Intelligent punching system and method for rolled film

Abstract

The invention discloses an intelligent punching system and method for a rolled film. The intelligent punching system comprises a rolling assembly, a data acquisition module, an image recognition assembly, a control module, a punching assembly and a rolling assembly. The data acquisition module and the image recognition assembly are connected with the control module through a data bus, and the control module is connected with the rolling assembly, the punching assembly and the rolling assembly through control channels. The material rolling assembly is used for transmitting a to-be-punched rolled film to a machining area, the data acquisition module is used for detecting position information of the film and synchronously transmitting the position information to the control module, the image recognition assembly is used for recognizing the film hole site, and the punching assembly is used for receiving a control signal of the control module and executing punching operation on a positioned film material. And the winding assembly is used for winding the punched film. The whole structure is simple, the problem that a traditional machine is insufficient in positioning precision is solved, and the punching precision is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of punching equipment, and in particular to a roll film intelligent punching system and method. Background Art

[0002] The roll film punching device is a device specially used for accurately punching roll film materials. It is widely used in packaging, agriculture, and industrial manufacturing. Such devices are mainly composed of a roll material unit, a data acquisition device, and a punching unit. The punching component punches holes in the film according to preset parameters to ensure that the hole spacing and hole diameter meet production requirements. Although the traditional roll film punching device can punch the film according to preset parameters, it has many defects and shortcomings. First, the accuracy of mechanical positioning and sensor detection is limited, and errors are prone to occur in complex hole positions and high-precision applications. Second, traditional equipment lacks dynamic adjustment capabilities. Once the punching parameters are set, the equipment cannot automatically correct according to real-time data, resulting in punching offset and hole spacing errors. In the punching process of multi-specification films, it takes a long time to prepare when changing specifications. At the same time, traditional equipment has poor adaptability and cannot intelligently identify different materials or hole patterns. Therefore, in order to solve the above problems, a roll film intelligent punching system and method is needed. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention proposes a roll film intelligent perforation system and method. Among them, the specific technical solution of a roll film intelligent perforation system is as follows:

[0004] A roll film intelligent punching system, characterized by:

[0005] It includes a coiling component, a data acquisition module, an image recognition component, a control module, a punching component and a winding component;

[0006] The data acquisition module and the image recognition component are connected to the control module via a data bus, and the control module is respectively connected to the coiling component, the punching component and the winding component via a control channel;

[0007] The coil assembly is used to transmit the roll film to be punched to the processing area, the data acquisition module is used to detect the position information of the film and synchronously transmit it to the control module, the image recognition assembly is used to identify the hole positions of the film, the punching assembly is used to receive the control signal of the control module, and perform the punching operation on the positioned film material, and the winding assembly is used to roll up the punched film.

[0008] In order to better implement the present invention, it can be further provided that: the punching assembly includes a punching head and a driving device, and the driving device receives a control signal from a control module to punch holes in the positioned film material.

[0009] A specific technical solution of the working method of a roll film intelligent punching system is as follows:

[0010] A working method of a roll film intelligent punching system, characterized in that:

[0011] The steps include:

[0012] S1: The roll material assembly transfers the roll film material to be punched to the processing area and keeps the film material stable;

[0013] S2: The data acquisition module detects the speed and position data of the film and transmits the data to the control module;

[0014] S3: the image recognition component collects the surface image of the film, analyzes the preset mark on the film, identifies the hole position of the film, and transmits the identification data to the control module;

[0015] S4: the control module generates a punching control signal and transmits it to the punching component;

[0016] S5: After receiving the control signal from the control module, the punching component performs a punching operation on the target position of the roll film material;

[0017] S6: The control module generates a synchronous control signal to the winding component, so that the winding speed of the winding component is synchronized with the punching operation to ensure that the film material is evenly rolled up.

[0018] Further:

[0019] The image recognition component includes an image preprocessing module and an image recognition model;

[0020] The image recognition model adopts a convolutional neural network model, which is used to enhance positioning accuracy and adopts the following training process:

[0021] The image recognition model takes N thin film images with preset labels as input data, and the marked positions are labels to form a data set.

[0022] Among them, T m is the mth image, y m The actual coordinates of the preset mark position;

[0023] The mapping function of the image recognition model is f α , the output prediction of the image recognition model The loss function of the image recognition model is obtained:

[0024]

[0025] The image recognition model is trained by minimizing the loss function L(α) by the gradient descent method to obtain the optimized model parameter α * , and obtain the trained image recognition model.

