Rendering method, equipment and medium of building appearance effect drawing based on three-dimensional image
Through the three-dimensional drawing-based building appearance rendering method, the problem of long construction time of building architectural rendering is solved, and the rapid construction of building appearance renderings that meet user needs is achieved.
Patent Information
- Application Number
- CN202410449453.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-15
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-04-15
AI Technical Summary
In the construction industry, the construction renderings in the existing technology take a long time to build. When the first renderings do not meet user needs, multiple rounds of communication are required to form the final renderings.
Using the three-dimensional diagram-based building appearance rendering method, by receiving the three-dimensional architectural design structure diagram and structural diagram description text input by the target user terminal, a query information set of component points, edges and frame edges is generated, query statements are executed and component diagrams are obtained. Query the results, fill them into the three-dimensional structural diagram for rendering, and quickly build the building appearance renderings.
The construction time of architectural renderings is shortened, and user needs are quickly identified through user voice descriptions and historical building appearance renderings, and building appearance renderings are generated to meet user preferences.
Smart Images

Figure CN118154802B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of building appearance rendering, and in particular to a method, device and medium for rendering a building appearance effect drawing based on a three-dimensional image. Background Art
[0002] In the construction industry, before constructing a building, a rendering of the building is usually designed for users to refer to. At present, when constructing a rendering of a building, the usual method is: the designer first constructs a rendering, and then modifies it after communicating with the customer to form the final building rendering. However, this method usually has the following disadvantages: when the initial rendering does not meet the user's needs, it may take multiple rounds of communication to form the final building rendering, resulting in a long construction time for the building rendering. Summary of the invention
[0003] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.
[0004] Some embodiments of the present disclosure propose a method for rendering a building appearance effect diagram based on a three-dimensional diagram, a computer device, and a computer-readable storage medium to solve the technical problems mentioned in the above background technology section.
[0005] In a first aspect, some embodiments of the present disclosure provide a method for rendering a building exterior effect drawing based on a three-dimensional image, the method comprising: receiving a three-dimensional building design structure drawing and a structure drawing description text input by a target user terminal, wherein the three-dimensional building design structure drawing comprises a building component block diagram group, and the structure drawing description text comprises: a building component description information group, wherein one building component block diagram corresponds to one building component description information; for each building component description information in the building component description information group, performing the following processing steps: generating a component point query information set corresponding to a component point structure, a component edge query information set corresponding to a component edge structure, and a group of corresponding component frame edge structures according to the building component description information; The invention relates to a method for preparing a component diagram of a building component. The method comprises the steps of: generating a component diagram query task information set corresponding to the component point query information set, the component edge query information set and the component framework edge query information set; generating a component diagram query statement set corresponding to the component diagram query task information set; executing the component diagram query statement set in a local building component diagram database to obtain a component diagram query result set; generating a target building component sub-diagram corresponding to the building component description information according to the component diagram query result set; filling each generated target building component sub-diagram into the corresponding building component block diagram in the three-dimensional building design structure diagram, and rendering the filled three-dimensional building design structure diagram to obtain a building appearance rendering.
[0006] In a second aspect, the present disclosure further provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the method described in any implementation of the first aspect is implemented.
[0007] In a third aspect, the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.
