Adhesive tape dynamic thickness detection method and system based on machine vision

Abstract

The embodiment of the invention discloses an adhesive tape dynamic thickness detection method and system based on machine vision. The method comprises the steps that an adhesive tape image shot by a camera module is acquired; performing image processing on the adhesive tape image to obtain a processed adhesive tape image; and analyzing the processed adhesive tape image through a thickness detection model to obtain a thickness detection result. According to the embodiment of the invention, the thickness of the adhesive layer can be effectively controlled through real-time monitoring and automatic deviation correction, so that the defects of the adhesive tape are reduced.

Description

Technical Field

[0001] The present application relates to the field of visual inspection technology, and in particular to a method and system for detecting dynamic thickness of an adhesive tape based on machine vision. Background Art

[0002] For a long time, quality inspection in the industrial field has been the most challenging task of computer vision, and industrial quality inspection is an indispensable part of the industrial inspection field. At present, in the actual application scenarios in the field of industrial quality inspection, most of them use traditional machine vision technology based on machine learning. This method is too sensitive to noise data, especially when the picture is not clear, the recognition effect is poor. In addition, this method requires manual extraction of calibration features and training and testing, which has a long development cycle and poor applicability. However, in industrial applications, adhesive tapes often need to be inspected for defects such as thickness, bubbles, and orange peel. However, due to the complex surface features of adhesive tapes, the appearance of the same defect varies greatly, and the accuracy of traditional machine vision algorithms is low. For example, traditional thickness detection methods, such as contact measurement or offline sampling, often have shortcomings such as slow response, low accuracy, and inability to provide real-time feedback, which not only affects product quality, but also limits the improvement of production efficiency. Summary of the invention

[0003] The embodiment of the present application discloses a method and system for dynamic thickness detection of an adhesive tape based on machine vision, which can effectively control the thickness of the adhesive layer through real-time monitoring and automatic deviation correction to reduce defects of the adhesive tape.

[0004] The embodiment of the present application discloses a method for detecting dynamic thickness of an adhesive tape based on machine vision, which is applied to an electronic device, wherein the electronic device is communicatively connected to a camera module, and the method comprises: Acquiring an image of the adhesive tape captured by the camera module; performing image processing on the adhesive tape image to obtain a processed adhesive tape image; The processed adhesive tape image is analyzed by a thickness detection model to obtain a thickness detection result.

[0005] As an optional implementation manner, performing image processing on the adhesive tape image to obtain a processed adhesive tape image includes: Performing noise reduction processing on the adhesive tape image to obtain a noise-reduced adhesive tape image; Performing edge point detection on the denoised adhesive tape image to extract candidate edge points in the denoised adhesive tape image; Candidate edge points that meet preset conditions are selected from a plurality of candidate edge points to form an edge point set, and the processed adhesive tape image is generated according to the edge point set.

[0006] As an optional implementation manner, the processed adhesive tape image is analyzed by a thickness detection model to obtain a thickness detection result, including: Obtaining an image feature matrix according to the processed adhesive tape image by using the thickness detection model; Expanding the image feature matrix to obtain a plurality of image feature vectors; Cascading the plurality of image feature vectors with the historical image feature vector to obtain a target feature vector; the historical image feature vector is an image feature vector generated based on the adhesive tape image collected in the previous time period; The target feature vector is classified to obtain the thickness detection result.

[0007] As an optional implementation manner, the image feature matrix is ​​obtained according to the processed adhesive tape image by the thickness detection model, including: Performing convolution processing on the adhesive tape image to generate a convolution feature map; Performing pooling processing on the convolutional feature map to generate a pooled feature map; Performing nonlinear activation on the pooled feature map to generate an activation feature map; Calculating the mean of each position of the activation feature map along the channel dimension to generate a spatial feature matrix; Calculate the Softmax function value of each position in the spatial feature matrix to obtain a spatial score matrix; The position-wise multiplication of the spatial feature matrix and the spatial score matrix is ​​calculated to obtain an image feature matrix.

[0008] As an optional implementation manner, the image feature matrix is ​​expanded to obtain a plurality of image feature vectors, including: The image feature matrix is ​​expanded along the row vector to obtain the multiple image feature vectors.

