Multi-brand automatic laser etching method and system combined with industrial vision, and medium

Abstract

The application discloses a multi-specification automatic laser etching method and system combined with industrial vision, and a medium, relates to the technical field of laser etching, and the method comprises the following steps: obtaining a product to be laser etched of an automatic laser etching machine; generating laser etching content and controlling and excavating a laser etching module in combination with product feature data; performing multi-dimensional risk prediction on a laser etching analysis scheme according to a laser etching risk predictor; if the laser etching risk prediction result does not satisfy a laser etching risk constraint, performing variation adjustment, establishing a candidate laser etching group; introducing a laser etching global optimization mechanism to perform optimization analysis and determine a laser etching optimization strategy; conveying the product to be laser etched to a predetermined position and synchronously starting a laser etching module to execute the laser etching optimization strategy. The technical problems of low laser etching efficiency of multi-specification products, lack of risk assessment optimization of a laser etching scheme and poor quality consistency of laser etched products in the prior art are solved, and the technical effects of improving the laser etching automation and production efficiency of multi-specification products and the quality consistency and stability of laser etched products are achieved.

Description

Technical Field

[0001] This application relates to the field of laser engraving technology, specifically to a multi-grade automated laser engraving method, system, and medium that combines industrial vision. Background Technology

[0002] Laser engraving technology is widely used in product identification and information engraving. With the increasing diversification of market demand, enterprises often involve the simultaneous processing of multiple product brands. During the laser engraving process, due to differences in the materials of different products and the complexity of the laser engraving content, there are risks such as uneven laser engraving depth and surface damage. Traditional laser engraving usually relies on manual programming and manual adjustment, which is difficult to meet the needs of large-scale and rapid production. Moreover, it is prone to problems such as errors in laser engraving content and positional deviations caused by operational errors, which seriously affect product quality and production efficiency. How to deeply integrate industrial vision, motion control and other technologies with laser engraving technology to achieve automated and intelligent laser engraving of multiple product brands has become a key issue in improving the production efficiency and product quality of the manufacturing industry.

[0003] Therefore, the current related technologies have technical problems such as low laser engraving efficiency for multiple product brands, lack of risk assessment and optimization of laser engraving solutions, and poor consistency in the quality of laser engraved products. Summary of the Invention

[0004] This application provides a multi-brand automated laser engraving method, system, and medium that combines industrial vision, solving the technical problems of low laser engraving efficiency, lack of risk assessment and optimization of laser engraving solutions, and poor consistency of laser engraved product quality in the prior art. It achieves the technical effect of improving the automation and production efficiency of multi-brand laser engraving, as well as the consistency and stability of product laser engraving quality.

[0005] This application provides a multi-brand automated laser engraving method combining industrial vision. The method includes: obtaining a product to be laser engraved by an automated laser engraving machine, the automated laser engraving machine including an industrial vision module, a motion control module, and a laser engraving module, the product to be laser engraved including multi-brand features; generating laser engraving content based on the multi-brand features; controlling and mining the laser engraving module based on the product feature data of the product to be laser engraved to generate a laser engraving analysis scheme; performing multi-dimensional risk prediction on the laser engraving analysis scheme based on a laser engraving risk predictor to determine the laser engraving risk prediction result; if the laser engraving risk prediction result does not meet the laser engraving risk constraint, performing mutation adjustment on the laser engraving analysis scheme based on the laser engraving risk predictor to establish a candidate laser engraving group; introducing a laser engraving global optimization mechanism to perform optimization analysis on the candidate laser engraving group to determine the laser engraving optimization strategy; transporting the product to be laser engraved to a predetermined position according to the industrial vision module and the motion control module, and synchronously starting the laser engraving module to execute the laser engraving optimization strategy.

[0006] In one possible implementation, the multi-brand automated laser engraving method combining industrial vision further performs the following processing: searching for laser engraving control schemes in the laser engraving module based on the laser engraving content and the product feature data to obtain a searched laser engraving scheme set; classifying variables based on the searched laser engraving scheme set to obtain multiple laser engraving parameter regions; evaluating the confidence level of each laser engraving parameter region to obtain each laser engraving confidence evaluation set; and optimizing the multiple laser engraving parameter regions by maximizing laser engraving confidence based on each laser engraving confidence evaluation set to generate the laser engraving analysis scheme.

[0007] In one possible implementation, the multi-grade automated laser engraving method combined with industrial vision further performs the following processing: setting laser engraving defect factors, which include basic laser engraving defects, key laser engraving defects, and process laser engraving defects; based on the laser engraving defect factors, performing defect fitting according to the laser engraving analysis scheme to obtain a laser engraving defect feature sequence, which includes basic defect fitting features, key defect fitting features, and process defect fitting features; inputting the laser engraving defect feature sequence into the laser engraving risk predictor to obtain the laser engraving risk prediction result, wherein the laser engraving risk predictor includes a basic defect risk analysis model, a key defect risk analysis model, and a process defect risk analysis model, and the laser engraving risk prediction result includes a basic defect risk coefficient, a key defect risk coefficient, and a process defect risk coefficient.

[0008] In one possible implementation, the multi-grade automated laser engraving method combining industrial vision further performs the following processing: modeling based on the laser engraving module to obtain a laser engraving model; performing laser engraving fitting on the product to be laser engraved based on the laser engraving model and the laser engraving analysis scheme to obtain a fitted laser engraved product; performing image unfolding on the fitted laser engraved product to obtain a fitted laser engraving image; performing deep learning on the laser engraving defect factors using a convolutional neural network to generate a laser engraving defect detection network, the laser engraving defect detection network including a basic defect detection network, a key defect detection network, and a process defect detection network; and inputting the fitted laser engraving image into the laser engraving defect detection network to generate the laser engraving defect feature sequence.

[0009] In one possible implementation, the multi-grade automated laser engraving method combined with industrial vision further performs the following processing: based on each laser engraving confidence evaluation set, multiple laser engraving parameter regions are selected according to the confidence evaluation threshold to establish multiple laser engraving confidence value sets; interval analysis is performed based on the multiple laser engraving confidence value sets to obtain adjustment constraints for each laser engraving variable; the laser engraving analysis scheme is mutated according to the adjustment constraints for each laser engraving variable to obtain a laser engraving adjustment scheme set; based on the laser engraving risk predictor, the laser engraving adjustment scheme set is traversed and optimized according to the laser engraving risk constraints to generate the candidate laser engraving group.

