Function area division method and device of scene area, storage medium and electronic device

Abstract

The application discloses a scene area function area division method and device, a storage medium and an electronic device, relates to the intelligent family technical field, and the method comprises the steps of inputting a target area into a target division model to obtain division information output by the target division model; a division strategy corresponding to the division information is matched from a division strategy set corresponding to the target area as a target division strategy, wherein the division strategy set comprises Q division strategies meeting a target division condition, each division strategy is used for indicating a division function area mode of the target area, and the number of target vertices located on the boundaries of each function area in the division mode indicated by the division strategy meeting the target division condition is greater than the number of target vertices, and the target vertices are vertices of the target area; and the target area is divided by using the target division strategy.

Description

Technical Field

[0001] This application relates to the field of smart home technology, and more specifically, to a method and apparatus for dividing a scene area into functional zones, a storage medium, and an electronic device. Background Technology

[0002] With the development of artificial intelligence technology, deep learning is widely used in various fields, serving and helping people to handle various tasks, bringing great convenience to life;

[0003] Currently, deep learning technology can help people divide scene regions into functional areas. For example, deep learning models can provide multiple division strategies based on the input scene region. Each division strategy can indicate how to divide the functional areas in the scene region, including the location and boundaries of the functional areas. However, the boundaries of each functional area in the division strategies output by current deep learning models often violate common sense, resulting in low practicality in dividing functional areas.

[0004] There is still no effective solution to the problem of poor functional area division in scene regions in related technologies. Summary of the Invention

[0005] This application provides a method and apparatus for dividing a scene area into functional areas, a storage medium, and an electronic device, to at least solve the problem of poor functional area division effect in related technologies.

[0006] According to one embodiment of this application, a method for dividing a scene area into functional zones is provided, including:

[0007] The target region is input into the target partitioning model to obtain the partitioning information output by the target partitioning model. The target region is the area of ​​functional areas to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional areas included in the target region.

[0008] The partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices. The target vertices are the vertices of the target region.

[0009] The target region is divided using the target segmentation strategy described above.

[0010] Optionally, before matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, the method further includes:

[0011] Obtain the target vertex information and associated vertex information corresponding to the target region, wherein the target vertex information is used to indicate the position of each vertex among the N target vertices of the target region, where N is a positive integer greater than or equal to 4, and the associated vertex information is used to indicate the position of each vertex among the M associated vertices in the target scene associated with the target region, where M is a positive integer greater than or equal to 1;

[0012] Based on the target vertex information and associated vertex information corresponding to the target region, Q partitioning strategies that satisfy the target partitioning conditions are generated, thus obtaining the partitioning strategy set.

[0013] Optionally, the step of generating Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, to obtain the partitioning strategy set, includes:

[0014] The N target vertices and the M associated vertices are projected onto the X-axis and Y-axis of the target Cartesian coordinate system, respectively, to obtain (N+M) first projection points on the X-axis and (N+M) second projection points on the Y-axis, wherein the target Cartesian coordinate system is a Cartesian coordinate system established on the plane where the target region is located;

[0015] A target partitioning vertex set is constructed based on (N+M) first projection points and (N+M) second projection points, wherein any target partitioning vertex in the target partitioning vertex set falls into the target region, and the x-coordinate of the target partitioning vertex is the same as the x-coordinate of at least one first projection point, and the y-coordinate of the target partitioning vertex is the same as the y-coordinate of at least one second projection point.

[0016] Based on the target partitioning vertex set, Q partitioning strategies that satisfy the target partitioning conditions are generated to obtain the partitioning strategy set, wherein in the partitioning method indicated by each partitioning strategy, the two diagonal vertices of the functional area divided in the target region fall into the target partitioning vertex set.

[0017] Optionally, constructing the target partitioning vertex set based on (N+M) first projection points and (N+M) second projection points includes:

[0018] A first dividing line perpendicular to the X-axis is generated through each of the first projection points, resulting in (N+M) first dividing lines. A second dividing line perpendicular to the Y-axis is generated through each of the second projection points, resulting in (N+M) second dividing lines.

[0019] The intersection point of any first dividing line and any second dividing line is defined as (N+M). 2 One initial partitioned vertex;

[0020] From (N+M) 2 The target partitioning vertices that fall within the target region are selected from the initial partitioning vertices to obtain the target partitioning vertex set.

[0021] Optionally, the step of generating Q partitioning strategies that satisfy the target partitioning conditions based on the target partitioning vertex set, to obtain the partitioning strategy set, includes:

[0022] The number of target functional areas in the target area is determined based on the area of ​​the target area.

[0023] Construct a rectangular region as a functional area within the target region by taking any two target partitioning vertices from the target partitioning vertex set as diagonal vertices, thus obtaining Q partitioning strategies as the partitioning strategy set, wherein each partitioning strategy includes the number of functional areas specified in the target functional area set.

[0024] Optionally, before obtaining the target vertex information and associated vertex information corresponding to the target region, the method further includes:

[0025] The vertices of each region other than the target region in the multiple regions included in the target scene are obtained as candidate vertices, thus obtaining a candidate vertex set;

[0026] The candidate vertices in the candidate vertex set whose distance from the boundary of the target region is less than the target distance are identified as the associated vertices.

