Multi-pattern layout optimization method for irregular blanks

Abstract

The invention relates to the technical field of irregular blank layout, in particular to a multi-pattern layout optimization method for irregular blanks. The method comprises the following steps: acquiring an irregular blank image in a region; performing contour detail reconstruction on the irregular blank image in the region to generate a contour reconstruction blank image; performing blank space distribution analysis on the contour reconstruction blank image to generate blank space distribution data; three-dimensional point cloud modeling in a region is carried out based on the blank space distribution data, and a full-region blank three-dimensional model is constructed; performing graph overlapping rate statistics on the full-area blank three-dimensional model to obtain a graph overlapping rate between the blanks; blank arrangement density evaluation is carried out based on the graph overlapping rate between the blanks, so that blank arrangement density evaluation data is generated; and carrying out dynamic layout planning simulation on the full-area blank three-dimensional model so as to generate a dynamic layout planning blank model. The space layout utilization rate of irregular blanks is increased, and efficient blank layout optimization is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of irregular blank layout, and in particular to a multi-graphic layout optimization method for irregular blanks. Background Art

[0002] With the high degree of automation and intelligence in modern industrial production, all walks of life have higher and higher requirements for material utilization and production efficiency. Especially for some irregularly shaped parts or products, how to achieve the best graphic layout within the limited raw material size and minimize material waste has always been an important challenge facing the manufacturing industry. Traditional graphic layout methods usually rely on experienced typesetters to improve material utilization through manual adjustment and optimization. However, this method is inefficient, the space utilization rate of layout optimization is low, and it is easily affected by human factors. It is difficult to systematically optimize complex irregular graphics. With the increasing development of computer technology, automated graphic layout methods based on algorithm optimization have emerged, such as genetic algorithm. Methods such as simulated annealing can improve material utilization to a certain extent. However, these methods still have some limitations when dealing with irregular graphic layout. First, the spatial relationship between irregular graphics is complex, and it is difficult to establish an accurate mathematical model. Second, these methods often assume that the graphics and billet shapes are known, and cannot cope with changes in billet size or graphic shape in actual production. Third, the calculation complexity is high and the optimization efficiency is low, which makes it difficult to meet the real-time production needs. Therefore, there is an urgent need for a new multi-graphic layout optimization method for irregular billets, which can accurately describe the spatial relationship between irregular graphics, dynamically adapt to changes in the production environment, and have good computing performance to meet the high efficiency, flexibility, and intelligence needs of modern manufacturing. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a multi-graphic layout optimization method for irregular blanks to solve at least one of the above technical problems.

[0004] To achieve the above object, the present invention provides a multi-pattern layout optimization method for irregular blanks, comprising the following steps: Step S1: Acquire an irregular blank image in a region; reconstruct contour details of the irregular blank image in the region to generate a contour reconstructed blank image; Step S2: Performing a blank spatial distribution analysis on the contour reconstructed blank image to generate blank spatial distribution data; performing a three-dimensional point cloud modeling in the region based on the blank spatial distribution data to construct a three-dimensional model of the blank in the entire region; Step S3: performing graphic overlap statistics on the three-dimensional model of the whole area blank to obtain the graphic overlap rate between blanks; performing blank arrangement density evaluation based on the graphic overlap rate between blanks, thereby generating blank arrangement density evaluation data; Step S4: Perform dynamic nesting planning simulation on the three-dimensional model of the full-area blank, thereby generating a dynamic nesting planning blank model; Step S5: performing topological constraint analysis between graphics on the dynamic layout planning blank model according to the blank layout density evaluation data, and performing topological layout optimization, so as to obtain a topological layout optimization blank model; Step S6: local packing defect identification is performed on the topological packing optimization blank model to obtain a packing defect blank area; global packing optimization is performed based on the packing defect blank area to generate an optimal blank packing optimization model to perform irregular blank packing optimization operations.

[0005] The present invention can clearly present the shape and structure of the blank by reconstructing the contour details of the irregular blank image, provide an accurate data basis for subsequent analysis and processing, generate a contour reconstructed blank image to accurately locate the morphological characteristics of the irregular blank, and provide necessary information for subsequent spatial distribution analysis and three-dimensional modeling. Based on the contour reconstructed blank image, the blank spatial distribution analysis can understand the distribution of the blank in the region, provide data support for subsequent modeling and planning, and construct a three-dimensional model of the blank in the entire region to realize the three-dimensional representation of the irregular blank, which provides a basis for subsequent layout planning and optimization. By performing graphic overlap rate statistics on the blank three-dimensional model, Evaluate the space utilization between billets, provide data basis for layout density evaluation, generate billet layout density evaluation data to understand the density distribution of billets, provide guidance for layout planning, and optimize the layout structure. Dynamic layout planning simulation can simulate the billet layout in different situations, provide intuitive display and comparison for layout optimization, and obtain the topological layout optimization billet model to optimize the layout structure of billets, improve space utilization and production efficiency, identify the layout defect billet area, which can help find layout problems, provide improvement direction for global layout optimization, and generate the optimal billet layout optimization model to achieve the best layout effect, improve production efficiency and reduce waste.

[0006] Preferably, step S1 comprises the following steps: Step S11: Acquire the image of the irregular blank in the area; Step S12: performing global brightness average calculation on the irregular blank image in the region to obtain an average brightness value of the image; Step S13: performing dark area recognition on the irregular blank image in the region and extracting the dark area image; Step S14: performing dark portion brightness balance enhancement processing on the dark portion image according to the average brightness value of the image, thereby obtaining a brightness enhanced blank image; Step S15: reconstructing the outline details of the brightness enhanced blank image to generate a contour reconstructed blank image.

[0007] The present invention acquires irregular blank images in a region, provides basic data, and enables subsequent processing steps to analyze and process the blanks. Global average brightness calculation of the irregular blank images can obtain the overall brightness level, and provide a reference value for subsequent brightness processing. The calculation of the average brightness value of the image understands the brightness and darkness of the overall image, and provides a reference basis for subsequent brightness enhancement processing. The dark region is identified and the dark image is extracted to separate the dark details in the blank image, and a target region is provided for brightness enhancement processing. The extraction of the dark image can make the subsequent brightness balance enhancement processing more targeted and improve the processing effect. Brightness balance enhancement processing of the dark image according to the average brightness value of the image can improve the visualization effect of the dark region, enhance the details and contrast of the image, and the brightness enhancement processing highlights the dark details, making the blank image clearer, and providing a better data basis for subsequent contour reconstruction. Contour detail reconstruction of the brightness enhanced blank image can accurately show the shape and details of the blank, and provide accurate shape information for subsequent multi-graphic nesting optimization. The contour reconstructed blank image is generated to accurately locate the shape characteristics of the blank, and provide necessary visual reference for subsequent nesting planning and optimization.

