Window class object aided design method, device and equipment based on neural network model
By using a window object-assisted design method based on a neural network model, the design of doors and windows is automatically processed, solving the problems of low efficiency and low accuracy in traditional manual design, and achieving efficient and accurate design output.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIULING (JIANGSU) DIGITAL INTELLIGENT TECH CO LTD
- Filing Date
- 2021-08-24
- Publication Date
- 2026-07-14
AI Technical Summary
Traditional door and window design relies on manual experience, resulting in low design efficiency, low accuracy, and high cost, and making comprehensive optimization impossible.
A window object-assisted design method based on a neural network model is adopted. By obtaining and annotating the sample apartment floor plan training set, a window object-assisted design model is obtained by training the neural network model, and the window design scheme and profile information are automatically output.
It improves design efficiency and accuracy, saves time and manpower, avoids the shortcomings of manual adjustments, and provides a more scientific and reasonable design solution.
Smart Images
Figure CN115718939B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of building-aided design technology, and in particular to a method, apparatus, computer device, and storage medium for window-type object-aided design based on a neural network model. Background Technology
[0002] Doors and windows are an important part of home decoration design. Since they directly affect the lighting of the house and the access of people, they are one of the most frequently used home decoration components in daily life. Therefore, whether the door and window design is scientific and reasonable is directly related to the overall effect of the home decoration design.
[0003] Traditional doors, windows, and curtain walls are designed manually. After the design is completed, it is optimized by people based on years of industry experience. For example, some regional features are added or removed to make the design more suitable for the local application, so that the design is more in line with people's needs.
[0004] However, in the traditional design optimization process, relying solely on manual experience for adjustments may result in low adjustment efficiency and incomplete optimization, leading to low design efficiency. Summary of the Invention
[0005] Therefore, it is necessary to provide a method, apparatus, computer device, and storage medium for window object auxiliary design based on a neural network model to address the above-mentioned technical problems.
[0006] In a first aspect, embodiments of this application provide a window object-aided design method based on a neural network model, including:
[0007] Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the diagrams within the training set;
[0008] The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model;
[0009] Obtain the floor plan of the apartment requiring design assistance and input it into the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme.
[0010] In one embodiment, the annotation of the sample apartment layout diagrams in the training set and the window object attributes in the diagrams includes:
[0011] Label the walls, doors, windows, glass curtain walls, and scale dimensions in the floor plan;
[0012] The category attributes of the wall are labeled, the location attributes of the doors and windows are labeled, and the profile attributes of the doors, windows, and glass curtain walls are labeled.
[0013] In one embodiment, the step of obtaining the floor plan diagram requiring design assistance and inputting it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan diagram requiring design assistance includes:
[0014] Feature extraction is performed on the floor plan image that requires auxiliary design to obtain the floor plan feature vector and window feature vector;
[0015] Cluster analysis is performed using the feature vectors of the apartment layout to obtain the design scheme information for the first window type; cluster analysis is performed using the feature vectors of the window type to obtain the design scheme information for the second window type.
[0016] The at least one window design scheme information is determined based on the first window design scheme information and the second window design scheme information.
[0017] In one embodiment, the profile information corresponding to each window design scheme is determined, wherein determining the profile information corresponding to each window design scheme includes:
[0018] Based on the determined window design scheme, the profile type, quantity, manufacturer, construction time, and cost of each window design scheme are matched according to preset rules to form the profile information.
[0019] The profile information and the window design scheme are output and displayed together.
[0020] In one embodiment, the step of obtaining the floor plan diagram requiring design assistance and inputting it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan diagram requiring design assistance includes:
[0021] Input the floor plan of the apartment that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan, and determine the functional area where the window object is located;
[0022] Based on the functional areas, determine the window type requirement information, and based on the window type requirement information, determine at least one window type design scheme information.
[0023] Secondly, embodiments of this application provide a window object-aided design device based on a neural network model, the device comprising:
[0024] The acquisition module is used to acquire a training set of sample apartment layout diagrams and to annotate the sample apartment layout diagrams and window object attributes in the diagrams in the training set.