[0009] As an optional implementation manner, the cascading process of the multiple image feature vectors and the historical image feature vectors to obtain the target feature vector includes: Inputting the plurality of image feature vectors and the historical image feature vectors into the transformer-based context encoder to obtain a plurality of context feature vectors; Performing feature aggregation optimization on the multiple context feature vectors to obtain multiple optimized context feature vectors; The multiple optimized context feature vectors are cascaded to obtain the target feature vector.

[0010] As an optional implementation, the electronic device is communicatively connected to the servo motor; after the processed adhesive tape image is analyzed by the thickness detection model to obtain the thickness detection result, the method further includes: If the thickness detection result indicates that the thickness is abnormal, the servo motor is controlled to adjust the position of the coating head and the glue flow rate to correct the thickness of the glue layer.

[0011] The embodiment of the present application discloses a dynamic thickness detection system for adhesive tape based on machine vision, which is applied to an electronic device, wherein the electronic device is communicatively connected with a camera module, and the system comprises: An image acquisition module, used to acquire the adhesive tape image captured by the camera module; An image processing module, used for performing image processing on the adhesive tape image to obtain a processed adhesive tape image; The thickness detection module is used to analyze the processed adhesive tape image through a thickness detection model to obtain a thickness detection result.

[0012] An embodiment of the present application discloses an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor implements any one of the machine vision-based dynamic thickness detection methods for adhesive tape disclosed in the embodiment of the present application.

[0013] The embodiment of the present application discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, any one of the machine vision-based dynamic thickness detection methods for adhesive tape disclosed in the embodiment of the present application is implemented.

[0014] Compared with the related art, the embodiments of the present application have the following beneficial effects: The embodiment of the present application provides a method and system for dynamic thickness detection of adhesive tape based on machine vision, which obtains the adhesive tape image captured by the camera module; performs image processing on the adhesive tape image to obtain the processed adhesive tape image; and analyzes the processed adhesive tape image through a thickness detection model to obtain a thickness detection result. By implementing the embodiment of the present application, image processing is performed on the adhesive tape image captured by the camera module to obtain the processed adhesive tape image, which can improve the accuracy of the thickness detection model in analyzing the processed adhesive tape image, thereby providing a technical basis for analyzing the processed adhesive tape image through the thickness detection model. The processed adhesive tape image is analyzed through the thickness detection model to obtain the thickness detection result, avoiding the problems of slow response and low detection accuracy caused by manual detection, and can effectively control the thickness of the adhesive layer through real-time monitoring and automatic deviation correction to reduce defects in the adhesive tape. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 It is a flow chart of a method for dynamic thickness detection of an adhesive tape based on machine vision disclosed in an embodiment of the present application; Figure 2 It is a schematic diagram of a process of analyzing a processed adhesive tape image by a thickness detection model to obtain a thickness detection result according to an embodiment of the present application; Figure 3 It is a structural schematic diagram of a machine vision-based dynamic thickness detection system for adhesive tape disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] It should be noted that the terms "including" and "having" and any variations thereof in the embodiments of the present application and the accompanying drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0019] The embodiments of the present application disclose a seat angle adjustment method, device, electronic device and storage medium, which can prevent users from feeling uncomfortable due to long-term sitting, thereby improving the comfort of the seat. Detailed descriptions are given below.

[0020] Figure 1 : is a flow chart of a method for dynamic thickness detection of adhesive tape based on machine vision disclosed in an embodiment of the present application. Figure 1 The described method for dynamic thickness detection of adhesive tape based on machine vision is applicable to electronic devices, which may include but are not limited to mobile phones, tablet computers, wearable devices, laptop computers, PCs (Personal Computers), etc., and the present application embodiment does not limit this. Figure 1As shown, the method for dynamic thickness detection of adhesive tape based on machine vision may include the following steps: Step S101, acquiring an adhesive tape image captured by a camera module.

[0021] In some embodiments, the electronic device may be connected to a camera module, which may be a linear array camera or a laser displacement sensor, so as to perform non-contact scanning of the continuously moving adhesive tape, so that the electronic device obtains an adhesive tape image. The adhesive tape image obtained by the electronic device may be taken at the current moment, and when the camera module takes the adhesive tape image, the shooting time may be recorded to obtain the adhesive tape image and the shooting time corresponding to the adhesive tape image.

[0022] Furthermore, the electronic device may control the camera module to take an image of the adhesive tape at intervals of a time period. For example, the electronic device may control the camera module to take an image of the adhesive tape at intervals of 0.5 seconds, so that the electronic device obtains an image of the adhesive tape with a shooting time interval of 0.5 seconds. The electronic device controls the camera module to take an image of the adhesive tape at intervals of a time period, thereby avoiding the amount of calculation caused by processing a large number of images and avoiding the influence of data redundancy on the accuracy of the thickness detection result.