[0010] In one possible implementation, the multi-grade automated laser engraving method combined with industrial vision further performs the following processing: extracting the nth laser engraving adjustment scheme from the laser engraving adjustment scheme set, where n is a positive integer; fitting laser engraving defects to the product to be laser engraved according to the nth laser engraving adjustment scheme to obtain the nth scheme defect feature sequence; inputting the nth scheme defect feature sequence into the laser engraving risk predictor to obtain the nth scheme laser engraving risk feature; determining whether the nth scheme laser engraving risk feature satisfies the laser engraving risk constraint; if the nth scheme laser engraving risk feature satisfies the laser engraving risk constraint, setting the nth laser engraving adjustment scheme as the nth candidate laser engraving scheme, and adding the nth candidate laser engraving scheme to the candidate laser engraving group.

[0011] In one possible implementation, the multi-grade automated laser engraving method combined with industrial vision further performs the following processing: weight analysis is performed on the multi-dimensional laser engraving risk indicators of the laser engraving risk predictor to obtain a global laser engraving risk analysis function, wherein the multi-dimensional laser engraving risk indicators include basic defect risk indicators, key defect risk indicators, and process defect risk indicators; global risk calculation is performed on the candidate laser engraving group according to the global laser engraving risk analysis function to obtain a global laser engraving risk distribution; global risk minimization optimization is performed on the candidate laser engraving group according to the global laser engraving risk distribution to generate the laser engraving optimization strategy.

[0012] In one possible implementation, the multi-grade automated laser engraving method combining industrial vision also performs the following processing: the multi-grade features include model features, grade features, batch features, and customized identification features.

[0013] This application also provides a multi-grade automated laser engraving system combining industrial vision. The system includes: a multi-dimensional feature information acquisition module for acquiring multi-dimensional feature information of low-voltage busbar trunking and using the multi-dimensional feature information as a constraint; a real-time operation information acquisition module for dynamically monitoring the low-voltage busbar trunking under the constraint conditions to obtain real-time operation information; a real-time operation fitness acquisition module for introducing a collaborative fitness comprehensive evaluation function to evaluate and analyze the real-time operation information to obtain real-time operation fitness; and a constraint condition adjustment and optimization module for adjusting and optimizing the constraint conditions if the real-time operation fitness does not reach a predetermined fitness threshold.

[0014] This application also provides a computer-readable storage medium, including: a computer program stored thereon that, when executed by a processor, implements a multi-grade automated laser engraving method incorporating industrial vision.

[0015] This application proposes a multi-brand automated laser engraving method, system, and medium combining industrial vision to obtain the product to be laser engraved by an automated laser engraving machine; generate laser engraving content and control and mine the laser engraving module based on product feature data; perform multi-dimensional risk prediction on the laser engraving analysis scheme according to a laser engraving risk predictor; if the laser engraving risk prediction result does not meet the laser engraving risk constraints, perform mutation adjustment and establish a candidate laser engraving group; introduce a global laser engraving optimization mechanism for optimization analysis and determine the laser engraving optimization strategy; transport the product to be laser engraved to a predetermined position and simultaneously start the laser engraving module to execute the laser engraving optimization strategy. This solves the technical problems of low efficiency in multi-brand laser engraving, lack of risk assessment and optimization of laser engraving schemes, and poor consistency in laser engraved product quality in existing technologies, achieving the technical effect of improving the automation and production efficiency of multi-brand laser engraving, and the consistency and stability of product laser engraving quality. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0017] Figure 1 This is a schematic diagram of a multi-grade automated laser engraving method combining industrial vision, provided in an embodiment of this application.

[0018] Figure 2 A schematic diagram of the structure of a multi-grade automated laser engraving system combining industrial vision, provided in an embodiment of this application.

[0019] Figure labeling: Unit 10 for obtaining laser-engraved product, Unit 20 for generating laser engraving analysis scheme, Unit 30 for multi-dimensional risk prediction, Unit 40 for establishing candidate laser engraving group, Unit 50 for determining laser engraving optimization strategy, and Unit 60 for executing laser engraving optimization strategy. Detailed Implementation

[0020] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below.

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" can be the same or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0023] This application provides an automated laser engraving method for multiple grades that combines industrial vision, such as... Figure 1 As shown, the method includes:

[0024] Step S100: Obtain the product to be laser-engraved by the automated laser engraving machine. The automated laser engraving machine includes an industrial vision module, a motion control module, and a laser engraving module. The product to be laser-engraved includes multiple brand features.

[0025] Step S100 further includes the following: the multiple grade characteristics include model characteristics, grade characteristics, batch characteristics, and customized identification characteristics.

[0026] Preferably, the multi-brand characteristics of the product to be laser-engraved refer to the differentiated identification information of the product in dimensions such as model, grade, batch, and personalized customization. Specifically, these include model characteristics, grade characteristics, batch characteristics, and customized identification characteristics. Among them, the model characteristic refers to the standardized model identifier of the product, used to distinguish different types, functions, or specifications of products. It is the basic identifier of the product identity, and the accuracy and clarity of the model characters must be ensured during laser engraving. The grade characteristic indicates the quality grade, performance grade, or usage grade of the product, helping users quickly determine the applicable scenarios of the product (such as food contact, high-strength mechanical structures, etc.). The laser-engraved grade identifier must comply with industry standards to avoid quality accidents caused by incorrect identification. The batch characteristic refers to the unique identifier of the product's production batch, which usually includes information such as production date, production line number, and batch number. The laser-engraved batch information must be unique and readable to meet the traceability needs of the upstream and downstream of the supply chain. The customized identification characteristic refers to the non-standardized identifier customized according to the customer's personalized needs, which may include graphics / logos, text information, QR codes, barcodes, etc. The laser-engraved customized identifier must be completely consistent with the customer's needs to ensure the accuracy and aesthetics of the identifier.

[0027] Preferably, the automated laser engraving machine works collaboratively with an industrial vision module, a motion control module, and a laser engraving module to achieve accurate identification, positioning, and laser engraving of multiple product brands. Specifically, the industrial vision module is used for product identification, feature extraction, and position calibration. It acquires images of the product to be laser engraved using an industrial camera (such as a CCD / CMOS camera), obtaining visual information such as the product's shape, size, surface features (such as texture and markers), and multiple brand identifiers (such as model, batch, and serial number). It analyzes the images using image processing algorithms (such as image filtering, edge detection, feature matching, and OCR text recognition), extracts key feature data of the product (such as brand type and laser engraving position reference point), and automatically matches the laser engraving content (such as text, patterns, and QR codes) corresponding to the product model by combining it with a preset product database, and generates laser engraving coordinate information.