[0008] Preferably, the specific steps of step S15 are: Performing edge contour visual recognition on the brightness enhanced blank image to extract the edge contour line of each blank; The thickness of each blank is identified one by one for the edge contour line of each blank, and the thickness shape parameters of each blank contour are generated; Based on each blank contour thickness and shape parameter, contour refinement filtering is performed on the brightness enhanced blank image to obtain a detail enhanced blank image; The detail-enhanced blank image is analyzed for abnormally protruding regions to obtain protruding contour sub-regions; Fitting local salient feature points to the salient contour sub-region to obtain multiple local feature points; Perform spline interpolation smoothing on multiple local feature points to obtain smooth feature contour lines; The detail enhanced blank image is reconstructed by performing contour superposition image reconstruction according to the smooth characteristic contour line to generate a contour reconstructed blank image.

[0009] The present invention can accurately extract the outer contour of each blank by identifying the edge contour of the brightness enhanced blank image, and provide basic data for subsequent processing. The thickness identification of each blank contour can generate the shape parameters of each blank contour, and provide a reference basis for subsequent contour refinement filtering. The thickness and shape parameters of the blank contour are generated to understand the morphological characteristics of the blank, and provide quantitative data support for subsequent processing. Contour refinement filtering based on the thickness and shape parameters of the blank contour can accurately extract the detailed information of the blank contour, making the image clearer and more realistic, and obtaining the contour features of the blank highlighted in the detail enhanced blank image, and providing a more accurate data basis for subsequent processing and analysis. The abnormal protruding areas of the detail enhanced blank image can be analyzed. Identify the special parts of the blank contour, provide the target area for the subsequent local feature point fitting, obtain the prominent contour sub-region to more accurately locate the key feature points of the blank, perform local feature point fitting and spline interpolation smoothing on the prominent contour sub-region to accurately extract the key feature points of the blank contour, retain the important details of the blank morphology, generate smooth feature contour lines to eliminate image noise and retain important contour features, and perform contour line superposition image reconstruction on the detail enhanced blank image based on the smooth feature contour line to clearly display the contour shape of the blank, providing an intuitive reference for the subsequent multi-graphic nesting optimization, and generating contour reconstructed blank images can provide more realistic and clear blank morphology information, optimize the layout structure and improve production efficiency. Preferably, the specific steps of step S2 are: Step S21: visually identifying the blank with the reconstructed contour, and marking it in sequence to obtain a blank number sequence; Step S22: segmenting the contour reconstructed blank image into blank bounding boxes one by one based on the blank number sequence to generate a plurality of blank bounding boxes; Step S23: Calculate the positions of the multiple blanks one by one for the bounding boxes of the blanks, and generate the coordinates of the position of each blank; Step S24: performing a blank space distribution analysis on the contour reconstructed blank image according to each blank position coordinate to generate blank space distribution data; Step S25: performing a three-dimensional morphological feature analysis on a plurality of blank boundary frames to generate a three-dimensional morphological feature of the blank; Step S26: Performing three-dimensional point cloud modeling in the region for the blank spatial distribution data according to the three-dimensional morphological characteristics of the blank, and constructing a three-dimensional model of the blank in the entire region.

[0010] The present invention can accurately identify each blank and assign a unique numbering sequence to it by performing blank visual recognition and sequential marking on the contour reconstructed blank image, segment the blank boundary box one by one on the contour reconstructed blank image according to the blank numbering sequence, effectively separate the area of ​​each blank, provide a separate blank space range for subsequent analysis, calculate the position of multiple blank boundary boxes one by one, generate the position coordinates of each blank, accurately locate the position of each blank in the overall image, and provide accurate data support for subsequent spatial distribution analysis, perform blank spatial distribution analysis on the contour reconstructed blank image according to each blank position coordinate, generate blank spatial distribution data, understand the distribution of the blank in the overall image, and provide a reference for optimizing layout, perform blank three-dimensional morphological feature analysis on multiple blank boundary boxes, and deeply understand the morphological features of each blank, providing a detailed data basis for subsequent modeling and analysis, perform regional three-dimensional point cloud modeling on the blank spatial distribution data according to the three-dimensional morphological features of the blank, construct a three-dimensional model of the blank in the entire region, and more intuitively display the spatial layout and morphological features of the blank, so as to provide more accurate data support for multi-graphic layout optimization.

[0011] Preferably, step S3 specifically comprises the following steps: Step S31: Calculate the space utilization rate of the three-dimensional model of the blank in the whole area to obtain the irregular space utilization rate; Step S32: identifying the blank intersection points on the full-area blank three-dimensional model to generate blank intersection points; Step S33: performing graphic overlap statistics according to the intersection points of the blanks to obtain the graphic overlap rate between the blanks; Step S34: calculating the regional gap area ratio of the three-dimensional model of the whole-region blank according to the utilization rate of the irregular space and the overlap rate of the patterns between the blanks, so as to generate regional space gap data; Step S35: Evaluate the blank arrangement density of the regional spatial gap data, thereby generating blank arrangement density evaluation data.

[0012] The present invention can accurately evaluate the utilization of irregular space by calculating the space utilization rate of the three-dimensional model of the whole area stock, provide data support for the stock arrangement, obtain the utilization rate of irregular space to understand the distribution of stock in space, provide a reference basis for subsequent arrangement optimization, identify the stock intersection points in the three-dimensional model of the whole area stock, accurately find out the contact points between stock, provide key data for the subsequent graphic overlap rate statistics, generate the stock intersection points to understand the contact situation between stock, and provide important information for the subsequent regional gap area ratio calculation, perform graphic overlap rate statistics according to the stock intersection points, quantify the overlap situation between stock, and provide a reference basis for the subsequent regional space optimization. The calculation of the gap data between the blanks provides a basis, and the overlapping rate of the graphics between the blanks is obtained to evaluate the compactness of the blank arrangement, which provides a key reference for the evaluation of the arrangement density. The calculation of the regional gap area ratio based on the irregular space utilization rate and the overlapping rate of the graphics between the blanks can quantify the gap situation in the region, and provide data support for the optimization of the blank arrangement. The regional space gap data is generated to understand the gap situation in the blank arrangement. The blank arrangement density evaluation of the regional space gap data can evaluate the density of the blank in the region, which provides a key reference for the optimization of the arrangement structure. The blank arrangement density evaluation data is generated to optimize the blank arrangement method and improve the space utilization efficiency and production efficiency.

[0013] Preferably, the specific steps of step S4 are: Step S41: Performing a position adjustment feasibility analysis on the three-dimensional morphological features of the blank, thereby obtaining position adjustment parameters for each blank; Step S42: adjusting the overlap loss calculation according to each blank position adjustment parameter to obtain blank overlap loss data; Step S43: performing iterative decision-making on the position adjustment of the three-dimensional model of the whole-region blank according to the blank overlap loss data, so as to obtain an optimized blank position adjustment scheme; Step S44: Based on the optimized blank position adjustment scheme, a dynamic nesting planning simulation is performed on the blank three-dimensional model of the entire region, thereby generating a dynamic nesting planning blank model.