[0008] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the building appearance rendering method based on three-dimensional images of some embodiments of the present disclosure, the construction time of the building rendering is shortened. First, the building description audio input by the target user terminal is received. The above-mentioned building description audio is a parameter descriptive audio describing the building appearance rendering. Thus, the building appearance rendering can be constructed according to the user's voice description. Secondly, according to the user's regional information corresponding to the above-mentioned building description audio, an audio recognition model corresponding to the above-mentioned user's regional information is selected from the pre-trained audio recognition model set as the target audio recognition model. Thus, it is convenient to recognize the audio. Thirdly, according to the above-mentioned target audio recognition model, the above-mentioned building description audio is subjected to audio text recognition to obtain the building description audio text, and the above-mentioned building description audio text is sent to the above-mentioned target user terminal. Thus, the user's building demand text can be identified. Then, in response to receiving the determination information corresponding to the above-mentioned building description audio text sent by the above-mentioned target user terminal, the above-mentioned building description audio text is subjected to parameter index extraction according to the preset parameter index comparison table to obtain a parameter index information group. Thus, each parameter index can be extracted, which is convenient for determining each component of the building rendering. Next, in response to determining that the target user is a historical target user, a group of historical building appearance renderings corresponding to the target user is obtained from a local database. Therefore, the user's preference for architectural style can be determined based on the user's historical building appearance renderings. Then, the historical building description text of each historical building appearance rendering in the above-mentioned historical building appearance rendering group is obtained to obtain a historical building description text group. Among them, the historical building description text in the above-mentioned historical building description text group includes: a standard parameter indicator information group. Therefore, it is convenient for the historical building description text to determine the building appearance sub-image corresponding to the parameter indicator. Finally, according to the above-mentioned parameter indicator information group and each standard parameter indicator information group, a target building appearance sub-image group is determined, wherein one parameter indicator information corresponds to one target building appearance sub-image; according to the above-mentioned target building appearance sub-image group, a building appearance rendering corresponding to the above-mentioned building description audio is constructed. Therefore, the building appearance rendering can be quickly constructed according to the building description audio provided by the user and the corresponding historical building appearance rendering. Thereby, the construction time of the building rendering is shortened. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.
[0010] Figure 1is a flow chart of some embodiments of a method for rendering a building appearance effect diagram based on a three-dimensional diagram according to the present disclosure;
[0011] Figure 2 It is a schematic block diagram of the structure of a computer device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION
[0012] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.
[0013] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.
[0014] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.
[0015] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".
[0016] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.
[0017] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.
[0018] Figure 1 The flowchart of some embodiments of the method for rendering a building appearance effect diagram based on a three-dimensional image according to the present disclosure is shown. The method for rendering a building appearance effect diagram based on a three-dimensional image according to some embodiments of the present disclosure is shown. The method for rendering a building appearance effect diagram based on a three-dimensional image comprises the following steps:
[0019] Step 101: receiving a three-dimensional architectural design structure diagram and a structure diagram description text input by a target user terminal.
[0020] In some embodiments, an execution subject (e.g., a computing device) of a method for rendering a building appearance rendering based on a three-dimensional image may receive a three-dimensional building design structure diagram and a structure diagram description text input by a target user terminal. The three-dimensional building design structure diagram includes a group of building component block diagrams, and the structure diagram description text includes: a group of building component description information, and one building component block diagram corresponds to one building component description information. The building component description information may refer to an architectural style requirement text describing the building component block diagram, which may include: size, style, door and window requirements, door and window style, floor height requirements, exterior wall color requirements, step style, etc. The structure diagram description text may be a style text describing the building appearance rendering. For example, the structure diagram description text may include: a luxurious villa located in a quiet, sun-kissed countryside, (ultra-detailed, high-rise: 1.2), (realistic, photo-realistic: 1.37), HDR, professional, vivid colors, defocused, oil painting style, warm and attractive palette, soft natural lighting.
[0021] Step 102: for each building component description information in the building component description information group, perform the following processing steps:
[0022] Step 1021, generating a component point query information set corresponding to the component point structure, a component edge query information set corresponding to the component edge structure, and a component frame edge query information set corresponding to the component frame edge structure according to the building component description information.
[0023] In some embodiments, the execution entity may generate a component point query information set corresponding to the component point structure, a component edge query information set corresponding to the component edge structure, and a component frame edge query information set corresponding to the component frame edge structure according to the building component description information.