[0028] Preferably, the motion control module is used to control the movement trajectory of the product and the laser engraving head to achieve precise positioning and dynamic processing. Based on the coordinate information output by the industrial vision module, the motion controller (such as PLC, motion control card) drives the actuator (such as servo motor, guide rail, lead screw) to transport the product to be laser engraved to the predetermined processing position (such as the coordinate origin of the laser engraving station). During the laser engraving process, the moving speed and direction of the product or the position of the laser engraving head are adjusted in real time according to the laser engraving analysis scheme (such as XYZ three-axis linkage) to ensure that the laser engraving path matches the product features (such as engraving along curved or arc surfaces). It supports multi-axis collaborative motion to meet the laser engraving needs of complex shaped products (such as cylinders and irregular parts).

[0029] Preferably, the laser engraving module is used to perform laser engraving operations. It mainly consists of a laser (such as a fiber laser or an ultraviolet laser), optical components (such as a galvanometer and a field lens), and control components. According to the laser engraving analysis scheme (such as parameters such as power, frequency, scanning speed, and engraving depth), the laser emits a high-energy laser beam, which is focused onto the product surface through the optical system, causing physical or chemical changes in the material (such as evaporation, carbonization, and discoloration), thereby forming the predetermined laser engraving content, such as text, patterns, and anti-counterfeiting marks. It supports the processing of multiple types of materials, such as metals (aluminum, copper, stainless steel), plastics, ceramics, and glass. By adjusting the laser parameters, it can adapt to the laser engraving needs of different materials, achieve non-contact processing, avoid damage to the product surface, and ensure high-speed and high-precision engraving to meet the efficiency requirements of mass production.

[0030] Step S200: Generate laser engraving content based on the multi-brand characteristics, and control and mine the laser engraving module in conjunction with the product characteristic data of the product to be laser engraved to generate a laser engraving analysis scheme.

[0031] Preferably, the industrial vision module extracts key features (such as model identification, batch number, and outline) from the product to be laser-engraved through image recognition and matches them with a preset product database. For different brands of products, it automatically calls the corresponding laser engraving content template and automatically generates laser engraving content, such as serial number and anti-counterfeiting code, according to preset business rules. At the same time, it automatically adjusts the font, size, and layout of the laser engraving content according to the product surface size and shape. It obtains product feature data of the product to be laser-engraved, which may include material properties and surface condition, and then uses the product feature data to control and mine the laser engraving module. That is, it automatically derives the optimal combination of laser engraving parameters, such as the combination of laser power, scanning speed, and laser frequency and pulse width, based on historical production data. It also monitors the working status of the laser engraving module in real time (such as laser temperature and galvanometer accuracy) and automatically compensates for equipment errors. Finally, based on the product's 3D model and laser engraving position, it calculates the motion path of the laser head or product, and optimizes the path to avoid fixtures and obstacles by combining the mechanical structure parameters of the motion control module, thus obtaining a laser engraving solution that significantly improves production flexibility and efficiency.

[0032] Furthermore, step S200 also includes step S210, performing a laser engraving control scheme retrieval on the laser engraving module based on the laser engraving content and the product feature data to obtain a retrieved laser engraving scheme set; step S220, classifying variables based on the retrieved laser engraving scheme set to obtain multiple laser engraving parameter areas; step S230, evaluating the confidence level of each laser engraving parameter area to obtain each laser engraving confidence evaluation set; and step S240, performing laser engraving confidence maximization optimization on the multiple laser engraving parameter areas based on each laser engraving confidence evaluation set to generate the laser engraving parsing scheme.

[0033] Preferably, the laser engraving content features (such as text / patterns, dimensional accuracy) and product feature data (material, surface curvature, reflectivity) are used as keywords to match and search for laser engraving control schemes in the historical scheme database. For example, when the product to be processed is "aluminum alloy mobile phone shell (thickness 1mm)" and the laser engraving content is "QR code (accuracy ≥3mil)", all historical laser engraving schemes that meet the conditions are automatically filtered to form a searchable laser engraving scheme set. Each historical scheme contains complete laser engraving control parameters (such as laser power, scanning speed, frequency, pulse width) and its corresponding processing effect evaluation (such as engraving depth, line width, surface roughness). Then, the variables are classified according to the searched laser engraving scheme set, that is, the parameters in the searched scheme set are classified to form multiple laser engraving parameter areas corresponding to multiple laser engraving control variables, such as laser parameter area (power, frequency, pulse width, duty cycle), motion parameter area (scanning speed, acceleration, galvanometer deflection angle), and process parameter area (defocusing amount, number of repetitions, filling spacing).

[0034] Preferably, a confidence evaluation is performed for each laser engraving parameter region. This involves statistically analyzing the frequency of each historical laser engraving control parameter in the search scheme set, using this frequency as the confidence coefficient corresponding to the historical laser engraving control parameter. A higher frequency indicates a more reliable parameter, thus forming a laser engraving confidence evaluation set for each parameter region, containing all historical laser engraving control parameters and their laser engraving confidence coefficients. Next, based on each laser engraving confidence evaluation set, a laser engraving confidence maximization optimization is performed for multiple laser engraving parameter regions. This involves using the product of the confidence coefficients of each parameter region as a comprehensive evaluation index to find the parameter combination that maximizes the product, thereby generating a laser engraving analysis scheme.

[0035] Step S300: Perform multi-dimensional risk prediction on the laser engraving analysis scheme based on the laser engraving risk predictor, and determine the laser engraving risk prediction result.

[0036] Step S300 further includes step S310, setting laser engraving defect factors, wherein the laser engraving defect factors include basic laser engraving defects, key laser engraving defects, and laser engraving process defects; step S320, based on the laser engraving defect factors, performing defect fitting according to the laser engraving analysis scheme to obtain a laser engraving defect feature sequence, wherein the laser engraving defect feature sequence includes basic defect fitting features, key defect fitting features, and process defect fitting features; step S330, inputting the laser engraving defect feature sequence into the laser engraving risk predictor to obtain the laser engraving risk prediction result, wherein the laser engraving risk predictor includes a basic defect risk analysis model, a key defect risk analysis model, and a process defect risk analysis model, and the laser engraving risk prediction result includes a basic defect risk coefficient, a key defect risk coefficient, and a process defect risk coefficient.

