A multi-data fusion laser pattern etching system and method

Abstract

The application relates to a multi-data fusion laser pattern etching system and method, which comprises the following: a non-contact three-dimensional visual scanning system, which is used for completing global curved surface data collection and modeling of a product to be processed and storing in the form of a standardized database; an intelligent adhering system, which realizes alignment of a protective film and a surface of the product to be processed, and realizes adhering of the protective film without stretching and wrinkling; and an intelligent laser etching system, which maps a two-dimensional pattern to a three-dimensional curved surface, eliminates pattern distortion through self-adaptive path planning of the three-dimensional curved surface, and completes non-distortion etching of a curved surface pattern of the product to be processed. Through cooperation of multi-data fusion, three-dimensional vision, intelligent control, laser processing and other technologies, the application realizes high-precision protective film adhering and non-distortion curved surface laser pattern etching of plane, curved surface and complex geometric shape products, improves processing precision, operation efficiency and product consistency of the protective film adhering and pattern etching, and meets actual needs of high-precision curved surface product protective film processing.

Description

Technical Field

[0001] This invention relates to the field of laser processing technology, and in particular to a laser patterning etching system and method with multi-data fusion. Background Technology

[0002] In existing laser patterning etching processes, the protective film lamination process is generally carried out manually or semi-automatically. Operators rely on visual judgment and simple tooling to cover the product surface with the protective film, and then manually remove air bubbles and compact it with tools such as scrapers. This process is highly dependent on the operator's subjective experience and skill level, resulting in significant fluctuations in lamination quality and making it difficult to achieve batch-to-batch consistency control. Especially for product surfaces with large curvature or complex geometric shapes, such as curved glass displays, automotive interior parts, or medical device shells, the protective film is prone to irreversible tensile deformation, wrinkle accumulation, or positional displacement due to localized stress concentration, resulting in poor adhesion between the film and the product surface, and making it difficult to completely eliminate residual air bubbles.

[0003] In the laser etching stage, traditional techniques process only two-dimensional drawing data, mapping planar graphics onto the product's curved surface through manual programming or simple projection. Due to a lack of precise perception and modeling capabilities of the product's actual three-dimensional form, geometric distortion inevitably occurs in the etched pattern on curved surfaces, manifesting as blurred edge contours, uneven etching depth distribution, and distorted pattern proportions. More importantly, traditional methods fail to establish a unified three-dimensional reference model, causing a disconnect between the geometric reference systems of the bonding and etching stages, resulting in the continuous accumulation and amplification of errors between processes.

[0004] During the bonding process, the system cannot dynamically adjust the film tension parameters and pressing path according to the curvature distribution of the surface, causing the film to fail to bond in high curvature areas. In the etching path planning, it cannot effectively compensate for the mapping distortion from the plane to the curved surface, nor can it optimize the laser energy parameters in real time according to the local curvature changes. This ultimately results in fluctuations in the width of the etching lines, poor consistency in depth, and pattern deformation defects, which seriously restricts the product yield and surface accuracy. Summary of the Invention

[0005] Therefore, the purpose of this invention is to provide a laser pattern etching system and method with multi-data fusion, which has the advantages of achieving stretch-free and wrinkle-free bonding of protective films, eliminating pattern distortion, ensuring uniform etching depth and edge accuracy, thereby improving product yield and surface accuracy.

[0006] To address the aforementioned technical problems, this invention provides a multi-data fusion laser pattern etching system, comprising: a non-contact 3D vision scanning system, an intelligent bonding system, and an intelligent laser etching system. The non-contact 3D vision scanning system is used to acquire and model the entire surface of the product to be processed, constructing a 3D reference model containing the surface features, dimensions, unevenness, and curvature distribution of the product, and storing it in a standardized database. The intelligent bonding system automatically selects a suitable protective film specification based on the 3D reference model of the product to be processed, and aligns the protective film with the surface of the product to be processed, achieving a non-stretching and wrinkle-free bonding of the protective film. The intelligent laser etching system maps 2D graphics to a 3D surface, eliminates graphic distortion through adaptive curvature path planning of the 3D surface, and adjusts laser parameters in real time to ensure uniform etching depth and edge accuracy, completing distortion-free etching of the surface graphics of the product to be processed.

[0007] The multi-data fusion laser pattern etching system of the present invention achieves high-precision protective film bonding and distortion-free laser pattern etching of planar, curved, and complex geometric products by combining multi-data fusion with technologies such as 3D vision, intelligent control, and laser processing. This improves the processing accuracy, work efficiency, and finished product consistency of protective film bonding and pattern etching, and meets the actual needs of high-precision curved product protective film processing.

[0008] In one embodiment of the present invention, an intelligent assisted excess film removal system is also included. After the intelligent laser etching system completes etching, an industrial camera is used to acquire images, identify and segment the waste areas that need to be removed, and map their physical coordinates onto a three-dimensional reference model. Then, the waste areas are projected onto the surface of the product to be processed in a highlighted form. Based on the geometric characteristics of each waste block, the optimal film removal starting point is recommended and marked to guide the operator to remove the waste areas in sequence.

[0009] This invention also provides a laser pattern etching method based on multi-data fusion, comprising the following steps: S1. Perform full-area surface 3D data acquisition and modeling processing: acquire full-area point cloud data of the product to be processed through a non-contact 3D vision scanning system, iteratively register the acquired point cloud data with the CAD template of the product to be processed, filter out outliers, and then use the NURBS surface fitting algorithm to construct an accurate 3D benchmark model from the processed point cloud data. S2. The intelligent bonding system utilizes the product size and surface curvature distribution information contained in the three-dimensional reference model to automatically select the most suitable protective film from the film material database. It calculates and sets the reference tension based on the elastic modulus of the protective film and the extreme value of the product curvature. At the same time, it collects the reference edge position of the protective film in real time and matches it with the three-dimensional reference model. It calculates the pose deviation between the protective film and the target bonding surface of the product to be processed through a spatial coordinate transformation algorithm. It performs real-time compensation based on the calculated pose deviation to ensure that the protective film is accurately aligned with the surface of the product to be processed. S3. The intelligent laser etching system receives a three-dimensional reference model and two-dimensional graphic data input by the user. It imports the two-dimensional graphic data into the intelligent laser etching system and uses the UV parametric spatial coordinate transformation algorithm to accurately map the two-dimensional graphic data onto the three-dimensional reference model, generating an initial three-dimensional etching path point set. Based on the curvature distribution in the three-dimensional reference model, the etching path is segmented and corrected, and smoothed by cubic spline interpolation. The laser power and scanning speed are adjusted in real time according to the curvature of each segment of the path. The width of the etching lines and the edge sharpness are monitored in real time, and the laser parameters are fine-tuned using a PID algorithm to form a closed-loop control, ultimately completing the etching of the curved surface graphic. S4. After etching is completed, the intelligent assisted excess film removal system acquires images through a multi-view industrial camera, identifies and segments the waste area that needs to be removed, maps its physical coordinates onto a three-dimensional reference model, highlights the waste outline on the surface of the product to be processed, recommends the best removal starting point, and guides the operator to remove the excess film with voice prompts.

