Glue spraying control method, process and system for question book rotary press, product and medium

Abstract

The invention provides a subject rotary press glue spraying control method, process and system, a product and a medium, and relates to the technical field of glue spraying image equipment. A basic data set is established by acquiring continuous multi-frame image data and extracting edge contour points and glue line feature points in the image data, a movement track is obtained by calculating displacement vectors of the edge contour points and the glue line feature points between adjacent image data, and spectral analysis is performed on an edge track and a glue line track to realize separation of vibration components. Then, a detection area is constructed according to the processing edge contour points and the corresponding processing glue line feature points by utilizing the geometric constraint relation kept in the deformation process of the small book edge contour; the detection area is adjusted according to the change relation between the edge track and the processed edge contour point, and real-time self-adaption of the detection area to product deformation is achieved. And by combining with an optimization strategy of physical characteristics, the stability of a detection reference and the accuracy of glue line detection are ensured.

Description

Technical Field

[0001] This application relates to the technical field of spray adhesive image equipment, and particularly relates to a spray adhesive control method, process, system, product and medium for a title rotary machine. Background Art

[0002] With the continuous improvement of the quality requirements for teaching materials and supplementary teaching products in the education field, during the production process of a title rotary machine for folding single-sided A4 paper to form a multi-page booklet product, the quality control of the back adhesive process becomes particularly important. Due to production efficiency requirements, the equipment usually needs to operate at a high speed of 300 - 600 pages per minute.

[0003] Currently, the back adhesive quality control of a title rotary machine mainly adopts a vision detection technology based on a preset reference point. This technology installs an industrial camera on the production line to collect the back image of the product, and performs feature extraction based on a pre-calibrated image reference point. The system judges whether the position and shape of the glue line meet the requirements according to the corresponding relationship between the extracted features and the preset reference point, so as to evaluate the quality state of the glue line.

[0004] However, during the spray adhesive process of a booklet using a title rotary machine, the arc surface of the booklet spine will undergo dynamic deformation due to equipment vibration, and at the same time, the displacement between pages will occur under the action of vibration in the stacked folding structure (because the size of the booklet product is small, this misalignment phenomenon is more prominent). Due to the superposition of these two unstable factors, the preset reference point of the system cannot accurately reflect the actual detection benchmark, resulting in a deviation between the feature extraction result and the real state, and affecting the reliability of the detection result. Summary of the Invention

[0005] This application provides a spray adhesive control method, process, system, product and medium for a title rotary machine to improve the reliability of detection results.

[0006] In a first aspect, the present application provides a method for controlling glue spraying of a question wheel rotary machine, including: acquiring continuous multi-frame image data, where the image data includes the glue line spraying area on the folded back of the product; extracting edge contour points and glue line feature points from the image data; forming an edge contour point sequence and a glue line feature point sequence; based on the edge contour point sequence, calculating the displacement vectors of the edge contour points between adjacent image data to generate an edge trajectory; based on the glue line feature point sequence, calculating the displacement vectors of the glue line feature points between adjacent image data to generate a glue line trajectory; respectively performing spectral analysis on the edge trajectory and the glue line trajectory, and separating vibration components based on a preset frequency threshold to obtain processed edge contour points and processed glue line feature points; constructing a detection area according to the processed edge contour points and the corresponding processed glue line feature points; adjusting the detection area according to the change relationship between the edge trajectory and the processed edge contour points to obtain a dynamically adjusted detection area over time; performing integrity detection on the glue line feature points inside the dynamically adjusted detection area, and outputting an evaluation result of the glue line quality status according to the integrity detection result.

[0007] By adopting the above technical solution, a basic data set is established by acquiring continuous multi-frame image data and extracting edge contour points and glue line feature points from the image data. Since the deformation of the glue line is continuous, while the equipment vibration and page displacement show jumpy changes. Based on this characteristic, the motion trajectories are obtained by calculating the displacement vectors of the edge contour points and the glue line feature points between adjacent image data, and the vibration components are separated by performing spectral analysis on the edge trajectory and the glue line trajectory. Then, by using the geometric constraint relationship maintained by the booklet edge contour during deformation, a detection area is constructed according to the processed edge contour points and the corresponding processed glue line feature points. By adjusting the detection area according to the change relationship between the edge trajectory and the processed edge contour points, the real-time adaptability of the detection area to the product deformation is realized. By combining the optimization strategy of physical characteristics, the stability of the detection benchmark and the accuracy of glue line detection are ensured.

[0008] Combined with some embodiments of the first aspect, in some embodiments, after the step of adjusting the detection area according to the change relationship between the edge trajectory and the processed edge contour points to obtain a dynamically adjusted detection area over time, the method further includes: determining a fitting plane according to the depth information of the edge contour points in the same image data; determining the transformation matrix between the fitting plane and the reference plane where the dynamically adjusted detection area is located; changing the dynamically adjusted detection area from the reference plane to the fitting plane according to the transformation matrix; projecting the dynamically adjusted detection area in the fitting plane to the reference plane to obtain an optimized detection area; the step of performing integrity detection on the glue line feature points inside the dynamically adjusted detection area and outputting an evaluation result of the glue line quality status according to the integrity detection result specifically includes: performing integrity detection on the glue line feature points inside the optimized detection area and outputting an evaluation result of the glue line quality status according to the integrity detection result.

[0009] By adopting the above technical solution, a fitting plane is determined based on the depth information of the edge contour points, and the transformation relationship with the reference plane is established, realizing the accurate conversion of the detection area between different planes. The dynamic detection area on the reference plane is converted to a fitting plane that better fits the actual surface of the product, and then projected back onto the reference plane to obtain an optimized detection area, enabling the detection area to more accurately fit the actual shape of the product. This region optimization method based on three-dimensional space effectively solves the detection deviation problem caused by the deformation of the product surface.

