A stone-like brick arrangement scheme generation method, device, equipment and storage medium

Abstract

The present application relates to the technical field of building facade design, and more particularly to a stone-like brick arrangement scheme generation method, device, equipment and storage medium; based on a preset deep learning model and a preliminary selection plane set, picture calibration data is reconstructed to obtain a quantitative facade drawing; according to a preset bottom starting point coordinate, a preset top end point coordinate and a preset stone-like brick specification parameter, the quantitative facade drawing is identified to obtain an initial paving range; according to a preset construction constraint, a preset multi-objective function and a preset heuristic model, the initial paving range and the quantitative facade drawing are solved to obtain an arrangement scheme; through multiple frames of pictures and benchmark algorithm calibration data, a quantitative drawing is quickly reconstructed by deep learning to replace manual work, paving range and arrangement scheme are accurately planned, doors and windows are avoided, waste is reduced, material utilization and aesthetics are improved, and engineering efficiency, quality and economy are comprehensively optimized.

Description

Technical Field

[0001] This invention relates to the field of architectural facade design technology, and in particular to a method, apparatus, equipment and storage medium for generating a layout scheme of imitation stone bricks. Background Technology

[0002] In building facade decoration projects, decorative bricks and imitation stone slabs are widely used due to their combination of aesthetics and practicality. However, existing design and construction processes still face specific technical bottlenecks. In particular, the layout and cutting schemes often rely on manual experience, making it difficult to simultaneously achieve the three core objectives of "minimum brick usage, minimum cutting waste, and consistency of texture and joint width." Furthermore, it fails to accurately avoid door and window openings, reducing material utilization and affecting the aesthetic appeal of the decoration.

[0003] Existing research and patents cover areas such as visual 3D reconstruction of building facades, laser 3D reconstruction, generation of wall cladding tile models, layout rules, and paving equipment and methods. However, no end-to-end solution has yet been formed. It is impossible to obtain facade data from ordinary photos without relying on existing BIM and CAD data, and then complete the combination and cutting optimization for given panel specifications. This has become a key factor restricting the efficiency and quality improvement of facade decoration projects. Summary of the Invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a method, apparatus, equipment and storage medium for generating a layout scheme of imitation stone bricks.

[0005] The first aspect of this invention provides a method for generating a layout scheme for imitation stone bricks, comprising: acquiring multi-frame captured image data, and calculating the multi-frame captured image data according to preset benchmark data, preset plane detection algorithm and preset filtering rules to obtain image calibration data and a preliminary set of planes; reconstructing the image calibration data based on a preset deep learning model and the preliminary set of planes to obtain a quantized elevation drawing; identifying the quantized elevation drawing according to preset bottom start coordinates, preset top end coordinates and preset imitation stone brick specification parameters to obtain an initial paving range; and solving the initial paving range and the quantized elevation drawing according to preset structural constraints, preset multi-objective functions and preset heuristic models to obtain a layout scheme.

[0006] Furthermore, the step of calculating the multi-frame captured image data according to preset benchmark data, preset planar detection algorithm, and preset filtering rules to obtain image calibration data and a preliminary set of planar features includes: analyzing the multi-frame captured image data according to the planar detection algorithm and benchmark data to obtain the total area and the set of facade planar features; calculating the total area according to a preset proportional coefficient to obtain an area threshold; and filtering the set of facade planar features according to the area threshold and filtering rules to obtain a preliminary set of planar features.

[0007] Furthermore, the step of analyzing multi-frame captured image data based on a plane detection algorithm and reference data to obtain the total area and the set of exterior facade planes includes: dimensional calibration of multi-frame captured image data based on reference data to obtain image calibration data; and detection of the image calibration data based on a plane detection algorithm, a preset distance threshold, and a preset minimum number of points to obtain the set of exterior facade planes. Calculate the area of ​​the set of exterior facade plans to obtain the total area.

[0008] Furthermore, the process of reconstructing the image calibration data based on a preset deep learning model and a preliminary set of planes to obtain quantized elevation drawings includes: performing secondary filtering on the preliminary set of planes according to preset plane parameters to obtain a reference plane; reconstructing the image calibration data based on a deep learning model to obtain building exterior elevation drawings; and annotating the building exterior elevation drawings according to the reference plane and preset door and window opening boundaries to obtain quantized elevation drawings.

[0009] Furthermore, the step of identifying the quantized elevation drawing based on preset bottom start coordinates, preset top end coordinates, and preset imitation stone brick specification parameters to obtain the initial paving range includes: generating a first horizontal straight line based on the bottom start coordinates; and generating a second horizontal straight line based on the top end coordinates. The vertical range is determined based on the first and second horizontal lines; the range of the quantified elevation drawings is analyzed to obtain the elevation range; it is determined whether the vertical range is less than or equal to the elevation range; if the vertical range is less than or equal to the elevation range, the quantified elevation drawings are identified based on the vertical range and the specifications of the imitation stone bricks to obtain the initial paving range.

[0010] Furthermore, the step of solving the initial paving range and quantized elevation drawings based on preset structural constraints, preset multi-objective functions, and preset heuristic models to obtain a layout scheme includes: performing strip decomposition on the initial paving range and quantized elevation drawings based on structural constraints and preset cutting dimensions to obtain a first dense point cloud and a second dense point cloud network set; solving the first dense point cloud network set and the second dense point cloud network set based on multi-objective functions, heuristic models, preset engineering constraints, and preset planning models to obtain a feasible solution; and optimizing the feasible solution based on preset comprehensive objectives to obtain a layout scheme.