[0014] The present invention evaluates the feasibility of adjusting the position of each blank by performing a position adjustment feasibility analysis on the three-dimensional morphological characteristics of the blank, and provides a basis for subsequent adjustment decisions. The analysis results of each blank position adjustment parameter can help determine the adjustment range and method of the blank position, and provide a reference for optimizing the layout. The overlap loss calculation based on each blank position adjustment parameter can quantify the overlap between the blanks, and provide data support for the optimization of the position adjustment scheme. The blank overlap loss data is obtained to evaluate the effects of different position adjustment schemes, and provide a basis for subsequent position adjustment decisions. The blank position adjustment scheme can be optimized and the layout efficiency can be improved by performing iterative decision-making on the position adjustment of the three-dimensional model of the blank in the entire region based on the blank overlap loss data. The optimized blank position adjustment scheme is obtained through iterative decision-making, and the overlap loss is reduced and the space utilization rate is improved, thereby optimizing the layout result. Dynamic layout planning simulation based on the optimized blank position adjustment scheme can generate a dynamic layout planning blank model, which shows the best layout effect. The generated dynamic layout planning blank model intuitively shows the optimal layout scheme, provides guidance for production implementation, and maximizes the space utilization rate and production efficiency.

[0015] Preferably, the specific steps of step S43 are: Performing position adjustment range analysis on multiple blank boundary boxes according to the three-dimensional morphological features of the blanks to obtain the position adjustment range of each blank; According to the blank overlap loss data, the collision range of each blank position adjustment range is excluded to obtain the collision exclusion position adjustment range; The collision elimination position adjustment range is used to make iterative decisions on the position adjustment of the full-area blank three-dimensional model, and multiple blank position adjustment schemes are obtained; The optimal convergence decision of the blank position is made for multiple blank position adjustment schemes to obtain the optimized blank position adjustment scheme.

[0016] The present invention can determine the position adjustment range of each billet by analyzing the three-dimensional morphological characteristics of the billet, provide a clear range limit for subsequent position adjustment, obtain the position adjustment range of each billet, maintain the stability and rationality of the billet during the position adjustment process, and provide a basis for layout optimization; perform collision range exclusion on the position adjustment range of each billet according to the billet overlap loss data, thereby avoiding collisions between billets and ensuring the accuracy of the layout result; obtain the collision exclusion position adjustment range to exclude potential collision situations, and provide a safer and more effective solution for position adjustment decision-making; perform iterative decision on position adjustment of the three-dimensional model of the billet in the entire region using the collision exclusion position adjustment range, thereby optimizing the position adjustment schemes of multiple billets and improving the layout efficiency; obtain the position adjustment schemes of multiple billets through iterative decision-making, adjust the billet positions, reduce overlapping losses, and improve space utilization; perform optimal convergence decision on the billet positions of multiple billet position adjustment schemes, thereby finding the best position adjustment scheme and ensuring the best effect of billet layout; obtain the optimized billet position adjustment scheme, improve the layout efficiency and the layout compactness, and provide the most optimized layout solution for production.

[0017] Preferably, the specific steps of step S5 are: Step S51: performing spatial voxel processing on the dynamic layout planning blank model to generate a regional blank grid; Step S52: Performing position and posture analysis on the regional stock grid to generate stock position and posture data; Step S53: performing inter-graphic topological constraint analysis on the blank position and posture data according to the blank arrangement density evaluation data to generate inter-blank topological constraint relationships; Step S54: topologically optimizing the dynamic nesting planning blank model based on the topological constraint relationship between blanks, thereby obtaining a topologically optimized blank model.

[0018] The present invention converts a complex stock model into a regular regional stock grid by performing spatial voxelization processing on a dynamic nesting planning stock model, simplifies subsequent position analysis and optimization processes, generates a regional stock grid that can accurately describe the spatial structure of the stock, and provides basic data for subsequent position and posture analysis, performs position and posture analysis on the regional stock grid, and generates stock position and posture data, which helps to understand the specific position and posture of each stock, and provides a basis for topological constraint analysis, generates stock position and posture data that accurately describes the spatial layout of the stock, and provides accurate parameters for subsequent topological nesting optimization, performs inter-graphic topological constraint analysis on stock position and posture data according to stock arrangement density evaluation data, and can establish a topological constraint relationship between stock materials, ensuring that the nesting process meets design requirements, generates a topological constraint relationship between stock materials to control the mutual position relationship between stock materials, and ensures that the final nesting result meets the design density and layout requirements, performs topological nesting optimization on the dynamic nesting planning stock model based on the topological constraint relationship between stock materials, and can optimize the position and posture of the stock materials, ensuring that the nesting result is optimal, and obtains a topological nesting optimized stock model to achieve compact arrangement and optimal space utilization of the stock materials, thereby improving nesting efficiency and production efficiency.

[0019] Preferably, the specific steps of step S6 are: Step S61: Calculate the global arrangement space filling rate of the topological layout optimization blank model, so as to obtain the blank arrangement space filling rate; Step S62: performing local layout defect identification on the topological layout optimization blank model based on the blank layout space filling rate to obtain a layout defect blank area; Step S63: performing iterative fine-tuning of the local posture of the defective blank area to generate local blank posture fine-tuning data; Step S64: performing global layout optimization on the topological layout optimization blank model based on the local blank posture fine-tuning data to generate an optimal blank layout optimization model; Step S65: Optimizing the irregular blank layout based on the optimal blank layout optimization model.

[0020] The present invention evaluates the space utilization efficiency of the blank in the overall layout by calculating the global layout space filling rate of the topological layout optimization blank model, provides guidance for subsequent layout optimization, and obtains the blank layout space filling rate, which can help analyze the layout efficiency and provide a reference basis for discovering layout defects. Based on the blank layout space filling rate, the topological layout optimization blank model is locally identified for layout defects to find irregular areas or defects in the layout, improve the layout quality, obtain the layout defect blank area to identify the local area that needs to be adjusted, provide positioning for subsequent posture fine-tuning, and perform local posture iterative fine-tuning on the layout defect blank area to finely adjust the position and posture of the local blank, so as to solve the layout problem. Defects can be eliminated, the quality of nesting can be improved, local blank posture fine-tuning data can be generated to accurately adjust the nesting defect area, and the nesting effect can be improved. Global nesting optimization of the topological nesting optimization blank model based on the local blank posture fine-tuning data can comprehensively consider the overall nesting effect, further optimize the position and posture of the blank, and generate the optimal blank nesting optimization model to achieve the best overall nesting effect, improve space utilization and production efficiency, and perform irregular blank nesting optimization operations based on the optimal blank nesting optimization model to achieve the final nesting optimization, ensure that the nesting results meet the design requirements, complete the irregular blank nesting optimization operation, improve production efficiency and product quality, and achieve optimization processing of multi-graphic nesting. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the steps of a multi-pattern layout optimization method for irregular blanks according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0022] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0023] The present application example provides a multi-graphic layout optimization method for irregular blanks. The execution subject of the multi-graphic layout optimization method for irregular blanks includes but is not limited to the following: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio image management system, an information management system, and a cloud data management system.

