes high-precision thickness calculation through multi-level hierarchical models and residual correction. Through multi-scale processing and residual correction of morphological data, the thickness dimension data of the gasket can be calculated more accurately, ensuring the accuracy and consistency of thickness adjustment, which can adapt to different morphological features and thickness changes, and improve the applicability and reliability of the system.

[0210] The control system of the present invention can perform real-time dynamic adjustment and feedback control according to the thickness dimension data to ensure the real-time and accuracy of the gasket thickness. By calculating the thickness deviation and generating the adjustment instruction, a fully automated thickness adjustment process, real-time dynamic adjustment and feedback control mechanism are realized, which significantly improves the production efficiency and consistency of product quality and reduces manual intervention and errors.

[0211] In order to ensure the correct position of the gasket on the slider seat, the present invention adopts an improved image processing algorithm to accurately detect and confirm the gasket position. Through edge detection and position matching, the position deviation is calculated and a position adjustment instruction is generated to ensure that the gasket position matches the standard position. This not only improves the accuracy of gasket position detection, but also reduces the error and uncertainty in position detection, ensuring the accuracy and consistency of thickness adjustment.

[0212] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for adjusting the thickness of a slider pad based on a line laser 3D camera. It is characterized in that The steps include: S1. Use a line laser 3D camera to scan the front, middle and rear sliders to obtain the topographic data of the three sliders; S2. Preprocess the acquired topographic data to eliminate dead angle interference and spike interference; S3, by analyzing the processed morphological data, using an improved multi-scale iterative reconstruction algorithm to generate the required gasket thickness dimension data; S4, transmitting thickness measurement data to a control system; S5. The control system determines whether the thickness and quantity of the gaskets need to be adjusted according to the thickness dimension data; S6, the control system controls the vacuum suction cup mechanism on the robot arm to suck the gasket of appropriate thickness; S7, placing the gasket on the slider seat, and confirming whether the gasket position is correct by improving the image processing algorithm; S8, the control system dynamically adjusts the gasket thickness according to the real-time feedback data; S9, repeat steps S1-S8 until the scan data of the three sliders meet the set requirements.

2. According to the method for adjusting the thickness of a slider gasket based on a line laser 3D camera according to claim 1, It is characterized in that The S1 specifically includes the following steps: S11, start the line laser 3D camera to scan the front, middle and rear sliders respectively, and the scanning path of each slider is from left to right and from top to bottom, forming a grid-like scanning pattern; S12, obtaining preliminary morphological data of each slider, the preliminary morphological data including height h 前 (x, y), h 中 (x, y), h 后 (x, y) and surface profile C 前 (x, y), C 中 (x, y), C 后 (x, y), where x and y represent the horizontal and vertical coordinates of the slider surface, respectively; S13, repeatedly scan the preliminary morphological data of each slider multiple times to eliminate the error that may occur in a single scan and form multiple repeated morphological data sets Where i represents the i-th scan, and n is the number of repeated scans; S14. For multiple repeated morphological data sets Perform data fusion and use weighted average algorithm to generate the final morphological data H 前 (x, y), H 中 (x, y), H 后 (x, y).