[0024] Specifically, the execution subject may perform a request analysis on the building component description information to generate component graph query requirement information. The component graph query requirement information may be requirement information for querying component graph structure data. The component graph structure data may be structure data of the component graph. The component graph query requirement information may include but is not limited to: the component graph data result to be queried, and the component graph structure data screening condition. The component graph data result to be queried may be a component graph. The component graph structure data screening condition may be a condition for how to screen out the component graph data to be queried. For example, the data screening condition may be to screen out edges with a length of **. The component point structure may be component graph structure data in the form of vertices in the building component. The component edge structure may be graph structure data in the form of edges. For example, it may be edge data in a building component subgraph. The component framework edge structure may be building component subgraph structure data in the form of component framework edges. For example, it may be edge data of a main frame in a building component subgraph. The component point query information may be query information for querying the building component subgraph structure data in the form of component point structure. For example, the component point query information may be "querying node data with a node identifier of "001"". The component edge query information may be query information for querying the building component subgraph structure data in the form of component edge structure. For example, the component edge query information may be "querying component edge data including the target node". The component framework edge query information may be query information for querying the building component subgraph structure data in the form of component framework edge structure. For example, the component framework edge query information may be "querying data of component framework edge length of 5".
[0025] Step 1022, generating a component graph query task information set corresponding to the component point query information set, the component edge query information set and the component framework edge query information set.
[0026] In some embodiments, the execution subject may generate a component graph query task information set corresponding to the component point query information set, the component edge query information set, and the component framework edge query information set. The component graph query task information in the component graph query task information set may be task information corresponding to the component graph query task. In practice, the task information may be a task identifier. The component graph query task may be a task for executing a direct query operation on the component graph.
[0027] In an actual application scenario, the execution subject may generate a component graph query task information set corresponding to the component point query information set, the component edge query information set, and the component framework edge query information set through the following steps:
[0028] The first step is to generate component graph query task information corresponding to each component point query information in the component point query information set, that is, directly generate corresponding component graph query task information for the component point query information of each component point structure.
[0029] In the second step, for each component edge query information in the component edge query information set, the following processing steps are performed:
[0030] 1. Determine the component edge description information corresponding to the component edge query information. The component edge description information may be description information describing the relevant structure of the component edge in the component edge query information. For example, the component edge description information may include: component edge node description information, component edge position description information, and component edge value description information. The component edge node description information may be description information corresponding to the node group associated with the component edge. The component edge position description information may be description information of the position of the component edge in the building component subgraph. The component edge value description information may be description information of the associated values between the nodes in the node group corresponding to the component edge.
[0031] 2. In response to determining that the component edge description information satisfies the first preset condition, generate component graph query task information for the component edge query information. The information completeness corresponding to the first preset condition is greater than the information completeness corresponding to the third preset condition. The information completeness corresponding to the first preset condition is (85%, 100%). Therefore, through the component graph query task, the component graph structure data query for the component edge query information can be realized.
[0032] 3. In response to determining that the component edge description information satisfies the third preset condition, generate target component graph query task information. The target component graph query task information is the component graph query task information corresponding to the component edge description information. The component edge query information may be information that the corresponding component edge description information satisfies the third preset condition. The third preset condition may be a characterization that the degree of completeness of the information corresponding to the component edge description information is in a first interval. The first interval may be [75%-85%]. The degree of completeness of the information in the first interval indicates that it is unclear whether to perform a component graph query operation. Therefore, for this type of component edge query information, under the premise of sufficient computing resources, target component graph query task information for the component edge query information may be generated at the same time.
[0033] In the third step, for each component framework edge query information in the component framework edge query information set, the following information processing steps are performed:
[0034] 1. Determine the component framework edge description information corresponding to the component framework edge query information. The component framework edge description information may be description information describing the relevant structure of the component framework edge in the component framework edge query information. For example, the component framework edge description information may include: component framework edge node description information, component framework edge position description information, and component framework edge numerical description information. The component framework edge node description information may be description information corresponding to the node group associated with the component framework edge. The component framework edge position description information may be description information of the position of a framework composed of multiple component framework edges in a building component subgraph. The component framework edge numerical description information may be edge lengths corresponding to multiple component framework edges.
[0035] 2. In response to determining that the component framework edge description information satisfies the second preset condition, generate component graph query task information for the component framework edge query information. The second preset condition corresponds to the information completeness degree of (75%, 100%]. The second preset condition can characterize that the description degree of the component framework edge description information is relatively complete. Therefore, through the component graph query task, the component graph structure data query for the component framework edge query information can be realized.