[0023] Step S102, performing image processing on the adhesive tape image to obtain a processed adhesive tape image.

[0024] In some embodiments, the electronic device may use an integrated high-resolution image processing algorithm to perform image processing on the acquired adhesive tape image to obtain a processed adhesive tape image. For example, the electronic device may perform edge detection processing or surface texture analysis processing on the adhesive tape image to accurately identify the boundary and thickness of the adhesive layer.

[0025] Furthermore, the electronic device may perform edge detection processing on the adhesive tape image by performing noise reduction processing on the adhesive tape image to obtain a noise-reduced adhesive tape image; performing edge point detection on the noise-reduced adhesive tape image to extract candidate edge points in the noise-reduced adhesive tape image; selecting candidate edge points that meet preset conditions from multiple candidate edge points to form an edge point set, and generating the processed adhesive tape image according to the edge point set. Among them, the electronic device may perform noise reduction processing on the adhesive tape image by performing mean filtering or median filtering on the adhesive tape image. The general principle of edge detection processing is to detect based on the grayscale change of the edge position. The electronic device may use one of Roberts, Prewitt, Sobel edge detection, Kirsch edge detection, log edge detection, and canny edge detection to perform edge detection processing on the adhesive tape image, which is not limited here.

[0026] The electronic device performs noise reduction processing on the adhesive tape image to obtain the noise-reduced adhesive tape image, performs edge point detection on the noise-reduced adhesive tape image to extract candidate edge points in the noise-reduced adhesive tape image, selects candidate edge points that meet preset conditions from multiple candidate edge points to form an edge point set, and generates the processed adhesive tape image according to the edge point set. The edge detection processing can be performed on the adhesive tape image to accurately identify the boundary and thickness of the adhesive layer in the adhesive tape image, thereby improving the accuracy of the thickness detection result obtained by analyzing the processed adhesive tape image through a thickness detection model.

[0027] Step S103, analyzing the processed adhesive tape image by using a thickness detection model to obtain a thickness detection result.

[0028] In recent years, deep learning and neural networks have been widely used in computer vision, natural language processing, text signal processing and other fields. In addition, deep learning and neural networks have also demonstrated a level close to or even beyond that of humans in image classification, object detection, semantic segmentation, text translation and other fields.

[0029] In some embodiments, the electronic device can use a convolutional neural network to establish a thickness detection model, and train the thickness detection model to be trained based on the sample adhesive tape image set, thereby obtaining a trained thickness detection model. The sample adhesive tape image set may include multiple sample adhesive tape images and target thickness detection results corresponding to each sample adhesive tape image.

[0030] The electronic device can input a set of sample adhesive tape images into a thickness detection model to be trained, analyze each sample adhesive tape image through the thickness detection model to be trained, obtain a predicted thickness detection result of each sample adhesive tape image, and adjust the model parameters of the thickness detection model to be trained according to the error between the predicted thickness detection result of each sample adhesive tape image and the target thickness detection result, until the error between the predicted thickness detection result of each sample adhesive tape image and the target thickness detection result is less than a preset error threshold, thereby obtaining a trained thickness detection model.

[0031] As an optional implementation, after the electronic device analyzes the processed adhesive tape image through the thickness detection model and obtains the thickness detection result, if the thickness detection result indicates that the thickness is abnormal, the servo motor is controlled to adjust the position of the coating head and the glue flow rate to correct the thickness of the adhesive layer. The servo motor may include a correction structure, and the correction structure may include a fine-tuning screw, a hydraulic cylinder or a magnetic suspension system to achieve high-precision dynamic control. If the thickness detection result indicates that the thickness is normal, the new adhesive tape image captured by the camera module is continuously acquired to analyze the new adhesive tape image through the thickness detection model to obtain a new thickness detection result, thereby achieving real-time detection of the thickness of the adhesive tape.

[0032] As an optional implementation, the electronic device can provide an intuitive graphical interface that displays the adhesive layer thickness distribution map, defect marks and production status for easy operator monitoring and adjustment. When serious defects are detected, it automatically triggers shutdown or marks unqualified sections to ensure the production quality of the adhesive tape.