[0024] See also Figures 1 to 4The present invention provides a multi-pattern layout optimization method for irregular blanks, comprising the following steps: Step S1: Acquire an irregular blank image in a region; reconstruct contour details of the irregular blank image in the region to generate a contour reconstructed blank image; Step S2: Performing a blank spatial distribution analysis on the contour reconstructed blank image to generate blank spatial distribution data; performing a three-dimensional point cloud modeling in the region based on the blank spatial distribution data to construct a three-dimensional model of the blank in the entire region; Step S3: performing graphic overlap statistics on the three-dimensional model of the whole area blank to obtain the graphic overlap rate between blanks; performing blank arrangement density evaluation based on the graphic overlap rate between blanks, thereby generating blank arrangement density evaluation data; Step S4: Perform dynamic nesting planning simulation on the three-dimensional model of the full-area blank, thereby generating a dynamic nesting planning blank model; Step S5: performing topological constraint analysis between graphics on the dynamic layout planning blank model according to the blank layout density evaluation data, and performing topological layout optimization, so as to obtain a topological layout optimization blank model; Step S6: local packing defect identification is performed on the topological packing optimization blank model to obtain a packing defect blank area; global packing optimization is performed based on the packing defect blank area to generate an optimal blank packing optimization model to perform irregular blank packing optimization operations.

[0025] The present invention can clearly present the shape and structure of the blank by reconstructing the contour details of the irregular blank image, provide an accurate data basis for subsequent analysis and processing, generate a contour reconstructed blank image to accurately locate the morphological characteristics of the irregular blank, and provide necessary information for subsequent spatial distribution analysis and three-dimensional modeling. Based on the contour reconstructed blank image, the blank spatial distribution analysis can understand the distribution of the blank in the region, provide data support for subsequent modeling and planning, and construct a three-dimensional model of the blank in the entire region to realize the three-dimensional representation of the irregular blank, which provides a basis for subsequent layout planning and optimization. By performing graphic overlap rate statistics on the blank three-dimensional model, Evaluate the space utilization between billets, provide data basis for layout density evaluation, generate billet layout density evaluation data to understand the density distribution of billets, provide guidance for layout planning, and optimize the layout structure. Dynamic layout planning simulation can simulate the billet layout in different situations, provide intuitive display and comparison for layout optimization, and obtain the topological layout optimization billet model to optimize the layout structure of billets, improve space utilization and production efficiency, identify the layout defect billet area, which can help find layout problems, provide improvement direction for global layout optimization, and generate the optimal billet layout optimization model to achieve the best layout effect, improve production efficiency and reduce waste.

[0026] In the embodiment of the present invention, refer to Figure 1, is a schematic flow chart of the steps of a multi-pattern layout optimization method for irregular blanks of the present invention. In this example, the steps of the multi-pattern layout optimization method for irregular blanks include: Step S1: Acquire an irregular blank image in a region; reconstruct contour details of the irregular blank image in the region to generate a contour reconstructed blank image; In this embodiment, a high-resolution camera or scanner is used to capture or scan an irregular blank image in the workpiece processing area to ensure that the image resolution is high enough to capture detailed contour information. The acquired original image is preprocessed by denoising, color correction and other preprocessing operations to improve the image quality. Common image preprocessing algorithms such as median filtering and Gaussian filtering can be used. Edge detection algorithms such as Canny edge detection and Sobel operator are used to extract the contour information of the blank from the preprocessed image. According to the threshold adjustment, the contour features with rich details are retained. The extracted contour edge is smoothed to reduce the impact of noise on the contour. Gaussian filtering, morphological filtering and other methods can be used. The contour edge is reconstructed by using interpolation or spline curve fitting and other methods to restore more detailed contour features. A machine learning model such as a convolutional neural network can be combined to intelligently reconstruct and repair the contour. The reconstructed contour features are fused with the original blank image to generate a blank image with richer contour details. Image synthesis technology can be used to superimpose the reconstructed contour information on the original image.

[0027] Step S2: Performing a blank spatial distribution analysis on the contour reconstructed blank image to generate blank spatial distribution data; performing a three-dimensional point cloud modeling in the region based on the blank spatial distribution data to construct a three-dimensional model of the blank in the entire region; In this embodiment, an image segmentation algorithm is used to identify and extract each billet block in the contour reconstructed billet image, and the contour and position information of each billet block are analyzed to obtain its spatial coordinates on the two-dimensional plane. According to the distribution of the billet blocks, a detailed billet spatial distribution data is generated, including the position, orientation, size and other information of each billet block. The above two-dimensional billet spatial distribution data is converted into three-dimensional point cloud data in combination with camera calibration parameters. Point cloud processing algorithms such as voxelization and smoothing filtering are used to pre-process, denoise and optimize the three-dimensional point cloud data. A three-dimensional reconstruction algorithm is used. For example, based on the Poisson surface reconstruction method, the optimized 3D point cloud is fitted into a smooth 3D mesh model, and finally a complete 3D blank model is obtained, which describes the 3D spatial distribution of all blanks in the entire processing area, removes overlapping areas, optimizes model details, and forms a complete, non-repetitive full-area 3D blank model. This 3D model not only contains the shape and position information of the blank, but also retains the material, texture and other details in the original contour reconstructed image. This full-area 3D blank model provides high-quality input data for subsequent cutting optimization, processing path planning, etc.

[0028] Step S3: performing graphic overlap statistics on the three-dimensional model of the whole area blank to obtain the graphic overlap rate between blanks; performing blank arrangement density evaluation based on the graphic overlap rate between blanks, thereby generating blank arrangement density evaluation data; In this embodiment, spatial geometric operations are performed on each pair of adjacent three-dimensional models of blanks to calculate the overlapping volume between them, and the overlapping volume of each pair of blanks is ratioed to their respective total volumes to obtain the graphic overlap rate between them. The overlapping rates between all blanks are statistically analyzed to form a complete set of graphic overlap rate data between blanks. The above-mentioned graphic overlap rate data between blanks are analyzed to calculate the average overlap rate of other blanks around each blank. The average overlap rate is used as an indicator to measure the density of blank arrangement. The higher the density, the greater the average overlap rate. According to the average overlap rate of blanks in different areas, the blank arrangement density of the entire processing area is evaluated, and the evaluation result is converted into a blank arrangement density heat map to intuitively reflect the blank arrangement density in different areas. The above-mentioned graphic overlap rate data between blanks and the blank arrangement density heat map are integrated into a complete set of blank arrangement density evaluation data, which not only includes the precise overlap rate between each pair of blanks, but also gives the distribution of the blank arrangement density in the entire processing area.