[0027] Optionally, before projecting the N target vertices and M associated vertices onto the X and Y axes of the target Cartesian coordinate system, the method further includes:

[0028] Obtain the minimum bounding rectangle of the target region;

[0029] A target Cartesian coordinate system is generated based on the minimum bounding rectangle, wherein the X-axis and Y-axis of the target Cartesian coordinate system are perpendicular or parallel to the sides of the minimum bounding rectangle.

[0030] Optionally, the step of inputting the target region into the target partitioning model to obtain the partitioning information output by the target partitioning model includes:

[0031] The target region is input into the target feature extraction layer to obtain the target geometric features and target attribute features output by the target feature extraction layer. The target geometric features are used to characterize the boundary, area and distance of the target region from the target scene exit, and the target attribute features are used to characterize the scene function of the target scene.

[0032] The target geometric features and the target attribute features are input into the target region segmentation layer to obtain the segmentation information output by the target region segmentation layer. The target segmentation model includes the target feature extraction layer and the target region segmentation layer connected in sequence. The target segmentation model is obtained by training the model parameters of the initial feature extraction layer and the initial region segmentation layer using multiple sets of region samples labeled with segmentation information.

[0033] Optionally, matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy includes:

[0034] Find P partitioning strategies from the set of partitioning strategies that are consistent with the partitioning information as reference partitioning strategies, where P is a positive integer greater than or equal to 1;

[0035] When P equals 1, the reference partitioning strategy is determined as the target partitioning strategy;

[0036] When P is greater than 1, the partitioning parameters of each of the reference partitioning strategies are obtained, wherein the partitioning parameters include the area ratio, the intersection ratio, and the number of overlapping points. In each functional area partitioning method, the area ratio is the ratio between the sum of the areas of the functional areas included in the target area and the area of ​​the target area. The intersection ratio is used to indicate the ratio between the union and intersection of the areas covered by each functional area. The number of overlapping points is used to indicate the number of vertices among the N target vertices that overlap with the vertices of each functional area. The reference partitioning strategy with the largest partitioning parameter is determined as the target partitioning strategy.

[0037] According to another embodiment of the present application, a functional area division device for a scene region is also provided, comprising:

[0038] The input module is used to input the target region into the target partitioning model and obtain the partitioning information output by the target partitioning model. The target region is the region of the functional area to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional area included in the target region.

[0039] The matching module is used to match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices. The target vertices are the vertices of the target region.

[0040] The segmentation module is used to segment the target region using the target segmentation strategy.

[0041] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to execute the above-mentioned functional area division method of the scene area when running.

[0042] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-mentioned scene area functional area division method through the computer program.

[0043] In this embodiment, the target region is input into the target partitioning model to obtain partitioning information output by the target partitioning model. The target region is the area in the target scene to be partitioned into functional areas, and the partitioning information indicates the location of the functional areas included in the target region. A partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy indicates a way to partition the target region into functional areas. In the partitioning methods indicated by the partitioning strategies that satisfy the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices, and the target vertices are the vertices of the target region. The target region is then partitioned using the target partitioning strategy. The process involves the target partitioning model outputting partitioning information corresponding to the target region. The target region is the area in the target scene to be partitioned into functional areas. The partitioning information indicates the location of the functional areas included in the target region. Then, a partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. Therefore, the obtained target partitioning strategy satisfies the target partitioning conditions. Since the number of target vertices on the boundaries of each functional area is greater than the number of target vertices in the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, and the target vertices are the vertices of the target region, the target partitioning strategy is finally used to partition the target region. By adopting the above technical solution, the problem of poor functional area partitioning effect in the scene area in related technologies is solved, and the technical effect of improving the functional area partitioning effect of the scene area is achieved. Attached Figure Description

[0044] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the hardware environment for a method of dividing a scene area into functional zones according to an embodiment of this application;

[0047] Figure 2 This is a flowchart of a method for dividing a scene area into functional zones according to an embodiment of this application;

[0048] Figure 3 This is a schematic diagram of a target scene and target area according to an embodiment of this application;

[0049] Figure 4 This is a schematic diagram of a reference partitioning strategy according to an embodiment of this application;

[0050] Figure 5 This is a schematic diagram of a target vertex and associated vertices according to an embodiment of this application;

[0051] Figure 6 This is a schematic diagram of a method for dividing a scene area into functional zones according to an embodiment of this application;

[0052] Figure 7 This is a schematic diagram of a functional area division framework for a scene region according to an embodiment of this application;

[0053] Figure 8 This is a structural block diagram of a scene area functional area division device according to an embodiment of this application. Detailed Implementation

[0054] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0055] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0056] According to one aspect of the embodiments of this application, a method for dividing a scene area into functional zones is provided. This method for dividing a scene area into functional zones is widely used in whole-house intelligent digital control application scenarios such as smart homes, smart home ecosystems, and smart house ecosystems. Optionally, in this embodiment, Figure 1This is a schematic diagram of a hardware environment for a method of dividing a scene area into functional areas according to an embodiment of this application. The above-described method of dividing a scene area into functional areas can be applied to, for example... Figure 1 The hardware environment shown consists of terminal device 102 and server 104. For example... Figure 1 As shown, server 104 is connected to terminal device 102 via a network and can be used to provide services (such as application services) to the terminal or clients installed on the terminal. A database can be set up on the server or independently of the server to provide data storage services for server 104. Cloud computing and / or edge computing services can be configured on the server or independently of the server to provide data processing services for server 104.