[0037] Preferably, laser engraving defect factors are defined as basic laser engraving defects, critical laser engraving defects, and laser engraving process defects. Basic laser engraving defects are common defects caused by equipment, materials, or basic operations, such as missing engraving, misalignment, and ghosting. The frequency of occurrence of each defect is statistically analyzed using historical production data, and weights are assigned for quantification. Critical laser engraving defects are defects that seriously affect the function or safety of the product, such as material damage and insufficient depth. High weights are assigned based on the degree of harm to product performance. Laser engraving process defects are defects caused by improper laser engraving parameters or process settings, such as unclear engraving and rough edges. Weights are assigned based on the correlation analysis between process parameters and defects.

[0038] Preferably, based on laser engraving defect factors, defect fitting is performed according to the laser engraving analysis scheme. That is, the laser engraving analysis scheme is combined with historical defect data to predict the types and degrees of defects that may occur in the current scheme, generating basic defect fitting features, key defect fitting features, and process defect fitting features, forming a laser engraving defect feature sequence. Specifically, laser engraving basic defect fitting compares the equipment parameters of the current scheme (such as galvanometer accuracy and laser head position) with the equipment status data of historical basic defects. This includes extracting cases of missed engraving and misalignment caused by equipment errors in historical data, analyzing the correlation between equipment parameters and defects (such as positional shift caused by galvanometer aging), comparing the equipment parameters of the current scheme with historical data, calculating the probability of basic defects occurring, and outputting basic defect fitting features, such as misalignment risk, high probability, possibly due to galvanometer calibration error.

[0039] Preferably, laser engraving critical defect fitting assesses the risk of material damage or insufficient depth based on product material, laser engraving depth requirements, and laser energy parameters. This includes establishing a material-laser parameter-defect relationship model, comparing the current solution's laser power, scanning speed, and material tolerance threshold, and outputting critical defect fitting characteristics. For example, material damage risk is considered medium-probability, and it is recommended to reduce the power to 12W. Laser engraving process defect fitting predicts the likelihood of process defects by combining process parameter combinations with historical engraving quality data. This includes analyzing historical data of cases with unclear engraving or rough edges, summarizing parameter combination patterns, comparing the current solution's parameters with historical defective parameters, and outputting process defect fitting characteristics. For example, unclear engraving risk is considered low-probability, but the fill spacing needs to be monitored.

[0040] Preferably, the laser engraving defect feature sequence is input into the laser engraving risk predictor. The laser engraving risk predictor converts the three types of defect fitting features into corresponding defect risk coefficients through a basic defect risk analysis model, a key defect risk analysis model, and a process defect risk analysis model. Specifically, a machine learning model (such as a decision tree or random forest) is trained based on historical basic defect data to obtain a basic defect risk analysis model. The basic defect fitting features are input for analysis, i.e., the basic defect risk coefficient is calculated based on the weights of historical data, ranging from 0 to 1, where 0 represents no risk and 1 represents extremely high risk. A rule engine combined with a Bayesian network is used to construct and train a key defect risk analysis model based on the causal relationship between material, parameters, and defects. The key defect fitting features are input for analysis, and the key defect risk coefficient is calculated, for example, a key defect risk coefficient of 0.7 (high risk). Using regression analysis or a neural network, a mapping relationship between parameter combinations and process defects is established to construct and obtain a process defect risk analysis model. The process defect fitting features are input, and the process defect risk coefficient is predicted based on historical process quality data, for example, a process defect risk coefficient of 0.2 (low risk). Finally, the basic defect risk coefficient, critical defect risk coefficient, and process defect risk coefficient are integrated as the laser engraving risk prediction result, which intuitively assesses the overall risk level of the laser engraving solution and ensures the accuracy of risk prediction.

[0041] Furthermore, step S320 also includes step S321, modeling based on the laser engraving module to obtain a laser engraving model; step S322, based on the laser engraving model, fitting the laser engraving product to be laser engraved according to the laser engraving analysis scheme to obtain a fitted laser engraved product; step S323, expanding the image based on the fitted laser engraved product to obtain a fitted laser engraved image; step S324, performing deep learning on the laser engraving defect factors using a convolutional neural network to generate a laser engraving defect detection network, the laser engraving defect detection network including a basic defect detection network, a key defect detection network, and a process defect detection network; step S325, inputting the fitted laser engraving image into the laser engraving defect detection network to generate the laser engraving defect feature sequence.

[0042] Preferably, the laser energy distribution and material interaction process of the laser engraving module are simulated through physical modeling. This includes constructing a spatial distribution model of the laser spot based on the optical component parameters of the laser engraving module, establishing a mathematical model of laser-material interaction by combining the thermodynamic properties of the material (such as melting point and thermal conductivity), and then integrating them to form a laser engraving model for predicting the actual engraving effect. Then, the parameters in the laser engraving analysis scheme are input into the laser engraving model to fit the laser engraving product. That is, the laser engraving model calculates the deposition process of laser energy on the product surface according to the parameters, simulates the material removal or discoloration effect, and generates a virtual three-dimensional model of the fitted laser engraving product, including the surface morphology after engraving, such as the depth of the indentation and the roughness of the pattern edge. Then, the fitted laser engraving product is image unfolded. That is, the three-dimensional model of the fitted laser engraving product is surface unfolded to map the engraving area of ​​the curved surface or complex shape into a two-dimensional planar image. Then, the fitted laser engraving image is generated based on the engraving depth and color change (such as the grayscale difference after metal oxidation) calculated by the model.

[0043] Preferably, a laser engraving defect detection network is generated by deep learning of laser engraving defect factors using convolutional neural networks. Specifically, a CNN architecture (such as ResNet or YOLO) is adopted, with convolutional layers extracting low-level features such as edges and textures from the image (e.g., the clarity of engraved lines), pooling layers reducing feature map resolution to improve the model's robustness to defect locations, and fully connected layers classifying features and outputting defect types and confidence scores. Based on the laser engraving defect factors, three sub-tasks are divided, corresponding to three sub-networks: a basic defect detection network (identifying basic defects such as missing engravings, misalignments, and ghosting), a key defect detection network (detecting defects affecting function such as material damage and insufficient depth), and a process defect detection network (identifying process problems such as unclear engravings and rough edges). Laser engraving images of various defects (e.g., manually annotated missing engraving areas) are extracted from historical production as positive samples, and laser engraving images of qualified products are used as negative samples to construct training data. The basic defect detection network, key defect detection network, and process defect detection network are then trained, and the network parameters are adjusted using a backpropagation algorithm to maximize the model's accuracy in identifying defect features.