[0033] In an embodiment of the present application, the electronic device obtains an adhesive tape image captured by a camera module; performs image processing on the adhesive tape image to obtain a processed adhesive tape image; and analyzes the processed adhesive tape image through a thickness detection model to obtain a thickness detection result. The electronic device performs image processing on the adhesive tape image captured by the camera module to obtain a processed adhesive tape image, which can improve the accuracy of the thickness detection model's analysis of the processed adhesive tape image, thereby providing a technical basis for analyzing the processed adhesive tape image through the thickness detection model. The processed adhesive tape image is analyzed through the thickness detection model to obtain a thickness detection result, thereby avoiding problems such as slow response and low detection accuracy caused by manual detection, and can effectively control the thickness of the adhesive layer through real-time monitoring and automatic deviation correction to reduce defects in the adhesive tape.

[0034] Figure 2 FIG. 1 is a flow chart of analyzing the processed adhesive tape image by a thickness detection model to obtain a thickness detection result according to an embodiment of the present application. Figure 2 As shown, the step analyzes the processed adhesive tape image through the thickness detection model to obtain the thickness detection result, and also includes the following steps: Step S201, obtaining an image feature matrix based on the processed adhesive tape image through a thickness detection model.

[0035] In some embodiments, considering that when performing adhesive tape thickness detection, attention should be paid to feature information related to thickness detection in spatial position while ignoring interference features unrelated to thickness detection, and given that the attention mechanism can select the focus position, a more distinguishable feature representation is generated, and the features after adding the attention module will change adaptively as the network deepens. Therefore, the electronic device can process the adhesive tape image through the first convolutional neural network model using spatial attention to extract the feature distribution information about thickness detection focused on the adhesive tape image in space, thereby obtaining an image feature matrix. Among them, the image feature matrix extracted by spatial attention reflects the weight of the feature difference in spatial dimension, which is used to suppress or strengthen the features of different spatial positions, so as to extract the feature information related to thickness detection focused on space.

[0036] Furthermore, the electronic device can perform convolution processing on the adhesive tape image to generate a convolution feature map; perform pooling processing on the convolution feature map to generate a pooling feature map; perform nonlinear activation on the pooling feature map to generate an activation feature map; calculate the mean of each position of the activation feature map along the channel dimension to generate a spatial feature matrix; calculate the Softmax-like function value of each position in the spatial feature matrix to obtain a spatial score matrix; calculate the positional dot product of the spatial feature matrix and the spatial score matrix to obtain an image feature matrix.

[0037] After the electronic device performs a series of processing on the adhesive tape image, an image feature matrix is ​​obtained, which can focus more on the feature information related to thickness detection, thus providing a technical basis for the accuracy of thickness detection of the adhesive tape.

[0038] Step S202: Expand the image feature matrix to obtain a plurality of image feature vectors.

[0039] Step S203, cascade processing is performed on multiple image feature vectors and historical image feature vectors to obtain a target feature vector; the historical image feature vector is an image feature vector generated based on the adhesive tape image collected in the previous time period.

[0040] In some embodiments, the electronic device may expand the image feature matrix along the row vector to obtain a plurality of image feature vectors, so as to more accurately obtain the image features of the adhesive tape image.

[0041] In some embodiments, a converter is used to capture the characteristics of long-distance contextual dependence, and multiple image feature vectors and historical image feature vectors are globally context-semantically encoded to obtain contextual semantic association feature representations with the overall semantic association of multiple image feature vectors and historical image feature vectors as the context background, that is, multiple contextual feature vectors. Among them, the historical image feature vector is an image feature vector generated based on the adhesive tape image collected in the previous time period. For example, if the adhesive tape image is collected at 15:20:10, and the image feature vector is generated based on the adhesive tape image, and the camera module collects the adhesive tape image once every 1 second, then the historical adhesive tape image can be collected at 15:20:09, and the historical image feature can be generated based on the historical adhesive tape image. The electronic device can obtain one or more historical adhesive tape images, and can generate multiple historical image feature vectors based on multiple historical adhesive tape images, so as to cascade the multiple image feature vectors and the multiple historical image feature vectors to obtain the target feature vector.

[0042] As an optional implementation, the electronic device inputs multiple image feature vectors and historical image feature vectors into a converter-based context encoder to obtain multiple context feature vectors; optimizes the feature aggregation degree of the multiple context feature vectors to obtain multiple optimized context feature vectors; and cascades the multiple optimized context feature vectors to obtain a target feature vector.