[0057] The aforementioned network may include, but is not limited to, at least one of the following: wired network, wireless network. The aforementioned wired network may include, but is not limited to, at least one of the following: wide area network, metropolitan area network, local area network. The aforementioned wireless network may include, but is not limited to, at least one of the following: Wi-Fi (Wireless Fidelity), Bluetooth. The terminal device 102 may not be limited to PC, mobile phone, tablet computer, smart air conditioner, smart range hood, smart refrigerator, smart oven, smart stove, smart washing machine, smart water heater, smart washing equipment, smart dishwasher, smart projector, smart TV, smart clothes rack, smart curtains, smart audio-visual equipment, smart socket, smart speaker, smart speaker box, smart fresh air equipment, smart kitchen and bathroom equipment, smart bathroom equipment, smart robot vacuum cleaner, smart window cleaning robot, smart mopping robot, smart air purifier, smart steam oven, smart microwave oven, smart water heater, smart air purifier, smart water dispenser, smart door lock, etc.

[0058] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0059] Machine Learning: Deep Learning (DL) is an interdisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. It is the core of artificial intelligence and the fundamental way to endow computers with intelligence.

[0060] Deep Learning (DL) is a new research direction in the field of machine learning. It was introduced into machine learning to bring it closer to its original goal—Artificial Intelligence (AI). Deep learning learns the inherent patterns and hierarchical representations of sample data. The information gained during this learning process greatly helps in interpreting data such as text, images, and sound. Its ultimate goal is to enable machines to have analytical and learning capabilities like humans, and to recognize data such as text, images, and sound.

[0061] Artificial rules: refer to specific rules formulated by people in the scene, such as the prior knowledge of designers. In the functional area division method of the scene area proposed in this application, the boundary of the functional area needs to meet certain artificial rules, that is, the number of target vertices on the boundary of each functional area is greater than the number of target vertices.

[0062] CNN: Convolutional Neural Networks, is a very common algorithm (model) in deep learning.

[0063] It should be noted that the functional area division method of the scene area proposed in this application is applicable to a wide range of scenarios, including but not limited to: office scenarios, warehouse scenarios, teaching scenarios, and home scenarios, etc. In the following embodiments, the functional area division process of the living and dining room area in the above-mentioned home scenario can be used as an example to illustrate the functional area division method of the scene area, but the application scenario is not limited.

[0064] This embodiment provides a method for dividing a scene area into functional zones, which is applied to the aforementioned device terminal. Figure 2 This is a flowchart of a method for dividing a scene area into functional zones according to an embodiment of this application, such as... Figure 2 As shown, the process includes the following steps:

[0065] Step S202: Input the target region into the target partitioning model to obtain the partitioning information output by the target partitioning model, wherein the target region is the region of functional areas to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional areas included in the target region;

[0066] Step S204: Match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices. The target vertices are the vertices of the target region.

[0067] Step S206: Divide the target region using the target segmentation strategy.

[0068] Through the above steps, the target partitioning model outputs partitioning information corresponding to the target region. The target region is the area in the target scene to be partitioned into functional areas. The partitioning information indicates the location of the functional areas included in the target region. Then, a partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. Therefore, the obtained target partitioning strategy satisfies the target partitioning conditions. Since the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices in the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the target vertices are the vertices of the target region. Finally, the target partitioning strategy is used to partition the target region. By adopting the above technical solution, the problem of poor functional area partitioning effect in scene areas in related technologies is solved, and the technical effect of improving the functional area partitioning effect of scene areas is achieved.

[0069] In the technical solution provided in step S202 above, the target scenario may include, but is not limited to, the aforementioned office scenario, warehouse scenario, teaching scenario, and home scenario, etc., and the target area is a sub-area of ​​the target scenario. Taking the home scenario as an example, the target scenario mentioned above is a sub-area of ​​the target scenario. Figure 3 This is a schematic diagram of a target scene and target area according to an embodiment of this application, such as... Figure 3As shown, a home scene may include multiple sub-areas, such as the bathroom, bedroom, kitchen, balcony, and public area (see the areas marked with thick lines in the diagram). Taking the public area as the target area in the target scene as an example, the public area is the area where people engage in public activities within the home scene, and it is the area to be divided into functional zones. Functional zones may include corridors, entryways, dining rooms, and living rooms. It is worth noting that the boundaries of the dining room and living room can be assumed to be divided into boundary 1 and boundary 2. Boundary 2 passes through one vertex (point A) of the target area, while boundary 1 does not pass through any vertex of the target area. Therefore, boundary 2 is a better division method than boundary 1. In the process of dividing the functional areas of the target area, the ideal boundary division method should try to pass through the vertices of the target area (which can be understood as the artificial rule mentioned above). Such boundary division results in functional areas that are more practical and closer to daily life, with better division effect.