[0011] Furthermore, the step of decomposing the initial tiling range and the quantized elevation drawing into strips according to the structural constraints and preset cutting dimensions to obtain a first dense point cloud and a second dense point cloud network set includes: identifying the quantized elevation drawing according to the structural constraints and vertical range to obtain multiple polygon reduction ranges; performing Boolean operations on the initial tiling range and the multiple polygon reduction ranges to obtain multiple tilable polygon regions; decomposing the quantized elevation drawing into strips according to the cutting dimensions to obtain a first dense point cloud; and decomposing the multiple tilable polygon regions into strips according to the cutting dimensions to obtain a second dense point cloud network set.

[0012] Furthermore, a device for generating a layout scheme for imitation stone bricks includes: a calculation module for acquiring multi-frame captured image data and calculating the multi-frame captured image data according to preset benchmark data, preset plane detection algorithm, and preset filtering rules to obtain image calibration data and a preliminary set of planes; a reconstruction module for reconstructing the image calibration data based on a preset deep learning model and the preliminary set of planes to obtain a quantized elevation drawing; a drawing recognition module for recognizing the quantized elevation drawing according to preset bottom start coordinates, preset top end coordinates, and preset imitation stone brick specification parameters to obtain an initial paving range; and a layout scheme generation module for solving the initial paving range and the quantized elevation drawing according to preset structural constraints, preset multi-objective functions, and preset heuristic models to obtain a layout scheme.

[0013] Furthermore, a device for generating a layout scheme for imitation stone bricks is provided, the device comprising: a memory and at least one processor, the memory storing instructions; at least one processor invokes the instructions in the memory to cause the computer device to execute the various steps of the method for generating a layout scheme for imitation stone bricks described above.

[0014] Furthermore, a computer-readable storage medium is provided, on which instructions are stored, which, when executed by a processor, implement the various steps of the above-described method for generating a layout scheme for imitation stone bricks.

[0015] In the technical solution of this invention, multiple frames of images are combined with benchmark data and algorithm processing to generate quantitative calibration data and a preliminary set of planes focusing on the construction area, laying a reliable data foundation. The deep learning model relies on accurate spatial benchmarks to quickly reconstruct quantified drawings containing facade shapes and door and window information, replacing manual drawing, shortening the cycle and reducing errors. The initial paving range based on coordinates and full-specification parameters of bricks is adapted to construction requirements and takes into account the reuse of leftover materials, reducing material waste. The layout scheme solved by structural constraints, multi-objective functions and heuristic models can accurately avoid doors and windows, balance the minimum amount of bricks used, the minimum cutting loss and the consistency of texture joint width, replace manual rough planning, improve material utilization and decorative aesthetics, and comprehensively optimize project efficiency, quality and economy. Attached Figure Description

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 A first flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 2 A second flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 3 The third flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 4 The fourth flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 5 The fifth flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 6 The sixth flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 7 The seventh flowchart of a method for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 8 A schematic diagram of a device for generating a layout scheme for imitation stone bricks provided in an embodiment of the present invention; Figure 9 This is a schematic diagram of a device for generating a layout scheme for imitation stone bricks, provided in an embodiment of the present invention. Detailed Implementation

[0017] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] A method, apparatus, equipment, and storage medium for generating a layout scheme for imitation stone bricks are disclosed, applicable to the facade renovation of low-rise irregularly shaped buildings. The method involves acquiring and scaling multi-view images of the facade to obtain a metric model with realistic scale. Based on this, the paving area is extracted and constraints such as joint width, texture direction, and rotation angle are set. A combined optimization model is constructed with the goals of minimizing board usage, cutting and waste, consistent joint width, and texture alignment. The model solves for the position, rotation angle, and cutting dimensions of each board. Numbered layout diagrams, cutting lists, and construction positioning data are generated, reducing measurement and layout costs and improving the efficiency and quality consistency of facade decoration construction. This method is particularly suitable for renovation projects of low-rise buildings and irregularly shaped facade decoration scenarios, enabling automation of board layout and standardization of the construction process, and has high engineering promotion value.