[0010] Combined with some embodiments of the first aspect, in some embodiments, the step of determining the fitting plane according to the depth information of the edge contour points in the same image data specifically includes: calculating the distances in depth between all the edge contour points and the plane respectively according to the depth information of the edge contour points; adjusting the plane to minimize the sum of the distances to obtain the fitting plane.

[0011] By adopting the above technical solution, when determining the fitting plane, by calculating the distances from all the edge contour points to the plane and minimizing them, the optimal fitting of the actual surface shape of the product is realized. It is ensured that the fitting plane can fit the actual surface state of the product to the greatest extent, providing a reliable spatial reference for the subsequent optimization of the detection area.

[0012] Combined with some embodiments of the first aspect, in some embodiments, before the step of determining the fitting plane according to the depth information of the edge contour points in the same image data, the method further includes: tracking the displacement changes of the edge contour points between adjacent frame image data; calculating the scale factor of the feature points based on the preset actual size of the edge contour; constructing a fundamental matrix according to the displacement changes, decomposing it in combination with the camera parameters to obtain a relative motion matrix, and applying the scale factor to the relative motion matrix to obtain a motion matrix; solving through triangulation to obtain the depth information of the edge contour points.

[0013] By adopting the above technical solution, basic data is obtained by tracking the displacement changes of the edge contour points, and the scale factor is calculated in combination with the preset actual size, ensuring the proportional accuracy of the depth calculation. By constructing a fundamental matrix and decomposing it in combination with the camera parameters to obtain a relative motion matrix, and then applying the scale factor to obtain a motion matrix with the true scale, and finally obtaining accurate depth information through triangulation, providing reliable three-dimensional space data for the construction of the fitting plane.

[0014] In some embodiments in combination with some embodiments of the first aspect, the steps of respectively performing spectral analysis on the edge trajectory and the glue line trajectory, separating vibration components based on a preset frequency threshold, and obtaining processed edge profile points and processed glue line feature points specifically include: respectively performing sliding window segmentation processing on the edge trajectory and the glue line trajectory to obtain a plurality of edge trajectory data segments and glue line trajectory data segments; respectively performing Fourier transform on the edge trajectory data segments and the glue line trajectory data segments to obtain corresponding frequency amplitude spectra; determining the edge trajectory data segments or glue line trajectory data segments with frequency amplitude spectra higher than the preset frequency threshold as vibration components; removing the vibration components, and merging the edge trajectory data segments and the glue line trajectory data segments after removing the vibration components to obtain processed edge profile points and processed glue line feature points.

[0015] By adopting the above technical solution, the sliding window segmentation processing method not only ensures the continuity of the data but also improves the accuracy of spectral analysis. The trajectory data is converted to the frequency domain through Fourier transform, the vibration components are accurately identified and separated according to the preset threshold, and then the vibration interference is effectively suppressed by merging the processed data segments, enabling the subsequent detection process to be based on stable feature points.

[0016] In some embodiments in combination with some embodiments of the first aspect, the steps of outputting an evaluation result of the glue line quality state according to the integrity detection result specifically include: classifying the products according to the integrity detection result; performing a glue filling operation on the products with grades within the glue line missing threshold range.

[0017] By adopting the above technical solution, the products are classified based on the integrity detection result, and a glue filling operation is performed on the products with glue line missing within the threshold range, realizing the closed-loop management of detection and control. This method of hierarchical processing and timely remedy not only ensures the product quality but also avoids unnecessary scrapping and improves the production efficiency.

[0018] In a second aspect, the present application provides a control process for spray gluing of a question wheel rotary machine, including: mixing colored glue; adding the colored glue into a glue gun and initializing the glue gun, a vision inspection device, and a conveying device; clamping and transporting the folded product to the spray gluing range of the glue gun through the conveying device; starting the glue gun to spray colored glue at the crease of the product; starting the vision inspection device to detect the integrity of the glue line; causing the vision inspection device to execute the method as in the first aspect; if the integrity inspection result shows that the glue line is complete, pressing the crease of the product for bonding; if the integrity inspection result shows that the glue line is incomplete, the vision inspection device determines the product as a defective product and sends the information of the defective product to the host; the host controls the picking mechanism to remove the defective product from the conveying device; evaluating and classifying the quality status of the glue line of the defective product; performing a glue replenishment action on the product within the missing glue line threshold range; starting the vision inspection device to detect the integrity of the glue line after glue replenishment for the defective product; if the integrity inspection result shows that the glue line is complete, moving it to the finished product conveying line by the picking mechanism again; if the integrity inspection result shows that the glue line is incomplete, performing a scrapping process.

[0019] In a third aspect, the present application provides a control system for spray gluing of a question wheel rotary machine. The control system for spray gluing of a question wheel rotary machine includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code. The computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the control system for spray gluing of a question wheel rotary machine to execute the method as in the first aspect.

[0020] In a fourth aspect, the present application provides a computer program product containing instructions. When the computer program product runs on the control system for spray gluing of a question wheel rotary machine, it causes the control system for spray gluing of a question wheel rotary machine to execute the method as in the first aspect.

[0021] In a fifth aspect, the present application provides a computer-readable storage medium, including instructions. When the instructions run on the control system for spray gluing of a question wheel rotary machine, it causes the control system for spray gluing of a question wheel rotary machine to execute the method as in the first aspect.