3. The method for adjusting the thickness of a slider pad based on a line laser 3D camera according to claim 2, It is characterized in that The S2 specifically includes the following steps: S21, the preliminary morphological data H of each slider is obtained 前 (x, y), H 中 (x, y), H 后 (x, y) performs boundary detection to identify and mark the boundary area in the slider topography data; S22, using the adaptive multi-angle fusion algorithm to process the identified boundary area, dividing the preliminary morphological data of each slider into several small areas, recorded as Where m is the number of segmented regions; For each small area R j Perform multi-angle scanning to obtain topographic data at multiple different angles Where θ represents different scanning angles, and p is the number of scanning angles; For each small region R j , apply the adaptive weight function W j (θ) Perform weighted fusion on the morphological data at different angles to generate morphological data H′ without blind spot interference j (x, y): Among them, W j (θ) is the adaptive weight, which is dynamically adjusted according to the scanning quality at each angle; All small areas R j No blind spot interference morphological data H′ j (x, y) splicing to generate the overall morphological data H′ after preliminary correction 前 (x, y), H′ 中 (x, y), H′ 后 (x, y); S23, the topographic data H′ after preliminary correction 前 (x, y), H′ 中 (x, y), H′ 后 (x, y) performs spike interference detection to identify and mark abnormal high or low value points; S24, apply the spike interference filtering algorithm to interpolate the marked abnormal points, use the bilinear interpolation method to perform weighted average of the values ​​of the surrounding normal points, and generate the final preprocessed morphological data Among them, W(u,v) is the interpolation weight, which indicates the contribution of each neighborhood point to the interpolation result in the interpolation process. The selection of weight is determined according to the distance of the surrounding points and the smoothness of the data. The weight of the closer point is larger and the weight of the farther point is smaller. k is the interpolation window size. H'(x+u,y+v) is the topographic data after preliminary correction, indicating the height value at the position (x+u,y+v). The result of the interpolation process H f (x, y) is the final preprocessed height value at position (x, y); S25, the final preprocessed morphological data Stored in the control system.

4. The method for adjusting the thickness of a slider pad based on a line laser 3D camera according to claim 3, It is characterized in that The S3 specifically includes the following steps: S31 preprocessed morphological data Perform multi-angle scanning to generate topographic data sets at multiple angles Where θ represents different scanning angles and p is the number of scanning angles: Among them, H θ (x, y) represents the topographic data at the scanning angle θ, x and y are the horizontal and vertical coordinates of the slider surface, respectively, I(x′, y′, z′) is the reflection intensity of the scanned surface at the spatial position (x′, y′, z′), which represents the reflection of the slider surface to the laser, G θ (xx′, yy′, zz′) is the scanning response function, which describes the response characteristics of laser scanning at different angles. The integral areas [a, b] and [c, d] are the boundaries of the scanning area, indicating the range covered by the scanning; S32 normalizes the topographic data at each angle to eliminate the influence of angle differences. The normalized topographic data is H′ θ (x, y); S33 uses an adaptive weight fusion algorithm to perform weighted fusion on the topography data from multiple angles to generate comprehensive topography data H c (x, y): Among them, H c (x, y) is the fused comprehensive morphological data, W(θ) is the adaptive weight, which is dynamically adjusted according to the quality of each angle data, μ θ and σ θ are the mean and standard deviation of the topographic data at angle θ, respectively, which are used to measure the central tendency and dispersion of the data; S34 for comprehensive morphological data H c (x, y) is processed at multiple scales to generate a multi-scale hierarchical model Among them, L k (x, y) is the hierarchical model at the kth level, n k is the window size, indicating the scale range of hierarchical processing, W k (i, j) is the weight function of the kth level, which is used to perform weighted processing on the comprehensive topographic data at different positions; At each scale level, S35 uses an adaptive residual correction algorithm based on a neural network to calculate the morphological residual R of the current level. k (x, y): Among them, R k (x, y) is the shape residual of the kth level, w i is the weight, which is dynamically adjusted according to the output of the neural network; S36 for the residual error R k (x, y) is subjected to adaptive smoothing based on wavelet transform, and the residual data after filtering is R k，f (x, y): R k，f (x,y)=W -1 {H w (u,v)·W{R k (x,y)}}; Among them, W and W -1 denote wavelet transform and inverse wavelet transform respectively, H w (u, v) is an adaptive wavelet filter function, which is used to smooth the residual data and eliminate noise; S37 smooths the residual data R k，f (x, y) is added to the hierarchical model L of the current level k (x, y), update the hierarchical model: L k+1 (x,y)=L k (x,y)+R k，f (x,y); Among them, L k+1 (x, y) is the updated hierarchical model; S38 repeats steps S35 to S37 until all hierarchical models at all levels have completed iterative reconstruction, generating the final multi-scale iterative reconstruction morphological data L q (x, y); S39 is used to reconstruct the final multi-scale iterative morphological data L q (x, y) for thickness dimension analysis, using thickness calculation formula T(x, y): T(x,y)=L q (x,y)-B(x,y); Among them, T(x, y) is the thickness dimension data of the required gasket generated, and B(x, y) is the reference surface data of the bottom of the slider, which represents the reference surface height of the slider at different positions.