[0025] The training module is used to input the labeled sample apartment layout diagram into a preset neural network model for training, so as to obtain a window object-assisted design model.
[0026] The auxiliary design module is used to obtain the floor plan of the apartment that needs auxiliary design and input it into the auxiliary design model of the window object to obtain at least one window design scheme information corresponding to the floor plan image that needs auxiliary design, as well as the profile information corresponding to each window design scheme information.
[0027] In one embodiment, the auxiliary design module includes:
[0028] The extraction unit is used to extract features from the floor plan image that requires auxiliary design, and obtain the floor plan feature vector and the window feature vector.
[0029] The analysis unit is used to perform cluster analysis using the apartment structure feature vector to obtain the first window design scheme information, and to perform cluster analysis using the window feature vector to obtain the second window design scheme information.
[0030] The determining unit is configured to determine the at least one window design scheme information based on the first window design scheme information and the second window design scheme information.
[0031] In one embodiment, the auxiliary design module includes:
[0032] Obtain the floor plan of the apartment requiring design assistance and input it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, including:
[0033] Input the floor plan of the apartment that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan, and determine the functional area where the window object is located;
[0034] Based on the functional areas, determine the window type requirement information, and based on the window type requirement information, determine at least one window type design scheme information.
[0035] Thirdly, embodiments of this application provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:
[0036] Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the diagrams within the training set;
[0037] The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model;
[0038] Obtain the floor plan of the apartment requiring design assistance and input it into the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme.
[0039] Fourthly, embodiments of this application provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:
[0040] Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the diagrams within the training set;
[0041] The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model;
[0042] Obtain the floor plan of the apartment requiring design assistance and input it into the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme.
[0043] This application provides a method, apparatus, computer device, and storage medium for window-type object-assisted design based on a neural network model. The method first takes a training set of sample floor plan diagrams to train a window-type object-assisted design model; then, it acquires the floor plan diagram requiring design assistance and inputs it into the window-type object-assisted design model to obtain at least one window design scheme information corresponding to the floor plan diagram, as well as profile information corresponding to each window design scheme. The window-type object-assisted design model is obtained by annotating and training the sample floor plan diagrams and window object attributes in the training set. It can extract design information and rules for different floor plan diagrams and corresponding window objects. Therefore, using the design information and rules, it can automatically output different window design scheme information and corresponding profile information for different floor plans. This avoids the problems of low efficiency, low accuracy, incomplete adjustments, and high learning costs associated with traditional manual design based on experience, greatly saving time and manpower. Attached Figure Description
[0044] Figure 1 A flowchart illustrating a window object-assisted design method based on a neural network model, as provided in one embodiment;
[0045] Figure 2 A flowchart illustrating a window object-assisted design method based on a neural network model, provided for another embodiment;
[0046] Figure 3 A flowchart illustrating a window object-assisted design method based on a neural network model, as provided in yet another embodiment;
[0047] Figure 4 A schematic diagram of a window object-assisted design device based on a neural network model is provided as one embodiment.
[0048] Figure 5 An internal structural diagram of a computer device provided in one embodiment. Detailed Implementation
[0049] Currently, when recognizing a designed 3D window object model, an existing image recognition model is usually used. During the recognition process, the image recognition model is adjusted and optimized based on experience. This may lead to problems such as low adjustment efficiency, low accuracy, incomplete adjustment, and high human learning costs, resulting in low efficiency and accuracy of window object assisted design based on neural network models.
[0050] The window object-aided design method, apparatus, computer device, and storage medium based on a neural network model provided in this application are specific processes for obtaining the window object model recognition result by inputting the two-dimensional model image corresponding to the window object model into a preset recognition network for recognition processing when a two-dimensional model image corresponding to the window object model is obtained. This not only reduces the cost of manpower and time, but also improves the accuracy and reliability of recognition.