[0022] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. A basic data set is established by acquiring consecutive multi-frame image data and extracting edge contour points and glue line feature points in the image data. Since the deformation of the glue line is continuous, while the equipment vibration and page displacement show jumpy changes. Based on this characteristic, the motion trajectories are obtained by calculating the displacement vectors of the edge contour points and the glue line feature points between adjacent image data, and the vibration components are separated by performing spectral analysis on the edge trajectories and the glue line trajectories. Then, by using the geometric constraint relationship maintained by the booklet edge contour during the deformation process, a detection region is constructed according to the processed edge contour points and the corresponding processed glue line feature points. By adjusting the detection region based on the variation relationship between the edge trajectory and the processed edge contour points, real-time self-adaptation of the detection region to the product deformation is achieved. Through an optimization strategy combining physical characteristics, the stability of the detection benchmark and the accuracy of glue line detection are ensured.

[0023] 2. The fitting plane is determined through the depth information of the edge contour points, and the transformation relationship with the reference plane is established, realizing the precise conversion of the detection region between different planes. The dynamic detection region in the reference plane is converted to a fitting plane that fits better with the actual surface of the product, and then projected back to the reference plane to obtain an optimized detection region, enabling the detection region to fit more accurately with the actual shape of the product. This region optimization method based on three-dimensional space effectively solves the detection deviation problem caused by the deformation of the product surface.

[0024] 3. When determining the fitting plane, the optimal fitting of the actual surface morphology of the product is achieved by calculating the distances from all edge contour points to the plane and minimizing them. It is ensured that the fitting plane can fit the actual surface state of the product to the greatest extent, providing a reliable spatial reference for the subsequent optimization of the detection region. Description of the Drawings

[0025] Figure 1 is a flowchart of a glue spraying control method for a rotary press in an embodiment of the present application; Figure 2 is another flowchart of a glue spraying control method for a rotary press in an embodiment of the present application; Figure 3 is Figure 2 another flowchart before step S201 of; Figure 4 is an exemplary hardware structure diagram of a glue spraying control system for a rotary press in an embodiment of the present application. Detailed Embodiments

[0026] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of the present application, the singular forms "a", "an", "the", "above-mentioned", "said", and "this" are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the terms and / or used in the present application refer to and include any and all possible combinations of one or more of the listed items.

[0027] Hereinafter, the terms "first" and "second" are only for descriptive purposes and cannot be construed as implying or suggesting relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0028] In the process of manufacturing a booklet product on A4 paper, it is necessary to spray glue at the folding points to ensure product quality. Due to the small size of the product and the narrow crease area, a fine glue spraying nozzle is required to achieve precise spraying. However, such a fine nozzle is prone to clogging problems, resulting in interrupted or uneven glue spraying. Especially during continuous production, the nozzle may gradually clog due to glue solidification or impurity accumulation, causing quality defects such as broken glue lines and insufficient glue volume, seriously affecting the binding firmness of the product.

[0029] Based on the above problems, the present application provides a spray glue process for a rotary press, including: Mixing colored glue; Adding a specific proportion of pigment to the conventional white glue to make the glue present a color that is easily recognizable by the vision system, improving the accuracy of visual inspection.

[0030] Adding the colored glue to the glue gun and initializing the glue gun, vision inspection device, and conveying device; Transporting the folded product to the glue spraying range of the glue gun by the conveying device through clamping; ensuring that each device is in a normal working state.

[0031] Stably transporting the folded product to the glue spraying station of the glue gun by the clamping device of the conveying device. The design of the clamping device ensures that the product maintains the correct posture during transportation, preventing deviation or deformation.

[0032] Starting the glue gun to spray colored glue at the crease of the product; The spraying parameters (such as pressure, speed, etc.) of the glue gun are preset according to the product characteristics.

[0033] Starting the vision inspection device to detect the integrity of the glue line; enabling the vision inspection device to execute a spray glue control method for a rotary press; If the integrity detection result shows that the glue line is intact, the crease of the laminated product is bonded; If the integrity detection result shows that the glue line is incomplete, the vision inspection device determines the product as a defective product and sends the information of the defective product to the host; The host controls the picking mechanism to remove the defective product from the conveying device; Classify the evaluation result of the glue line quality status of the defective product; Perform a glue filling operation on the product whose glue line is within the missing glue line threshold range; After glue filling, start the vision inspection device to detect the integrity of the glue line of the defective product; If the integrity detection result shows that the glue line is intact, it is moved again by the picking mechanism into the finished product conveying line; If the integrity detection result shows that the glue line is incomplete, it is discarded.

[0034] It can be seen that by adding a specific pigment to the glue for dyeing, the sprayed glue line has clear visual characteristics. During the glue spraying process, the vision detection system monitors the status of the glue line in real time. Once it detects that the width of the glue line is abnormal or the spraying area deviates from the preset range, the system immediately gives an alarm prompt. It can timely detect the nozzle blockage problem and clean it, effectively preventing the generation of batch defective products and improving production efficiency.

[0035] S101. Obtain continuous multi-frame image data, where the image data includes the glue line spraying area on the folded back of the product; Among them, the continuous multi-frame image data refers to a series of image sequences collected by an industrial camera at continuous time points, and each frame of image represents the state information of the product at a certain moment. The folded back of the product refers to the spine part formed after the booklet product is folded, which is the target area for glue spraying.

[0036] S102. Extract the edge contour points and glue line feature points from the image data; form an edge contour point sequence and a glue line feature point sequence; Among them, the edge contour points refer to the characteristic pixel points on the edge of the folded back of the product, which are used to represent the contour shape and spatial position of the product in the image. The glue line feature points represent the key points on the glue spraying line, which are used to describe the shape and distribution characteristics of the glue line.