[0070] Optionally, in this embodiment, the partitioning information is used to indicate the location of the functional areas included in the target area. It can be understood that the partitioning information corresponding to the target area can indicate which functional areas are included in the target area, and the location of each functional area in the target area. The target partitioning model can output matching partitioning information based on the characteristics of the target area. For example, if the area of ​​the target area is large enough, the corresponding partitioning information may indicate that the target area includes more functional areas, while if the area of ​​the target area is small, the corresponding partitioning information may indicate that the target area includes fewer functional areas. Unlike the current related technologies where the model directly gives the boundaries of functional areas, in this application, the model may not give the boundaries of functional areas. It can only refer to the location of the functional areas provided by the model and combine it with manual rules to determine the boundaries, thus avoiding the unreasonable functional area boundaries given by the model in the related technologies, which affect the partitioning effect of functional areas.

[0071] In an exemplary embodiment, the target region can be input into a target partitioning model to obtain partitioning information output by the target partitioning model, but not limited to the following methods: inputting the target region into a target feature extraction layer to obtain target geometric features and target attribute features output by the target feature extraction layer, wherein the target geometric features are used to characterize the boundary, area, and distance from the target region to the target scene exit of the target region, and the target attribute features are used to characterize the scene function of the target scene; inputting the target geometric features and the target attribute features into a target region partitioning layer to obtain the partitioning information output by the target region partitioning layer, wherein the target partitioning model includes the target feature extraction layer and the target region partitioning layer connected in sequence, and the target partitioning model is obtained by training the model parameters of the initial feature extraction layer and the initial region partitioning layer using multiple sets of region samples labeled with partitioning information labels.

[0072] Optionally, in this embodiment, the target geometric feature is one of the features of the target area. It can be a feature of a set category, and can be, but is not limited to, representing the boundary, area, and distance of the target area to the exit of the target scene. The division information output by the model matches the target geometric feature. For example, the boundary and area of ​​the target area can be used as a reference for how many functional areas the target area is divided into, and the distance of the target area to the exit of the target scene can be used as a basis for dividing the location of each functional area. The exit of the target scene can include, but is not limited to, doorways, windows, balconies, etc., which connect the target scene to the outside world. Typically, the functional area near the doorway in the target area can be set as the entrance hall, and the area near the balcony can be set as the living room.

[0073] Optionally, in this embodiment, the target attribute features are used to characterize the scene function of the target scene, and can also serve as a reference for dividing the target area into how many functional areas. For example, when the target attribute features characterize the scene function of the target scene as a home scene, the functional areas in the target area are functional areas related to the home scene, such as corridors, living rooms, dining rooms, and entrance halls. When the target attribute features characterize the scene function of the target scene as a teaching scene, the functional areas in the target area are functional areas related to the teaching scene, such as teacher offices, study rooms, teaching areas, and laboratories.

[0074] Optionally, in this embodiment, the target scene can be any scene with boundaries, such as a room, a building, or any area with boundaries, such as a park, garden, or scenic spot. The target area is the functional area to be divided in the corresponding target scene. For example, if the target scene is a scenic spot, the functional area may include a ticket office, a play area, a shopping area, and a rest area.

[0075] In the technical solution provided in step S204 above, in the division method indicated by the division strategy that satisfies the target division condition, the number of target vertices located on the boundary of each functional area is greater than the number of target vertices. This can be understood as the above-mentioned artificial rule. The target division condition can be set by the user according to the requirements. Under normal circumstances, the boundary of each functional area passes through the target vertices as much as possible.

[0076] In an exemplary embodiment, a partitioning strategy corresponding to the partitioning information can be matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy in the following manner, but not limited to: finding P partitioning strategies consistent with the partitioning information from the partitioning strategy set as reference partitioning strategies, where P is a positive integer greater than or equal to 1; when P equals 1, determining the reference partitioning strategy as the target partitioning strategy; when P is greater than 1, obtaining the partitioning parameters of each reference partitioning strategy, wherein the partitioning parameters include area ratio, intersection ratio, and number of overlapping points, in each functional area partitioning method, the area ratio is the ratio between the sum of the areas of each functional area included in the target region and the area of ​​the target region, the intersection ratio is used to indicate the ratio between the union and intersection of the areas covered by each functional area, and the number of overlapping points is used to indicate the number of vertices among the N target vertices that overlap with the vertices of each functional area; and determining the reference partitioning strategy with the largest partitioning parameter as the target partitioning strategy.

[0077] Optionally, in this embodiment, when P is greater than 1, the partitioning parameters of each reference partitioning strategy are obtained, and the reference partitioning strategy with the largest partitioning parameters is determined as the target partitioning strategy. This can be understood as the reference partitioning strategy with the largest partitioning parameters being a better reference partitioning strategy. Figure 4 This is a schematic diagram of a reference partitioning strategy according to an embodiment of this application, such as... Figure 4 As shown, the set of partitioning strategies corresponding to the target region includes multiple partitioning strategies. Two partitioning strategies that are consistent with the partitioning information are selected from the set of partitioning strategies as reference partitioning strategies. Then, the reference partitioning strategy with the largest partitioning parameter is determined as the target partitioning strategy.