[0010] In one embodiment of the present invention, in step S1 above, the collected point cloud data is iteratively registered with the CAD template of the product to be processed, and the registration error is calculated using the iterative nearest point algorithm: ; Where N is the number of point clouds; The three-dimensional coordinate point is obtained by projecting the i-th point on the two-dimensional graphic. The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions; The preset allowable threshold for graphic mapping error.

[0011] In one embodiment of the present invention, in step S1 above, a three-dimensional reference model is constructed using the NURBS surface fitting algorithm, and the fitting accuracy is measured by the root mean square error. ; in, A three-dimensional baseline model; The preset allowable threshold for surface fitting accuracy.

[0012] In one embodiment of the present invention, in step S2 above, real-time compensation is performed based on the calculated pose deviation: ; in, To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; This refers to the rotational angle deviation of the protective film around the X-axis; This refers to the rotational angle deviation of the protective film around the Y-axis; This is to protect the film from rotational angle deviation around the Z-axis.

[0013] In one embodiment of the present invention, in step S3 above, the input is the three-dimensional reference model output in step S1. and user-provided two-dimensional graphic data Two-dimensional graphic data is transformed using a spatial coordinate transformation algorithm based on UV parameterization. Mapping to a 3D baseline model Above, generate the initial three-dimensional etching path point set. .

[0014] In one embodiment of the present invention, the two-dimensional graphic data Mapping to a 3D baseline model The mapping error is: ; in, The maximum mapping error after mapping a two-dimensional graphic to a three-dimensional surface; To make the second two-dimensional graphic a graphic point The three-dimensional coordinate points obtained after projection transformation; The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions.

[0015] In one embodiment of the present invention, in step S3 above, the density of path points is dynamically adjusted according to the curvature of the surface, and the spacing between basic points is set. At the radius of curvature of the surface Less than the threshold The area uses encryption factors. Reduce the point spacing to ; During the etching parameter generation stage, the intelligent laser etching system determines the etching parameters based on the curvature of each path segment. Real-time adjustment of laser power and scanning speed To maintain a consistent etching depth, Using an empirical formula: ; in, The laser frequency, The pulse width. and For material constants, Target etching depth; This represents the actual etching depth. During the etching process, the intelligent laser etching system monitors the width of the etched lines in real time. And edge sharpness, when width deviation is detected At the same time, a PID algorithm is used to fine-tune the etching power and speed, forming a closed-loop feedback control, and the output is the optimized three-dimensional etching path. and the corresponding laser parameter sequence It is stored in standard CNC G-code format.

[0016] In one embodiment of the present invention, in step S4 above, after the etching process is completed, the intelligent assisted excess film removal system uses a multi-view industrial camera to acquire multi-view images of the product surface with the protective film, obtaining an image sequence of the etched area. The complete film surface image is reconstructed using an image stitching algorithm, and the cutting boundaries formed by laser etching are extracted using an edge detection operator to construct a binary mask of the etching contour. Binarization mask Compared with the theoretical three-dimensional etching path generated in step S3 Perform registration and comparison, calculate the positional deviation of the etching boundary, and ensure that all cutting lines are completed and without breakage; During the redundant membrane region identification stage, the system will use a binarized mask. The enclosed area within the shape is defined as the reserved shape area. The entire area covered by the protective film minus The resulting collection of excess film areas that need to be removed is then obtained. ; For the collection of excess membrane regions The intelligent assisted excess membrane removal system performs connected component analysis and divides it into... A separate waste block ( Each block is equipped with geometric feature parameters, including the center coordinates, area, and boundary length of the minimum bounding rectangle. The image coordinates are converted to the physical coordinates corresponding to the product's 3D reference model through coordinate system calibration, and the conversion error is controlled within ≤0.2mm.

[0017] The multi-data fusion laser pattern etching system and method of the present invention have the following advantages compared with the prior art: 1. This application automatically acquires the point cloud of the product surface using a non-contact 3D vision scanning system and performs high-precision registration with a CAD template. After effectively eliminating scanning noise, an advanced surface fitting algorithm is used to construct a 3D benchmark model that perfectly matches the actual shape of the product. This model fully includes surface features, curvature distribution, and dimensional information, ensuring the geometric accuracy of protective film bonding and laser etching path planning. It avoids processing errors caused by data deviations from the source, and is especially suitable for the precision manufacturing of complex curved surface products.

[0018] 2. This application achieves intelligent bonding of the protective film without stretching or wrinkles, significantly improving bonding quality and consistency. The system automatically matches the optimal film specifications based on product size and surface characteristics, and precisely adjusts the conveying tension to prevent film deformation. A vision positioning system monitors the film's posture in real time, and a multi-degree-of-freedom fine-tuning mechanism minimizes alignment errors between the film and the product surface. A flexible pressing mechanism applies controllable pressure and speed to uniformly press the film along the surface normal from the initial point, dynamically monitoring the contact force and adjusting the pressing path to ensure a tight, bubble-free bond between the film and the surface, providing a smooth and uniform foundation for subsequent laser etching.

[0019] 3. This application achieves distortion-free etching of curved surfaces, ensuring the precision and uniformity of pattern edges and depth. The system accurately maps the user-input 2D graphic to a 3D curved surface using a spatial coordinate transformation algorithm, eliminating projection distortion from a plane to a curved surface. Adaptive encryption and smoothing processing of the etching path is applied based on the surface curvature distribution, continuously optimizing the laser head's trajectory. Laser power and scanning speed are adjusted in real-time according to the curvature of different regions to ensure a constant etching depth; simultaneously, a coaxial vision sensor monitors the etching line status in real-time, providing closed-loop feedback to control parameters, ensuring sharp edges and consistent depth of the etched pattern, thus completely solving the problem of graphic distortion in curved surface etching.

[0020] 4. After etching is completed, the system uses a high-resolution industrial camera to acquire images from multiple perspectives, automatically identifies and segments the waste area to be removed, and accurately converts its coordinates to the physical space of the product. The waste outline is then highlighted on the product surface using laser projection or augmented reality technology, and the optimal removal starting point is intelligently recommended based on geometric features. Operators no longer need to painstakingly search for boundaries, allowing for quick and accurate removal of excess film, avoiding accidental damage to the retained pattern, and significantly improving the efficiency and reliability of the removal process.