[0029] Step S4: Perform dynamic nesting planning simulation on the three-dimensional model of the full-area blank, thereby generating a dynamic nesting planning blank model; In this embodiment, the prepared three-dimensional model of the full-area blanks is input into the dynamic nesting planning algorithm for processing. The algorithm will automatically analyze the characteristics of the blanks and generate an optimal dynamic nesting plan based on the set optimization goals. The plan not only includes the final position of each blank, but also gives the order and time arrangement of their placement. Through simulation calculation of the dynamic nesting planning process, a complete dynamic nesting planning blank model is finally obtained. The blank arrangement plan obtained by the dynamic planning algorithm is mapped back to the constructed full-area three-dimensional model. Using 3D rendering technology, such as OpenGL, WebGL, etc., the planned blank placement is visualized, supporting observation from different perspectives and levels, and providing dynamic interactive functions to facilitate analysis and evaluation of planning results.

[0030] Step S5: performing topological constraint analysis between graphics on the dynamic layout planning blank model according to the blank layout density evaluation data, and performing topological layout optimization, so as to obtain a topological layout optimization blank model; In this embodiment, the geometric information and relative position relationship of each blank are extracted from the dynamic layout planning blank model. According to the aforementioned blank arrangement density evaluation data, the area with too high arrangement density is identified. For these high-density areas, the spatial relationship between adjacent blanks, such as proximity, overlap, etc., is analyzed, and a topological constraint model is established. Heuristic algorithms or optimization algorithms, such as genetic algorithms, taboo search, etc., are used to optimize the placement and direction of the blanks. The algorithm goal is to minimize the congestion in high-density areas and improve the overall space utilization while meeting various constraints. During the algorithm iteration process, the position and direction of the blanks are dynamically adjusted according to the topological constraint model to achieve the optimal arrangement effect.

[0031] Step S6: local packing defect identification is performed on the topological packing optimization blank model to obtain a packing defect blank area; global packing optimization is performed based on the packing defect blank area to generate an optimal blank packing optimization model to perform irregular blank packing optimization operations.

[0032] In this embodiment, the topological layout optimization blank model is subjected to layout space identification, and the places with large space gaps are detected and marked as defective areas, which need to be re-optimized. A global optimization algorithm, such as a genetic algorithm, a simulated annealing algorithm, etc., is used to optimize the layout defective areas. The algorithm goal is to maximize the space utilization of the area and eliminate local layout defects under the premise of meeting various constraints. The optimization algorithm will adjust the position and direction of the blank in the defective area until the requirements are met. At the same time, the optimization algorithm will also consider the global layout effect to avoid the overall layout quality degradation caused by local optimization.

[0033] In this embodiment, refer to Figure 2, is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Step S11: Acquire the image of the irregular blank in the area; Step S12: performing global brightness average calculation on the irregular blank image in the region to obtain an average brightness value of the image; Step S13: performing dark area recognition on the irregular blank image in the region and extracting the dark area image; Step S14: performing dark portion brightness balance enhancement processing on the dark portion image according to the average brightness value of the image, thereby obtaining a brightness enhanced blank image; Step S15: reconstructing the outline details of the brightness enhanced blank image to generate a contour reconstructed blank image.

[0034] In this embodiment, a high-resolution camera or scanner is used to photograph or scan the irregular blanks in the processing area to obtain the original blank image data, ensure good lighting conditions, avoid overexposure or overdarkness of the image, and for blanks with complex shapes and large surface undulations, it is necessary to photograph or scan from multiple angles to obtain comprehensive image information. A global brightness analysis is performed on the acquired blank image to calculate the average brightness value of the entire image. An image processing library, such as the cv2.mean() function of OpenCV, can be used to implement statistical calculation of global brightness. The average brightness value will be used as a reference standard for subsequent brightness equalization processing. Each pixel point of the blank image is traversed to identify areas with brightness values ​​below a certain threshold and determine them as dark areas. Image segmentation technology, such as Otsu's method, is used to separate the image data of the dark area as the subsequent bright area. The target area with brightness balance is selected, and the dark area information of the original image is retained to prepare for the next step of brightness enhancement processing. The gap between the overall brightness value of the dark image and the aforementioned global average brightness value is calculated. According to this gap, the dark image is subjected to linear or nonlinear brightness enhancement processing to improve the overall brightness of the dark area. Technical means such as histogram equalization and gamma correction can be used to achieve brightness balance enhancement of the dark image. After this step of processing, the billet image with brightness enhancement can be obtained. Image processing algorithms such as edge detection and contour extraction can be used to reconstruct clear contour details from the billet image with brightness enhancement. Methods such as Canny operator and Sobel operator can be used to enhance the edge contour information of the billet image, and finally a billet image with contour reconstruction is generated, which provides a good visual basis for subsequent processing path planning and control.

[0035] In this embodiment, the specific steps of step S15 are: Performing edge contour visual recognition on the brightness enhanced blank image to extract the edge contour line of each blank; The thickness of each blank is identified one by one for the edge contour line of each blank, and the thickness shape parameters of each blank contour are generated; Based on each blank contour thickness and shape parameter, contour refinement filtering is performed on the brightness enhanced blank image to obtain a detail enhanced blank image; The detail-enhanced blank image is analyzed for abnormally protruding regions to obtain protruding contour sub-regions; Fitting local salient feature points to the salient contour sub-region to obtain multiple local feature points; Perform spline interpolation smoothing on multiple local feature points to obtain smooth feature contour lines; The detail enhanced blank image is reconstructed by performing contour superposition image reconstruction according to the smooth characteristic contour line to generate a contour reconstructed blank image.