[0037] Specifically, preprocess each frame of image, including image enhancement, noise suppression, and contrast adjustment, to improve the accuracy of feature extraction. Then, for the product edge area, use an edge detection algorithm to extract contour feature points, and at the same time use the color feature of the dyed glue to extract the key points on the glue line. For each feature point, the system records its two-dimensional coordinate information and related attributes (such as gray value, gradient direction, etc.). Finally, organize the same type of feature points into a sequence according to spatial position or time order, and establish the association relationship between the feature points.

[0038] S103. Calculate the displacement vectors of the edge contour points between adjacent image data based on the edge contour point sequence to generate an edge trajectory. Specifically, the system establishes a matching relationship for the corresponding edge contour points in two adjacent frames of images to ensure that the same physical point is being tracked. Then, the spatial position differences between each pair of matching points are calculated to obtain the displacement vectors in the two-dimensional plane. These displacement vectors not only contain information about the displacement magnitude and direction but also reflect the deformation characteristics of the product during movement. By connecting the displacement vectors that are continuous in time, a trajectory curve describing the edge movement pattern is formed.

[0039] S104. Calculate the displacement vectors of the glue line feature points between adjacent image data based on the glue line feature point sequence to generate a glue line trajectory. Specifically, establish the correspondence of the glue line feature points between adjacent frames to ensure that the feature points at the same physical position are being tracked. Calculate the position offset between the corresponding feature points to obtain the displacement vectors reflecting the local deformation of the glue line. These displacement vectors not only contain the movement information of the glue line but also reflect the deformation characteristics during the glue spraying process. By performing temporal correlation and spatial connection on the displacement vectors, a continuous trajectory describing the dynamic behavior of the glue line is constructed.

[0040] S105. Perform spectral analysis on the edge trajectory and the glue line trajectory respectively, and separate the vibration components based on a preset frequency threshold to obtain processed edge contour points and processed glue line feature points. This step is executed after obtaining the motion trajectories and is used to eliminate the interference caused by environmental vibration and equipment jitter. Specifically, the system first performs Fourier transform on the edge trajectory and the glue line trajectory to convert the time-domain motion signals into frequency-domain representations. Then, analyze the spectral characteristics to identify the low-frequency components belonging to normal motion and the high-frequency components caused by vibration. Based on a pre-set frequency threshold, the system filters out the vibration components above the threshold and retains the low-frequency signals reflecting the actual motion characteristics. Finally, the processed signals are transformed back to the time domain through inverse transformation to obtain the positions of the feature points with the vibration effects eliminated.

[0041] In some specific embodiments, after continuously sampling the X and Y coordinate data of the edge trajectory and the glue line trajectory, the edge trajectory and the glue line trajectory can be segmented. A preset length can be used, or an index of the overlapping rate of adjacent segments can be preset. Perform FFT calculations on each data segment, set the preset frequency threshold as the cut-off frequency, retain the basic motion trajectories in the low-frequency band, and filter out the vibration components in the high-frequency band. During reconstruction, set the components above the preset frequency threshold in the frequency-domain data to 0, retain the complete amplitudes within the preset frequency threshold, and then perform IFFT calculations. Smoothly transition the overlapping regions of the reconstructed data segments.

[0042] In some other specific embodiments, step S105 specifically includes: S1051. Perform sliding window segmentation processing on the edge trajectory and the glue line trajectory respectively to obtain a number of edge trajectory data segments and glue line trajectory data segments; Among them, the sliding window refers to a fixed-length data segment that moves in the time series and is used to divide the continuous signal into local intervals. The segmentation processing means dividing the continuous trajectory data into multiple overlapping data segments.

[0043] Specifically, the system first determines the size of the sliding window, which requires considering the balance between the periodic characteristics of the signal and the computational efficiency. Then, set the overlap rate of the window, and usually choose an overlap degree of about 50% to ensure the continuity of the data. For each sliding position, the system extracts the trajectory data within the current window and applies a window function to reduce spectral leakage.

[0044] In some specific embodiments, first set a fixed window length to accurately define the scope of data processing on this basis. Then, calculate the window overlap parameter to ensure that the connection between different windows and the data coverage meet the expected requirements, so as to achieve a more reasonable data processing layout. Subsequently, use the window function for weighting operation. In this way, the importance of specific parts in the data is highlighted, and it also helps to improve the overall characteristics of the data, making it more conducive to subsequent analysis and processing.

[0045] In addition, an adaptive window method can also be selected to further optimize the entire process, enabling it to dynamically adjust the window-related parameters according to the characteristics of the actual data, better adapt to different data situations, and enhance the flexibility and adaptability of data processing.

[0046] S1052. Perform Fourier transform on the edge trajectory data segment and the glue line trajectory data segment respectively to obtain the corresponding frequency amplitude spectrum; Among them, the Fourier transform refers to a mathematical method for converting a time-domain signal into a frequency-domain representation and is used to analyze the frequency composition of the signal.

[0047] Specifically, the system first preprocesses each data segment, including mean removal and trend elimination, to reduce the DC component and low-frequency interference in spectral analysis. Then, apply the fast Fourier transform (FFT) to the processed data segment to convert the time-domain data into a complex-form spectral representation.

[0048] S1053. Determine the edge trajectory data segment or the glue line trajectory data segment with a frequency amplitude spectrum higher than the preset frequency threshold as the vibration component; Specifically, analyze the spectral characteristics of each data segment and detect the frequency components exceeding the threshold. The system needs to consider the effects of spectral leakage and noise. For the data segments identified as vibration components, the system also needs to record their characteristic parameters to provide a basis for subsequent signal reconstruction.