[0051] It should be noted that the window object-assisted design method based on a neural network model provided in this application can be executed by a window object-assisted design device based on a neural network model. The window object includes at least one of walls, doors and windows, and glass curtain walls. This window object-assisted design device based on a neural network model can be implemented as part or all of a computer device through software, hardware, or a combination of both. Optionally, the computer device can be a personal computer (PC), portable device, server, or other electronic device with data processing capabilities, or it can be an embedded device, intelligent device, etc. This embodiment does not limit the specific form of the computer device. The execution subject of the following method embodiments is described using a computer device as an example.
[0052] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0053] Figure 1 This is a flowchart illustrating a window-based object-assisted design method based on a neural network model, as provided in one embodiment. This embodiment relates to the specific process by which a computer device outputs assisted design results based on a window-based object. Figure 1 As shown, the method includes:
[0054] Step S11: Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the training set.
[0055] Specifically, the sample floor plan training set includes various floor plans and a collection of door and window object attributes designed within each floor plan. These door and window object attributes characterize the type, structure, size, type, and material of doors and windows. Door and window type refers to the variety of doors and windows; for example, windows can include ordinary windows, bathroom windows, balcony windows, and bay windows. The opening direction of a window is determined by its location, such as left-opening, right-opening, top-opening, bottom-opening, or double-opening windows. The opening direction is usually related to the window type attribute and other attributes. Windows typically open to the left or right, in which case the axis connecting the window to the wall will be positioned on the left or right side; double-opening windows will have the window axis positioned on both the left and right sides. Of course, the requirement is to label the functional areas of each floor plan, such as bedrooms, living rooms, and kitchens.
[0056] Furthermore, the walls, doors, windows, glass curtain walls, and their proportional dimensions in the floor plan are labeled; the category attributes of the walls are labeled, the location attributes of the doors and windows are labeled, and the material attributes of the doors, windows, and glass curtain walls are labeled.
[0057] Specifically, the proportional dimensions include building height, floor height, and area, where the area refers to the unit's floor area. If the window type is a glass curtain wall, the area is the required area to cover the glass curtain wall. The floor height is used to account for environmental factors such as noise and dust, while the building height and floor height are used to account for lighting factors. The wall structure includes wall distribution data, load-bearing wall distribution data, etc.
[0058] The location attributes of the doors and windows are mainly labeled according to functional areas, such as ordinary windows, bathroom windows, balcony windows, and bay windows, which correspond to functional areas such as bathrooms, balconies, and bedrooms. Additionally, the labeling information may include the material properties of the doors, windows, and glass curtain walls.
[0059] It should be noted that the above-mentioned annotation information includes annotations made on the floor plan of the apartment requiring design assistance, as well as attribute annotation information stored in the 3D building model. This attribute annotation information can be stored in the attribute database of the 3D building model. When inputting the floor plan requiring design assistance into the neural network model, the floor plan and the image annotated with its relevant attribute information can be input simultaneously into the neural network model. The relevant attribute information can be identified and extracted using a semantic segmentation model.
[0060] It should be understood that the method provided by this invention can also be applied to other types of curtain walls, not limited to glass curtain walls.
[0061] In actual processing, the information carried by the floor plan diagram that requires auxiliary design includes: window object type, material type information, window object profile appearance information, window object profile size information, window object stone type information, etc. For example, when the window object is a glass curtain wall, the information includes: the material properties of the columns and beams (aluminum profile), the width or narrowness of the columns or beams, and the stone type information such as granite; when the window object is a door, the information includes: the thickness of the door and the size of the door; when the window object is a window, the window model can carry information including: the window material, the window position, and the window thickness.
[0062] Step S12: Input the labeled sample apartment layout diagram into a preset neural network model for training to obtain a window object-assisted design model.