[0021] 5. This application optimizes the system architecture to enable parallel and collaborative key processes: noise removal and surface fitting are performed simultaneously during the 3D scanning stage; tension delivery and visual positioning are executed synchronously during the bonding stage, with dynamic adjustment of pressing and force feedback; graphic mapping and path planning are completed synchronously during the etching stage, with etching execution, visual monitoring, and parameter control working in parallel. This locally parallel processing mode fully utilizes system resources, achieving seamless connection between processes, significantly shortening the processing cycle time compared to traditional serial operations, and meeting the efficiency requirements of large-scale production. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of the multi-data fusion laser patterning etching system of the present invention; Figure 2 This is a flowchart of the multi-data fusion laser pattern etching method of the present invention; Figure 3 This is a flowchart of the intelligent bonding system of the present invention. Detailed Implementation

[0024] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0025] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] It should be noted that the following description covers various aspects of embodiments within the scope of the appended claims. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0027] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. The drawings only show the components related to this application and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0028] Additionally, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that practice can be carried out without these specific details.

[0029] Traditional protective film lamination often employs manual or semi-automatic methods, with lamination quality heavily reliant on operator experience. For complex curved surfaces, this can easily lead to film stretching, wrinkling, or misalignment. Laser etching typically relies on two-dimensional drawings, lacking accurate perception of the product's actual three-dimensional shape. This results in distortion of the etched pattern on curved areas, making it difficult to guarantee edge precision and depth uniformity. Due to the lack of real-time acquisition and utilization of product three-dimensional data, traditional methods cannot establish an accurate product reference model, leading to inconsistent geometric references between lamination and etching, and significant cumulative errors.

[0030] In response, this application proposes a multi-data fusion laser patterning etching system, referring to... Figure 1 and 3As shown, it includes: a non-contact 3D vision scanning system, an intelligent bonding system, and an intelligent laser etching system; the non-contact 3D vision scanning system is used to complete the acquisition and modeling of the entire surface of the product to be processed, constructing a 3D reference model containing the surface features, size information, concavity and convexity changes, and curvature distribution of the product to be processed, and storing it in a standardized database; the intelligent bonding system automatically selects the appropriate protective film specification according to the 3D reference model of the product to be processed, and realizes the alignment of the protective film with the surface of the product to be processed, achieving a non-stretching and wrinkle-free bonding of the protective film; the intelligent laser etching system maps the 2D graphics to the 3D surface, eliminates graphic distortion through the curvature adaptive path planning of the 3D surface, and adjusts the laser parameters in real time to ensure uniform etching depth and edge accuracy, completing the distortion-free etching of the surface graphics of the product to be processed. Figure 1 In the diagram, the green lines represent the core geometric and dimensional data transmitted from the 3D reference model to the etching system. This data maps 2D graphics to 3D surfaces, supporting path planning and etching precision control. The red lines represent the surface curvature distribution data transmitted from the 3D reference model to the etching system. This data adaptively adjusts laser parameters to ensure uniform etching depth on large surfaces. The blue lines represent the dimensional and curvature data transmitted from the 3D reference model to the bonding system. This data automatically matches protective film specifications, calculates tension and bonding paths, achieving wrinkle-free and precise bonding of large workpieces. The solid black lines represent the serial execution flow within the system, showing the sequence of steps and data transmission relationships between modules. The dashed black lines represent the feedback / iteration relationships within the system.

[0031] The non-contact 3D vision scanning system mainly consists of a 3D data acquisition module, a point cloud registration module, a noise removal module, a surface fitting module, and a 3D benchmark model construction module; the intelligent bonding system mainly consists of a protective film specification automatic matching module, a constant tension conveying module, a visual positioning module, a six-degree-of-freedom pose fine-tuning module, and a flexible pressing module; the intelligent laser etching system mainly consists of a 2D graphic import module, a 3D surface mapping module, a curvature adaptive path planning module, a laser parameter real-time adjustment module, and a coaxial vision closed-loop feedback module. The non-contact 3D vision scanning system, intelligent bonding system, and intelligent laser etching system collaborate through data interaction to achieve fully automated operation of protective film bonding and laser pattern etching. During the operation of the non-contact 3D vision scanning system, 3D point cloud data acquisition and product geometric feature recognition are performed simultaneously, while noise data removal and surface fitting calculations are processed in parallel. During the operation of the intelligent bonding system, constant tension delivery of the protective film and visual positioning pose detection are performed simultaneously, while flexible pressing bonding and real-time feedback dynamic adjustment of contact force are coordinated. During the operation of the intelligent laser etching system, 3D surface pattern mapping and etching path planning are completed simultaneously, while laser etching execution, coaxial vision real-time monitoring, and parameter closed-loop feedback control work in parallel. Through local collaborative parallel operation, the real-time performance and processing efficiency of the operation are improved.

[0032] Therefore, the multi-data fusion laser pattern etching system proposed in this application effectively solves the problems of protective film bonding quality fluctuations, easy stretching and wrinkling, and laser etching pattern distortion and uneven depth in curved areas in traditional methods by integrating non-contact 3D visual scanning, intelligent bonding, and intelligent laser etching functions. This system can provide accurate 3D geometric references, achieve stretch-free and wrinkle-free bonding of the protective film, and ensure distortion-free etching of curved patterns, effectively improving the processing accuracy and consistency of complex curved surface products.

[0033] In the embodiments described above in this application, an intelligent laser etching system is proposed to complete distortion-free etching of curved patterns. However, during its implementation, when removing excess film after etching, the operator may have difficulty accurately identifying waste areas, and improper selection of the starting point may lead to low film removal efficiency or residual waste. To address this, this application further proposes, based on the aforementioned multi-data fusion laser pattern etching system, an intelligent assisted excess film removal system. This system is used to acquire images using an industrial camera after the intelligent laser etching system has completed etching, identify and segment the waste areas to be removed, map their physical coordinates onto a three-dimensional reference model, and then project the waste areas in a highlighted form onto the surface of the product to be processed. Based on the geometric characteristics of each waste block, the system recommends and marks the optimal film removal starting point, guiding the operator to remove the waste areas sequentially.

[0034] Specifically, the intelligent assisted excess film removal system is an automated module integrating visual recognition, spatial positioning, projection guidance, and human-machine interaction. This system can consist of one or more industrial cameras, an image processing unit, a projector, and a control unit, or it can be integrated into the post-processing station of a laser etching equipment as a standalone post-processing unit. The industrial cameras can be high-resolution CCD or CMOS cameras, equipped with suitable lenses and light sources to ensure image clarity and color accuracy, facilitating subsequent image processing algorithms; or a multi-view industrial camera array can be used to acquire images from different angles to address complex curved surfaces and occlusion issues, ensuring the integrity of the entire image data area.