[0049] S1054. Remove the vibration components, and merge the edge trajectory data segment and the glue line trajectory data segment after removing the vibration components to obtain the processed edge contour points and the processed glue line characteristic points.

[0050] Specifically, the system designs a frequency-domain filter to suppress or remove the identified vibration frequency components. Perform an inverse transform on the filtered spectrum to restore the time-domain signal. Considering the boundary effects caused by segmented processing, the system needs to perform smooth transition processing in the overlapping regions of the data segments. Finally, recombine the processed data segments in chronological order to construct a complete motion trajectory. During the merging process, it is necessary to ensure the continuity and smoothness of the trajectory.

[0051] It can be seen that the sliding window segmented processing method not only ensures the continuity of the data but also improves the accuracy of spectral analysis. The trajectory data is transformed into the frequency domain through Fourier transform, the vibration components are accurately identified and separated according to the preset threshold, and then by merging the processed data segments, the effective suppression of vibration interference is achieved, enabling the subsequent detection process to be based on stable characteristic points.

[0052] S106. Construct a detection area based on the processed edge contour points and the corresponding processed glue line characteristic points; Specifically, the system first establishes a spatial mapping relationship between the processed edge contour points and the glue line characteristic points to determine their relative positions. Then, based on the distribution characteristics of these characteristic points, an envelope area is constructed as the detection range. This detection area needs to completely cover the glue line distribution area.

[0053] In some specific embodiments, for the convex hull boundary of the glue line characteristic points, outline the approximate range contour formed by the glue line characteristic points; construct the minimum circumscribed rectangle based on the determined convex hull boundary to simplify the shape representation of this area, which is more in line with the actual glue segment.

[0054] Calculate the perpendicular distance di from the processed edge contour points to the four sides of the detection area; find the side corresponding to the shortest distance dmin as the reference side L; record the projection point Qi of the contour point on the reference side; represent the position by the distance si from the starting point of the reference side to the projection point; establish the mapping relationship: Mi = (L, si, dmin).

[0055] S107. Adjust the detection area according to the change relationship between the edge trajectory and the processed edge contour points to obtain a dynamically adjusted detection area that changes over time; Specifically, the system compares the position deviation between the edge trajectory and the processed contour points in real time, and calculates the deformation degree and displacement trend between them. Based on these change characteristics, the system dynamically updates the shape and position parameters of the detection area, enabling the detection area to adapt to the movement and deformation of the product.

[0056] Continuing with the above example, according to the change relationship between the edge trajectory and the processed edge contour points, the projection point Qi is modified accordingly according to the mapping relationship, and then the projection points are smoothly transitioned. At this time, a new detection area, namely the dynamic detection area, is obtained.

[0057] For example: the displacement vector of the edge trajectory is Δv = (Δx, Δy), where Δx is the difference between the edge trajectory and the processed edge contour points on the x-axis, and Δy is the difference between the edge trajectory and the processed edge contour points on the y-axis; the projection point Qi is displaced according to the same displacement vector Δv.

[0058] S108. Perform integrity detection on the glue line feature points inside the dynamic detection area, and output the evaluation result of the glue line quality status according to the integrity detection result.

[0059] Specifically, extract the morphological features of the glue line within the dynamic detection area, including parameters such as width, continuity, and edge smoothness. Then, compare these features with the preset quality standards, and calculate the overall quality score through the set scoring mechanism. Finally, output the quality grade or specific defect description according to the score, providing a decision-making basis for production control.

[0060] It can be seen that a basic data set is established by obtaining continuous multi-frame image data and extracting edge contour points and glue line feature points in the image data. Since the deformation of the glue line is continuous, while the equipment vibration and page displacement show jumpy changes. Based on this characteristic, the motion trajectory is obtained by calculating the displacement vectors between the edge contour points and the glue line feature points in adjacent image data, and the vibration components are separated by performing spectral analysis on the edge trajectory and the glue line trajectory. Then, by using the geometric constraint relationship maintained by the booklet edge contour during the deformation process, the detection area is constructed according to the processed edge contour points and the corresponding processed glue line feature points. By adjusting the detection area according to the change relationship between the edge trajectory and the processed edge contour points, the real-time self-adaptation of the detection area to the product deformation is realized. By combining the optimization strategy of physical characteristics, the stability of the detection benchmark and the accuracy of the glue line detection are ensured.

[0061] In some specific embodiments, step S108 specifically includes: S1081. Classify the products into different grades according to the integrity detection results; Specifically, based on the various indicators obtained from the detection, such as glue line continuity, uniformity, width consistency, etc., and then compare them with multiple grade intervals and the corresponding quality standards to determine the quality grade to which the product belongs.

[0062] S1082. Perform the glue filling action on the products whose grades are within the range of the glue line missing threshold.

[0063] Specifically, the system first evaluates the specific situation of the glue line missing, including the missing position, range, and degree. Then, it compares the evaluation result with the preset repairable threshold to determine whether the glue filling condition is met. For the products that meet the conditions, the system needs to generate detailed glue filling instructions, including parameters such as the glue filling position, dosage, and speed. This process needs to consider the state of the original glue line to ensure perfect fusion with the original glue line after glue filling. The system also needs to monitor the glue filling process in real time to ensure the remedial effect.

[0064] It can be seen that grading the products based on the integrity detection results and performing the glue filling operation on the products with glue line missing within the threshold range realizes the closed-loop management of detection and control. This way of grading processing and timely remedial measures not only ensures the product quality but also avoids unnecessary scrapping and improves the production efficiency.