[0019] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the method for generating a layout scheme for imitation stone bricks according to the present invention includes: 101. Acquire multi-frame captured image data, and calculate the multi-frame captured image data according to preset reference data, preset plane detection algorithm and preset filtering rules to obtain image calibration data and initial set of planes; In this embodiment, multiple images are taken around each target facade to form multi-frame image data; relying on preset benchmark data, plane detection algorithm and screening rules, the image calibration data and the initial selected plane set are accurately obtained. The image calibration data has quantitative attributes, and the initial selected plane set focuses on the effective construction area, providing reliable support for subsequent steps such as imitation stone brick layout and material calculation. 102. Based on the preset deep learning model and the initial set of planes, the image calibration data is reconstructed to obtain quantized elevation drawings; In this embodiment, the initial set of planes provides a precise spatial benchmark, laying a high-quality foundation for model reconstruction. The deep learning model, with its strong feature extraction capabilities, efficiently restores the three-dimensional geometric details of the facade, generating quantified facade drawings. These drawings possess both quantitative attributes and engineering adaptability, allowing direct integration with subsequent stages such as imitation stone brick layout and construction line setting, replacing traditional manual drafting. This significantly shortens the preliminary preparation cycle, reduces labor costs and engineering errors, and provides accurate and efficient data support for facade decoration projects. 103. Identify the quantized elevation drawing based on the preset bottom start coordinates, preset top end coordinates, and preset imitation stone brick specification parameters to obtain the initial paving range. In this embodiment, the specifications of the imitation stone bricks include the length, width, thickness, grout width, and texture direction of the imitation stone bricks, as well as optional data on the reuse of surplus materials and inventory. The longitudinal boundary is anchored by the bottom start coordinates and the top end coordinates. Combined with the specifications of the imitation stone bricks, the accurate identification of the quantified facade drawings is achieved. This ensures that the initial paving range is adapted to the brick type and construction requirements, reducing the cutting of non-whole bricks and material waste. It also takes into account the utilization of surplus materials and inventory optimization, reducing costs. The efficient range definition provides an accurate basis for subsequent imitation stone brick layout, material calculation, and other links, improving the economy and construction efficiency of the facade decoration project. 104. Solve the initial paving range and quantified elevation drawings based on the preset structural constraints, preset multi-objective functions and preset heuristic models to obtain the layout scheme; In this embodiment, the quantified elevation drawings show the shape of the building's exterior facade, and the quantified elevation drawings indicate the facade shape, door and window positions and dimensions. The clearly marked facade shape, door and window positions and dimensions on the quantified elevation drawings provide a reliable data foundation for solving the problem, construct constraints to ensure construction compliance, balance cost and quality through multi-objective functions, and efficiently output the optimal layout scheme through heuristic models. This scheme can accurately avoid doors and windows, adapt to the facade form, replace manual rough planning with a scheme, reduce brick cutting waste and paving rework, provide accurate basis for construction, improve layout efficiency and project quality, and adapt to the needs of facade decoration. In this embodiment, multiple frames of images are combined with benchmark data and algorithm processing to generate quantitative calibration data and a preliminary set of planes focusing on the construction area, laying a reliable data foundation. The deep learning model relies on accurate spatial benchmarks to quickly reconstruct quantified drawings containing facade shapes and door and window information, replacing manual drawing, shortening the cycle and reducing errors. The initial paving range based on coordinates and full-specification parameters of the bricks is adapted to construction requirements and takes into account the reuse of leftover materials, reducing material waste. The layout scheme solved by structural constraints, multi-objective functions and heuristic models can accurately avoid doors and windows, balance the minimum amount of bricks used, the minimum cutting loss and the consistency of texture and joint width, replace manual rough planning, improve material utilization and decorative aesthetics, and comprehensively optimize project efficiency, quality and economy.

[0020] Please see Figure 2 In a second embodiment of the method for generating a layout scheme for imitation stone bricks according to the present invention, step 101 specifically includes: 201. Analyze the multi-frame captured image data based on the planar detection algorithm and reference data to obtain the total area and the set of exterior facade plans; In this embodiment, the planar surface of the entire building facade is efficiently extracted through a planar detection algorithm, and the total area and planar set are obtained simultaneously. This fully covers the main walls and key areas, providing a reliable foundation for subsequent planar screening, imitation stone brick layout and material calculation, reducing labor costs and engineering errors, and adapting to various building facade scenarios. 202. Calculate the total area according to the preset ratio coefficient to obtain the area threshold; In this embodiment, the area threshold = total area × proportional coefficient; the proportional coefficient is not a fixed value and needs to be adjusted according to the building type, construction specifications, and brick characteristics: Building type: For residential buildings, the main wall of the exterior facade accounts for a high proportion, so the proportional coefficient is set to 0.01-0.02 (the threshold is 1%-2% of the total area); For commercial buildings, there are many decorative components, so the proportional coefficient is increased to 0.03-0.05 to avoid mis-screening the main floor plan; Construction requirements: The smallest paving unit of imitation stone bricks is usually ≥0.2㎡, and the coefficient is calculated by combining the total area (e.g., for a 1000㎡ residential building, a coefficient of 0.02 corresponds to a threshold of 20㎡, which can filter out fragmented floor plans); Specification reference: It meets the requirements for continuous operation areas on walls in the "Standard for Acceptance of Construction Decoration and Renovation Engineering Quality"; 203. Filter the set of facade plans according to the area threshold and filtering rules to obtain a preliminary set of plans; In this embodiment, the filtering rule is to traverse the set of planes and remove planes with an area lower than the area threshold. The initial set of planes obtained after filtering focuses on the main areas with construction value, simplifies the data scale, reduces the calculation cost of subsequent benchmark plane determination and imitation stone brick layout, provides high-quality basic data for subsequent engineering stages, reduces manual input and engineering errors, and is suitable for various building facade scenarios. In this embodiment, relying on planar detection algorithms and benchmark data, the planar dimensions of the entire building facade are efficiently extracted and the total area is obtained simultaneously, covering the main body and key areas. The data reliability is high, and the scaling factor can be flexibly adjusted according to the building type, construction needs, and specification requirements to avoid mis-screening. By traversing and filtering to remove fragmented planes below the threshold, the initial selection set focuses on the main areas with construction value, simplifying the data scale, reducing the calculation cost of subsequent benchmark plane determination and imitation stone brick layout, reducing manual input and engineering errors, adapting to various building facade scenarios, providing accurate support for subsequent brick layout and material calculation, and achieving cost reduction, efficiency improvement, and engineering quality assurance.