[0078] In an exemplary embodiment, before matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, the method may include, but is not limited to, the following: obtaining target vertex information and associated vertex information corresponding to the target region, wherein the target vertex information is used to indicate the position of each vertex among the N target vertices of the target region, where N is a positive integer greater than or equal to 4, and the associated vertex information is used to indicate the position of each vertex among the M associated vertices in the target scene associated with the target region, where M is a positive integer greater than or equal to 1; generating Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, thereby obtaining the partitioning strategy set.

[0079] Optionally, in this embodiment, the partitioning strategy that satisfies the target partitioning conditions is generated based on the target vertex information and associated vertex information.

[0080] In an exemplary embodiment, Q partitioning strategies satisfying the target partitioning conditions can be generated based on the target vertex information and associated vertex information corresponding to the target region, to obtain the partitioning strategy set: N target vertices and M associated vertices are projected onto the X-axis and Y-axis of the target Cartesian coordinate system, respectively, to obtain (N+M) first projection points on the X-axis and (N+M) second projection points on the Y-axis, wherein the target Cartesian coordinate system is a Cartesian coordinate system established on the plane where the target region is located; based on the (N+M) first projection points and (N+M) associated vertices... A target partitioning vertex set is constructed using the second projection points, wherein any target partitioning vertex in the target partitioning vertex set falls within the target region, and the x-coordinate of the target partitioning vertex is the same as the x-coordinate of at least one first projection point, and the y-coordinate of the target partitioning vertex is the same as the y-coordinate of at least one second projection point; Q partitioning strategies that satisfy the target partitioning conditions are generated based on the target partitioning vertex set, resulting in the partitioning strategy set, wherein in each partitioning strategy, the two diagonal vertices of the functional areas partitioned in the target region fall within the target partitioning vertex set.

[0081] Optionally, in this embodiment, in the partitioning method indicated by each partitioning strategy, the two diagonal vertices of the functional area partitioned in the target region fall into the target partitioning vertex set. Since the x-coordinate of the target partitioning vertex is the same as the x-coordinate of at least one first projection point, and the y-coordinate of the target partitioning vertex is the same as the y-coordinate of at least one second projection point, the boundary of the functional area partitioned in the target region passes through the target vertex or associated vertex, thus satisfying the target partitioning condition.

[0082] In one exemplary embodiment, the target partitioning vertex set can be constructed based on (N+M) first projection points and (N+M) second projection points in the following manner: A first partitioning line perpendicular to the X-axis is generated through each first projection point, resulting in (N+M) first partitioning lines; a second partitioning line perpendicular to the Y-axis is generated through each second projection point, resulting in (N+M) second partitioning lines; the intersection point of any first partitioning line and any second partitioning line is determined as (N+M). 2 There are (N+M) initial partitioned vertices; 2 The target partitioning vertices that fall within the target region are selected from the initial partitioning vertices to obtain the target partitioning vertex set.

[0083] Optionally, in this embodiment, the first dividing line and the second dividing line can be understood as dividing grid lines in the target area, and each grid line passes through the target vertex or associated vertex.

[0084] In one exemplary embodiment, the partitioning strategy set can be obtained by generating Q partitioning strategies that satisfy the target partitioning conditions based on the target partitioning vertex set in the following manner, but not limited to: determining the number of target functional areas in the target region based on the area of ​​the target region; constructing a rectangular region as a functional area in the target region by taking any two target partitioning vertices in the target partitioning vertex set as diagonal vertices, thereby obtaining Q partitioning strategies as the partitioning strategy set, wherein each partitioning strategy includes the number of target functional areas.

[0085] Optionally, in this embodiment, each of the division strategies includes a number of target functional areas. The number of target functional areas can be determined based on the characteristics of the target area. For example, taking the target geometric features of the target area as an example, when the area of ​​the target area in the target geometric features is large, the target area can be divided into four functional areas: corridor, living room, dining room and entrance hall. However, when the area of ​​the target area in the target geometric features is small, the target area can be divided into two functional areas: corridor and entrance hall.

[0086] In an exemplary embodiment, before obtaining the target vertex information and associated vertex information corresponding to the target region, the method may include, but is not limited to, the following: obtaining vertices of each region other than the target region in the multiple regions included in the target scene as candidate vertices to obtain a candidate vertex set; and determining the candidate vertices in the candidate vertex set whose distance from the boundary of the target region is less than the target distance as the associated vertices.

[0087] Optionally, in this embodiment, Figure 5 This is a schematic diagram of a target vertex and associated vertices according to an embodiment of this application, such as... Figure 5 As shown, the target area in the target scene includes target vertices (A, B, C, D, E, F, G, H, I, J). In addition to the target area, the target scene also includes a bathroom, bedroom, balcony, and kitchen. Vertices in each of the above areas are selected as candidate vertices, such as L and K. Among the candidate vertices, the candidate vertex whose distance from the boundary of the target area is less than the target distance (e.g., K) is determined as the associated vertex. The distance between L and the boundary of the target area is greater than or equal to the target distance, so it is not selected as a candidate vertex.