[0036] Step 1023, generating a component graph query statement set corresponding to the component graph query task information set.
[0037] In some embodiments, the execution subject may generate a component graph query statement set corresponding to the component graph query task information set. The component graph query statements in the component graph query statement set correspond to the component graph query task information in the component graph query task information set. The component graph query statement may be a database statement for executing a component graph query operation.
[0038] In an actual application scenario, the execution subject may perform the following processing steps for each component graph query task information in the component graph query task information set:
[0039] The first step is to determine the component diagram query complexity corresponding to the component diagram query task information. The component diagram query complexity can represent the query difficulty of the task corresponding to the component diagram query task information. For example, the component diagram query complexity can be represented by a numerical value, and the larger the numerical value, the higher the query difficulty. The above-mentioned execution entity can determine the component diagram query complexity corresponding to the above-mentioned component diagram query task information according to the schema of the building component diagram database.
[0040] In the second step, according to the component diagram query complexity, the component diagram structure corresponding to the component diagram query task information and the query condition, the component diagram construction tool information corresponding to the component diagram query task information is obtained. Among them, the query condition can be the query condition in the information content corresponding to the component diagram query task information. The query condition can represent the logic of how to perform information query. The tool corresponding to the component diagram construction tool information can be a commander tool. For example, the component diagram construction tool information can be HasVidCommander tool information. First, the above-mentioned execution subject can determine whether the component diagram query complexity is greater than the first value. Then, in response to determining that the component diagram query complexity is greater than the first value, according to the query condition and the component diagram structure, the component diagram query task information is divided into information to generate a component diagram query task information subset corresponding to the component diagram query complexity less than or equal to the above-mentioned first value. Then, the corresponding component diagram construction tool is assigned to the above-mentioned component diagram query task information subset to obtain the component diagram construction tool information. Among them, the specific allocation method of the first construction tool can be preset. For example, for the component diagram query task, the HasVidCommander tool is used to generate subsequent query statements.
[0041] The third step is to use the component graph construction tool corresponding to the component graph construction tool information to generate a component graph query statement for the component graph query task information.
[0042] Step 1024, execute the component graph query statement set in the local building component graph database to obtain a component graph query result set.
[0043] In some embodiments, the above-mentioned execution entity can execute the component graph query statement set in the local building component graph database to obtain a component graph query result set.
[0044] In an actual application scenario, the execution subject may perform the following processing steps in response to determining that the target component graph query statement exists in the component graph query statement set: in response to determining that the execution speed of the execution result corresponding to the target component graph query statement is faster than a preset speed, determine the task execution result corresponding to the target component graph query statement as the component graph query result corresponding to the component edge description information. The component graph query result in the component graph query result set may be the execution result of the component graph query task.
[0045] Step 1025: Generate a target building component subgraph corresponding to the building component description information according to the component graph query result set.
[0046] In some embodiments, the execution subject may generate a target building component subgraph corresponding to the building component description information according to the component graph query result set. That is, the building component subgraph corresponding to each component graph query result in the component graph query result set may be determined to obtain a building component subgraph set. Thereafter, the similarity between the description information corresponding to each building component subgraph and the building component description information may be determined. Finally, the building component subgraph with the greatest similarity may be determined as the target building component subgraph. The similarity between the description information corresponding to each building component subgraph and the building component description information may be determined by the cosine similarity formula.
[0047] Step 103, filling each generated target building component sub-graph into the corresponding building component block diagram in the three-dimensional building design structure diagram, and rendering the filled three-dimensional building design structure diagram to obtain a building appearance rendering.
[0048] In some embodiments, the above-mentioned execution subject can fill each generated target building component sub-graph into the corresponding building component block diagram in the three-dimensional building design structure diagram, and render the filled three-dimensional building design structure diagram to obtain a building appearance rendering. For example, first, the building component block diagram corresponding to each target building component sub-graph can be determined. Then, the target building component sub-graph is filled into the building component block diagram. Finally, the filled three-dimensional building design structure diagram can be rendered by an image renderer to obtain a building appearance rendering. It should be noted that the building component sub-graphs are all architectural color pictures. For example, the image renderer can be 3ds Max, Maya, SketchUp, Rhino, etc.