[0036] In this embodiment, the Canny edge detection algorithm is used to extract the edge of the billet image after brightness enhancement to obtain the contour information of the image. Through the connected domain analysis, each independent billet contour is segmented to extract the independent edge contour of each billet. For the billet with complex shape, there are multiple edge contours, and further contour merging processing is required. Each extracted billet edge contour is traversed to calculate its geometric feature parameters such as width, length, and area. According to these geometric parameters, a quantitative description model of the billet contour thickness is established to obtain the contour thickness shape parameters of each billet. These parameters will provide a basis for subsequent contour refinement filtering and feature point extraction. According to the obtained billet contour thickness parameters, the brightness enhanced image is subjected to targeted contour refinement filtering. For the billet area with rough contour, high-pass filtering is used to enhance the edge; for the area with fine contour, low-pass filtering is used to protect the details. After this step, the billet image with detail enhancement can be obtained, which is ready for the subsequent feature point extraction. The local difference analysis is performed to identify the areas with prominent contours. A gradient-based saliency detection algorithm or an abnormal area extraction method based on region growing can be used. Through this step, the local sub-region with the most prominent contour in the billet image can be locked. For the extracted prominent contour sub-region, corner point detection, key point extraction and other algorithms are used to obtain local key feature points. These feature points reflect the key inflection points or turning positions of the billet contour line, providing a basis for subsequent contour line smoothing. Through the extraction of multiple local feature points, the overall shape of the billet contour can be described more accurately. The contour line is smoothed by using the spline function interpolation method. Through spline interpolation, smooth and natural contour line segments can be generated between feature points to avoid sharp corners. Finally, the smoothed feature contour line is obtained, which is ready for the next step of contour line superposition and reconstruction. The obtained smooth feature contour line is superimposed on the detail-enhanced billet image. Through the precise reconstruction of the contour line, the areas lacking details in the billet image can be supplemented and repaired to generate a final billet image with rich contour details.

[0037] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Step S21: visually identifying the blank with the reconstructed contour, and marking it in sequence to obtain a blank number sequence; Step S22: segmenting the contour reconstructed blank image into blank bounding boxes one by one based on the blank number sequence to generate a plurality of blank bounding boxes; Step S23: Calculate the positions of the multiple blanks one by one for the bounding boxes of the blanks, and generate the coordinates of the position of each blank; Step S24: performing a blank space distribution analysis on the contour reconstructed blank image according to each blank position coordinate to generate blank space distribution data; Step S25: performing a three-dimensional morphological feature analysis on a plurality of blank boundary frames to generate a three-dimensional morphological feature of the blank; Step S26: Performing three-dimensional point cloud modeling in the region for the blank spatial distribution data according to the three-dimensional morphological characteristics of the blank, and constructing a three-dimensional model of the blank in the entire region.

[0038] In the present embodiment, image processing and machine vision technology are used to segment and identify the contour reconstruction blank image, and edge detection, region growing and other methods can be used to segment each independent blank region, and the segmented blank regions are numbered and marked by connected region analysis and other methods to generate a blank numbering sequence, and the unique numbering of each blank can be determined. According to the blank numbering sequence obtained in the previous step, bounding box extraction is performed on each numbered blank region in turn on the contour reconstruction blank image, and a target detection algorithm such as YOLO, Faster R-CNN, etc. can be used to perform bounding box regression on the blank regions of the image one by one, and the circumscribed rectangular frame coordinates of each blank are obtained, and the bounding box coordinates of all blanks are saved to form a set of multiple blank bounding boxes. Traverse the multiple blank bounding boxes obtained in the previous step, calculate the center point coordinates of each bounding box, you can directly take the center point coordinates (x, y) of the bounding box, or calculate the average value of the four vertices of the bounding box as the center point coordinates, save the position coordinates (x, y) of each blank, and form a blank position coordinate set. According to the blank position coordinate set, draw the position points of each blank on the contour reconstructed blank image. By analyzing the spatial distribution of these position points, you can get the overall distribution of the blank on the blank, such as distribution density, aggregation degree, etc. Save these spatial distribution feature data to form blank spatial distribution data. Based on multiple blanks, The bounding box is combined with 3D scanning or other 3D imaging technology to model and analyze the 3D shape of each blank, and the geometric features such as size, volume, surface area, etc. of the blank are extracted from the 3D model as the 3D morphological features of the blank. These 3D morphological feature data are saved to form a set of 3D morphological features of the blank. The 3D geometric model of each blank is reconstructed in 3D space using the blank spatial distribution data and 3D morphological features. The 3D models of these blanks are assembled together to construct a full-area 3D model of the blank with the entire contour reconstructed. This full-area 3D model contains the position and 3D morphological information of each blank, which can be used for subsequent analysis and application.

[0039] In this embodiment, reference Figure 4 The above is a schematic flow chart of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Step S31: Calculate the space utilization rate of the three-dimensional model of the blank in the whole area to obtain the irregular space utilization rate; Step S32: identifying the blank intersection points on the full-area blank three-dimensional model to generate blank intersection points; Step S33: performing graphic overlap statistics according to the intersection points of the blanks to obtain the graphic overlap rate between the blanks; Step S34: calculating the regional gap area ratio of the three-dimensional model of the whole-region blank according to the utilization rate of the irregular space and the overlap rate of the patterns between the blanks, so as to generate regional space gap data; Step S35: Evaluate the blank arrangement density of the regional spatial gap data, thereby generating blank arrangement density evaluation data.

[0040] In the present embodiment, the total volume of the entire area is calculated using a three-dimensional geometric calculation method, the sum of the volumes of all stock three-dimensional models is counted, and compared with the total volume to obtain space utilization, all stock in the three-dimensional model of the stock in the entire area are traversed, and whether there is geometric overlap or contact between each stock is detected, for stock pairs with overlap or contact, their intersection point coordinates are identified, and methods such as a geometric intersection algorithm or a minimum distance calculation can be used to save all identified stock intersection point coordinates to form a stock intersection point set, and overlapping area calculation is performed for each pair of stock with an intersection point, and the ratio of the overlapping area of ​​each pair of stock to each area is calculated to obtain the graphic overlap rate between the stock, and all The graphic overlap rate of the blank pair is saved to form the overall graphic overlap rate data between the blanks, and the overall regional spatial gap ratio is calculated. Specifically, the total volume occupied by the blanks can be calculated first, and then the spatial gap volume can be obtained by subtracting the blank volume from the total volume. Finally, the ratio of the spatial gap volume to the total volume is calculated as the regional spatial gap data. The obtained regional spatial gap data is combined with the irregular space utilization rate to evaluate the overall blank arrangement density. The higher the spatial gap ratio, the lower the arrangement density; the lower the irregular space utilization rate, the lower the arrangement density. Combining these two indicators, an overall blank arrangement density evaluation data is given, which reflects the filling degree of the blank in the entire area.

[0041] In this embodiment, step S4 includes the following steps: Step S41: Performing a position adjustment feasibility analysis on the three-dimensional morphological features of the blank, thereby obtaining position adjustment parameters for each blank; Step S42: adjusting the overlap loss calculation according to each blank position adjustment parameter to obtain blank overlap loss data; Step S43: performing iterative decision-making on the position adjustment of the three-dimensional model of the whole-region blank according to the blank overlap loss data, so as to obtain an optimized blank position adjustment scheme; Step S44: Based on the optimized blank position adjustment scheme, a dynamic nesting planning simulation is performed on the blank three-dimensional model of the entire region, thereby generating a dynamic nesting planning blank model.