[0088] In an exemplary embodiment, before projecting the N target vertices and the M associated vertices onto the X-axis and Y-axis of the target Cartesian coordinate system, the method may include, but is not limited to, obtaining the minimum bounding rectangle of the target region; generating the target Cartesian coordinate system based on the minimum bounding rectangle, wherein the X-axis and Y-axis of the target Cartesian coordinate system are perpendicular or parallel to the sides of the minimum bounding rectangle.

[0089] Optionally, in this embodiment, the relationship between the coordinate axes of the target Cartesian coordinate system and the boundary of the target region is as perpendicular or parallel as possible.

[0090] In the technical solution provided in step S206 above, after determining the target partitioning strategy, the target area can be partitioned using the target partitioning strategy. At this time, the partitioned functional areas refer to the location of the functional areas provided by the model, and the boundaries are determined by manual rules, which avoids the problem of unreasonable functional area boundaries directly given by the model in related technologies, which affects the partitioning effect of the functional areas.

[0091] To better understand the process of dividing the functional areas of the above-mentioned scene area, the following description, in conjunction with optional embodiments, further illustrates the process of dividing the functional areas of the above-mentioned scene area, but is not intended to limit the technical solutions of the embodiments of this application.

[0092] This embodiment provides a method for dividing a scene area into functional zones. Figure 6 This is a schematic diagram of a method for dividing a scene area into functional zones according to an embodiment of this application, as shown below. Figure 6 As shown, it includes the following steps:

[0093] 1) Constructing a set of candidate functional zones corresponding to the target area based on manually generated rules;

[0094] 2) The maximum inscribed rectangle is greater than the minimum area of ​​the restaurant (2.25m²). 2 Alternatively, if the length / width is greater than 1.5m, the target area includes the living room and / or the dining room; otherwise, the target area includes the entryway and / or the hallway (passageway).

[0095] 3) If the target area includes a living room and / or a dining room, determine the number and location of functional areas. If there is only one living room (dining room), classify the living room and dining room and use a recursive backtracking method to obtain the optimal combination of functional areas. If there are both a living room and a dining room, use a recursive backtracking method to obtain the optimal combination of functional areas.

[0096] 4) When the target area includes an entryway and / or a corridor (passageway), classify the entryway and passageway. Determine the candidate entryway set based on area ratio, aspect ratio, boundary percentage, number of vertices, and entrance door percentage. The elements in the candidate entryway set can meet the following requirements: 1. Area ratio (0.02~0.5); 2. Aspect ratio <7; 3. Boundary percentage >0.499; 4. Number of vertices >2.5; 5. Entrance door percentage >0.8. The area ratio can be the ratio of the area of ​​the candidate entryway functional area to the area of ​​the target area; the aspect ratio can be the aspect ratio of the candidate entryway functional area; the boundary percentage can be the ratio of the length of the boundary of the candidate entryway functional area to the length of the boundary of the target area; the number of vertices can be the number of target vertices that the boundary of the candidate entryway functional area passes through; and the entrance door percentage can be the ratio of the boundary of the candidate entryway functional area and the entrance door to the width of the entrance door. If the candidate set for the entryway is empty, the entire area is designated as the passageway area; otherwise, the smallest area is selected from the candidate set as the entryway area, and the remaining area is designated as the passageway area.

[0097] It is worth noting that the above steps can be handled using the following framework:

[0098] Figure 7 This is a schematic diagram of a functional area division framework for a scene region according to an embodiment of this application, as shown below. Figure 7 As shown, there are two processing branches in the target region of the framework:

[0099] The first approach uses a CNN model (center prediction model) to predict the center positions of the living room, dining room, entrance hall, and corridor in the target area. The layout termination prediction model determines whether to stop prediction. For example, in the case of a small area, the layout termination prediction model stops prediction when the entrance hall is identified in the target area. When the layout termination prediction model determines to continue prediction, it will predict the center of the next functional area until the final functional area position of the target area is output (corresponding to the above division information).

[0100] The second branch calculates candidate boxes based on the target area, resulting in four candidate functional area division methods (corresponding to the above set of division strategies);

[0101] The two branches converge, and a partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. If there are multiple partitioning strategies that match the partitioning information in the partitioning strategy set, recursive backtracking is adopted to select the partitioning strategy with the largest area, the largest intersection ratio, and the largest number of container vertices (corresponding to the above partitioning parameters) as the target partitioning strategy.

[0102] The functional area division method for a scene area proposed in this application, on the one hand, can make full use of the apartment layout data by automatically extracting apartment features using deep learning. On the other hand, by utilizing the designer's prior knowledge to calculate the alternative set of functional areas, the method can incorporate the designer's requirements for the boundaries of functional areas. Furthermore, this invention can consider a default handling method for living and dining room areas that are too small. Therefore, this invention is robust and can handle virtually all situations.

[0103] This invention utilizes deep learning models to extract apartment layout features, thereby predicting the number and central location of functional areas. Therefore, it can fully leverage apartment layout data to obtain more reasonable central locations for functional areas. Simultaneously, it employs prior knowledge to calculate a candidate set of functional areas, resulting in more reasonable boundaries for these areas. Finally, it considers different processing procedures under various circumstances. Therefore, the proposed living and dining room zoning framework is more robust, yields a more rational layout, and can be applied in practice at a higher speed.