[0063] For example, when training door and window recognition, the floor plan diagram is input into the initial neural network model. Feature extraction can be performed on the floor plan sample images using relevant image recognition algorithms. For instance, contour extraction algorithms can be used to extract the contours of doors and windows from the floor plan sample images, and then feature differentiation can be performed on the extracted results. During feature extraction, multi-dimensional detection of doors and windows can be achieved through target region detection and semantic segmentation detection. After the door and window sample images with completed feature extraction are input into the neural network model, relevant operations are used to change the relevant parameters of the model, thereby improving the model's recognition accuracy. During the model training process, the door and window sample images with completed feature extraction are input into the preset neural network model to obtain the output results. The output results are judged to determine whether the model's performance meets the requirements. For example, the training process of the model can be judged based on the value of the loss function. When the value of the loss function reaches a level that satisfies the output results, the model's performance is considered to meet the requirements, and the model training can be stopped, resulting in a model for window object detection.
[0064] Step S13: Obtain the floor plan diagram that needs to be assisted in the design and input it into the window object assisted design model to obtain at least one window design scheme information corresponding to the floor plan diagram that needs to be assisted in the design, as well as the profile information corresponding to each window design scheme information.
[0065] Specifically, when the computer device determines the recognition result of the window object model, it can output it in real time, so that the user can promptly know the design scheme of the window object in the current floor plan and the profiles and related information required by the design scheme.
[0066] In one embodiment, based on the determined at least one window design scheme, the profile type, quantity, manufacturer, construction time, and cost of each window design scheme are matched according to preset rules to form the profile information. Then, the profile information and the window design scheme are output and displayed together.
[0067] For example, if the input floor plan image requiring design assistance shows a three-bedroom, two-living room, one-kitchen, one-bathroom, and one-balcony layout, different design schemes will be provided based on the various parameters of the floor plan. One design scheme might include a window in both the kitchen and bathroom, with the bathroom window being a double (or triple) insulated and soundproof window. If the balcony is large and on a lower floor, three to four soundproof windows can be designed, requiring good sealing. Alternatively, casement windows could be used. Of course, other design schemes are also possible, such as those with low overall design costs or those offering the best sound insulation and lighting.
[0068] Once the window design scheme for the apartment type represented by the suitable apartment structure image is output, the glass window profiles to be or possibly used can be determined based on the design scheme, thereby providing customers with better reference information, eliminating the need for other cost calculations, and making profile procurement more convenient.
[0069] A profile database can be set up. When soundproof windows are needed, the database searches for windows with corresponding soundproofing effects. Specifically, based on window size, functional (soundproofing) requirements, and material information, it matches the profile models, quantities, and manufacturers used in each window design scheme according to preset rules (query rules). With this information, construction time and cost can be calculated based on construction experience. This construction experience can be obtained by looking up past experience in tables (with significant errors) or by setting up a calculation plugin, eliminating the need for dedicated personnel.
[0070] In summary, this application first uses a training set of sample apartment floor plan diagrams to train a window-type object-assisted design model; then it acquires the apartment floor plan diagram requiring design assistance and inputs it into the window-type object-assisted design model to obtain at least one window design scheme information corresponding to the apartment floor plan diagram requiring design assistance, as well as the profile information corresponding to each window design scheme. The window-type object-assisted design model is obtained by annotating and training the sample apartment floor plan diagrams and the window object attributes in the diagrams in the training set. It can uncover design information and rules for different apartment floor plan diagrams and corresponding window objects. Therefore, using the design information and rules, different window design scheme information and corresponding profile information can be automatically output for different apartment floor plans. This avoids the problems of low efficiency, low accuracy, incomplete adjustments, and high human learning costs associated with traditional manual design based on experience, greatly saving time and manpower.
[0071] The above embodiments disclose the specific process by which a computer device obtains the design-aided result for the window object based on the neural network model according to the window object model. It should be noted that the following methods are for illustrative purposes only and are not intended to limit the scope of this application.