[0044] Preferably, the fitted laser engraving image is input into the laser engraving defect detection network. Each sub-network outputs basic defect detection results (coordinates of missing engraving areas, misalignment offset, etc.), key defect detection results (material damage area, distribution of insufficient depth areas), and process defect detection results (edge ​​roughness, blurred line areas). Finally, the three types of defect detection results are integrated and transformed into a structured data sequence to form a laser engraving defect feature sequence containing fitting features of basic / key / process defects, thereby ensuring the level of intelligence and quality reliability of the laser engraving process for multiple product grades.

[0045] Step S400: If the laser engraving risk prediction result does not meet the laser engraving risk constraint, the laser engraving analysis scheme is mutated and adjusted according to the laser engraving risk predictor to establish a candidate laser engraving group.

[0046] Step S400 further includes step S410, selecting multiple laser engraving parameter regions based on each laser engraving confidence evaluation set and according to the confidence evaluation threshold, and establishing multiple laser engraving confidence value sets; step S420, performing interval analysis based on the multiple laser engraving confidence value sets to obtain adjustment constraints for each laser engraving variable; step S430, mutating the laser engraving analysis scheme according to the adjustment constraints for each laser engraving variable to obtain a laser engraving adjustment scheme set; step S440, based on the laser engraving risk predictor, performing traversal optimization on the laser engraving adjustment scheme set according to the laser engraving risk constraints to generate the candidate laser engraving group.

[0047] Preferably, based on product quality requirements and historical data, acceptable upper limits are set for different types of defect risks, such as a basic defect risk coefficient ≤ 0.4 and a critical defect coefficient ≤ 0.3, serving as laser engraving risk constraints. If the laser engraving risk prediction result does not meet the laser engraving risk constraints, the laser engraving analysis scheme is adjusted according to the laser engraving risk predictor. Specifically, minimum confidence standards are set for different types of laser engraving parameters, such as a basic parameter confidence level ≥ 0.6 and a critical parameter confidence level ≥ 0.8. Based on each laser engraving confidence evaluation set, multiple laser engraving parameter regions are selected according to the confidence evaluation threshold. That is, for each laser engraving parameter region, if the parameter with the highest confidence coefficient meets the threshold requirement, the parameter region is retained to filter out parameters with unstable historical performance, thereby forming several high-confidence laser engraving parameter regions. Then, multiple confidence coefficient parameter values ​​are selected from each retained parameter region, and the parameter values ​​are cross-combined to form a multi-dimensional parameter combination set, i.e., the laser engraving confidence value set.

[0048] Preferably, interval analysis is performed based on multiple sets of laser engraving confidence values. This involves analyzing the correlation between parameter intervals, including analyzing the physical dependencies between parameters (e.g., power and speed cannot simultaneously exceed the equipment's upper limit), establishing constraint rules, and summarizing the correlation constraints between parameter combinations and defects based on historical process data. For example, excessively low power and excessively high speed can lead to unclear engraving. This generates adjustment constraints for various laser engraving variables, such as power not exceeding the laser's maximum output and the optimal ratio range between speed and power. Then, the laser engraving analysis scheme is mutated based on these adjustment constraints. Specifically, based on the set of confidence values, the parameters are slightly fluctuated (e.g., ±5%) to generate more diverse parameter combinations. These combinations are then cross-referenced and replaced to obtain new combinations not based on historical experience. Finally, mutated parameter combinations that do not meet the conditions are eliminated based on the adjustment constraints, resulting in a final set of laser engraving adjustment schemes.

[0049] Preferably, the parameters (such as power and speed) of each adjustment scheme are input into the laser engraving risk predictor. The predictor outputs the corresponding risk coefficient based on the defect factor model (basic / critical / process defects). Then, for all schemes in the adjustment scheme set, it is checked one by one whether they meet all risk constraints, and schemes that do not meet any constraint are eliminated. Then, for the schemes that pass the risk assessment, they are clustered according to parameter characteristics (such as power level and speed). One to two representative schemes are selected from each category to avoid homogenization of schemes within the candidate group. Finally, the schemes that meet the risk constraints are summarized to generate a candidate laser engraving group, that is, schemes with different parameter tendencies, such as high-efficiency schemes and high-stability schemes, to adapt to the dynamic needs of multiple grades of products (such as different materials and different precision requirements), and to ensure that the candidate schemes have both historical reliability and risk controllability.

[0050] Further, step S440 also includes step S441, extracting the nth laser engraving adjustment scheme according to the laser engraving adjustment scheme set, where n is a positive integer; step S442, performing laser engraving defect fitting on the product to be laser engraved according to the nth laser engraving adjustment scheme to obtain the defect feature sequence of the nth scheme; step S443, inputting the defect feature sequence of the nth scheme into the laser engraving risk predictor to obtain the laser engraving risk feature of the nth scheme; step S444, determining whether the laser engraving risk feature of the nth scheme satisfies the laser engraving risk constraint; step S445, if the laser engraving risk feature of the nth scheme satisfies the laser engraving risk constraint, setting the nth laser engraving adjustment scheme as the nth candidate laser engraving scheme, and adding the nth candidate laser engraving scheme to the candidate laser engraving group.

[0051] Preferably, the nth laser engraving adjustment scheme is randomly extracted from the set of laser engraving adjustment schemes, where n is a positive integer representing the number of laser engraving adjustment schemes. Then, laser engraving defects are fitted to the product to be laser engraved based on the nth laser engraving adjustment scheme. That is, based on the laser engraving model (such as the laser-material interaction model), the parameters of the nth scheme are input, the product state after engraving is simulated, and the basic defect fitting features, key defect fitting features, and process defect fitting features corresponding to the nth scheme are generated to form the nth scheme defect feature sequence. Then, the nth scheme defect feature sequence is input into the laser engraving risk predictor, including analysis through the basic defect risk analysis model, the key defect risk analysis model, and the process defect risk analysis model to obtain the basic defect risk coefficient, the key defect risk coefficient, and the process defect risk coefficient, and integrated to generate the nth scheme laser engraving risk features. Finally, the laser engraving risk characteristics of the nth scheme are compared with the laser engraving risk constraints. If the laser engraving risk characteristics of the nth scheme meet the laser engraving risk constraints, the laser engraving adjustment scheme n is set as the nth candidate laser engraving scheme, and the nth candidate laser engraving scheme is added to the candidate laser engraving group. The candidate laser engraving group is gradually accumulated as qualified schemes are screened, until all laser engraving adjustment schemes are traversed.