[0049] In an actual application scenario, the above execution subject can render the filled three-dimensional building design structure diagram through the following steps to obtain the building appearance rendering:
[0050] In the first step, the filled three-dimensional architectural design structure diagram is input into a pre-trained three-dimensional architectural drawing detection model to obtain a three-dimensional architectural drawing detection result. The three-dimensional architectural drawing detection result can indicate whether there are defects in the filling of the three-dimensional architectural design structure diagram. The three-dimensional architectural drawing detection model can be a pre-trained neural network model that takes the filled three-dimensional architectural design structure diagram as input and the three-dimensional architectural drawing detection result as output. For example, the three-dimensional architectural drawing detection result can indicate that a sub-image of a target building component in the three-dimensional architectural design structure diagram is filled crookedly. For example, the three-dimensional architectural drawing detection model can be a machine learning model structure using various target detections, such as YOLOv5 or YOLOv8.
[0051] In the second step, in response to determining that the three-dimensional building image detection result indicates that there are no defects, the filled three-dimensional building design structure diagram is input into a pre-trained image optimization network model to obtain an optimized three-dimensional building design structure diagram as a building appearance rendering. The image optimization network model can be a pre-trained neural network model that takes the filled three-dimensional building design structure diagram as input and takes the optimized three-dimensional building design structure diagram as output. For example, the image optimization network model can be a trained Diffusion Model diffusion model.
[0052] Furthermore, the 3D building image detection model can be trained by the following steps:
[0053] The first step is to obtain a 3D architectural design structure diagram training sample set, wherein the 3D architectural design structure diagram training samples in the 3D architectural design structure diagram training sample set include 3D architectural design structure sample diagrams.
[0054] The second step is to select a target 3D architectural design structure diagram training sample from the 3D architectural design structure diagram training sample set. A 3D architectural design structure diagram training sample can be randomly selected from the 3D architectural design structure diagram training sample set as the target 3D architectural design structure diagram training sample.
[0055] In the third step, the three-dimensional architectural design structure sample images included in the target three-dimensional architectural design structure training sample are input into the feature extraction network included in the initial three-dimensional architectural image detection model to perform hierarchical feature extraction processing on the three-dimensional architectural design structure sample images to obtain the features of the three-dimensional architectural design structure sample images of each layer. The feature extraction network can be a convolutional neural network based on a deep learning method. As an example, the feature extraction network can be a backbone network improved by adding an attention mechanism on the basis of the backbone feature extraction network used by the target detection algorithm. For example, on the basis of the backbone network (CSPDarknet), the SE (Squeeze and Excitation) attention mechanism is added to obtain an improved SE-CSPDarknet backbone network. The feature extraction network can include multiple convolution modules, multiple attention mechanism convolution modules, and pooling modules. It can be used to extract the image features of the sample images of the appearance effects of each layer of the building. For example, the backbone feature extraction network includes 6 convolution modules, 4 CSP-SE convolution modules (i.e., attention mechanism convolution modules), and 1 SPPF module (i.e., pooling module). It should be noted that the specific parameters (such as input feature map size, output feature map size, network parameters, etc.) and structures of each module can also be set according to actual conditions. The above-mentioned attention mechanism convolution module may include: a first convolution submodule, which is used to perform a two-dimensional convolution operation on the feature map input to the attention mechanism convolution module to obtain a first convolution feature map; a splitting submodule, which is used to split the first convolution feature map according to channel two to obtain a split feature map; a plurality of sequentially connected residual convolution submodules, which are used to perform convolution and residual connection processing on the input feature map to obtain a residual convolution feature map; a residual connection submodule, which is used to superimpose the split feature map with multiple residual convolution feature maps according to the channel to obtain a residual feature map; a second convolution submodule, which is used to perform a two-dimensional convolution operation on the residual feature map to obtain a second convolution feature map; an attention mechanism submodule, which is used to process the second convolution feature map to obtain an output feature map of the attention mechanism convolution module. For example, the residual convolution submodule may include two convolution modules (with a convolution kernel of 3×3, a step size of 1, and a padding of 1), and a residual connection module. The residual connection module can add the input feature map and the feature map output by the second convolution module according to the channel to obtain the output feature map of the residual convolution submodule. The attention mechanism submodule can include a global average pooling operation, two two-dimensional convolution operations (the convolution kernel is 1×1, the step size is 1, and the padding is 0), and two activation functions (such as Relu and Sigmod), and then the input feature map can be multiplied with the corresponding feature map weight according to the feature channel to obtain the output feature map of the attention mechanism submodule.Among them, the SPPF module can include a two-dimensional convolution operation (Conv2d, convolution kernel is 1x1, step size is 1, padding is 0), 3 maximum pooling operations (Max Pooling, pooling kernel is 5), a residual connection module, a channel addition module, and another two-dimensional convolution operation (Conv2d, convolution kernel is 1x1, step size is 1, padding is 0).