[0072] Figure 2 This is a flowchart illustrating a window-type object-assisted design method based on a neural network model, as provided in another embodiment. This embodiment involves inputting a floor plan diagram requiring design assistance into the window-type object-assisted design model to obtain at least one window-type design scheme information corresponding to the floor plan diagram. Optionally, based on the above embodiment, as follows... Figure 2 As shown, step S13 can be achieved through the following sub-steps:
[0073] Step S131: Extract features from the floor plan image that requires auxiliary design to obtain floor plan feature vector and window feature vector.
[0074] Specifically, features contained in the image are extracted through a neural network model to form feature vectors, which are mainly divided into apartment structure feature vectors and window feature vectors. For example, feature vectors of apartment lines and enclosed areas, and window feature vectors of living room windows, master bedroom windows, etc.
[0075] Step S132: Perform cluster analysis using the apartment structure feature vector to obtain the first window design scheme information, and perform cluster analysis using the window feature vector to obtain the second window design scheme information.
[0076] The apartment layout feature vector and window type feature vector are used to calculate the distance to the cluster center using a clustering analysis algorithm, thus achieving cluster partitioning. Two design schemes are then created using the two feature vectors respectively; each design scheme can have multiple similar or replaceable design scheme information.
[0077] Step S133: Determine the at least one window design scheme information based on the first window design scheme information and the second window design scheme information.
[0078] The design scheme information of the first window type is combined with the design scheme information of the second window type, and the combination score is calculated. The scheme with the higher combination score is taken as the optimal output design scheme information, and the scheme with the lower score is taken as the supplementary design scheme. Specifically, the output and display can be based on the combination score results.
[0079] Specifically, the computer device inputs the feature-annotated image of each window object into the initial model for training. The initial model can be an existing window object auxiliary design model based on a neural network model. When the initial model receives the feature-annotated image of the window object, it will be trained. Specifically, it uses the object detection and segmentation technology of Mask R-CNN and TensorFlow to perform deep learning on the feature-annotated image of the window object, thereby obtaining the recognition model including the design rules of the window object.
[0080] It should be noted that the design rules for window objects can include three metrics: support, confidence, and lift. Support: The support of X → Y represents the probability that {X,Y} appears in the total itemset. Confidence: The confidence of X → Y represents the probability that Y can be derived from the rule X → Y given that the prerequisite X has occurred; that is, the probability that Y may also exist if X exists. Generally, rules with a confidence of 95% or higher are considered mandatory, while those with a confidence of 60-70% are considered rules to be satisfied as much as possible. Lift: The lift of X → Y represents the ratio of the probability that Y exists given X to the overall probability of Y occurring, which can be expressed as P(Y|X) / P(Y). Optionally, the above association relationships can broadly include the following four categories: Boolean association rules, quantification rules, one-dimensional and multi-dimensional rules, and single-level and multi-level association rules. For example, a Boolean association rule could be that if X exists, Y must also exist, meaning that if Y does not exist, the requirement is not met; a quantification rule could be a rule for the quantification value of a certain indicator of a window object's characteristics.
[0081] Specifically, the computer device inputs the aforementioned window object feature-annotated image into the initial model. The recognition network can also perform hierarchical classification of the window object features and other entity objects in the aforementioned window object feature-annotated image based on a decision tree approach, and extract the correlation relationships of different window object features based on the hierarchical structure. Then, based on these correlation relationships, frequent itemsets are obtained, and frequent items with a probability of occurrence higher than 90% in the frequent itemsets are taken as strong rules, thereby obtaining multiple strong rules. The neural network including the multiple strong rules is determined as the recognition network including window object feature design rules.
[0082] Classification using decision trees can involve a tree structure built according to a series of rules for classification and prediction. The top node of the decision tree is the root node, and each node forms a new node downwards. Nodes without branches are leaf nodes, and each leaf node corresponds to a decision, i.e., a possible classification result. During computation, the tree is traversed from the root node downwards. Each node corresponds to an attribute, and different branches are selected for different attribute values, finally reaching a leaf node to complete the classification. Decision tree algorithms have a simple structure, high classification accuracy, and good robustness to noisy data.