[0052] Step S500: Introduce a global laser engraving optimization mechanism to perform optimization analysis on the candidate laser engraving group and determine the laser engraving optimization strategy.

[0053] Step S500 further includes step S510, performing weight analysis on the multi-dimensional laser engraving risk indicators of the laser engraving risk predictor to obtain a global laser engraving risk analysis function, wherein the multi-dimensional laser engraving risk indicators include basic defect risk indicators, key defect risk indicators, and process defect risk indicators; step S520, performing global risk calculation on the candidate laser engraving group based on the global laser engraving risk analysis function to obtain a global laser engraving risk distribution; step S530, performing global risk minimization optimization on the candidate laser engraving group based on the global laser engraving risk distribution to generate the laser engraving optimization strategy.

[0054] Preferably, a global laser engraving optimization mechanism is introduced to perform optimization analysis on the candidate laser engraving group. This involves selecting the scheme or parameter combination that minimizes the overall risk from the candidate group. Specifically, weighted analysis is performed based on the multi-dimensional laser engraving risk indicators (basic defect risk indicator, critical defect risk indicator, and process defect risk indicator) of the laser engraving risk predictor. The basic defect risk indicator measures the probability of occurrence of fundamental problems such as misalignment and incomplete engraving; the critical defect risk indicator quantifies core risks such as material damage and functional failure; and the process defect risk indicator assesses the risk of engraved surface quality (such as clarity and roughness). Weights are dynamically allocated based on product characteristics, and then coupled to construct a global laser engraving risk analysis function. ,in, Indicates the global risk coefficient. =1, where =1 represents the weights of the basic defect risk index, critical defect risk index, and process defect risk index, respectively. These represent the basic defect risk coefficient, critical defect risk coefficient, and process defect risk coefficient, respectively. A global risk calculation is performed on the candidate laser engraving group based on the global risk analysis function. This involves substituting each laser engraving adjustment scheme in the candidate group into the global risk analysis function to calculate the comprehensive risk value, and then sorting and visualizing the risk values ​​of all schemes to form a risk distribution, thus obtaining the global laser engraving risk distribution. Finally, based on the global risk distribution, a global risk minimization optimization is performed on the candidate laser engraving group to generate a laser engraving optimization strategy. When a significantly low-risk scheme exists (e.g., scheme D has a significantly lower risk than others), it is directly selected as the laser engraving optimization strategy. When the risks of multiple schemes are similar, a strategy library is generated with additional application conditions, such as using scheme E (risk 0.15, moderate efficiency) for normal production and scheme F (risk 0.18, more robust parameters) for equipment aging.

[0055] Step S600: The product to be laser-engraved is transported to a predetermined position according to the industrial vision module and the motion control module, and the laser engraving module is simultaneously started to execute the laser engraving optimization strategy.

[0056] Preferably, the industrial vision module and motion control module transport the product to be laser-engraved to a predetermined position. The industrial vision module (camera, light source) is installed above the loading station or laser engraving station to acquire product images in real time. The motion control module (servo motor, guide rail) drives the conveyor belt or robotic arm to handle and position the product. The laser engraving module (laser, galvanometer) is fixed at the processing station and receives control commands to execute engraving. Specifically, the servo motor drives the conveyor belt or robotic arm to move the product from the loading station to the laser engraving station, so that the laser-engraved surface faces the laser head. During the movement, the position is fed back in real time by a grating ruler or encoder, and deviations are dynamically corrected. Then, the laser engraving module is started synchronously to execute the laser engraving optimization strategy, that is, the process parameters are extracted from the laser engraving optimization strategy and sent to the laser engraving module controller for laser engraving. This includes the laser starting according to the power parameters, the optical components scanning according to the path planning (such as straight line, curve, dot matrix), and the motion control module synchronously controlling the movement of the product or the laser head to achieve dynamic engraving. This realizes fully automated intelligent production from recognition to processing and ensures the consistency and stability of production efficiency and laser engraved product quality.

[0057] In the above text, refer to Figure 1 A multi-grade automated laser engraving method incorporating industrial vision, according to embodiments of the present invention, is described in detail. Next, reference will be made to... Figure 2 This invention describes a multi-grade automated laser engraving system incorporating industrial vision, according to an embodiment of the present invention.

[0058] The automated laser engraving system for multiple grades of products, incorporating industrial vision, according to embodiments of the present invention, addresses the technical problems in existing technologies, such as low laser engraving efficiency for multiple grades of products, lack of risk assessment and optimization of laser engraving schemes, and poor consistency in laser engraved product quality. It achieves the technical effect of improving the automation and production efficiency of laser engraving for multiple grades of products, as well as the consistency and stability of product laser engraving quality. Figure 2 As shown, the multi-brand automated laser engraving system combined with industrial vision includes: a product acquisition unit 10, a laser engraving analysis scheme generation unit 20, a multi-dimensional risk prediction unit 30, a candidate laser engraving group establishment unit 40, a laser engraving optimization strategy determination unit 50, and a laser engraving optimization strategy execution unit 60.

[0059] The product acquisition unit 10 is used to acquire the product to be laser-engraved by the automated laser engraving machine. The automated laser engraving machine includes an industrial vision module, a motion control module, and a laser engraving module. The product to be laser-engraved includes multiple brand features. The laser engraving analysis scheme generation unit 20 is used to generate laser engraving content based on the multiple brand features, and to control and mine the laser engraving module in conjunction with the product feature data of the product to be laser-engraved to generate a laser engraving analysis scheme. The multi-dimensional risk prediction unit 30 is used to perform multi-dimensional risk prediction on the laser engraving analysis scheme based on the laser engraving risk predictor, and to determine the laser engraving risk prediction result. The system includes: a candidate laser engraving group establishment unit 40, used to establish a candidate laser engraving group by adjusting the laser engraving analysis scheme according to the laser engraving risk predictor if the laser engraving risk prediction result does not meet the laser engraving risk constraints; a laser engraving optimization strategy determination unit 50, used to introduce a global laser engraving optimization mechanism to perform optimization analysis on the candidate laser engraving group and determine the laser engraving optimization strategy; and a laser engraving optimization strategy execution unit 60, used to simultaneously start the laser engraving module to execute the laser engraving optimization strategy when the product to be laser engraved is transported to a predetermined position by the industrial vision module and the motion control module.