[0056] As an example, the backbone feature extraction network in this embodiment may include ten feature extraction stages, that is, ten feature extraction modules connected in sequence, the first stage: convolution module-input feature map size 320x320x64; the second stage: CSP-SE convolution module-input feature map size 160x160x128; the third stage: convolution module-input feature map size 160x160x128; the fourth stage: CSP-SE convolution module-input feature map size 80x80x256; the fifth stage: convolution module-input feature map size 80x80x256; the sixth stage: CSP-SE convolution module-input feature map size 40x40x512; the seventh stage: convolution module-input feature map size 40x40x512; the eighth stage: CSP-SE convolution module-input feature map size 20x20x512; the ninth stage: SPPF pooling module-input feature map size 20x20x512.
[0057] In the fourth step, the features of the sample images of the three-dimensional architectural design structure of each layer are input into the feature fusion network included in the initial three-dimensional architectural image detection model to obtain the fusion features of each sample architectural design structure image. The feature fusion network can be set according to actual needs, such as a bottom-up upsampling feature fusion structure, or a top-down downsampling feature fusion structure. Here, the upsampling feature fusion structure can first upsample the feature map output by the ninth stage by a specified multiple to match the feature map output by the sixth stage and perform channel superposition. Then, the superimposed feature map can be processed by the attention mechanism convolution module to output the first structure map fusion feature map. After that, the first fusion feature map can be upsampled by a specified multiple to match the feature map output by the fourth stage and perform channel superposition. Finally, the superimposed feature map can be processed by the attention mechanism convolution module to output the second structure map fusion feature map.
[0058] The fifth step is to input the fusion features of each sample architectural design structure diagram into the positioning detection network included in the initial three-dimensional architectural drawing detection model to obtain the positioning detection result. The positioning detection network may include two parts: a classification structure and a positioning structure. It is understandable that the specific structure of the positioning detection network can also be set according to actual needs. As an example, the classification structure in the classification positioning network may include a convolution module (convolution kernel is 3, step size is 1, and padding is 0), a two-dimensional convolution operation (Conv2d, convolution kernel is 1×1, step size is 1, and padding is 0). The positioning structure may include a convolution module (convolution kernel is 3, step size is 1, and padding is 0), a two-dimensional convolution operation (Conv2d, convolution kernel is 1×1, step size is 1, and padding is 0). The execution subject may input the fusion features of each sample architectural design structure diagram into the positioning detection network in the initial three-dimensional architectural drawing detection model, thereby obtaining the classification data and positioning data of the fusion features of each sample architectural design structure diagram. Next, the classification data of the fusion features of each sample architectural design structure diagram are dimensionally superimposed to obtain the total classification prediction data of the three-dimensional architectural design structure sample diagram. Among them, the classification prediction dimension can include the number of predicted detection frames and the predicted value of each detection frame belonging to each labeled category. In addition, the positioning data of the fusion features of each sample architectural design structure diagram can be dimensionally superimposed to obtain the total positioning prediction data of the three-dimensional architectural design structure sample diagram. Among them, the positioning prediction dimension can include the number of predicted detection frames and the predicted position of each detection frame in the three-dimensional architectural design structure sample diagram.