[0026] Further: S3-1: the image preprocessing module performs grayscale processing on the image of the film surface to convert it into a single-channel image;

[0027] S3-2: The image preprocessing module performs smoothing filtering on the grayscale image to remove noise, make the edge of the preset mark clearer, and obtain the image filter P1(u,v):

[0028]

[0029] Among them, Q(a,c) is the filter kernel, and R(u,v) is the original grayscale image;

[0030] S3-3: The image preprocessing module performs Canny edge detection on the smoothed image P1 (u, v), extracts edge information in the image, and forms an edge image P2 (u, v);

[0031] S3-4: The image recognition model identifies the preset hole position (u1, v1) based on the edge image P2 (u, v) as input recognition;

[0032] S3-5: The image recognition model inputs the preset hole position (u1, v1) into the control module.

[0033] Further: the training process of the convolutional neural network uses a film image dataset with a preset identifier, the network is set with a learning rate of 0.001, a maximum number of iterations of 2000 times, and a batch size of 64;

[0034] The convolutional neural network uses a gradient descent algorithm to continuously update network parameters by minimizing the loss function to optimize the accuracy of the model.

[0035] Furthermore: the convolutional neural network uses the Adam optimizer to accelerate convergence and reduce oscillations during training, and ultimately obtains model parameters that can accurately identify the preset marking position on the film.

[0036] The beneficial effects of the present invention are as follows: the present invention has a simple overall structure, and the position, speed and hole position of the film are monitored and accurately located in real time through the data acquisition module and the image recognition component, thereby solving the problem of insufficient positioning accuracy of traditional machinery and greatly improving the accuracy of punching. The image recognition component in the system can identify the preset logo on the film and accurately determine the hole position, making the processing of films of different specifications more convenient. The convolutional neural network uses the Adam optimizer, which can adaptively adjust the learning rate of each parameter to achieve rapid convergence, and is insensitive to hyperparameters, providing stable performance. Combined with momentum and root mean square propagation, it can effectively reduce oscillations in parameter updates, thereby quickly approaching the target. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is the overall structure diagram of the present invention;

[0038] Figure 2 It is the work flow chart of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] like Figure 1 As shown:

[0041] A roll film intelligent punching system, comprising a roll material component, a data acquisition module, an image recognition component, a control module, a punching component and a winding component;

[0042] The data acquisition module and the image recognition component are connected to the control module via a data bus, and the control module is respectively connected to the coiling component, the punching component and the winding component via a control channel;

[0043] The coil assembly is used to transmit the roll film to be punched to the processing area, the data acquisition module is used to detect the position information of the film and synchronously transmit it to the control module, the image recognition assembly is used to identify the hole positions of the film, the punching assembly is used to receive the control signal of the control module, and perform the punching operation on the positioned film material, and the winding assembly is used to roll up the punched film.

[0044] The punching assembly includes a punching head and a driving device, and the driving device receives a control signal from a control module to punch holes in the positioned film material.

[0045] like Figure 2As shown, a working method of a roll film intelligent punching system includes the following steps:

[0046] S1: The roll material assembly transfers the roll film material to be punched to the processing area and keeps the film material stable;

[0047] S2: The data acquisition module detects the speed and position data of the film and transmits the data to the control module;

[0048] S3: The image recognition component collects the surface image of the film, analyzes the preset mark on the film, identifies the hole position of the film, and transmits the identification data to the control module.

[0049] Wherein, the image recognition component comprises an image preprocessing module and an image recognition model;

[0050] The image recognition model adopts a convolutional neural network model, which is used to enhance positioning accuracy and adopts the following training process:

[0051] The image recognition model takes N thin film images with preset labels as input data, and the marked positions are labels to form a data set.

[0052] Among them, T m is the mth image, y m The actual coordinates of the preset mark position;

[0053] The mapping function of the image recognition model is f α , the output prediction of the image recognition model The loss function of the image recognition model is obtained:

[0054]

[0055] The image recognition model is trained by minimizing the loss function L(α) by the gradient descent method to obtain the optimized model parameter α * , and obtain the trained image recognition model.

[0056] The training process of the convolutional neural network uses a film image dataset with preset labels. The network is set with a learning rate of 0.001, a maximum number of iterations of 2000 times, and a batch size of 64;

[0057] The convolutional neural network uses a gradient descent algorithm to continuously update network parameters by minimizing the loss function to optimize the accuracy of the model.

[0058] The convolutional neural network uses the Adam optimizer to accelerate convergence and reduce oscillations during training, and ultimately obtains model parameters that can accurately identify the preset marking position on the film.