[0104] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0105] Figure 8 This is a structural block diagram of a scene area functional area division device according to an embodiment of this application; as shown below. Figure 8 As shown, it includes:

[0106] The input module 802 is used to input the target region into the target partitioning model and obtain the partitioning information output by the target partitioning model, wherein the target region is the region of the functional area to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional area included in the target region;

[0107] The matching module 804 is used to match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices. The target vertices are the vertices of the target region.

[0108] The segmentation module 806 is used to segment the target region using the target segmentation strategy.

[0109] Through the above embodiments, the target partitioning model outputs partitioning information corresponding to the target region. The target region is the area of ​​the functional area to be partitioned in the target scene. The partitioning information is used to indicate the location of the functional areas included in the target region. Then, the partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. Therefore, the obtained target partitioning strategy satisfies the target partitioning condition. Since the number of target vertices on the boundaries of each functional area is greater than the number of target vertices in the partitioning method indicated by the partitioning strategy that satisfies the target partitioning condition, the target vertices are the vertices of the target region. Finally, the target partitioning strategy is used to partition the target region. By adopting the above technical solution, the problem of poor functional area partitioning effect in the scene area in the related technology is solved, and the technical effect of improving the functional area partitioning effect of the scene area is achieved.

[0110] In one exemplary embodiment, the apparatus further includes:

[0111] The first acquisition module is used to acquire target vertex information and associated vertex information corresponding to the target region before matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The target vertex information is used to indicate the position of each vertex among the N target vertices of the target region, where N is a positive integer greater than or equal to 4. The associated vertex information is used to indicate the position of each vertex among the M associated vertices in the target scene associated with the target region, where M is a positive integer greater than or equal to 1.

[0112] The first generation module is used to generate Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, thereby obtaining the partitioning strategy set.

[0113] In one exemplary embodiment, the first generation module includes:

[0114] The projection unit is used to project N target vertices and M associated vertices onto the X-axis and Y-axis of the target Cartesian coordinate system, respectively, to obtain (N+M) first projection points on the X-axis and (N+M) second projection points on the Y-axis, wherein the target Cartesian coordinate system is a Cartesian coordinate system established on the plane where the target region is located;

[0115] A construction unit is configured to construct a target partitioning vertex set based on (N+M) first projection points and (N+M) second projection points, wherein any target partitioning vertex in the target partitioning vertex set falls into the target region, and the x-coordinate of the target partitioning vertex is the same as the x-coordinate of at least one first projection point, and the y-coordinate of the target partitioning vertex is the same as the y-coordinate of at least one second projection point.

[0116] The generation unit is used to generate Q partitioning strategies that satisfy the target partitioning conditions based on the target partitioning vertex set, thereby obtaining the partitioning strategy set, wherein in the partitioning method indicated by each partitioning strategy, the two diagonal vertices of the functional area partitioned in the target region fall into the target partitioning vertex set.

[0117] In one exemplary embodiment, the building unit is further configured to:

[0118] A first dividing line perpendicular to the X-axis is generated through each of the first projection points, resulting in (N+M) first dividing lines. A second dividing line perpendicular to the Y-axis is generated through each of the second projection points, resulting in (N+M) second dividing lines.

[0119] The intersection point of any first dividing line and any second dividing line is defined as (N+M). 2 One initial partitioned vertex;

[0120] From (N+M) 2 The target partitioning vertices that fall within the target region are selected from the initial partitioning vertices to obtain the target partitioning vertex set.

[0121] In one exemplary embodiment, the generating unit is further configured to:

[0122] The number of target functional areas in the target area is determined based on the area of ​​the target area.

[0123] Construct a rectangular region as a functional area within the target region by taking any two target partitioning vertices from the target partitioning vertex set as diagonal vertices, thus obtaining Q partitioning strategies as the partitioning strategy set, wherein each partitioning strategy includes the number of functional areas specified in the target functional area set.

[0124] In one exemplary embodiment, the apparatus further includes:

[0125] The second acquisition module is used to acquire vertices of each region other than the target region in the multiple regions included in the target scene as candidate vertices before acquiring the target vertex information and associated vertex information corresponding to the target region, thereby obtaining a candidate vertex set;

[0126] The determination module is used to determine the candidate vertices in the candidate vertex set whose distance from the boundary of the target region is less than the target distance as the associated vertices.

[0127] In one exemplary embodiment, the apparatus further includes:

[0128] The third acquisition module is used to acquire the minimum bounding rectangle of the target region before projecting the N target vertices and the M associated vertices onto the X-axis and Y-axis of the target Cartesian coordinate system, respectively.

[0129] The second generation module is used to generate a target Cartesian coordinate system based on the minimum bounding rectangle, wherein the X-axis and Y-axis of the target Cartesian coordinate system are perpendicular or parallel to the sides of the minimum bounding rectangle.

[0130] In one exemplary embodiment, the input module includes:

[0131] The first input unit is used to input the target region into the target feature extraction layer to obtain the target geometric features and target attribute features output by the target feature extraction layer. The target geometric features are used to characterize the boundary, area and distance of the target region from the target scene exit, and the target attribute features are used to characterize the scene function of the target scene.