[0060] The specific configuration of the laser engraving analysis scheme generation unit 20 will be described in detail below. The laser engraving analysis scheme generation unit 20 further includes: retrieving laser engraving control schemes for the laser engraving module based on the laser engraving content and the product feature data to obtain a retrieved laser engraving scheme set; classifying variables based on the retrieved laser engraving scheme set to obtain multiple laser engraving parameter areas; evaluating the confidence level of each laser engraving parameter area to obtain each laser engraving confidence evaluation set; and optimizing the multiple laser engraving parameter areas to maximize laser engraving confidence based on each laser engraving confidence evaluation set to generate the laser engraving analysis scheme.

[0061] The specific configuration of the multidimensional risk prediction unit 30 will be described in detail below. The multidimensional risk prediction unit 30 further includes: setting laser engraving defect factors, which include basic laser engraving defects, key laser engraving defects, and laser engraving process defects; based on the laser engraving defect factors, performing defect fitting according to the laser engraving analysis scheme to obtain a laser engraving defect feature sequence, which includes basic defect fitting features, key defect fitting features, and process defect fitting features; inputting the laser engraving defect feature sequence into the laser engraving risk predictor to obtain the laser engraving risk prediction result, wherein the laser engraving risk predictor includes a basic defect risk analysis model, a key defect risk analysis model, and a process defect risk analysis model, and the laser engraving risk prediction result includes a basic defect risk coefficient, a key defect risk coefficient, and a process defect risk coefficient.

[0062] The specific configuration of the multi-dimensional risk prediction unit 30 will be described in detail below. The multi-dimensional risk prediction unit 30 further includes: modeling based on the laser engraving module to obtain a laser engraving model; performing laser engraving fitting on the product to be laser engraved according to the laser engraving analysis scheme based on the laser engraving model to obtain a fitted laser engraved product; performing image unfolding on the fitted laser engraved product to obtain a fitted laser engraved image; performing deep learning on the laser engraving defect factors using a convolutional neural network to generate a laser engraving defect detection network, the laser engraving defect detection network including a basic defect detection network, a key defect detection network, and a process defect detection network; and inputting the fitted laser engraving image into the laser engraving defect detection network to generate the laser engraving defect feature sequence.

[0063] The specific configuration of the candidate laser engraving group establishment unit 40 will be described in detail below. The candidate laser engraving group establishment unit 40 further includes: selecting multiple laser engraving parameter regions based on each laser engraving confidence evaluation set and according to the confidence evaluation threshold, and establishing multiple laser engraving confidence value sets; performing interval analysis based on the multiple laser engraving confidence value sets to obtain adjustment constraints for each laser engraving variable; mutating the laser engraving analysis scheme according to the adjustment constraints for each laser engraving variable to obtain a laser engraving adjustment scheme set; and, based on the laser engraving risk predictor, traversing and optimizing the laser engraving adjustment scheme set according to the laser engraving risk constraints to generate the candidate laser engraving group.

[0064] The specific configuration of the candidate laser engraving group establishment unit 40 will be described in detail below. The candidate laser engraving group establishment unit 40 further includes: extracting the nth laser engraving adjustment scheme from the laser engraving adjustment scheme set, where n is a positive integer; performing laser engraving defect fitting on the product to be laser engraved according to the nth laser engraving adjustment scheme to obtain the nth scheme defect feature sequence; inputting the nth scheme defect feature sequence into the laser engraving risk predictor to obtain the nth scheme laser engraving risk feature; determining whether the nth scheme laser engraving risk feature satisfies the laser engraving risk constraint; if the nth scheme laser engraving risk feature satisfies the laser engraving risk constraint, setting the nth laser engraving adjustment scheme as the nth candidate laser engraving scheme, and adding the nth candidate laser engraving scheme to the candidate laser engraving group.

[0065] The specific configuration of the laser engraving optimization strategy determination unit 50 will be described in detail below. The laser engraving optimization strategy determination unit 50 further includes: performing weight analysis on the multi-dimensional laser engraving risk indicators of the laser engraving risk predictor to obtain a global laser engraving risk analysis function, wherein the multi-dimensional laser engraving risk indicators include basic defect risk indicators, critical defect risk indicators, and process defect risk indicators; performing global risk calculation on the candidate laser engraving group based on the global laser engraving risk analysis function to obtain a global laser engraving risk distribution; and performing global risk minimization optimization on the candidate laser engraving group based on the global laser engraving risk distribution to generate the laser engraving optimization strategy.

[0066] The specific configuration of the product acquisition unit 10 to be laser-engraved will be described in detail below. The product acquisition unit 10 to be laser-engraved further includes: the multi-brand features include model features, grade features, batch features, and customized identification features.

[0067] The multi-grade automated laser engraving system combining industrial vision provided in the embodiments of the present invention can execute the multi-grade automated laser engraving method combining industrial vision provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0068] Based on the foregoing embodiments, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the multi-grade automated laser engraving method combining industrial vision as described in any of the preceding embodiments.