[0083] Optionally, when a computer device performs deep learning on window-type object feature-annotated images using the object detection and segmentation techniques of Mask R-CNN and TensorFlow, it can obtain a trained model based on the window-type object feature-annotated images.
[0084] To describe the technical solutions provided in this application in more detail, a specific embodiment is described herein, such as... Figure 3 As shown, step S13 includes the following steps:
[0085] Step S134: Input the floor plan diagram that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan diagram, and determine the functional area where the window object is located.
[0086] When performing image recognition, the neural network model can accurately identify window-like objects in the floor plan, thereby achieving localization. Localization can be understood as the position of the window-like object within the floor plan or the wall it occupies. Functional areas in the floor plan are enclosed by walls; these walls are used to determine the functional areas. The neural network model identifies each wall within the enclosed area, and after training, it can determine the corresponding functional area based on the position of the window-like object.
[0087] Step S135: Determine window type requirement information based on the functional area, and determine at least one window type design scheme based on the window type requirement information.
[0088] As mentioned in the above embodiments, windows in different functional areas have different functional requirements. Windows in the master bedroom require sound insulation, while windows on the balcony require good sealing and sound insulation; single and double-opening windows are also available. Since the needs of each functional area are different, and glass curtain walls may have various shapes, the window type, profile, etc., need to be determined according to each shape.
[0089] It should be understood that, although Figures 1-3 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figures 1-3 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
[0090] like Figure 4 As shown, this application also provides a window object-aided design device based on a neural network model, the device comprising:
[0091] The acquisition module 410 is used to acquire a training set of sample apartment layout diagrams and to annotate the sample apartment layout diagrams and window object attributes in the diagrams in the training set.
[0092] Training module 420 is used to input the labeled sample house structure diagram into a preset neural network model for training, so as to obtain a window object-assisted design model;
[0093] The auxiliary design module 430 is used to obtain the floor plan of the house that needs to be assisted in design and input the window object auxiliary design model to obtain at least one window design scheme information corresponding to the floor plan image that needs to be assisted in design, as well as the profile information corresponding to each window design scheme information.
[0094] In one embodiment of this application, the auxiliary design module 430 includes:
[0095] The extraction unit is used to extract features from the floor plan image that requires auxiliary design, and obtain the floor plan feature vector and the window feature vector.
[0096] The analysis unit is used to perform cluster analysis using the apartment structure feature vector to obtain the first window design scheme information, and to perform cluster analysis using the window feature vector to obtain the second window design scheme information.
[0097] The determining unit is configured to determine the at least one window design scheme information based on the first window design scheme information and the second window design scheme information.
[0098] In one embodiment of this application, the auxiliary design module 430 further includes:
[0099] Obtain the floor plan of the apartment requiring design assistance and input it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, including:
[0100] Input the floor plan of the apartment that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan, and determine the functional area where the window object is located;
[0101] Based on the functional areas, determine the window type requirement information, and based on the window type requirement information, determine at least one window type design scheme information.
[0102] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0103] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:
[0104] Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the diagrams within the training set;
[0105] The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model;
[0106] Obtain the floor plan of the apartment requiring design assistance and input it into the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme.
[0107] It should be clear that the process of the processor executing the computer program in the embodiments of this application is consistent with the execution process of each step in the above method, as can be seen in the description above.
[0108] In one embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, performs the following steps:
[0109] Obtain a training set of sample apartment layout diagrams, and annotate the sample apartment layout diagrams and window object attributes in the diagrams within the training set;
[0110] The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model;
[0111] Obtain the floor plan of the apartment requiring design assistance and input it into the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme.
[0112] It should be clear that the process of the processor executing the computer program in the embodiments of this application is consistent with the execution process of each step in the above method, as can be seen in the description above.
[0113] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0114] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0115] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.