[0021] Please see Figure 3 The third embodiment of the method for generating a layout scheme for imitation stone bricks in this invention includes step 201, which specifically includes: 301. Based on the reference data, perform size calibration on the multi-frame captured image data to obtain image calibration data; In this embodiment, multi-point planar references (composed of three or more feature points with known three-dimensional coordinates on the facade (such as wall corners, door and window corners, and axis intersections), whose core function is to locate the spatial posture of the facade, and correct the visual error of facade tilt caused by shooting angle deviation through the coordinate association of feature points, so as to ensure that the plane detected later is parallel to the actual facade and avoid spatial dimension distortion) and length references (such as tape measure scales, corner targets, and standard dimensions of doors and windows) are used to perform external parameter and scale calibration on each frame of captured image data, providing a quantitative basis for subsequent area calculation; 302. Detect the image calibration data according to the plane detection algorithm, the preset distance threshold and the preset minimum number of points to obtain the set of exterior facade planes; In this embodiment, the plane detection algorithm is the RANSAC (Random Sample Consensus) algorithm. This algorithm fits the optimal plane from the point cloud converted from image calibration data through an iterative process of random sampling, model fitting, and error verification. It randomly selects a small number of point clouds to fit the initial plane, calculates the distance between other points and the plane, and retains points with the required distance to form an interior point set. This iteration is repeated until the plane model with the largest interior point set is obtained. A preset maximum distance threshold (e.g., 5mm) between the point cloud and the fitted plane is used. Discrete points exceeding the distance threshold (point clouds corresponding to wall stains or shooting noise) are judged as noise and removed to ensure that the fitted plane only reflects the true flatness of the wall and avoids... To mitigate the distortion of the planar model caused by noise, a minimum number of point clouds is set to constitute an effective plane (e.g., 100 points, corresponding to an actual area of ​​approximately 0.2㎡). This filters out small, fragmented planes such as decorative lines and equipment supports, retaining only the main wall planes with potential for stone-like brick paving. This results in the final set of exterior facade planes, ensuring the practicality of the planar plane set for engineering applications. The detected exterior facade plane set consists of independent planes corresponding to different areas (e.g., walls on both sides of doors and windows, and walls in corner areas). This feature of regional extraction provides accurate basic data for subsequent identification of paving areas (excluding door and window openings) and regional layout optimization, avoiding the inclusion of non-construction areas in the area calculation. 303. Calculate the area of ​​the exterior facade plan set to obtain the total area; In this embodiment, the area of ​​each plane is accurately calculated based on the geometric parameters of the facade plane (such as the length and width of a rectangular plane, which are calculated by combining the pixel span with the quantization relationship of 1px=2mm) or the area formula of the polygon enclosed by the point cloud. For irregular planes (such as corner arc areas), the accuracy is ensured by calculating and summing the points after dense sampling of the point cloud. The total area of ​​the facade is obtained by summing the area of ​​each independent plane in the set of facade planes. In this embodiment, the image extrinsic parameters and scale calibration are achieved through benchmark data, laying the foundation for quantitative calculation. The set of facade planes in different regions provides accurate data for subsequent construction. The area calculation based on geometric parameters and point cloud algorithms can ensure accuracy even in regular areas, adapting to complex scenarios and providing data support for all aspects such as material procurement and construction planning, thereby achieving cost reduction and efficiency improvement.

[0022] Please see Figure 4 The fourth embodiment of the method for generating a layout scheme for imitation stone bricks in this invention includes step 102, which specifically includes: 401. Perform a secondary screening of the initial set of planes based on the preset plane parameters to obtain the reference plane; In this embodiment, the planar parameters include spatial attitude parameters (e.g., the angle between the plane's normal vector and the vertical direction ≤ 5°, ensuring it is a facade rather than a roof or ground plane), geometric accuracy parameters (e.g., plane flatness error ≤ 2mm, meeting the accuracy requirements for facade construction), and functional correlation parameters (e.g., the positional relationship between the plane and door / window openings, and the distance threshold from the building axis). These parameters are derived from building construction specifications and facade surveying standards, providing a clear basis for selection. The secondary selection logic is as follows: first, invalid planes are eliminated based on spatial attitude and geometric accuracy parameters; then, the correlation between the plane and the main building structure is verified (e.g., whether it is parallel to the building axis); finally, the plane is sorted by flatness and dimensional integrity, and the plane ranked first is selected as the reference plane. 402. Reconstruct the image labeling data based on a deep learning model to obtain building facade drawings; In this embodiment, a pre-trained deep learning model (such as NeRF, PointRCNN, or an improved version of MVSNet) is used to predict the 3D coordinates of the image calibration data. Based on the predicted 3D coordinate point set, a dense point cloud is filled to represent the wall surface and component details (density adapted to the "10mm×10mm" precision in the scheme), generating a complete 3D geometric reconstruction of the building facade, including details such as wall flatness, the depth of door and window openings, and the spatial angles of corners. The pre-trained deep learning model reconstructs the image calibration data to obtain the building facade drawings, which are in CAD format. The pre-training data for the deep learning model includes multiple frames of high-definition images (resolution ≥1920×1080) covering different building types (residential, commercial, public facilities, old buildings) and different facade styles (modern minimalist, European classical, Chinese brick, industrial). Each building's image must contain a complete sequence of front, side, and oblique views (typically 10-30 frames). The model can be used in multiple frames, and shooting conditions are diverse (different lighting: sunny, cloudy, evening; different environments: whether there is vegetation or construction facilities; different shooting equipment: mobile phone, SLR, drone) to ensure the model's generalizability. 403. Mark the exterior facade drawings of the building according to the reference plane and the preset door and window opening boundaries to obtain quantized facade drawings; In this embodiment, a local coordinate system is established with the geometric center of the reference plane as the origin, and the three-dimensional coordinates of the reconstructed drawings are transformed into two-dimensional coordinates that can be measured in construction, ensuring that the dimensions of the drawings match the physical space on site; combined with the boundaries of door and window openings (building design specifications and on-site measurement data), their positions and dimensions are accurately marked in the drawings, providing clear functional constraints for subsequent avoidance of imitation stone brick paving and polygon reduction generation; In this embodiment, the deep learning model relies on the strong generalization of training with diverse data to reconstruct the image calibration data, accurately restore details such as wall flatness and door and window depth, and quickly output CAD format drawings to improve efficiency. Combined with the local coordinate system established by the reference plane and the door and window boundary annotation, it achieves accurate matching between the drawing size and the site space, providing clear constraints for subsequent tiling avoidance and range identification, and adapting to various building scenarios.