[0035] Through the above technical solutions, the intelligent assisted excess film removal system can automatically identify and accurately segment the etched waste areas, solving the problems of inaccurate identification and low efficiency in traditional manual operations. By accurately mapping the physical coordinates of the waste areas onto a three-dimensional reference model and projecting them in a highlighted form onto the surface of the product to be processed, the system provides operators with intuitive and accurate visual guidance, greatly reducing operational difficulty and human error. Simultaneously, the system intelligently recommends and marks the optimal film removal starting point based on the geometric characteristics of the waste area. Combined with sequential guidance, this optimizes the film removal process, effectively avoiding waste residue and film tearing, significantly improving the efficiency and accuracy of film removal, and ensuring the final processing quality and consistency of the product.

[0036] Based on the above system, the present invention also provides a laser pattern etching method with multi-data fusion, referring to... Figure 2 As shown, it includes the following steps: S1. Perform full-area surface 3D data acquisition and modeling processing: Acquire full-area point cloud data of the product to be processed through a non-contact 3D vision scanning system. Iteratively register the acquired point cloud data with the CAD template of the product to be processed and filter out outliers. Then, use the NURBS surface fitting algorithm to construct an accurate 3D reference model from the processed point cloud data. This model will include the surface features, size information, concavity and convexity changes, and curvature distribution of the product. Noise data generated during the scanning process will be removed. Through a high-precision surface fitting algorithm, effective 3D product data will be obtained, forming a standardized database to provide data support for subsequent laser engraving path and film bonding control.

[0037] S2. The intelligent bonding system utilizes product dimensions and surface curvature distribution information contained in the 3D reference model to automatically select the most suitable protective film from the film material database. It calculates and sets the reference tension based on the protective film's elastic modulus and the product's curvature extreme values. Simultaneously, it collects the reference edge position of the protective film in real time and matches it with the 3D reference model. Through a spatial coordinate transformation algorithm, it calculates the pose deviation between the protective film and the target bonding surface of the product to be processed. Based on the calculated pose deviation, it performs real-time compensation to ensure the protective film is precisely aligned with the surface of the product to be processed. This ensures the film surface is free of stretching and wrinkles, preventing displacement during bonding. A flexible pressing mechanism pre-presses at the initial bonding point, ensuring the film is uniformly bonded to the product surface.

[0038] The S3 intelligent laser etching system receives a 3D reference model and user-input 2D graphic data. The 2D graphic data is imported into the system, and a UV parametric spatial coordinate transformation algorithm is used to accurately map the 2D graphic data onto the 3D reference model, generating an initial 3D etching path point set. Based on the curvature distribution in the 3D reference model, the etching path is segmented and corrected, and smoothed using cubic spline interpolation. The laser power and scanning speed are adjusted in real-time according to the curvature of each segment of the path. The width of the etched lines and edge sharpness are monitored in real-time, and a PID algorithm is used to fine-tune the laser parameters, forming a closed-loop control to ultimately complete the etching of the curved surface. Combining the changes in surface curvature, the mapped etching path is segmented and corrected to eliminate graphic distortion from planar to curved surface etching, ensuring that the laser etching head can move seamlessly along the optimized path. For areas with different curvatures, the laser power and speed are adjusted to ensure the accuracy of the etched pattern. Based on the optimized path and parameters, the laser actuator is activated for precise etching. The system monitors the etching status in real-time and adjusts the laser parameters based on feedback to ensure the clarity and edge precision of the etched pattern.

[0039] S4. After etching is completed, the intelligent assisted excess film removal system acquires images through a multi-view industrial camera, identifies and segments the waste area that needs to be removed, maps its physical coordinates onto a three-dimensional reference model, highlights the waste outline on the surface of the product to be processed, recommends the best removal starting point, and guides the operator to remove the excess film with voice prompts.

[0040] Through the above technical solutions, this application achieves fully automated processing driven by 3D data. By establishing a precise 3D reference model and integrating it throughout the bonding and etching processes, the consistency of geometric references at each step is ensured. Dynamic tension control and real-time pose compensation based on curvature distribution effectively prevent film stretching and wrinkling. Furthermore, the curvature-adaptive path correction and laser parameter closed-loop adjustment mechanism fundamentally solve the technical challenges of etching pattern distortion and uneven depth. Overall, this method, through multi-data fusion and closed-loop control strategies, systematically improves the bonding accuracy of protective films and the quality of laser etching for complex curved surface products, providing reliable technical support for high-consistency curved surface processing.

[0041] Furthermore, in step S1 above, the collected point cloud data is iteratively registered with the CAD template of the product to be processed, and the registration error is calculated using the iterative nearest point algorithm. ; Where N is the number of point clouds, representing the total number of point cloud data points involved in the registration error calculation; The three-dimensional coordinate point is obtained by projecting the i-th point on the two-dimensional graphic. The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions; The preset allowable threshold for graphic mapping error.

[0042] The Iterative Closest Point Algorithm (ILP) is a widely used algorithm for point cloud registration. Its core idea is to iteratively find the closest point pairs between two point sets and calculate the transformation matrix based on these pairs to minimize the distance between the two point sets. During registration, it can quantify the geometric deviation between the point cloud data and the CAD template, providing an objective numerical metric to evaluate the accuracy of the registration.

[0043] Through the above technical solution, this application provides an objective and quantitative registration accuracy evaluation mechanism. The introduction of the iterative nearest-point algorithm enables precise calculation of the geometric deviation between point cloud data and the CAD template, avoiding the uncertainty caused by subjective judgment or experience-based reliance in traditional methods. This error calculation formula comprehensively considers all point cloud points, ensuring the representativeness of the evaluation. Furthermore, by comparing with a preset threshold, the system can automatically determine whether the registration result meets the accuracy requirements. When the error exceeds the threshold, the system can promptly issue an alarm or automatically trigger a re-registration process, effectively avoiding subsequent graphic mapping distortion and etching quality degradation caused by inaccurate model construction. This precise registration process provides a high-precision three-dimensional reference model for the subsequent intelligent bonding system to select a protective film and for the intelligent laser etching system to perform distortion-free etching, significantly improving the overall processing accuracy and reliability of the entire multi-data fusion laser graphic etching system.

[0044] Furthermore, in step S1 above, a three-dimensional benchmark model is constructed using the NURBS surface fitting algorithm, and the fitting accuracy is measured by the root mean square error. ; in, A three-dimensional baseline model; The preset allowable threshold for surface fitting accuracy.

[0045] The model includes surface feature parameters such as curvature distribution. Dimensional information and convexity / concave variations; the final output is a standardized database, consisting of point cloud data and a 3D baseline model. The format storage provides precise data support for subsequent laser etching pattern path planning and film bonding control.

[0046] The NURBS surface fitting algorithm is used to construct a 3D benchmark model. Its core lies in converting discrete point cloud data into a continuous, smooth mathematical surface representation. This algorithm can flexibly handle various complex surfaces and accurately capture the geometric features and curvature distribution of the product to be processed. In practical applications, NURBS surface fitting can be implemented in several ways. For example, one approach is to adjust the control points and weights of the NURBS surface through iterative optimization to minimize the distance between the surface and the point cloud data. Another approach is to divide the point cloud data into multiple local regions, perform NURBS surface fitting independently on each region, and then stitch these local surfaces together into a complete global 3D benchmark model through continuity constraints.