[0042] In this embodiment, based on the three-dimensional model of the full-area billet generated previously, the three-dimensional morphological characteristics of each billet are analyzed one by one, including size, shape, position, etc. For each billet, it is evaluated whether there is overlap or overlap in the current position, as well as the direction and range that can be adjusted. According to the geometric characteristics and mutual position relationship of the billets, the position adjustment parameters of each billet are given, including the adjustable direction and amplitude. According to the position adjustment parameters of each billet, an attempt is made to fine-tune the position of the billet three-dimensional model. For the adjusted billet model, the overall space utilization rate, the overlap rate between billets and other indicators are recalculated, and these geometric indicators before and after the adjustment are compared to obtain the overlapping loss data generated in the adjustment process. The overlapping loss data reflects the decrease in space utilization and the increase in regional gaps caused by the position adjustment process. To solve the problem, the three-dimensional model of the whole area blank is optimized and adjusted. The heuristic algorithm or the group optimization algorithm can be used. Through iteration, the position of each blank is continuously adjusted to find the solution with the least loss. In each iteration, it is necessary to recalculate the space utilization rate, overlap rate and other indicators, and update the overlapping loss data. When the loss data converges to the minimum value, the position adjustment plan is output as the optimization result. According to the optimized position adjustment plan, the three-dimensional model of the whole area blank is adjusted, and the adjusted blank model is input into the dynamic nesting planning algorithm to simulate the placement process of the blank in the production process. The dynamic nesting planning algorithm will generate an optimal dynamic nesting plan according to the geometric characteristics of the blank and the relative position relationship, and output this dynamic nesting plan to form the final dynamic nesting planning blank model.

[0043] In this embodiment, the specific steps of step S43 are: Performing position adjustment range analysis on multiple blank boundary boxes according to the three-dimensional morphological features of the blanks to obtain the position adjustment range of each blank; According to the blank overlap loss data, the collision range of each blank position adjustment range is excluded to obtain the collision exclusion position adjustment range; The collision elimination position adjustment range is used to make iterative decisions on the position adjustment of the full-area blank three-dimensional model, and multiple blank position adjustment schemes are obtained; The optimal convergence decision of the blank position is made for multiple blank position adjustment schemes to obtain the optimized blank position adjustment scheme.

[0044] In this embodiment, according to the position and spatial relationship of each billet, the direction and range in which it can be adjusted are evaluated, and the position adjustment range of each billet is given. The position adjustment range needs to consider the geometric characteristics of the billet itself and the relationship with other billets. The three-dimensional model of the billet is tried to be fine-tuned. For the adjusted billet model, the overall space utilization rate, the overlap rate between billets and other indicators are calculated to obtain the overlapping loss data. According to the overlapping loss data, the position adjustment range of each billet is trimmed and excluded to obtain the adjustment range that can avoid collision. A heuristic algorithm or a group optimization algorithm is used to continuously adjust the position of each billet in an iterative manner to find a solution with the least loss. In each iteration, it is necessary to recalculate the space utilization rate, the overlap rate and other indicators, and update the overlapping loss data. When the loss data converges to the minimum value, the position adjustment scheme is output as a feasible optimization result. Through multiple iterations, multiple different billet position adjustment schemes can be obtained. The multiple position adjustment schemes obtained in the previous step are compared to evaluate their differences in overall performance indicators. The indicators include overall space utilization rate, overlapping loss, production efficiency, etc. The scheme with the best comprehensive performance is selected as the final optimized billet position adjustment scheme.

[0045] In this embodiment, the specific steps of step S5 are: Step S51: performing spatial voxel processing on the dynamic layout planning blank model to generate a regional blank grid; Step S52: Performing position and posture analysis on the regional stock grid to generate stock position and posture data; Step S53: performing inter-graphic topological constraint analysis on the blank position and posture data according to the blank arrangement density evaluation data to generate inter-blank topological constraint relationships; Step S54: topologically optimizing the dynamic nesting planning blank model based on the topological constraint relationship between blanks, thereby obtaining a topologically optimized blank model.

[0046] In this embodiment, the entire working area is divided into a plurality of small three-dimensional cubic units (voxels). Each three-dimensional model of the stock is discretized and mapped to the corresponding voxel unit. A three-dimensional stock mesh model representing the entire working area is obtained. For each voxel unit, the stock information contained therein is analyzed, and the position coordinates and rotation angle of the stock are recorded. The position and posture data of all the stock are sorted into a data table as input for subsequent processing. According to the distribution density of the stock in the working area, the mutual relationship and restrictions between them are analyzed. For example, there are problems of extrusion or interference between adjacent stock, and certain position constraints need to be imposed. Through topological analysis, the mutual influence relationship between each stock is determined, and corresponding constraints are established. The aforementioned stock position and posture data and topological constraint relationship are used as input. Using a heuristic algorithm or an optimization solution method, try to adjust the position and posture of the stock to meet the topological constraints. The goal is to maximize the overall space utilization while ensuring that no stock interference occurs. Through iterative optimization, a layout optimization result that meets all topological constraints is finally obtained.

[0047] In this embodiment, the specific steps of step S6 are: Step S61: Calculate the global arrangement space filling rate of the topological layout optimization blank model, so as to obtain the blank arrangement space filling rate; Step S62: performing local layout defect identification on the topological layout optimization blank model based on the blank layout space filling rate to obtain a layout defect blank area; Step S63: performing iterative fine-tuning of the local posture of the defective blank area to generate local blank posture fine-tuning data; Step S64: performing global layout optimization on the topological layout optimization blank model based on the local blank posture fine-tuning data to generate an optimal blank layout optimization model; Step S65: Optimizing the irregular blank layout based on the optimal blank layout optimization model.

[0048] In this embodiment, the space filling rate of the stock model after topological layout optimization in the entire working area is calculated. The filling rate can be obtained by counting the volume percentage occupied by the stock in all voxel units. This indicator reflects the overall effect of the layout optimization and provides an important basis for subsequent steps. The filling rate of each local area is analyzed, and the areas with low filling rate are identified. There are some layout defects in these areas, such as stock overlap, excessive stock gaps, etc. These local defective areas are marked as the focus of the next step of optimization. For the identified defective areas, the stock therein is locally adjusted and optimized, and the position and rotation angle of the stock can be fine-tuned to reduce the size of the stock. The gap between materials is reduced and overlapping is eliminated. The nesting effect of local areas is gradually improved through iterative optimization. The fine-tuning data of these local blank postures are recorded. The obtained local posture fine-tuning data is applied to the topological nesting optimization model. Through global optimization and adjustment, the overall space utilization and nesting quality are further improved. Finally, an optimal blank nesting optimization model that meets various constraints is obtained. The generated optimal nesting model is applied to the actual irregular blank nesting operation. According to the blank position and posture information of the model, the production equipment is arranged for precise blank placement. Through this optimized nesting, the production efficiency and resource utilization can be greatly improved.