[0065] However, in the actual production process, due to the dynamic deformation of the arc surface of the booklet spine caused by equipment vibration, and at the same time, the displacement between pages occurs under the action of vibration in the stacked folding structure, the side of the product often shows irregular deformation. In this case, if the vertical reference plane is still used for detection, it will affect the detection accuracy.

[0066] Therefore, in some preferred embodiments, after step S107, it further includes: S201. Determine the fitting plane according to the depth information of the edge contour points in the same image data; Among them, the depth information refers to the Z-axis coordinate value of the point in the three-dimensional space, which is used to represent the spatial position information of the object surface. The fitting plane represents the best matching spatial plane calculated by mathematical methods and is used to approximately describe the actual deformed surface.

[0067] Specifically, use the least squares method or other optimization algorithms to calculate the parametric equation of the best fitting plane, including the normal vector and intercept of the plane. This fitting plane needs to be as close as possible to the actual product surface and have sufficient smoothness to eliminate the influence of local noise.

[0068] In some specific embodiments: Step S201 specifically includes: S2011. Calculate the distances between all the edge contour points and the plane in terms of depth according to the depth information of the edge contour points. Specifically, an initial plane equation is established. Usually, the plane where the centroid of the point cloud lies is selected as the starting plane. Then, for each edge contour point, the signed distance from the point to the current plane is calculated. The sign of the distance indicates whether the point is on the positive or negative side of the plane. These distance values will serve as an important basis for subsequent plane optimization.

[0069] S2012. Adjust the plane to minimize the sum of distances to obtain the fitted plane.

[0070] Specifically, the system first constructs a least - squares optimization objective function, taking the sum of the squares of the distances from all points to the plane as the optimization index. Then, through an iterative optimization algorithm, the normal vector and intercept parameters of the plane are continuously adjusted to gradually reduce the value of the objective function.

[0071] It can be seen that when determining the fitted plane, by calculating the distances from all edge contour points to the plane and minimizing them, the optimal fitting of the actual surface shape of the product is achieved. This ensures that the fitted plane can best conform to the actual surface state of the product, providing a reliable spatial reference for subsequent detection area optimization.

[0072] In some embodiments, before step S201, it further includes: S301. Track the displacement changes of edge contour points between adjacent frame image data; Specifically, edge contour feature points are extracted from each frame of the image to ensure the stability and recognizability of the feature points. The corresponding relationship of feature points is established between adjacent frames, which requires considering the local apparent features and spatial distribution constraints of the feature points. At the same time, the motion trajectory of each feature point needs to be recorded, including displacement direction and amplitude information. These displacement information will provide an important basis for subsequent motion estimation and depth calculation.

[0073] S302. Calculate the scale factor of the feature points based on the preset actual size of the edge contour; Among them, the preset actual size of the edge contour refers to the true measurement value of the product edge contour in the physical space, which is used to provide an absolute scale reference. The scale factor represents the proportional coefficient from the image space to the physical space and is used to convert relative motion into actual motion. The actual size represents the true length unit in the physical world. Preset is used to represent a pre - known or measured standard value.

[0074] This step is executed after feature point tracking to solve the scale uncertainty problem in motion recovery. Specifically, the system first obtains the standard physical size information of the product edge contour. Then, it measures the pixel distances between corresponding feature points in the image space to establish the correspondence between the image space and the physical space. The system needs to consider the imaging characteristics of the camera and the influence of perspective effects, and select appropriate feature point pairs to calculate the scale factor. At the same time, it also needs to handle measurement errors and uncertainties, and improve the accuracy of scale estimation through statistical analysis of multiple sets of measurement data.

[0075] S303. Construct a fundamental matrix based on the displacement change, decompose it in combination with camera parameters to obtain a relative motion matrix, and apply the scale factor to the relative motion matrix to obtain a motion matrix. Among them, the fundamental matrix refers to a 3×3 matrix that describes the epipolar geometry relationship between two images and is used to represent the algebraic constraints of camera motion. Camera parameters represent the internal imaging characteristics of the camera, including focal length, principal point, distortion coefficient, etc. The relative motion matrix refers to the rotation and translation matrix that describes the change in camera pose. The motion matrix is used to represent the camera motion transformation with actual scale. Construct means the process of establishing a matrix model through mathematical methods. Decompose refers to the process of decomposing the fundamental matrix into camera motion parameters. Apply is used to represent the operation of integrating scale information into motion estimation.

[0076] Specifically, the system first constructs a fundamental matrix using the matched feature point pairs, which requires at least 8 corresponding point pairs, and usually uses more point pairs to improve stability. Then, in combination with the known internal camera parameters, the fundamental matrix is decomposed into a rotation matrix and a translation vector. This process may produce multiple solutions, and the correct solution needs to be selected through geometric constraints. Finally, the scale factor calculated previously is applied to the translation vector to obtain a motion matrix that reflects the actual motion scale.

[0077] In some specific embodiments, construct the fundamental matrix E: construct the fundamental matrix using displacement change information; decompose it in combination with the internal camera parameter K; obtain the relative motion matrix, including the rotation matrix R and the unit translation vector t.

[0078] Calculate the true scale: measure the projected size L_image of the edge contour in the image; obtain the actual size L_real of the edge contour; calculate the scale factor: s = L_real / L_image; Obtain the motion matrix of the true scale: apply the scale factor to the translation vector: T = s * t; construct the complete camera motion matrix [R|T].

[0079] S304. Solve through triangulation to obtain the depth information of the edge contour points.

[0080] Specifically, the system first establishes the projection equations of the feature points in the two images, which contain the internal and external parameters of the camera and the image coordinates of the feature points. Then, based on the known camera motion and internal parameters, a least-squares optimization problem is constructed to solve the three-dimensional coordinates of the feature points.