[0059] Specifically: including the following processes:

[0060] S3-1: The image preprocessing module performs grayscale processing on the image of the film surface to convert it into a single-channel image;

[0061] S3-2: The image preprocessing module performs smoothing filtering on the grayscale image to remove noise, make the edge of the preset mark clearer, and obtain the image filter P1(u,v):

[0062]

[0063] Among them, Q(a,c) is the filter kernel, and R(u,v) is the original grayscale image;

[0064] S3-3: The image preprocessing module performs Canny edge detection on the smoothed image P1 (u, v), extracts edge information in the image, and forms an edge image P2 (u, v);

[0065] S3-4: The image recognition model identifies the preset hole position (u1, v1) based on the edge image P2 (u, v) as input recognition;

[0066] S3-5: The image recognition model inputs the preset hole position (u1, v1) into the control module.

[0067] S4: the control module generates a punching control signal and transmits it to the punching component;

[0068] S5: After receiving the control signal from the control module, the punching component performs a punching operation on the target position of the roll film material;

[0069] S6: The control module generates a synchronous control signal to the winding component, so that the winding speed of the winding component is synchronized with the punching operation to ensure that the film material is evenly rolled up.

[0070] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0071] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A roll film intelligent punching system, characterized by: It includes a coiling component, a data acquisition module, an image recognition component, a control module, a punching component and a winding component; The data acquisition module and the image recognition component are connected to the control module via a data bus, and the control module is respectively connected to the coiling component, the punching component and the winding component via a control channel; The coil assembly is used to transmit the roll film to be punched to the processing area, the data acquisition module is used to detect the position information of the film and synchronously transmit it to the control module, the image recognition assembly is used to identify the hole positions of the film, the punching assembly is used to receive the control signal of the control module, and perform the punching operation on the positioned film material, and the winding assembly is used to roll up the punched film.

2. According to claim 1, a roll film intelligent punching system is characterized by: The punching assembly includes a punching head and a driving device. The driving device receives a control signal from a control module and punches holes in the positioned film material.

3. A working method of the intelligent perforating system for roll film according to claim 2, characterized in that: The steps include: S1: The roll material assembly transfers the roll film material to be punched to the processing area and keeps the film material stable; S2: The data acquisition module detects the speed and position data of the film and transmits the data to the control module; S3: the image recognition component collects the surface image of the film, analyzes the preset mark on the film, identifies the hole position of the film, and transmits the identification data to the control module; S4: the control module generates a punching control signal and transmits it to the punching component; S5: After receiving the control signal from the control module, the punching component performs a punching operation on the target position of the roll film material; S6: The control module generates a synchronous control signal to the winding component, so that the winding speed of the winding component is synchronized with the punching operation to ensure that the film material is evenly rolled up.

4. The working method of the intelligent perforating system for roll film according to claim 3, characterized in that: S3-1: The image recognition component includes an image preprocessing module and an image recognition model; The image recognition model adopts a convolutional neural network model, which is used to enhance positioning accuracy and adopts the following training process: The image recognition model takes N thin film images with preset labels as input data, and the marked positions are labels to form a data set. Among them, T m is the mth image, y m The actual coordinates of the preset mark position; The mapping function of the image recognition model is f α , the output prediction of the image recognition model The loss function of the image recognition model is obtained: The image recognition model is trained by minimizing the loss function L(α) by the gradient descent method to obtain the optimized model parameter α * , and obtain the trained image recognition model.

5. The working method of the intelligent perforating system for roll film according to claim 4, characterized in that: S3-1: The image preprocessing module performs grayscale processing on the image of the film surface to convert it into a single-channel image; S3-2: The image preprocessing module performs smoothing filtering on the grayscale image to remove noise, make the edge of the preset mark clearer, and obtain the image filter P1(u,v): Among them, Q(a,c) is the filter kernel, and R(u,v) is the original grayscale image; S3-3: The image preprocessing module performs Canny edge detection on the smoothed image P1 (u, v), extracts edge information in the image, and forms an edge image P2 (u, v); S3-4: The image recognition model identifies the preset hole position (u1, v1) based on the edge image P2 (u, v) as input recognition; S3-5: The image recognition model inputs the preset hole position (u1, v1) into the control module.

6. The working method of the intelligent perforating system for roll film according to claim 5, characterized in that: The training process of the convolutional neural network uses a film image dataset with preset labels. The network is set with a learning rate of 0.001, a maximum number of iterations of 2000 times, and a batch size of 64; The convolutional neural network uses a gradient descent algorithm to continuously update network parameters by minimizing the loss function to optimize the accuracy of the model.

7. The working method of the film roll intelligent punching system according to claim 6, characterized in that: The convolutional neural network uses the Adam optimizer to accelerate convergence and reduce oscillations during training, and ultimately obtains model parameters that can accurately identify the preset marking position on the film.

Images