[0049] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0050] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A multi-pattern layout optimization method for irregular blanks, characterized in that: The following steps are involved: Step S1: Acquire an irregular blank image in a region; reconstruct contour details of the irregular blank image in the region to generate a contour reconstructed blank image; Step S2: Performing a blank spatial distribution analysis on the contour reconstructed blank image to generate blank spatial distribution data; performing a three-dimensional point cloud modeling in the region based on the blank spatial distribution data to construct a three-dimensional model of the blank in the entire region; Step S3: performing graphic overlap statistics on the three-dimensional model of the whole area blank to obtain the graphic overlap rate between blanks; performing blank arrangement density evaluation based on the graphic overlap rate between blanks, thereby generating blank arrangement density evaluation data; Step S4: Perform dynamic nesting planning simulation on the three-dimensional model of the full-area blank, thereby generating a dynamic nesting planning blank model; Step S5: performing topological constraint analysis between graphics on the dynamic layout planning blank model according to the blank layout density evaluation data, and performing topological layout optimization, so as to obtain a topological layout optimization blank model; Step S6: local pattern arrangement defect identification is performed on the topological pattern arrangement optimization blank model to obtain the pattern arrangement defect blank area; global pattern arrangement optimization is performed based on the pattern arrangement defect blank area to generate the optimal blank pattern arrangement optimization model to perform irregular blank pattern arrangement optimization operations.

2. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: Acquire the image of the irregular blank in the area; Step S12: performing global brightness average calculation on the irregular blank image in the region to obtain an average brightness value of the image; Step S13: performing dark area recognition on the irregular blank image in the region and extracting the dark area image; Step S14: performing dark portion brightness balance enhancement processing on the dark portion image according to the average brightness value of the image, thereby obtaining a brightness enhanced blank image; Step S15: reconstructing the outline details of the brightness enhanced blank image to generate a contour reconstructed blank image.

3. The multi-pattern layout optimization method for irregular blanks according to claim 2, characterized in that: The specific steps of step S15 are: Performing edge contour visual recognition on the brightness enhanced blank image to extract the edge contour line of each blank; The thickness of each blank is identified one by one for the edge contour line of each blank, and the thickness shape parameters of each blank contour are generated; Based on each blank contour thickness and shape parameter, contour refinement filtering is performed on the brightness enhanced blank image to obtain a detail enhanced blank image; The detail-enhanced blank image is analyzed for abnormally protruding regions to obtain protruding contour sub-regions; Fitting local salient feature points to the salient contour sub-region to obtain multiple local feature points; Perform spline interpolation smoothing on multiple local feature points to obtain smooth feature contour lines; The detail enhanced blank image is reconstructed by performing contour superposition image reconstruction according to the smooth characteristic contour line to generate a contour reconstructed blank image.

4. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: visually identifying the blank with the reconstructed contour, and marking it in sequence to obtain a blank number sequence; Step S22: segmenting the contour reconstructed blank image into blank bounding boxes one by one based on the blank number sequence to generate a plurality of blank bounding boxes; Step S23: Calculate the positions of the multiple blanks one by one for the bounding boxes of the blanks, and generate the coordinates of the position of each blank; Step S24: performing a blank space distribution analysis on the contour reconstructed blank image according to each blank position coordinate to generate blank space distribution data; Step S25: performing a three-dimensional morphological feature analysis on a plurality of blank boundary frames to generate a three-dimensional morphological feature of the blank; Step S26: Performing three-dimensional point cloud modeling in the region for the blank spatial distribution data according to the three-dimensional morphological characteristics of the blank, and constructing a three-dimensional model of the blank in the entire region.

5. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: Calculate the space utilization rate of the three-dimensional model of the blank in the whole area to obtain the irregular space utilization rate; Step S32: identifying the blank intersection points on the full-area blank three-dimensional model to generate blank intersection points; Step S33: performing graphic overlap statistics according to the intersection points of the blanks to obtain the graphic overlap rate between the blanks; Step S34: calculating the regional gap area ratio of the three-dimensional model of the whole-region blank according to the utilization rate of the irregular space and the overlap rate of the patterns between the blanks, so as to generate regional space gap data; Step S35: Evaluate the blank arrangement density of the regional spatial gap data, thereby generating blank arrangement density evaluation data.

6. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S5 are: Step S41: Performing a position adjustment feasibility analysis on the three-dimensional morphological features of the blank, thereby obtaining position adjustment parameters for each blank; Step S42: adjusting the overlap loss calculation according to each blank position adjustment parameter to obtain blank overlap loss data; Step S43: performing iterative decision-making on the position adjustment of the three-dimensional model of the whole-region blank according to the blank overlap loss data, so as to obtain an optimized blank position adjustment scheme; Step S44: Based on the optimized blank position adjustment scheme, a dynamic nesting planning simulation is performed on the blank three-dimensional model of the entire region, thereby generating a dynamic nesting planning blank model.

7. The multi-pattern layout optimization method for irregular blanks according to claim 6, characterized in that: The specific steps of step S43 are: Performing position adjustment range analysis on multiple blank boundary boxes according to the three-dimensional morphological features of the blanks to obtain the position adjustment range of each blank; According to the blank overlap loss data, the collision range of each blank position adjustment range is excluded to obtain the collision exclusion position adjustment range; The collision elimination position adjustment range is used to make iterative decisions on the position adjustment of the full-area blank three-dimensional model, and multiple blank position adjustment schemes are obtained; The optimal convergence decision of the blank position is made for multiple blank position adjustment schemes to obtain the optimized blank position adjustment scheme.

8. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: performing spatial voxel processing on the dynamic layout planning blank model to generate a regional blank grid; Step S52: Performing position and posture analysis on the regional stock grid to generate stock position and posture data; Step S53: performing inter-graphic topological constraint analysis on the blank position and posture data according to the blank arrangement density evaluation data to generate inter-blank topological constraint relationships; Step S54: topologically optimizing the dynamic nesting planning blank model based on the topological constraint relationship between blanks, thereby obtaining a topologically optimized blank model.

9. The multi-pattern layout optimization method for irregular blanks according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: Calculate the global arrangement space filling rate of the topological layout optimization blank model, so as to obtain the blank arrangement space filling rate; Step S62: performing local layout defect identification on the topological layout optimization blank model based on the blank layout space filling rate to obtain a layout defect blank area; Step S63: performing iterative fine-tuning of the local posture of the defective blank area to generate local blank posture fine-tuning data; Step S64: performing global layout optimization on the topological layout optimization blank model based on the local blank posture fine-tuning data to generate an optimal blank layout optimization model; Step S65: Optimizing the irregular blank layout based on the optimal blank layout optimization model.

Images