[0059] Step 6: Based on the positioning detection results, the output result of the initial 3D building image detection model is generated. For the positioning detection results of the fusion features of each sample building design structure image of the same 3D building design structure sample image, the execution subject can calculate the average value as the output result of the model.
[0060] The seventh step is to determine the loss value between the sample label corresponding to the target three-dimensional architectural design structure diagram training sample and the output result. In practice, the above-mentioned execution subject can determine the loss value between the sample label corresponding to the target three-dimensional architectural design structure diagram training sample and the output result through a hinge loss function or a cross entropy loss function.
[0061] In the eighth step, in response to determining that the loss value is less than or equal to a preset loss value, the initial three-dimensional building image detection model is determined as a trained three-dimensional building image detection model.
[0062] Figure 2 A schematic block diagram of the structure of a computer device provided in an embodiment of the present disclosure. The computer device may be a terminal.
[0063] like Figure 2As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.
[0064] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the methods for rendering a building appearance effect diagram based on a three-dimensional image.
[0065] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.
[0066] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any method for rendering a building appearance effect diagram based on a three-dimensional image.
[0067] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 2 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present disclosure, and does not constitute a limitation on the computer device to which the scheme of the present disclosure is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0068] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0069] In one embodiment, the processor is used to run a computer program stored in a memory to implement the following steps: receiving a three-dimensional building design structure diagram and a structure diagram description text input by a target user terminal, wherein the three-dimensional building design structure diagram includes a building component block diagram group, and the structure diagram description text includes: a building component description information group, one building component block diagram corresponds to one building component description information; for each building component description information in the building component description information group, performing the following processing steps: generating a component point query information set corresponding to a component point structure, a component edge query information set corresponding to a component edge structure, and a group of corresponding component frame edge structures according to the building component description information; The invention relates to a method for preparing a component diagram of a building component. The method comprises the steps of: generating a component diagram query task information set corresponding to the component point query information set, the component edge query information set and the component framework edge query information set; generating a component diagram query statement set corresponding to the component diagram query task information set; executing the component diagram query statement set in a local building component diagram database to obtain a component diagram query result set; generating a target building component sub-diagram corresponding to the building component description information according to the component diagram query result set; filling each generated target building component sub-diagram into the corresponding building component block diagram in the three-dimensional building design structure diagram, and rendering the filled three-dimensional building design structure diagram to obtain a building appearance rendering.
[0070] The embodiments of the present disclosure also provide a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions. The method implemented when the program instructions are executed can refer to the various embodiments of the method for rendering a building appearance effect diagram based on a three-dimensional image in the present disclosure.
[0071] The computer-readable storage medium may be an internal storage unit of the computer device in the aforementioned embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., provided on the computer device.
[0072] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.
[0073] The serial numbers of the embodiments of the present disclosure are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific implementation mode of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be based on the protection scope of the claims.