[0047] By introducing the root mean square error (RMSE) as a quantitative measure of NURBS surface fitting accuracy through the above technical solution, an objective and verifiable quality control mechanism can be provided for the construction of the 3D benchmark model. Given that the point cloud data has already been registered and outlier filtered using the iterative nearest point algorithm in step S1, the RMSE is used to strictly control the accuracy of NURBS surface fitting, ensuring that the constructed 3D benchmark model can highly accurately reflect the true geometric shape of the product to be processed. This accurate model construction avoids geometric reference inconsistencies caused by inaccurate models, thus effectively suppressing the accumulation of errors in subsequent intelligent bonding and intelligent laser etching systems. For example, a high-precision 3D benchmark model can provide accurate surface curvature information for the intelligent bonding system, enabling it to more accurately calculate the reference tension of the protective film and achieve precise alignment between the protective film and the product surface without stretching or wrinkles. Meanwhile, for intelligent laser etching systems, a precise three-dimensional reference model is the foundation for achieving distortion-free mapping of two-dimensional graphics and curvature adaptive path planning, ensuring uniform etching depth and edge accuracy, and ultimately significantly improving the processing quality and product consistency of the entire laser pattern etching system.

[0048] To achieve step S2 above, a smart bonding system integrating multi-sensor feedback and dynamic tension adjustment is first constructed to integrate the physical properties of the protective film with the three-dimensional curved surface model of the product. Real-time association, the specific implementation process is as follows: The intelligent bonding system is based on the product dimensions and surface curvature distribution output in the first step. Based on a pre-set membrane material database, the system automatically selects protective film specifications that meet the coverage requirements and have a matching elongation. During the conveying stage, a constant tension conveying mechanism is used and a reference tension is set. reference tension The elastic modulus of the membrane material and the extreme curvature of the product are calculated using an algorithm to prevent tensile deformation; simultaneously, the visual positioning system acquires the reference edge position of the protective film in real time and compares it with... Matching is performed, and the pose deviation between the current attitude of the membrane and the target bonding surface is calculated using a spatial coordinate transformation algorithm. The control system drives a six-degree-of-freedom fine-tuning platform to compensate for the membrane's position and attitude errors in real time. Ensure the film is aligned with the product surface and free of wrinkles. Among other things, To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; This refers to the rotational angle deviation of the protective film around the X-axis; This refers to the rotational angle deviation of the protective film around the Y-axis; This is to protect the film from rotational angle deviation around the Z-axis.

[0049] At the initial bonding point, the flexible pressing mechanism performs pre-compression with controlled pressure and speed, causing the film to move from the initial point along... The film is uniformly bonded in the normal direction. Throughout the process, the contact force feedback between the film and the curved surface is monitored in real time to dynamically adjust the bonding path, and the output is the coordinates of the bonding start point. and membrane materials Aligned standardized control parameter set.

[0050] Through the above technical solution, this application provides a quantitative and comprehensive real-time pose deviation compensation mechanism. This mechanism, by clearly defining the positional deviations of the protective film in the X, Y, and Z axes and the rotational angle deviations around these axes, incorporates them into the compensation formula, achieving precise control of the six-degree-of-freedom pose of the protective film. This enables the intelligent bonding system to dynamically adjust based on actual measurement data, avoiding errors caused by subjective judgment in traditional methods and significantly improving the reliability and accuracy of compensation. During the bonding process, the system can promptly correct minute pose deviations, effectively preventing error accumulation, thereby ensuring precise alignment between the protective film and the surface of the product to be processed without stretching or wrinkles. This comprehensive compensation capability greatly improves the bonding quality and consistency of the protective film, especially for products with complex curved surfaces or large curvatures. It effectively solves problems such as film stretching, wrinkles, and positional misalignment that easily occur in traditional bonding methods, providing a high-precision foundation for subsequent intelligent laser etching.

[0051] Furthermore, in step S3 above, the input is the three-dimensional reference model output in step S1. and user-provided two-dimensional graphic data Two-dimensional graphic data is transformed using a spatial coordinate transformation algorithm based on UV parameterization. Mapping to a 3D baseline model Above, generate the initial three-dimensional etching path point set. The three-dimensional reference model is a high-precision digital representation of the surface of the product to be processed, containing key information such as the product's geometry, dimensions, unevenness, and curvature distribution. This model provides a precise geometric reference for subsequent operations such as protective film lamination, two-dimensional graphic mapping, and laser etching. The two-dimensional graphic data provided by the user is a digital representation of the patterns or text that the user expects to etch onto the surface of the product. This data is typically in the form of vector graphics (e.g., DXF, SVG format) or high-resolution bitmaps (e.g., PNG, BMP format), defining the original geometric information and topological structure of the pattern to be etched. The UV-parameterized spatial coordinate transformation algorithm is the core technology for achieving accurate mapping from two-dimensional graphics to three-dimensional surfaces. UV parametricization is a technique that maps a three-dimensional surface locally or globally to a two-dimensional parameter space (usually the UV plane). This algorithm utilizes the property that each point on the surface can be uniquely determined by a pair of UV parameters, operating the two-dimensional graphic data in the UV parameter space and then mapping it back to the three-dimensional surface. Through the aforementioned UV-parameterized spatial coordinate transformation algorithm, the initial three-dimensional etching path point set is finally generated. This point set is a series of discrete three-dimensional coordinate points formed by precisely mapping two-dimensional graphic data. These points together constitute the trajectory that the etching laser head should follow on the three-dimensional curved surface. This point set serves as the basic input for subsequent etching path correction, smoothing, and laser parameter adjustment, providing precise geometric path guidance for laser etching.

[0052] Through the above technical solution, this application effectively solves the problems of graphic distortion and accuracy loss that may occur when mapping two-dimensional graphic data to a three-dimensional reference model. Specifically, by introducing a spatial coordinate transformation algorithm based on UV parameterization, the precise surface information contained in the three-dimensional reference model can be fully utilized to seamlessly and geometrically map two-dimensional graphic data onto complex three-dimensional surfaces. This mapping method avoids graphic stretching, compression, or distortion caused by traditional simple projection or manual programming, ensuring that the original geometric shape of the etching pattern on the surface is accurately preserved. The resulting initial three-dimensional etching path point set has high accuracy and low distortion, laying a solid foundation for subsequent etching path segmentation correction, smoothing processing, and adaptive adjustment of laser power and scanning speed. This significantly improves the edge accuracy and overall quality of the surface graphic etching, ensuring that the etching effect of the final product meets the design requirements.