5. The method for adjusting the thickness of a slider pad based on a line laser 3D camera according to claim 4, It is characterized in that The S5 specifically includes the following steps: S51, the calculated gasket thickness dimension data T 前 (x, y), T 中 (x, y), T 后 (x, y) is transmitted to the control system; S52, the control system performs statistical analysis on the gasket thickness dimension data at each position and calculates the average thickness at each position Where N is the number of measurement points, T(x i ,y i ) is the position (x i ,y i ) at the thickness dimension data; S53, the control system compares the calculated average thickness with the preset standard thickness T 标准 Compare and calculate thickness deviation ΔT 前 , ΔT 中 , ΔT 后 : Among them, ΔT is the thickness deviation, which indicates the difference between the current thickness and the standard thickness; S54. Based on the thickness deviation ΔT, the control system determines whether the thickness and quantity of the gaskets need to be adjusted, as follows: If |ΔT|≤∈, the current thickness is considered to be within the allowable error range and no adjustment is required; If |ΔT|>∈, the thickness or quantity of the gasket needs to be adjusted, where ∈ is the allowable thickness error range; S55, the control system determines the adjustment scheme according to the positive and negative thickness deviation: If ΔT>0, it means the current thickness is too large and the thickness or quantity of the gasket needs to be reduced; If ΔT<0, it means the current thickness is insufficient and the thickness or quantity of the gasket needs to be increased; S56, the control system generates an adjustment instruction, including the thickness or quantity of the gasket to be adjusted, and the adjustment instruction includes a specific adjustment parameter ΔT 调整 : Where, ΔT 调整 is the thickness difference that needs to be adjusted.

6. The method for adjusting the thickness of a slider pad based on a line laser 3D camera according to claim 5, It is characterized in that The S7 specifically comprises the following steps: S71, moving the gasket sucked by S6 to above the slider seat, and adjusting the posture of the robot arm so that the gasket is aligned with the predetermined position of the slider seat; S72, using an improved image processing algorithm to perform positioning detection on the gasket, and obtaining the initial position data P of the gasket on the slider seat 初 (x, y), where x and y are the horizontal and vertical coordinates of the pad on the slider seat; S73, initial position data P of gasket 初 (x, y) performs edge detection to identify the edge contour E(x, y) of the gasket: in, Respectively represent the gradients of the pad position data in the x and y directions; S74, compare the detected gasket edge profile E(x, y) with the standard edge profile E(x, y) of the slider seat. 标准 (x, y) is matched and the position deviation ΔP(x, y) is calculated: ΔP(x,y)=E 标准 (x, y)-E(x, y); Among them, ΔP(x, y) is the deviation of the shim position from the standard position, which represents the difference between the actual position of the shim on the slider seat and the expected position; S75. Based on the position deviation ΔP (x, y), the control system determines whether the position of the gasket is correct, as follows: If |ΔP(x,y)|≤∈ 1 , then the gasket position is considered correct; If |ΔP(x, y)|>∈ 1 , then the position of the gasket needs to be adjusted, where ∈ 1 is the allowable deviation range; S76, the control system generates a position adjustment instruction according to the position deviation ΔP (x, y), and the adjustment instruction includes a specific adjustment parameter ΔP 调整 (x, y): ΔP 调整 (x, y)=ΔP(x, y); Where ΔP 调整 (x, y) is the position deviation value that needs to be adjusted; S77, sending the position adjustment instruction to the actuator, the actuator performs the actual position adjustment operation according to the instruction, ensuring that the gasket position matches the standard position, and finally confirming whether the gasket position is correct.

Images