[0081] It can be seen that by tracking the displacement changes of the edge contour points to obtain the basic data and combining with the preset actual size to calculate the scale factor, the proportional accuracy of the depth calculation is ensured. By constructing the fundamental matrix and decomposing it in combination with the camera parameters to obtain the relative motion matrix, and then applying the scale factor to obtain the motion matrix of the true scale, and finally obtaining the accurate depth information through triangulation, which provides reliable three-dimensional space data for the construction of the fitting plane.

[0082] S202. Determine the transformation matrix between the fitting plane and the reference plane where the dynamic detection area is located; Specifically, the system first determines the parametric equation of the reference plane, and usually selects a plane parallel to the image plane as the reference. Then, analyze the geometric relationship between the fitting plane and the reference plane, including the included angle of the normal vectors of the two planes and the relative position offset. Based on these geometric relationships, a transformation matrix containing rotation and translation components is constructed.

[0083] S203. Change the dynamic detection area from the reference plane to the fitting plane according to the transformation matrix; It should be noted that the "change" here is not the same as the "projection" below. Specifically, the "change" here actually refers to the transformation of the plane, that is, it involves the change operation of the plane in space or a certain set scenario. And the related dynamic detection area will be closely associated with the transformed plane and change accordingly with the transformed plane. There is such a co-varying relationship between the two.

[0084] Specifically, the dynamic detection area originally on the reference plane is accurately transferred to the fitting plane according to the set transformation matrix, thus completing the change process from one plane to another to meet the specific requirements for the plane position and other aspects in subsequent related processing or analysis.

[0085] S204. Project the dynamic detection area in the fitting plane to the reference plane to obtain the optimized detection area; It should be noted that projection is a specific geometric mapping operation. Specifically, it is an operation in an existing space or set scenario where a figure, region (such as the dynamic detection region) in a plane (such as the fitting plane mentioned here) is projected onto another plane (such as the reference plane) according to corresponding projection rules (such as different projection methods like orthographic projection and oblique projection, each with its specific projection direction and angle requirements), presenting a corresponding image. That is, the shape, position, and other information contained in the dynamic detection region on the fitting plane are transferred to the reference plane through projection, thus generating a new region that meets specific requirements (such as the optimized detection region here).

[0086] Different from the changes mentioned earlier, the changes focus on the transformation of the entire plane itself and the synchronous changes of related regions, which is an overall change operation in aspects such as the position and state of the plane; while projection focuses on mapping the elements in one plane to another plane according to specific rules, emphasizing more the process of transferring and presenting graphic information from one plane to another. The two have essentially different operation connotations and purposes.

[0087] More precisely, the rules applied are different. One is a transformation matrix, and the other is an orthographic projection method.

[0088] It should be understood that the reason for carrying out such a series of operations is that initially, the dynamic detection region is above the reference plane. However, as mentioned earlier, the reference plane is not the most ideal and suitable plane for subsequent operations. In view of this, the practice of copying the reference plane is adopted, and then the azimuth angle of the copied plane is changed. The purpose is to make it better fit the requirements for subsequent data processing, detection, etc. through changing its azimuth angle. And the closely related dynamic detection region, due to its cooperative relationship with the copied plane, will also change accordingly with the change of the azimuth angle of the copied plane. But in actual observation, the work needs to be carried out from the specific angle of the reference plane. Therefore, in order to be able to maintain the original observation angle and make good use of the advantages brought by the plane after the azimuth angle change, it is necessary to project the dynamically changed detection region during the change of the copied plane back onto the reference plane through this projection operation, so that it is possible to view and process relevant content from the familiar angle of the reference plane and, at the same time, make better use of the optimization effect brought by the previous change of the plane azimuth angle to better complete the entire process.

[0089] Replace S108 with S205, perform integrity detection on the glue line feature points inside the optimized detection region, and output the evaluation result of the glue line quality status according to the integrity detection result.

[0090] It can be seen that by determining the fitting plane based on the depth information of the edge contour points and establishing the transformation relationship with the reference plane, the precise conversion of the detection area between different planes is achieved. The dynamic detection area on the reference plane is converted to a fitting plane that better conforms to the actual surface of the product, and then projected back onto the reference plane to obtain an optimized detection area, enabling the detection area to more accurately conform to the actual shape of the product. This region optimization method based on three-dimensional space effectively solves the problem of detection deviation caused by surface deformation of the product.

[0091] The following introduces the exemplary problem rotary machine glue spraying control system 400 provided by the embodiments of the present application. Figure 4 It is an exemplary hardware structure diagram of the problem rotary machine glue spraying control system 400 provided by the embodiments of the present application.

[0092] In some embodiments, the problem rotary machine glue spraying control system 400 is a computer device or the problem rotary machine glue spraying control system 400 includes a computer device. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with other external terminals or servers through a network connection. In some embodiments, the network interface can be a wired network interface, and in some embodiments, the network interface can also be a wireless network interface. The computer program, when executed by the processor, implements the method in the embodiments of the present application.

[0093] Those skilled in the art can understand that Figure 4 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0094] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0095] As used in the foregoing embodiments, depending on the context, the term "when..." may be construed to mean "if...", or "after...", or "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if detecting (the stated condition or event)" may be construed to mean "if determining...", or "in response to determining...", or "when detecting (the stated condition or event)", or "in response to detecting (the stated condition or event)".

[0096] In the foregoing embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented by software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more integrated available media. The available medium may be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid-state drive), etc.