[0023] Please see Figure 5 The fifth embodiment of a method for generating a layout scheme for imitation stone bricks according to the present invention includes step 103 specifically comprising: 501. Generate the first horizontal straight line based on the coordinates of the bottom starting point; 502. Generate a second horizontal straight line based on the coordinates of the top endpoint; In this embodiment, a first horizontal line (L1) is generated by the preset bottom starting coordinates (such as the paving starting point 30cm above the ground), and a second horizontal line (L2) is generated by the top ending coordinates (such as the paving ending point below the waistline). The coordinate data usually comes from the construction design requirements or on-site measurement and calibration, transforming the abstract paving start and end heights into specific geometric boundaries (line L1, line L2) in the quantified drawings. 503. Determine the vertical range based on the first and second horizontal lines; In this embodiment, L1 and L2 are used as the upper and lower boundaries to form a closed vertical range (i.e., the height range of the paving, such as 2.3m-6.8m). This range directly corresponds to the longitudinal working range of the actual construction. The quantitative definition of the vertical range provides basic data for subsequent material usage estimation and construction team division (such as paving in sections according to the vertical range), while avoiding the paving range from covering unnecessary areas and reducing ineffective construction. 504. Perform a range analysis on the quantified elevation drawings to obtain the elevation range; In this embodiment, a global range analysis is performed on the quantified facade drawings (including facade shape and dimension annotations) to extract the facade range, that is, the longitudinal height range of the building facade itself (e.g., from 0m to 8m from the ground), which includes the overall height boundary of all functional areas (doors, windows, walls, openings, etc.). 505. Determine whether the vertical range is less than or equal to the elevation range; In this embodiment, invalid paving plans that exceed the physical boundaries of the building can be directly filtered out through engineering feasibility verification, avoiding material waste (such as unusable cut bricks) and construction rework (such as needing to be removed after paving on the roof) caused by scope violations, thus ensuring the feasibility of the solution. 506. When the vertical range is less than or equal to the facade range, the quantized facade drawing is identified based on the vertical range and the imitation stone brick specification parameters to obtain the initial paving range. In this embodiment, based on the specifications of the imitation stone bricks, the quantified facade drawings are used to identify the regions. That is, from the facade range, the corresponding horizontal region (such as the horizontal span of the wall is 3m-15m) is extracted from the vertical range, and obvious non-paving areas (such as large glass curtain walls) are excluded, thus forming the initial paving range (closed polygonal region). In this embodiment, double horizontal straight lines are generated based on construction requirements or on-site coordinates, transforming the abstract paving height into a precise geometric boundary on the drawing. This avoids the coarse errors of manual line drawing. The vertical range determined in this way quantifies the longitudinal work area, providing reliable data for material estimation and team division of labor. It avoids invalid construction by paving over unnecessary areas. By extracting the actual height boundary of the building to analyze the facade range, and filtering out out-of-bounds plans through compliance judgment, it avoids material waste and construction rework risks. Suitable horizontal areas are extracted from the facade range and non-paving areas are excluded to form a closed initial paving range, improving the efficiency of early preparation and laying a compliant and suitable foundation for subsequent layout optimization.