[0053] Furthermore, two-dimensional graphic data Mapping to a 3D baseline model The mapping error is: ; in, The maximum mapping error after mapping a two-dimensional graphic to a three-dimensional surface; To make the second two-dimensional graphic a graphic point The three-dimensional coordinate points obtained after projection transformation; The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions; The preset allowable threshold for graphic mapping error can be set according to the actual processing accuracy requirements, with a range of 0.005 mm - 0.05 mm.

[0054] Through the above technical solution, this application introduces a precise mapping error quantification mechanism in the process of mapping two-dimensional graphic data onto a three-dimensional reference model to generate an initial three-dimensional etching path point set. By calculating the maximum mapping error after mapping the two-dimensional graphic to the three-dimensional surface and comparing the three-dimensional coordinate points after projection transformation with the three-dimensional coordinate points of the three-dimensional reference model at the corresponding UV parameter positions, the system can evaluate the mapping accuracy in real time and objectively. This quantization method makes the mapping process no longer a "black box" operation, but a monitorable and optimizable one. When the mapping error is detected to exceed a preset threshold, the system can adjust the mapping parameters or algorithm in a timely manner, thereby effectively avoiding the etching path from deviating from the actual surface position, significantly eliminating graphic distortion, and greatly improving the edge accuracy of the etching. This not only solves the problem of graphic distortion caused by the lack of a precise quantization mechanism in traditional methods, but also ensures the foundation for high-precision, distortion-free etching of complex surfaces by the subsequent intelligent laser etching system, providing a solid guarantee for the final realization of high-quality surface graphic etching.

[0055] In step S3 above, the density of path points is dynamically adjusted according to the curvature of the surface, and the spacing between basic points is set. At the radius of curvature of the surface Less than the threshold The area uses encryption factors. Reduce the point spacing to Simultaneously, cubic spline interpolation is used to smooth the path, ensuring the laser head's trajectory is continuous and distortion-free. The corrected path point set is denoted as... .

[0056] During the etching parameter generation stage, the intelligent laser etching system determines the etching parameters based on the curvature of each path segment. Real-time adjustment of laser power and scanning speed To maintain a consistent etching depth, Using an empirical formula: ; in, The laser frequency, The pulse width. and For material constants, Target etching depth; This represents the actual etching depth. During the etching process, the intelligent laser etching system monitors the width of the etched lines in real time. And edge sharpness, when width deviation is detected At the same time, a PID algorithm is used to fine-tune the etching power and speed, forming a closed-loop feedback control, and the output is the optimized three-dimensional etching path. and the corresponding laser parameter sequence It is stored in standard CNC G-code format to ensure the accuracy and consistency of pattern etching.

[0057] Through the above technical solution, this application effectively solves the problems of insufficient accuracy and uneven depth caused by curvature changes in complex surface etching. First, by dynamically adjusting the path point density according to the surface curvature, it ensures denser path points in high-curvature regions, thus more accurately conforming to surface changes and avoiding etching distortion caused by sparse path points, significantly improving the geometric accuracy of etching. Second, in the etching parameter generation stage, the laser power and scanning speed are adjusted in real time according to the curvature of each path segment, and empirical formulas are used to maintain consistent etching depth. This scientific calculation and adaptive parameter adjustment effectively compensates for the uneven energy distribution in surface etching, ensuring uniform etching depth in different curvature regions. Furthermore, during the etching process, by real-time monitoring of the etching line width and edge sharpness, and using a PID algorithm for closed-loop feedback control, immediate detection and correction of etching deviations are achieved, greatly improving the accuracy of the etching edges and overall stability, effectively avoiding etching quality degradation caused by external disturbances or material inhomogeneity. Finally, the optimized 3D etching path and corresponding laser parameter sequence are output and stored in standard CNC G-code format. This not only facilitates subsequent execution and ensures process consistency but also provides a basis for quality traceability. These technical features work synergistically to significantly enhance the adaptability and control precision of the laser etching process, making it possible to achieve high-precision, distortion-free pattern etching on complex curved surfaces.

[0058] In step S4 above, after the etching process is completed, the intelligent assisted excess film removal system uses a multi-view industrial camera to acquire multi-view images of the product surface with the protective film, obtaining an image sequence of the etched area. The complete film surface image is reconstructed using an image stitching algorithm, and the cutting boundaries formed by laser etching are extracted using an edge detection operator to construct a binary mask of the etching contour. Binarization mask Compared with the theoretical three-dimensional etching path generated in step S3 Perform registration and comparison, calculate the positional deviation of the etching boundary, and ensure that all cutting lines are completed and without breakage.

[0059] During the redundant membrane region identification stage, the system will use a binarized mask. The enclosed area within the shape is defined as the reserved shape area. The entire area covered by the protective film minus The resulting collection of excess film areas that need to be removed is then obtained. .

[0060] For the collection of excess membrane regions The intelligent assisted excess membrane removal system performs connected component analysis and divides it into... A separate waste block ( Each block is equipped with geometric feature parameters, including the center coordinates, area, and boundary length of the minimum bounding rectangle. The image coordinates are converted to the physical coordinates corresponding to the product's 3D reference model through coordinate system calibration, and the conversion error is controlled within ≤0.2mm.

[0061] To achieve intelligent guidance, the system uses laser projection devices or augmented reality displays embedded above the product workstations to... The outline is projected or overlaid on the corresponding position of the product in a bright, dynamic flashing manner. The projection color can be set to green to indicate the area to be torn off. Simultaneously, to assist the operator in quickly removing the film, the system automatically recommends the optimal tearing starting point based on the geometric characteristics of each waste block. The recommendation logic is based on boundary convex hull analysis: if the block... If a point of extreme curvature or a sharp corner exists, that point is selected as the starting point; if no significant sharp corner exists, the point on its boundary closest to the edge of the workstation is selected as the starting point. The starting point position is indicated by a red arrow or a light dot in the guidance interface, and the number and area of ​​the area to be torn are displayed in real time. Furthermore, the system can integrate voice prompts to guide the operator step-by-step through the tearing process. Each block is processed, and the removal progress is monitored in real time. If an area is not completely removed, a visual reminder is issued. The output of this step is a guide file containing the physical coordinates, contour geometry information, and recommended starting points of all excess membrane areas. It is used to achieve efficient and complete tearing operations.