[0043] If each image feature vector and the historical image feature vector are regarded as the contextual semantic feature representation of a single node, and the nodes have a topological association relationship of the context encoding of the converter, the class probability feature aggregation between the nodes can be improved based on the overall topological structure of the nodes. Specifically, the class probability matching feature vector between each context feature vector is calculated, which can be shown as follows: in, are the context feature vectors, is the distance between each context feature vector, represents the distance threshold, Indicates that the distance between each context feature vector is less than the preset distance threshold. is a weighted hyperparameter.

[0044] Step S204, classifying the target feature vector to obtain a thickness detection result.

[0045] In some embodiments, the electronic device can classify the target feature vector through the Softmax classification function to obtain the thickness detection result. In an embodiment of the present application, the electronic device obtains an image feature matrix based on the processed adhesive tape image through a thickness detection model, expands the image feature matrix to obtain multiple image feature vectors, cascades the multiple image feature vectors with historical image feature vectors to obtain target feature vectors, classifies the target feature vectors, and obtains a thickness detection result. Through real-time monitoring and automatic correction, the thickness of the adhesive layer can be effectively controlled to reduce defects in the adhesive tape.

[0046] See also Figure 3 , Figure 3 1 is a schematic diagram of a structure of a dynamic thickness detection system for adhesive tape based on machine vision disclosed in an embodiment of the present application. The system can be applied to the above-mentioned electronic equipment. Figure 3 As shown, a machine vision-based dynamic thickness detection system 300 for adhesive tape may include: an image acquisition module 301 , an image processing module 302 , and a thickness detection module 303 .

[0047] An image acquisition module 301 is used to acquire the adhesive tape image captured by the camera module; An image processing module 302 is used to perform image processing on the adhesive tape image to obtain a processed adhesive tape image; The thickness detection module 303 is used to analyze the processed adhesive tape image through a thickness detection model to obtain a thickness detection result.

[0048] In one embodiment, the image processing module 302 is also used to perform noise reduction processing on the adhesive tape image to obtain a noise-reduced adhesive tape image; perform edge point detection on the noise-reduced adhesive tape image to extract candidate edge points in the noise-reduced adhesive tape image; select candidate edge points that meet preset conditions from multiple candidate edge points to form an edge point set, and generate a processed adhesive tape image based on the edge point set.

[0049] In one embodiment, the thickness detection module 303 further includes a matrix calculation unit, a matrix expansion unit, a vector cascade unit, and a vector classification unit: A matrix calculation unit, used for obtaining an image feature matrix according to the processed adhesive tape image through a thickness detection model; A matrix expansion unit, used for expanding the image feature matrix to obtain a plurality of image feature vectors; A vector cascade unit, used for cascading a plurality of image feature vectors with a historical image feature vector to obtain a target feature vector; the historical image feature vector is an image feature vector generated based on an adhesive tape image collected in a previous time period; The vector classification unit is used to classify the target feature vector to obtain the thickness detection result.

[0050] In one embodiment, the matrix calculation unit is also used to perform convolution processing on the adhesive tape image to generate a convolution feature map; perform pooling processing on the convolution feature map to generate a pooling feature map; perform nonlinear activation on the pooling feature map to generate an activation feature map; calculate the mean of each position of the activation feature map along the channel dimension to generate a spatial feature matrix; calculate the Softmax-like function value of each position in the spatial feature matrix to obtain a spatial score matrix; calculate the positional dot product of the spatial feature matrix and the spatial score matrix to obtain an image feature matrix.

[0051] In one embodiment, the matrix expansion unit is further used to expand the image feature matrix along the row vector to obtain multiple image feature vectors.

[0052] In one embodiment, the vector cascade unit is further used to input multiple image feature vectors and historical image feature vectors into a converter-based context encoder to obtain multiple context feature vectors; optimize the feature aggregation degree of the multiple context feature vectors to obtain multiple optimized context feature vectors; and cascade the multiple optimized context feature vectors to obtain a target feature vector.

[0053] In one embodiment, a machine vision-based dynamic thickness detection system 300 for adhesive tape further includes a control module: The control module is used to control the servo motor to adjust the position of the coating head and the flow rate of glue to correct the thickness of the glue layer if the thickness detection result indicates that the thickness is abnormal.

[0054] An embodiment of the present application discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by the processor, the processor implements any one of the machine vision-based dynamic thickness detection methods for adhesive tapes disclosed in the embodiment of the present application.