[0024] Please see Figure 6 The sixth embodiment of a method for generating a layout scheme for imitation stone bricks according to an embodiment of the present invention, specifically includes step 104: 601. Based on the structural constraints and preset cutting dimensions, the initial tiling range and quantized elevation drawings are decomposed into strips to obtain the first dense point cloud and the second dense point cloud network set. In this embodiment, the first dense point cloud covers the entire facade, providing a precise coordinate reference and ensuring the matching of local and overall positions; the second dense point cloud network focuses on the paving area, laying a solid foundation for accurate paving and construction layout. 602. Solve the first and second dense point cloud network sets based on the multi-objective function, heuristic model, preset engineering constraints and preset planning model to obtain a feasible solution. In this embodiment, the planning model includes an integer linear programming model (ILP) and a mixed integer programming model (MIP). The second dense point cloud network set (10mm×10mm gridded paving area) is the core solution carrier: each point cloud unit is defined as a decision variable for whether to pave bricks, the size of the paving bricks, and the rotation state (e.g., The binary decision variable representing whether the point cloud cell in the i-th row and j-th column is tiled. This represents the cutting dimensions of the brickwork in the i-th row and j-th column of the point cloud unit. The rotation angle (0° / 90°) of the point cloud element in row i and column j is represented. The first dense point cloud provides a global coordinate reference: mapping the local coordinates (x, y) of the point cloud element to the global building coordinates, and converting them into construction layout coordinate variables. (e.g., representing the on-site layout coordinates of the point cloud unit in the i-th row and j-th column), ensuring that the local layout matches the overall facade position; ILP only processes integer variables (e.g., The total number of boards used for the point cloud unit in row i and column j), suitable for discrete decisions such as the number of bricks and their placement; MIP is compatible with integer variables and continuous variables (such as... Cutting edge length, (Waste area), which can accurately quantify continuous indicators such as cutting precision and seam width deviation, adapting to complex layout scenarios; transforming engineering constraints into mathematical inequalities / equalities (such as rotation angle constraints). seam continuity constraint (Adjacent unit seam alignment), the contribution value of the i-th point cloud unit under the k-th sub-target, and the priority constraint of the entire plate at the window edge. (The entire panel of point cloud window edge cells in row i and column j uses weights, with a weight threshold of a)), defining the feasible range of variable values; using a multi-objective function As an optimization target This represents the total number of boards used. This serves as an indicator of texture alignment and seam width consistency. As the first objective weight, As the weight of the second objective, As the third objective weight, The fourth objective weight is defined by the weights of each objective (which can be set manually or adjusted through adaptive iteration). The model iteratively calculates and, under the premise of satisfying all constraints, uses a heuristic model to find the combination of variables that minimizes F (i.e., the optimal solution that uses fewer boards, saves cutting, minimizes waste, and improves quality), and outputs a feasible solution. The feasible solution includes multiple layout diagrams, a cutting list (location, original board number, cutting size, rotation angle, texture direction, and destination), and construction layout coordinate data. 603. Optimize the feasible solution based on the preset comprehensive objectives to obtain the layout scheme; In this embodiment, the feasible solution typically includes multiple candidate solutions that meet engineering constraints (such as layout diagrams and cutting lists under different weight combinations). First, based on the core optimization objective (minimizing F) in the comprehensive objective, 1-2 optimal basic solutions (i.e., the solutions with the best cost-quality balance) are selected from the candidate solutions. For the selected optimal basic solutions (such as a layout diagram), each brick is refined according to the construction adaptation objective: ① Numbering regularization: Brick numbers are generated according to "region number, row number, column number" (e.g., "Area A-01-05" represents the 5th brick in the 1st row of Area A). The number needs to be marked in the corresponding position on the layout diagram to facilitate construction personnel to pick up and lay bricks according to the number; ② Coordinate precision: Based on the global coordinate reference of the first dense point cloud, the local point cloud coordinates (x, y) of each brick are converted into two-dimensional coordinates for construction (e.g., (1050mm, 2300mm)) to ensure that the coordinates are compatible with the on-site layout equipment (total station, laser instrument); ③ Layout optimization: Overlay "brick number, coordinate label, and area division line" on the layout diagram, remove redundant information, and improve construction visualization; ④ Link brick number to cutting list: Add a brick number field to the cutting list to clarify the specific brick number corresponding to each cutting record (e.g., "original board number B001 corresponds to cutting size 600×300mm and destination A area"); ⑤ Link brick number to construction layout coordinates: Generate a brick number-two-dimensional coordinate lookup table, allowing construction personnel to quickly look up the corresponding layout position by number, avoiding coordinate confusion; ⑥ Generate a unique QR code for each brick, containing the brick number, key information from the cutting list, and layout coordinates, which can be quickly verified on-site by scanning with a mobile phone; According to the comprehensive target requirements, export the optimized information to multiple formats: DXF format: including numbered layout diagram and layout coordinate labels, adapting to the CAD drawing needs of the construction team; CSV / Excel Format: Includes a material statistics table (total board usage, waste material, and surplus material), a cutting list, and a brick number-coordinate reference table; QR code labels: generated in batches according to brick number, which can be directly printed and affixed to bricks or material packaging; Error visualization report generation: records key deviation data during the optimization process (such as point cloud coordinate mapping error, joint width consistency deviation, and texture alignment deviation), and presents them intuitively through charts (such as error distribution heatmaps and deviation trend lines) to ensure that all deviations are within the allowable range for construction; if there are deviations exceeding the standard, the process is traced back to the model solution stage to adjust parameters (such as optimizing the point cloud coordinate mapping accuracy), regenerates a feasible solution, and then optimizes again; Final solution integration: integrates the numbered layout diagram, standardized cutting list, coordinate reference table, QR code labels, and error visualization report into a complete layout plan, which is delivered to relevant parties such as construction, processing, and material management. In this embodiment, the first dense point cloud provides a global coordinate reference to ensure position matching, while the second dense point cloud focuses on the paving area to achieve refined support. Relying on the quantitative advantages of ILP and MIP models, and combining heuristic models to improve efficiency, the optimal balance between board usage, cutting loss, waste area, and construction quality is achieved through multi-objective functions under engineering constraints. The comprehensive objective optimization further refines the solution. Through regularized numbering, precise coordinate transformation, full-chain data association, and QR code verification, the visualization and convenience of construction are improved. Multi-format file output adapts to the needs of different positions, and error visualization reports ensure that deviations are controllable. The overall solution reduces material waste and construction rework rate, reduces cross-link communication costs, and achieves refined management of the entire process from design to construction. It is suitable for various building facade scenarios and combines economy and practicality.