[0062] Based on the aforementioned laser etching system and method, this application further proposes an intelligent assisted system for removing excess film. By introducing automated image processing and data analysis, it effectively solves the problems of low efficiency, error-proneness, and inaccurate judgment of etching integrity associated with traditional manual film removal. Specifically, after the etching process is completed, the system uses a multi-view industrial camera to acquire images. Multi-angle coverage ensures comprehensive perception of the etched area on the product surface, avoiding omissions or obstructions that may occur with a single view. Subsequently, an image stitching algorithm seamlessly integrates these local images into a complete film surface image, providing a global view for subsequent precise analysis. Next, an edge detection operator extracts the cutting boundaries formed by laser etching and constructs a binary mask, achieving a precise digital representation of the actual etching contour. By registering and comparing this binary mask with the theoretical three-dimensional etching path, the system can automatically calculate the positional deviation of the etching boundaries, thereby verifying in real time whether all cutting lines are complete and unbroken. This mechanism significantly improves the efficiency and accuracy of etching quality detection, avoids errors and omissions from manual visual inspection, and ensures product quality. In the excess film area identification stage, the system accurately defines the retention area based on a binary mask and separates the set of excess film areas to be removed from the entire protective film coverage area through set operations. Furthermore, connected component analysis is used to segment these excess film areas into independent waste blocks, and each block is assigned geometric feature parameters, such as the center coordinates, area, and boundary length of the minimum bounding rectangle. These parameters provide a quantitative basis for subsequent film removal strategy formulation and operation guidance, making the film removal process more targeted and intelligent. Finally, through high-precision coordinate system calibration, the two-dimensional coordinates of the waste blocks identified in the image are accurately converted to the physical coordinates corresponding to the product's three-dimensional reference model, with the conversion error controlled to ≤0.2mm. This high-precision physical coordinate positioning, combined with subsequent projection or voice prompts, accurately guides the operator to find the optimal film removal starting point and sequence, greatly improving the efficiency and accuracy of film removal, reducing the risk of missed or incorrect removal of waste, and thus improving the overall automation level and product yield.

[0063] The following example will provide a more detailed explanation of the above technical solution: A multi-data fusion laser pattern etching method for protective film lamination is disclosed. A large-size curved panel is used as the workpiece, with dimensions of 320mm × 220mm × 12mm. The workpiece has a hyperboloid structure with a minimum edge curvature radius of 30mm and a planar transition in the central area. A 0.15mm thick flexible PET protective film is selected, possessing a tensile strength ≥150MPa, an elongation ≤2%, and a high temperature resistance of 100℃. The protective film lamination and laser pattern etching are performed according to this method.

[0064] The first step involves acquiring and modeling 3D data of the entire curved surface of the product. A non-contact 3D vision scanning system is used to acquire point cloud data of the entire workpiece at a scanning resolution of 0.01mm. The geometric features of planes, curved surfaces and transition areas are automatically identified. The acquired point cloud is iteratively registered with the standard CAD template of the vehicle center console to remove noise points and outliers caused by curvature abrupt changes. A 3D reference model that perfectly matches the workpiece is constructed using the NURBS surface fitting algorithm. The model contains complete information on surface curvature distribution, size and concavity / convexity changes, forming a standardized database to provide accurate data support for subsequent bonding control and laser path planning of large-size workpieces. The second step involves intelligent lamination of the large-size protective film based on the three-dimensional reference model. The system matches the required PET protective film according to the workpiece's size of 320mm×220mm and the curvature distribution of the entire surface. The film is conveyed by a constant tension conveying mechanism based on the reference tension calculated according to the elastic modulus of the film and the curvature of the surface, avoiding loosening or stretching deformation of the large-size film. The vision positioning system collects the relative pose data between the reference edge of the protective film and the curved surface of the workpiece in real time. The six-degree-of-freedom fine-tuning platform compensates for position and angle deviations in real time to ensure that the protective film is precisely aligned with the surface of the large-size workpiece. The flexible pressing mechanism starts from the center of the workpiece and gradually pre-presses outward along the normal direction of the curved surface at a constant pressure of 8N and a speed of 15mm / s, so that the protective film is uniformly, wrinkle-free, and bubble-free, and completely adheres to the surface of the curved panel. The system outputs the coordinates of the lamination start point and a standardized set of control parameters. The third step involves automatic laser etching and path planning for large curved surfaces. The input 40mm×25mm brand logo and 280mm×180mm functional area border 2D vector graphics are mapped to the 3D reference model of the curved panel through a UV parametric spatial coordinate transformation algorithm. Combined with the curvature distribution of the surface, the etching path is segmented and adaptively corrected. The path point density is increased in areas with large edge curvature, while the basic point spacing is maintained in planar areas. Cubic spline interpolation is used to smooth the path, eliminating graphic distortion caused by converting large-size planar surfaces to curved surfaces. At the same time, the laser power, scanning speed, laser frequency and pulse width are adjusted in real time according to the curvature of each segment to keep the etching depth uniform. During the etching process, the line width and edge integrity are monitored in real time by a coaxial vision sensor. The laser parameters are finely adjusted in a closed loop using a PID algorithm to generate a 3D etching path and parameter sequence in standard G-code format, completing high-precision, distortion-free etching of the surface graphics of large workpieces. The fourth step involves intelligent assisted removal of excess film from large-sized workpieces. A multi-view industrial camera captures a complete image of the etched protective film. The image is then stitched together and edge detection is used to extract the etched contour. The area to be retained and the area of ​​excess film waste are separated. The image coordinates of the waste area are converted into the physical coordinates corresponding to the three-dimensional reference model of the workpiece. The excess film contour is then projected directly onto the surface of the large-sized workpiece in a green dynamic highlight form using a laser projection device. The system automatically recommends the best tearing start point based on the geometric characteristics of each waste area, either at the sharp corner of the boundary or near the edge of the workpiece, and marks it with a red arrow. Voice prompts guide the operator to tear off the waste in sequence, while the tearing status is monitored in real time to ensure that no excess part of the large-sized protective film is missed or left behind. After completing the above steps, the entire workpiece, especially the curved grooves and corners, is first cleaned non-contactly to remove residual dust and tiny film debris from the laser etching. Then, the exposed surface to be coated is activated to improve surface energy and coating adhesion. After visual inspection confirms that there is no dust, no residual adhesive, and no protective film residue, the workpiece is sent to the coating station for surface coating to ensure that the coating is uniform, has clear edges, and is firmly bonded, meeting the appearance and performance requirements of large-sized curved workpieces.

[0065] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A laser patterning etching system with multi-data fusion, characterized in that, include: Non-contact 3D vision scanning system, intelligent bonding system, and intelligent laser etching system; The non-contact 3D vision scanning system is used to complete the acquisition and modeling of the full-area curved surface data of the product to be processed, construct a 3D reference model containing the surface features, size information, concavity and convexity changes and curvature distribution of the product to be processed, and store it in a standardized database format. The intelligent bonding system automatically selects the appropriate protective film specifications based on the three-dimensional reference model of the product to be processed, and aligns the protective film with the surface of the product to be processed, achieving a bonding of the protective film without stretching or wrinkles. The intelligent laser etching system maps two-dimensional graphics onto a three-dimensional surface, eliminates graphic distortion through curvature adaptive path planning of the three-dimensional surface, and adjusts laser parameters in real time to ensure uniform etching depth and edge accuracy, thus completing distortion-free etching of the surface graphics of the product to be processed.