[0055] The embodiment of the present application discloses a computer program product, including a computer program, and the computer program can be executed by a processor to implement a machine vision-based dynamic thickness detection method for an adhesive tape as described in the above embodiments.

[0056] It should be understood that the "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. Those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required for the present application.

[0057] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the above-mentioned processes does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0058] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0059] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0060] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a memory and includes several requests for a computer device (which can be a personal computer, server or network device, etc., specifically a processor in a computer device) to perform some or all of the steps of the above-mentioned methods of various embodiments of the present application.

[0061] A person skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0062] The above is a detailed introduction to a method and system for dynamic thickness detection of adhesive tape based on machine vision disclosed in the embodiment of the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A method for dynamic thickness detection of adhesive tape based on machine vision, characterized in that: Applied to an electronic device, the electronic device is communicatively connected with a camera module, and the method comprises: Acquiring an image of the adhesive tape captured by the camera module; performing image processing on the adhesive tape image to obtain a processed adhesive tape image; The processed adhesive tape image is analyzed by a thickness detection model to obtain a thickness detection result.

2. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 1, characterized in that: The step of performing image processing on the adhesive tape image to obtain a processed adhesive tape image includes: Performing noise reduction processing on the adhesive tape image to obtain a noise-reduced adhesive tape image; Performing edge point detection on the denoised adhesive tape image to extract candidate edge points in the denoised adhesive tape image; Candidate edge points that meet preset conditions are selected from a plurality of candidate edge points to form an edge point set, and the processed adhesive tape image is generated according to the edge point set.

3. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 1, characterized in that: The processing of the adhesive tape image by using the thickness detection model to obtain the thickness detection result includes: Obtaining an image feature matrix according to the processed adhesive tape image by using the thickness detection model; Expanding the image feature matrix to obtain a plurality of image feature vectors; Cascading the plurality of image feature vectors with the historical image feature vector to obtain a target feature vector; the historical image feature vector is an image feature vector generated based on the adhesive tape image collected in the previous time period; The target feature vector is classified to obtain the thickness detection result.

4. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 3, characterized in that: The method of obtaining an image feature matrix according to the processed adhesive tape image by the thickness detection model includes: Performing convolution processing on the adhesive tape image to generate a convolution feature map; Performing pooling processing on the convolutional feature map to generate a pooled feature map; Performing nonlinear activation on the pooled feature map to generate an activation feature map; Calculating the mean of each position of the activation feature map along the channel dimension to generate a spatial feature matrix; Calculate the Softmax function value of each position in the spatial feature matrix to obtain a spatial score matrix; The position-wise multiplication of the spatial feature matrix and the spatial score matrix is ​​calculated to obtain an image feature matrix.

5. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 3, characterized in that: The image feature matrix is ​​expanded to obtain a plurality of image feature vectors, including: The image feature matrix is ​​expanded along the row vector to obtain the multiple image feature vectors.

6. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 3, characterized in that: The step of performing cascade processing on the plurality of image feature vectors and the historical image feature vectors to obtain a target feature vector includes: Inputting the plurality of image feature vectors and the historical image feature vectors into the transformer-based context encoder to obtain a plurality of context feature vectors; Performing feature aggregation optimization on the multiple context feature vectors to obtain multiple optimized context feature vectors; The multiple optimized context feature vectors are cascaded to obtain the target feature vector.

7. The method for dynamic thickness detection of adhesive tape based on machine vision according to claim 1, characterized in that: The electronic device is communicatively connected with the servo motor; After analyzing the processed adhesive tape image by the thickness detection model to obtain the thickness detection result, the method further includes: If the thickness detection result indicates that the thickness is abnormal, the servo motor is controlled to adjust the position of the coating head and the glue flow rate to correct the thickness of the glue layer.

8. A dynamic thickness detection system for adhesive tape based on machine vision, characterized in that: Applied to electronic equipment, the electronic equipment is communicatively connected with a camera module, and the system comprises: An image acquisition module, used to acquire the adhesive tape image captured by the camera module; An image processing module, used for performing image processing on the adhesive tape image to obtain a processed adhesive tape image; The thickness detection module is used to analyze the processed adhesive tape image through a thickness detection model to obtain a thickness detection result.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor implements a method for dynamic thickness detection of an adhesive tape based on machine vision as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for dynamic thickness detection of an adhesive tape based on machine vision as described in any one of claims 1 to 7 is implemented.

Images