[0097] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the foregoing embodiments can be completed by computer programs instructing relevant hardware. The programs can be stored in a computer-readable storage medium. When the programs are executed, they can include the processes of the foregoing method embodiments. The foregoing storage media include various media that can store program codes, such as ROM or random access memory RAM, magnetic disks, or optical discs.

Claims

1. A control method for the glue spraying of a question paper rotary machine, characterized in that, Including: Obtain consecutive multi-frame image data, where the image data includes the glue line spraying area on the folded back of the product; Extract edge contour points and glue line feature points from the image data; Form an edge contour point sequence and a glue line feature point sequence; Based on the edge contour point sequence, calculate the displacement vectors of the edge contour points between adjacent image data to generate an edge trajectory; Based on the glue line feature point sequence, calculate the displacement vectors of the glue line feature points between adjacent image data to generate a glue line trajectory; Perform spectral analysis on the edge trajectory and the glue line trajectory respectively, and separate the vibration components based on a preset frequency threshold to obtain processed edge contour points and processed glue line feature points; Construct a detection area according to the processed edge contour points and the corresponding processed glue line feature points; Adjust the detection area according to the change relationship between the edge trajectory and the processed edge contour points to obtain a dynamically adjusted detection area over time; Perform integrity detection on the glue line feature points inside the dynamically adjusted detection area, and output a glue line quality status evaluation result according to the integrity detection result.

2. The method according to claim 1, wherein After the step of adjusting the detection area according to the change relationship between the edge trajectory and the processed edge contour points to obtain a dynamically adjusted detection area over time, the method further includes: Determine a fitting plane according to the depth information of the edge contour points in the same image data; Determine the transformation matrix between the fitting plane and the reference plane where the dynamically adjusted detection area is located; Change the dynamically adjusted detection area from the reference plane to the fitting plane according to the transformation matrix; Project the dynamically adjusted detection area in the fitting plane to the reference plane to obtain an optimized detection area; The step of performing integrity detection on the glue line feature points inside the dynamically adjusted detection area and outputting a glue line quality status evaluation result according to the integrity detection result specifically includes: Perform integrity detection on the glue line feature points inside the optimized detection area, and output a glue line quality status evaluation result according to the integrity detection result.

3. The method according to claim 2, wherein The step of determining a fitting plane according to the depth information of the edge contour points in the same image data specifically includes: Calculate the distances in depth between all the edge contour points and the plane respectively according to the depth information of the edge contour points; Adjust the plane to minimize the sum of the distances to obtain the fitting plane.

4. The method according to claim 2, wherein Before the step of determining a fitting plane according to the depth information of the edge contour points in the same image data, the method further includes: Track the displacement changes of the edge contour points between adjacent frames of the image data; Calculate the scale factor of the feature points based on a preset actual size of the edge contour; Construct a fundamental matrix according to the displacement changes, decompose it in combination with the camera parameters to obtain a relative motion matrix, and apply the scale factor to the relative motion matrix to obtain a motion matrix; Solve through triangulation to obtain the depth information of the edge contour points.

5. The method according to claim 1, characterized in that, The step of performing spectral analysis on the edge trajectory and the glue line trajectory respectively, and separating the vibration components based on a preset frequency threshold to obtain processed edge contour points and processed glue line feature points specifically includes: Perform sliding window segmentation processing on the edge trajectory and the glue line trajectory respectively to obtain a number of edge trajectory data segments and glue line trajectory data segments; Perform Fourier transform on the edge trajectory data segments and the glue line trajectory data segments respectively to obtain the corresponding frequency amplitude spectra; Determine the edge trajectory data segments or glue line trajectory data segments with frequency amplitude spectra higher than the preset frequency threshold as vibration components; Remove the vibration components, and merge the edge trajectory data segments and the glue line trajectory data segments after removing the vibration components to obtain the processed edge contour points and the processed glue line feature points.

6. The method according to claim 1, wherein The step of outputting the evaluation result of the glue line quality state according to the integrity detection result specifically includes: Classify the products according to the integrity detection result; Perform a glue replenishment operation on the products within the glue line missing threshold range.

7. A process for spraying glue on a question wheel rotary machine, characterized in that, Include: Mix the colored glue; Add the colored glue to the glue gun, and initialize the glue gun, the vision detection device, and the conveying device; Clamp and transport the folded product through the conveying device to the glue spraying range of the glue gun; Start the glue gun to spray colored glue at the crease of the product; Start the vision detection device to detect the integrity of the glue line; so that the vision detection device executes the method described in any one of claims 1-6; If the integrity detection result is that the glue line is complete, press the crease of the product for bonding; If the integrity detection result is that the glue line is incomplete, the vision detection device determines the product as a defective product and sends the information of the defective product to the host; The host controls the picking mechanism to remove the defective product from the conveying device; Classify the evaluation result of the glue line quality state of the defective product; Perform a glue replenishment operation on the products classified within the glue line missing threshold range; Start the vision detection device to detect the integrity of the glue line for the defective product after glue replenishment; If the integrity detection result is that the glue line is complete, move it to the finished product conveying line again by the picking mechanism; If the integrity detection result is that the glue line is incomplete, perform a scrapping process.

8. A control system for spray gluing of a question paper rotary machine, characterized in that, The rotary press glue spraying control system of the present invention includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the rotary press glue spraying control system to execute the method described in any one of claims 1-6.

9. A computer program product containing instructions, characterized in that, When the computer program product runs on the rotary press glue spraying control system, enable the rotary press glue spraying control system to execute the method described in any one of claims 1-6.

10. A computer-readable storage medium, comprising instructions, characterized in that, When the instruction runs on the rotary press glue spraying control system, enable the rotary press glue spraying control system to execute the method described in any one of claims 1-6.

Images