[0069] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0070] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A multi-grade automated laser engraving method combining industrial vision, characterized in that, The method includes: The product to be laser-engraved by the automated laser engraving machine includes an industrial vision module, a motion control module, and a laser engraving module. The product to be laser-engraved includes multiple grade features. Based on the multi-brand characteristics, laser engraving content is generated. Combined with the product characteristic data of the product to be laser engraved, the laser engraving module is controlled and mined to generate a laser engraving analysis scheme. The laser engraving risk prediction result is determined by performing multi-dimensional risk prediction on the laser engraving analysis scheme based on the laser engraving risk predictor. If the laser engraving risk prediction result does not meet the laser engraving risk constraint, the laser engraving analysis scheme is mutated and adjusted according to the laser engraving risk predictor to establish a candidate laser engraving group. A global laser engraving optimization mechanism is introduced to perform optimization analysis on the candidate laser engraving group and determine the laser engraving optimization strategy. The industrial vision module and the motion control module transport the product to be laser-engraved to the predetermined position, and the laser engraving module is simultaneously activated to execute the laser engraving optimization strategy. Based on the laser engraving risk predictor, a multi-dimensional risk prediction is performed on the laser engraving analysis scheme to determine the laser engraving risk prediction results, including: The laser engraving defect factors are set, including basic laser engraving defects, key laser engraving defects, and laser engraving process defects. Based on the laser engraving defect factor, defect fitting is performed according to the laser engraving analysis scheme to obtain a laser engraving defect feature sequence, which includes basic defect fitting features, key defect fitting features and process defect fitting features. The laser engraving defect feature sequence is input into the laser engraving risk predictor to obtain the laser engraving risk prediction result. The laser engraving risk predictor includes a basic defect risk analysis model, a key defect risk analysis model, and a process defect risk analysis model. The laser engraving risk prediction result includes a basic defect risk coefficient, a key defect risk coefficient, and a process defect risk coefficient. Based on the laser engraving defect factor, defect fitting is performed according to the laser engraving analysis scheme to obtain a laser engraving defect feature sequence, including: Modeling is performed based on the laser engraving module to obtain a laser engraving model; Based on the laser engraving model, the laser engraving product to be laser engraved is fitted and laser engraved according to the laser engraving analysis scheme to obtain the fitted laser engraved product. The fitted laser-engraved product is used to expand the image to obtain the fitted laser-engraved image; Deep learning is performed on the laser engraving defect factors using convolutional neural networks to generate a laser engraving defect detection network, which includes a basic defect detection network, a key defect detection network, and a process defect detection network. The fitted laser-engraved image is input into the laser-engraved defect detection network to generate the laser-engraved defect feature sequence. The laser engraving global optimization mechanism includes: The laser engraving risk multidimensional index of the laser engraving risk predictor is weighted and analyzed to obtain the laser engraving global risk analysis function. The laser engraving risk multidimensional index includes basic defect risk index, key defect risk index and process defect risk index. The global risk of the candidate laser engraving group is calculated based on the laser engraving global risk analysis function to obtain the global risk distribution of laser engraving. Based on the global risk distribution of laser engraving, the candidate laser engraving group is optimized by minimizing global risk, and the laser engraving optimization strategy is generated.

2. The method as described in claim 1, characterized in that, Based on the multi-brand characteristics, laser engraving content is generated. Combined with the product characteristic data of the product to be laser engraved, the laser engraving module is controlled and mined to generate a laser engraving parsing scheme, including: Based on the laser engraving content and the product feature data, the laser engraving module is searched for laser engraving control schemes to obtain a set of laser engraving schemes. Based on the retrieved laser engraving scheme set, variables are classified to obtain multiple laser engraving parameter areas; Confidence evaluation is performed on each laser engraving parameter area to obtain each laser engraving confidence evaluation set; Based on the laser engraving confidence evaluation sets, the laser engraving parameter regions are optimized by maximizing laser engraving confidence to generate the laser engraving analysis scheme.

3. The method as described in claim 1, characterized in that, If the laser engraving risk prediction result does not meet the laser engraving risk constraints, the laser engraving analysis scheme is mutated and adjusted according to the laser engraving risk predictor to establish a candidate laser engraving group, including: Based on each laser engraving confidence evaluation set, multiple laser engraving parameter regions are selected according to the confidence evaluation threshold to establish multiple laser engraving confidence value sets; Based on the multiple laser engraving confidence value sets, interval analysis is performed to obtain the adjustment constraints of each laser engraving variable; The laser engraving analysis scheme is mutated according to the laser engraving variable adjustment constraints to obtain a laser engraving adjustment scheme set. Based on the laser engraving risk predictor, the laser engraving adjustment scheme set is traversed and optimized according to the laser engraving risk constraints to generate the candidate laser engraving group.

4. The method as described in claim 3, characterized in that, Based on the laser engraving risk predictor, the laser engraving adjustment scheme set is traversed and optimized according to the laser engraving risk constraints to generate the candidate laser engraving group, including: Based on the laser engraving adjustment scheme set, extract the nth laser engraving adjustment scheme, where n is a positive integer; Based on the laser engraving adjustment scheme n, the laser engraving defect fitting is performed on the product to be laser engraved to obtain the defect feature sequence of scheme n; The defect feature sequence of the nth scheme is input into the laser engraving risk predictor to obtain the laser engraving risk feature of the nth scheme; Determine whether the laser engraving risk characteristics of the nth scheme meet the laser engraving risk constraints; If the laser engraving risk characteristics of the nth scheme satisfy the laser engraving risk constraint, the laser engraving adjustment nth scheme is set as the nth candidate laser engraving scheme, and the nth candidate laser engraving scheme is added to the candidate laser engraving group.

5. The method as described in claim 1, characterized in that, The multi-brand characteristics include model characteristics, grade characteristics, batch characteristics, and customized identification characteristics.

6. A multi-grade automated laser engraving system combining industrial vision, characterized in that: The system is used to implement the automated laser engraving method for multiple grades combined with industrial vision as described in any one of claims 1 to 5, the system comprising: A product acquisition unit is used to acquire a product to be laser-engraved by an automated laser engraving machine. The automated laser engraving machine includes an industrial vision module, a motion control module, and a laser engraving module. The product to be laser-engraved includes multiple grade features. The laser engraving analysis scheme generation unit is used to generate laser engraving content based on the multi-brand characteristics, and to control and mine the laser engraving module in combination with the product characteristic data of the product to be laser engraved, so as to generate a laser engraving analysis scheme. A multi-dimensional risk prediction unit is used to perform multi-dimensional risk prediction on the laser engraving analysis scheme based on the laser engraving risk predictor, and determine the laser engraving risk prediction result. The candidate laser engraving group establishment unit is used to establish a candidate laser engraving group by performing variation adjustment on the laser engraving analysis scheme according to the laser engraving risk predictor if the laser engraving risk prediction result does not meet the laser engraving risk constraint. The laser engraving optimization strategy determination unit is used to introduce a global laser engraving optimization mechanism to perform optimization analysis on the candidate laser engraving group and determine the laser engraving optimization strategy. The laser engraving optimization strategy execution unit is used to transport the product to be laser engraved to a predetermined position according to the industrial vision module and the motion control module, and simultaneously start the laser engraving module to execute the laser engraving optimization strategy.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the multi-grade automated laser engraving method incorporating industrial vision as described in any one of claims 1-5.

Images