[0025] Please see Figure 7 The seventh embodiment of a method for generating a layout scheme for imitation stone bricks according to the present invention, step 601 specifically includes: 701. Identify the quantized elevation drawings based on structural constraints and vertical ranges to obtain multiple polygon reduction ranges; In this embodiment, the construction constraints include boundary constraints, avoidance constraints, process constraints, and texture constraints. Boundary constraint rules guide the establishment of a local coordinate system with the lower left corner of the quantized elevation drawing as the origin, calibrating the correspondence between the drawing coordinates and the actual construction coordinates. Vertical range parameters (bottom starting point coordinate Y1, top ending point coordinate Y2) are extracted, and horizontal constraint lines L1 (Y=Y1) and L2 (Y=Y2) are generated in the drawing to define the effective longitudinal range. Based on avoidance constraints, door and window openings, equipment holes, and drip lines are identified one by one, and the identified markers are named "Avoidance Type-Number" (e.g., "Door / Window-01"). Based on process constraints, elevation corner edges are identified, the corner angles are calculated, and independent continuous areas are divided and numbered (e.g., "Area A, Area B"). Based on texture constraints, texture direction arrows and baselines are marked in each independent continuous area, associated with the area number, and multiple independent polygon reduction ranges are divided based on the above constraints. 702. Perform Boolean operations on the initial tiling area and multiple polygon reduction areas to obtain multiple tiling polygon regions; In this embodiment, each paving polygon area represents a stone-like brick. The joint width of the stone-like brick is basically fixed (5mm-10mm), which is determined by the precision of the stone-like brick itself. The texture direction is determined according to the pattern of the stone-like brick. Boolean operations can automatically eliminate all unpaving areas in the initial range (these unpaving areas are determined by multiple subtracted polygon ranges). Each generated paving polygon area meets three conditions: boundary compliance, no functional conflict, and process feasibility. For example, the initial range of a facade includes door and window openings and corners. After calculation, it will be split into two independent paving polygons: the area to the left of the door and window and the area to the right of the door and window. This avoids the problems of continuous corner paving and opening coverage. The polygon boundaries generated by Boolean operations are seamlessly connected and fit together. When arranging them later, only the gap needs to be set according to the fixed joint width to meet the precision requirements. At the same time, the independent division of polygons also provides a basis for matching the texture direction by region (the texture direction matching the pattern can be set separately for each polygon). 703. Decompose the quantized elevation drawing into strips according to the cutting dimensions to obtain the first dense point cloud; In this embodiment, the entire quantized facade drawing is decomposed into strips based on the cutting size (e.g., 10mm×10mm). The resulting first dense point cloud covers the entire facade area and contains all geometric information of both paving and non-paving areas. Its core function is to serve as a global positioning benchmark, providing a reference for the coordinate calibration and regional association of the subsequent second dense point cloud, ensuring that the position of the local paving area matches the overall facade (e.g., global coordinate correspondence during construction layout). 704. Decompose multiple tileable polygonal regions into strips according to the cutting size to obtain a second dense point cloud network set; In this embodiment, the second dense point cloud network is a mesh type formed by strip decomposition of the paving polygon region. The paving polygon region is decomposed based on the second clipping size (e.g., 10mm × 10mm), and the generated point cloud network is a meshed local fine data. The 10mm mesh accuracy is much higher than the size accuracy of the imitation stone brick (usually in centimeters). It can not only completely preserve the boundary details of the paving region (e.g., the irregular edges of the polygon), but also transform the continuous region into discrete point cloud units. This structured data can be directly input into the subsequent heuristic model and planning model (ILP / MIP), supporting fine calculations for grid-by-grid optimization and slab-by-slab arrangement. In this embodiment, the quantified elevation drawings are identified based on structural constraints and vertical ranges to generate clearly marked polygonal areas, ensuring that the paving boundaries and functional areas are identified without deviation. The paving polygonal areas are automatically purified through Boolean operations, avoiding conflicts at corners and openings. The paving accuracy is adapted to the boundary and fixed joint width requirements, and it supports matching texture directions by region, taking into account both aesthetics and standardization. Two types of dense point clouds are generated by strip decomposition and quantified elevation drawings and multiple paving polygonal areas. The first point cloud provides a global positioning benchmark to ensure coordinate matching, while the second point cloud transforms the region into refined structured data, which can be directly connected to heuristic and planning models and supports optimization calculations for each section. The entire process automates the process from drawing recognition to data modeling, reduces human error, improves layout efficiency, reduces material waste, and provides reliable support for construction layout and precise paving.

[0026] The above describes a method for generating a layout scheme for imitation stone bricks according to an embodiment of the present invention. The following describes a device for generating a layout scheme for imitation stone bricks according to an embodiment of the present invention. Please refer to [link / reference]. Figure 8 One embodiment of the device for generating a layout scheme for imitation stone bricks according to an embodiment of the present invention includes: The calculation module 1 is used to acquire multi-frame captured image data and perform calculations on the multi-frame captured image data according to preset reference data, preset plane detection algorithm and preset filtering rules to obtain image calibration data and a preliminary set of planes. Reconstruction Module 2 is used to reconstruct the image calibration data based on a preset deep learning model and a preliminary set of planes to obtain quantized elevation drawings. The drawing recognition module 3 is used to recognize the quantized facade drawing based on the preset bottom start coordinates, the preset top end coordinates, and the preset imitation stone brick specification parameters, so as to obtain the initial paving range. The layout scheme generation module 4 is used to solve the initial paving range and quantified elevation drawings based on preset structural constraints, preset multi-objective functions and preset heuristic models to obtain the layout scheme. In this embodiment, multiple frames of images are combined with benchmark data and algorithm processing to generate quantitative calibration data and a preliminary set of planes focusing on the construction area, laying a reliable data foundation. The deep learning model relies on accurate spatial benchmarks to quickly reconstruct quantified drawings containing facade shapes and door and window information, replacing manual drawing, shortening the cycle and reducing errors. The initial paving range based on coordinates and full-specification parameters of the bricks is adapted to construction requirements and takes into account the reuse of leftover materials, reducing material waste. The layout scheme solved by structural constraints, multi-objective functions and heuristic models can accurately avoid doors and windows, balance the minimum amount of bricks used, the minimum cutting loss and the consistency of texture and joint width, replace manual rough planning, improve material utilization and decorative aesthetics, and comprehensively optimize project efficiency, quality and economy.

[0027] Figure 9 This is a schematic diagram of a device for generating a layout scheme for imitation stone bricks according to an embodiment of the present invention. This device 900 can vary significantly depending on its configuration or performance. It may include one or more central processing units (CPUs) 910 (e.g., one or more processors) and a memory 920, and one or more storage media 930 (e.g., one or more mass storage devices) for storing application programs 933 or data 932. The memory 920 and storage media 930 can be temporary or persistent storage. The program stored in the storage media 930 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the device 900. Furthermore, the processor 910 may be configured to communicate with the storage media 930 and execute the series of instruction operations in the storage media 930 on the device 900 to implement the steps of the method for generating a layout scheme for imitation stone bricks provided in the above-described embodiments.