2. The laser patterning etching system with multi-data fusion according to claim 1, characterized in that: It also includes an intelligent assisted excess film removal system, which, after the intelligent laser etching system has completed etching, uses an industrial camera to capture images, identify and segment the waste areas that need to be removed, and maps their physical coordinates onto a three-dimensional reference model. Then, the waste areas are projected onto the surface of the product to be processed in a highlighted form. Based on the geometric characteristics of each waste block, the system recommends the best film removal starting point and marks it, guiding the operator to remove the waste areas in sequence.

3. A laser pattern etching method with multi-data fusion, characterized in that: Includes the following steps: S1. Perform full-area surface 3D data acquisition and modeling processing: acquire full-area point cloud data of the product to be processed through a non-contact 3D vision scanning system, iteratively register the acquired point cloud data with the CAD template of the product to be processed, filter out outliers, and then use the NURBS surface fitting algorithm to construct an accurate 3D benchmark model from the processed point cloud data. S2. The intelligent bonding system utilizes the product size and surface curvature distribution information contained in the three-dimensional reference model to automatically select the most suitable protective film from the film material database. It calculates and sets the reference tension based on the elastic modulus of the protective film and the extreme value of the product curvature. At the same time, it collects the reference edge position of the protective film in real time and matches it with the three-dimensional reference model. It calculates the pose deviation between the protective film and the target bonding surface of the product to be processed through a spatial coordinate transformation algorithm. It performs real-time compensation based on the calculated pose deviation to ensure that the protective film is accurately aligned with the surface of the product to be processed. S3. The intelligent laser etching system receives a three-dimensional reference model and two-dimensional graphic data input by the user. It imports the two-dimensional graphic data into the intelligent laser etching system and uses the UV parametric spatial coordinate transformation algorithm to accurately map the two-dimensional graphic data onto the three-dimensional reference model, generating an initial three-dimensional etching path point set. Based on the curvature distribution in the three-dimensional reference model, the etching path is segmented and corrected, and smoothed by cubic spline interpolation. The laser power and scanning speed are adjusted in real time according to the curvature of each segment of the path. The width of the etching lines and the edge sharpness are monitored in real time, and the laser parameters are fine-tuned using a PID algorithm to form a closed-loop control, ultimately completing the etching of the curved surface graphic. S4. After etching is completed, the intelligent assisted excess film removal system acquires images through a multi-view industrial camera, identifies and segments the waste area that needs to be removed, maps its physical coordinates onto a three-dimensional reference model, highlights the waste outline on the surface of the product to be processed, recommends the best removal starting point, and guides the operator to remove the excess film with voice prompts.

4. The laser pattern etching method with multi-data fusion according to claim 3, characterized in that: In step S1 above, the collected point cloud data is iteratively registered with the CAD template of the product to be processed, and the registration error is calculated using the iterative nearest point algorithm. ； Where N is the number of point clouds; The three-dimensional coordinate point is obtained by projecting the i-th point on the two-dimensional graphic. The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions; The preset allowable threshold for graphic mapping error.

5. The laser pattern etching method with multi-data fusion according to claim 4, characterized in that: In step S1 above, a three-dimensional benchmark model is constructed using the NURBS surface fitting algorithm, and the fitting accuracy is measured by the root mean square error. ； in, A three-dimensional baseline model; The preset allowable threshold for surface fitting accuracy.

6. The laser pattern etching method with multi-data fusion according to claim 3, characterized in that: In step S2 above, real-time compensation is performed based on the calculated pose deviation: ； in, To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; To protect the positional deviation of the film in the X-axis direction; This refers to the deviation of the protective film's rotation angle around the X-axis; This refers to the rotational angle deviation of the protective film around the Y-axis; This is to protect the film from rotational angle deviation around the Z-axis.

7. The laser pattern etching method with multi-data fusion according to claim 5, characterized in that: In step S3 above, the input is the three-dimensional reference model output in step S1. and user-provided two-dimensional graphic data Two-dimensional graphic data is transformed using a spatial coordinate transformation algorithm based on UV parameterization. Mapping to a 3D baseline model Above, generate the initial three-dimensional etching path point set. .

8. The laser patterning etching method with multi-data fusion according to claim 7, characterized in that: The two-dimensional graphic data Mapping to a 3D baseline model The mapping error is: ； in, The maximum mapping error after mapping a two-dimensional graphic to a three-dimensional surface; To make the second two-dimensional graphic a graphic point The three-dimensional coordinate points obtained after projection transformation; The three-dimensional coordinates of the three-dimensional reference model of the product to be processed at the corresponding UV parameter positions.

9. The laser pattern etching method with multi-data fusion according to claim 3, characterized in that: In step S3 above, the density of path points is dynamically adjusted according to the curvature of the surface, and the spacing between basic points is set. At the radius of curvature of the surface Less than the threshold The area uses encryption factors. Reduce the point spacing to ; During the etching parameter generation stage, the intelligent laser etching system determines the etching parameters based on the curvature of each path segment. Real-time adjustment of laser power and scanning speed To maintain a consistent etching depth, Using an empirical formula: ； in, The laser frequency, The pulse width. and For material constants, Target etching depth; This represents the actual etching depth. During the etching process, the intelligent laser etching system monitors the width of the etched lines in real time. And edge sharpness, when width deviation is detected At the same time, a PID algorithm is used to fine-tune the etching power and speed, forming a closed-loop feedback control, and the output is the optimized three-dimensional etching path. and the corresponding laser parameter sequence It is stored in standard CNC G-code format.

10. The laser patterning etching method with multi-data fusion according to claim 9, characterized in that: In step S4 above, after the etching process is completed, the intelligent assisted excess film removal system uses a multi-view industrial camera to acquire multi-view images of the product surface with the protective film, obtaining an image sequence of the etched area. The complete film surface image is reconstructed using an image stitching algorithm, and the cutting boundaries formed by laser etching are extracted using an edge detection operator to construct a binary mask of the etching contour. Binarization mask Compared with the theoretical three-dimensional etching path generated in step S3 Perform registration and comparison, calculate the positional deviation of the etching boundary, and ensure that all cutting lines are completed and without breakage; During the redundant membrane region identification stage, the system will use a binarized mask. The enclosed area within the shape is defined as the reserved shape area. The entire area covered by the protective film minus The resulting collection of excess film areas that need to be removed is then obtained. ; For the collection of excess membrane regions The intelligent assisted excess membrane removal system performs connected component analysis and divides it into... A separate waste block ( Each block is equipped with geometric feature parameters, including the center coordinates, area, and boundary length of the minimum bounding rectangle. The image coordinates are converted to the physical coordinates corresponding to the product's 3D reference model through coordinate system calibration, and the conversion error is controlled to ≤0.2mm.

Images