Claims
1. A window-type object-aided design method based on a neural network model, characterized in that, The method includes: Obtain a training set of sample floor plan diagrams, and annotate the sample floor plan diagrams and window object attributes in the training set; the window object attributes include the position attributes of doors and windows, and the position attributes of doors and windows are annotated according to the functional areas in the sample floor plan diagrams; The labeled sample apartment layout diagram is input into a preset neural network model for training to obtain a window object-assisted design model; Obtain the floor plan of the apartment requiring design assistance and input the window object design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, as well as the profile information corresponding to each window design scheme information; the window design scheme information is used to indicate the design of window objects in the floor plan.
2. The method according to claim 1, characterized in that, The annotation of the sample apartment layout diagrams and window object attributes in the training set includes: Label the walls, doors, windows, glass curtain walls, and scale dimensions in the floor plan; The category attributes of the wall are labeled, the location attributes of the doors and windows are labeled, and the profile attributes of the doors, windows, and glass curtain walls are labeled.
3. The method according to claim 1, characterized in that, The process of obtaining the floor plan diagram requiring design assistance and inputting it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan diagram requiring design assistance includes: Feature extraction is performed on the floor plan image that requires auxiliary design to obtain the floor plan feature vector and window feature vector; Cluster analysis is performed using the feature vectors of the apartment layout to obtain the design scheme information for the first window type; cluster analysis is performed using the feature vectors of the window type to obtain the design scheme information for the second window type. The at least one window design scheme information is determined based on the first window design scheme information and the second window design scheme information.
4. The method according to claim 1, characterized in that, Determine the profile information corresponding to each window design scheme, including: Based on the determined window design scheme, the profile type, quantity, manufacturer, construction time, and cost of each window design scheme are matched according to preset rules to form the profile information. The profile information and the window design scheme are output and displayed together.
5. The method according to claim 1, characterized in that, The process of obtaining the floor plan diagram requiring design assistance and inputting it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan diagram requiring design assistance includes: Input the floor plan of the apartment that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan, and determine the functional area where the window object is located; Based on the functional areas, determine the window type requirement information, and based on the window type requirement information, determine at least one window type design scheme information.
6. A window-type object-aided design device based on a neural network model, characterized in that, The device includes: The acquisition module is used to acquire a training set of sample apartment layout diagrams and to annotate the sample apartment layout diagrams and window object attributes in the diagrams in the training set; the window object attributes include the position attributes of doors and windows, and the position attributes of doors and windows are annotated according to the functional areas in the sample apartment layout diagrams; The training module is used to input the labeled sample apartment layout diagram into a preset neural network model for training, so as to obtain a window object-assisted design model. The auxiliary design module is used to obtain the floor plan of the apartment requiring auxiliary design and input the window object auxiliary design model to obtain at least one window design scheme information corresponding to the floor plan image requiring auxiliary design, as well as the profile information corresponding to each window design scheme information; the window design scheme information is used to indicate the design of window objects in the floor plan.
7. The apparatus according to claim 6, characterized in that, The auxiliary design module includes: The extraction unit is used to extract features from the floor plan image that requires auxiliary design, and obtain the floor plan feature vector and the window feature vector. The analysis unit is used to perform cluster analysis using the apartment structure feature vector to obtain the first window design scheme information, and to perform cluster analysis using the window feature vector to obtain the second window design scheme information. The determining unit is configured to determine the at least one window design scheme information based on the first window design scheme information and the second window design scheme information.
8. The apparatus according to claim 6, characterized in that, The auxiliary design module includes: Obtain the floor plan of the apartment requiring design assistance and input it into the window object-based design assistance model to obtain at least one window design scheme information corresponding to the floor plan image requiring design assistance, including: Input the floor plan of the apartment that needs to be assisted in the design into the window object assisted design model, locate the window object in the floor plan, and determine the functional area where the window object is located; Based on the functional areas, determine the window type requirement information, and based on the window type requirement information, determine at least one window type design scheme information.
9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor is used to implement the steps of any one of claims 1-5 when executing the computer program.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of any one of claims 1-5.