[0028] A device 900 for generating a layout scheme for imitation stone bricks may further include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input / output interfaces 960, and / or one or more operating systems 931, such as Windows Server, MacOSX, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 9 The illustrated structure of the device for generating a layout scheme of imitation stone bricks does not constitute a limitation on the device for generating a layout scheme of imitation stone bricks. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0029] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of a method for generating a layout scheme of imitation stone bricks.

[0030] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0031] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0032] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for generating a layout scheme for imitation stone bricks, characterized in that, include: Acquire multi-frame captured image data, and perform calculations on the multi-frame captured image data according to preset reference data, preset plane detection algorithm and preset filtering rules to obtain image calibration data and a preliminary set of planes; Based on a pre-set deep learning model and a preliminary set of planes, the image calibration data is reconstructed to obtain quantized elevation drawings; The quantized elevation drawing is identified based on the preset bottom start coordinates, the preset top end coordinates, and the preset imitation stone brick specification parameters to obtain the initial paving range. The initial tiling range and quantified elevation drawings are solved based on the preset structural constraints, preset multi-objective functions, and preset heuristic models to obtain the layout scheme.

2. The method for generating a layout scheme for imitation stone bricks as described in claim 1, characterized in that, The step of calculating multi-frame captured image data based on preset reference data, a preset plane detection algorithm, and preset filtering rules to obtain image calibration data and a preliminary set of planes includes: Based on the planar detection algorithm and benchmark data, the multi-frame captured image data is analyzed to obtain the total area and the set of exterior facade plans; The total area is calculated based on a preset ratio coefficient to obtain an area threshold. The set of facade planes is filtered according to area thresholds and filtering rules to obtain a preliminary set of planes.

3. The method for generating a layout scheme for imitation stone bricks as described in claim 2, characterized in that, The analysis of multiple frames of captured image data based on a planar detection algorithm and reference data to obtain the total area and the set of exterior facade plans includes: The dimensions of multiple frames of captured images are calibrated based on the reference data to obtain image calibration data; The image calibration data is detected based on a plane detection algorithm, a preset distance threshold, and a preset minimum number of points to obtain a set of exterior facade planes; Calculate the area of ​​the set of exterior facade plans to obtain the total area.

4. The method for generating a layout scheme for imitation stone bricks as described in claim 1, characterized in that, The process of reconstructing image calibration data based on a preset deep learning model and a preliminary set of planes to obtain quantized elevation drawings includes: The initial set of planes is further filtered based on preset plane parameters to obtain the reference plane; The image labeling data is reconstructed based on a deep learning model to obtain the building facade drawings; The exterior facade drawings of the building are annotated according to the reference plane and the pre-set boundaries of door and window openings to obtain quantified facade drawings.

5. The method for generating a layout scheme for imitation stone bricks as described in claim 1, characterized in that, The process of identifying the quantized elevation drawing based on preset bottom start coordinates, preset top end coordinates, and preset imitation stone brick specification parameters to obtain the initial paving range includes: Generate the first horizontal straight line based on the coordinates of the bottom starting point; Generate a second horizontal line based on the coordinates of the top endpoint; The vertical range is determined based on the first horizontal line and the second horizontal line. Perform a range analysis on the quantified elevation drawings to determine the elevation range; Determine whether the vertical range is less than or equal to the elevation range; When the vertical range is less than or equal to the facade range, the quantized facade drawing is identified based on the vertical range and the specifications of the imitation stone bricks to obtain the initial paving range.

6. The method for generating a layout scheme for imitation stone bricks as described in claim 5, characterized in that, The process of solving the initial tiling range and quantified elevation drawings based on preset construction constraints, preset multi-objective functions, and preset heuristic models to obtain the layout scheme includes: Based on the structural constraints and preset cutting dimensions, the initial tiling range and quantized elevation drawings are decomposed into strips to obtain the first dense point cloud and the second dense point cloud network set. The first and second dense point cloud network sets are solved based on the multi-objective function, heuristic model, preset engineering constraints and preset planning model to obtain a feasible solution. The feasible solution is optimized based on the preset comprehensive objectives to obtain the layout scheme.

7. The method for generating a layout scheme for imitation stone bricks as described in claim 6, characterized in that, The process of decomposing the initial tiling area and quantized elevation drawings into strips based on structural constraints and preset cutting dimensions to obtain a first dense point cloud and a second dense point cloud network set includes: The quantized elevation drawings are identified based on structural constraints and vertical ranges to obtain multiple polygon reduction ranges; Perform Boolean operations on the initial tiling area and multiple reduced polygon areas to obtain multiple tiling polygon regions; The quantized elevation drawing is decomposed into strips according to the cutting dimensions to obtain the first dense point cloud; Based on the cutting dimensions, multiple tileable polygon regions are decomposed into strips to obtain a second dense point cloud network set.

8. A device for generating a layout scheme for imitation stone bricks, characterized in that, include: The calculation module is used to acquire multi-frame captured image data and perform calculations on the multi-frame captured image data according to preset benchmark data, preset plane detection algorithm and preset filtering rules to obtain image calibration data and a preliminary set of planes. The reconstruction module is used to reconstruct the image calibration data based on a preset deep learning model and a preliminary set of planes to obtain quantified elevation drawings. The drawing recognition module is used to recognize the quantized facade drawing based on the preset bottom start coordinates, the preset top end coordinates, and the preset imitation stone brick specification parameters, so as to obtain the initial paving range. The layout scheme generation module is used to solve the initial paving range and quantified elevation drawings based on preset structural constraints, preset multi-objective functions and preset heuristic models to obtain the layout scheme.

9. A device for generating a layout scheme for imitation stone bricks, characterized in that, include: A memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the imitation stone brick layout scheme generation device to perform the steps of the imitation stone brick layout scheme generation method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement each step of the method for generating a layout scheme for imitation stone bricks as described in any one of claims 1-7.

Images