Claims
1. A method for rendering a building appearance effect diagram based on a three-dimensional image, characterized in that: The method comprises: Receiving a three-dimensional building design structure diagram and a structure diagram description text input by a target user terminal, wherein the three-dimensional building design structure diagram includes a building component block diagram group, and the structure diagram description text includes: a building component description information group, where one building component block diagram corresponds to one building component description information; For each building component description information in the building component description information group, the following processing steps are performed: Generate, according to the building component description information, a component point query information set corresponding to the component point structure, a component edge query information set corresponding to the component edge structure, and a component frame edge query information set corresponding to the component frame edge structure; Generate a component graph query task information set corresponding to the component point query information set, the component edge query information set and the component framework edge query information set; Generate a component graph query statement set corresponding to the component graph query task information set; Execute the component graph query statement set in the local building component graph database to obtain a component graph query result set; Generate a target building component subgraph corresponding to the building component description information according to the component graph query result set; Filling each generated target building component sub-graph into the corresponding building component block diagram in the three-dimensional building design structure diagram, and rendering the filled three-dimensional building design structure diagram to obtain a building appearance rendering; The generating of the component graph query task information set corresponding to the component point query information set, the component edge query information set and the component framework edge query information set comprises: Generate component graph query task information corresponding to each component point query information in the component point query information set; For each component edge query information in the component edge query information set, the following processing steps are performed: Determine component edge description information corresponding to the component edge query information; In response to determining that the component edge description information satisfies a first preset condition, generating component graph query task information for the component edge query information; In response to determining that the component edge description information satisfies a third preset condition, generating target component graph query task information, wherein the target component graph query task information is the component graph query task information corresponding to the component edge description information; For each component frame edge query information in the component frame edge query information set, the following information processing steps are performed: Determine the component framework edge description information corresponding to the component framework edge query information; In response to determining that the component framework edge description information satisfies a second preset condition, generating component graph query task information for the component framework edge query information; The step of rendering the filled three-dimensional architectural design structure diagram to obtain the architectural appearance rendering includes: Inputting the filled three-dimensional architectural design structure drawing into a pre-trained three-dimensional architectural drawing detection model to obtain a three-dimensional architectural drawing detection result, wherein the three-dimensional architectural drawing detection result indicates whether there are defects in the filling of the three-dimensional architectural design structure drawing; In response to determining that the three-dimensional building image detection result indicates that there are no defects, inputting the filled three-dimensional building design structure diagram into a pre-trained image optimization network model to obtain an optimized three-dimensional building design structure diagram as a building appearance rendering; Wherein, generating a component graph query statement set corresponding to the component graph query task information set includes: Determine the component graph query complexity corresponding to the component graph query task information, wherein the component graph query complexity represents the query difficulty of the task corresponding to the component graph query task information; According to the complexity of the component graph query, the component graph structure corresponding to the component graph query task information and the query condition, the component graph construction tool information corresponding to the component graph query task information is obtained; and a component graph query statement for the component graph query task information is generated using the component graph construction tool corresponding to the component graph construction tool information; Among them, the 3D building image detection model is trained through the following steps: Acquire a three-dimensional architectural design structure diagram training sample set, wherein the three-dimensional architectural design structure diagram training samples in the three-dimensional architectural design structure diagram training sample set include three-dimensional architectural design structure sample diagrams; Selecting a target three-dimensional building design structure drawing training sample from the three-dimensional building design structure drawing training sample set; Inputting the three-dimensional architectural design structure sample images included in the target three-dimensional architectural design structure image training samples into a feature extraction network included in the initial three-dimensional architectural image detection model to perform hierarchical feature extraction processing on the three-dimensional architectural design structure sample images to obtain features of the three-dimensional architectural design structure sample images of each layer; Inputting the sample image features of the three-dimensional architectural design structure of each layer into the feature fusion network included in the initial three-dimensional architectural image detection model to obtain the fusion features of each sample architectural design structure image; The fusion features of each sample architectural design structure drawing are input into the positioning detection network included in the initial three-dimensional architectural drawing detection model to obtain a positioning detection result; Based on the positioning detection result, generating an output result of the initial three-dimensional building image detection model; Determine a loss value between a sample label corresponding to the target three-dimensional architectural design structure drawing training sample and the output result; In response to determining that the loss value is less than or equal to a preset loss value, the initial three-dimensional building image detection model is determined as a trained three-dimensional building image detection model.
2. The method for rendering a building appearance effect diagram based on a three-dimensional diagram according to claim 1, characterized in that: The executing the component graph query statement set in the local building component graph database to obtain the component graph query result set includes: In response to determining that the target component graph query statement exists in the component graph query statement set, the following processing steps are performed: In response to determining that the execution speed of the execution result corresponding to the target component graph query statement is faster than a preset speed, the task execution result corresponding to the target component graph query statement is determined as the component graph query result corresponding to the component edge description information.
3. A computer device, wherein: The computer device comprises a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 2 are implemented.
4. A computer-readable storage medium, wherein: The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 2 are implemented.