[0132] The second input unit is used to input the target geometric features and the target attribute features into the target region segmentation layer to obtain the segmentation information output by the target region segmentation layer. The target segmentation model includes the target feature extraction layer and the target region segmentation layer connected in sequence. The target segmentation model is obtained by training the model parameters of the initial feature extraction layer and the initial region segmentation layer connected in sequence using multiple sets of region samples labeled with segmentation information labels.

[0133] In one exemplary embodiment, the matching module includes:

[0134] The search unit is used to search for P partitioning strategies that are consistent with the partitioning information from the set of partitioning strategies as reference partitioning strategies, where P is a positive integer greater than or equal to 1.

[0135] The first determining unit is used to determine the reference partitioning strategy as the target partitioning strategy when P equals 1.

[0136] The second determining unit is used to obtain the partitioning parameters of each of the reference partitioning strategies when P is greater than 1. The partitioning parameters include the area ratio, the intersection ratio, and the number of overlapping points. In each functional area partitioning method, the area ratio is the ratio between the sum of the areas of the functional areas included in the target area and the area of ​​the target area. The intersection ratio is used to indicate the ratio between the union and intersection of the areas covered by each functional area. The number of overlapping points is used to indicate the number of vertices among the N target vertices that overlap with the vertices of each functional area. The reference partitioning strategy with the largest partitioning parameter is determined as the target partitioning strategy.

[0137] Embodiments of this application also provide a storage medium including a stored program, wherein the program executes any of the methods described above when it is run.

[0138] Optionally, in this embodiment, the storage medium may be configured to store program code for performing the following steps:

[0139] S1, Input the target region into the target partitioning model to obtain the partitioning information output by the target partitioning model, wherein the target region is the region of functional areas to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional areas included in the target region;

[0140] S2, Match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, wherein the partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1, each partitioning strategy is used to indicate a way of dividing the target region into functional areas, and in the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices, and the target vertices are the vertices of the target region;

[0141] S3, the target region is divided using the target segmentation strategy.

[0142] Embodiments of this application also provide an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0143] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0144] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0145] S1, Input the target region into the target partitioning model to obtain the partitioning information output by the target partitioning model, wherein the target region is the region of functional areas to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional areas included in the target region;

[0146] S2, Match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, wherein the partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1, each partitioning strategy is used to indicate a way of dividing the target region into functional areas, and in the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than the number of target vertices, and the target vertices are the vertices of the target region;

[0147] S3, the target region is divided using the target segmentation strategy.

[0148] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0149] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0150] Obviously, those skilled in the art should understand that the modules or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0151] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for dividing a scene area into functional zones, characterized in that, include: The target region is input into the target partitioning model to obtain the partitioning information output by the target partitioning model. The target region is the area of ​​functional areas to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional areas included in the target region. The partitioning strategy corresponding to the partitioning information is matched from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than a threshold for the number of target vertices. The target vertex is the vertex of the target region. The target region is divided using the target segmentation strategy described above; Before matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, the method further includes: obtaining target vertex information and associated vertex information corresponding to the target region, wherein the target vertex information is used to indicate the position of each vertex among the N target vertices of the target region, where N is a positive integer greater than or equal to 4, and the associated vertex information is used to indicate the position of each vertex among the M associated vertices in the target scene associated with the target region, where M is a positive integer greater than or equal to 1; generating Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, thereby obtaining the partitioning strategy set; Wherein, the partitioning strategy that satisfies the target partitioning condition means that the boundary of the functional area obtained by partitioning the target region according to the partitioning strategy passes through the target vertex or the associated vertex; The step of matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy includes: finding P partitioning strategies consistent with the partitioning information from the partitioning strategy set as reference partitioning strategies, where P is a positive integer greater than or equal to 1; when P equals 1, determining the reference partitioning strategy as the target partitioning strategy; when P is greater than 1, obtaining the partitioning parameters of each reference partitioning strategy, where the partitioning parameters include area ratio, intersection ratio, and number of overlapping points; and determining the reference partitioning strategy with the largest partitioning parameter as the target partitioning strategy.

2. The method according to claim 1, characterized in that, The step of generating Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, and obtaining the partitioning strategy set, includes: The N target vertices and M associated vertices are projected onto the X-axis and Y-axis of the target Cartesian coordinate system, respectively, to obtain (N+M) first projection points on the X-axis and (N+M) second projection points on the Y-axis, wherein the target Cartesian coordinate system is a Cartesian coordinate system established on the plane where the target region is located; A target partitioning vertex set is constructed based on (N+M) first projection points and (N+M) second projection points, wherein any target partitioning vertex in the target partitioning vertex set falls into the target region, and the x-coordinate of the target partitioning vertex is the same as the x-coordinate of at least one first projection point, and the y-coordinate of the target partitioning vertex is the same as the y-coordinate of at least one second projection point. Based on the target partitioning vertex set, Q partitioning strategies that satisfy the target partitioning conditions are generated to obtain the partitioning strategy set, wherein in the partitioning method indicated by each partitioning strategy, the two diagonal vertices of the functional area divided in the target region fall into the target partitioning vertex set.

3. The method according to claim 2, characterized in that, The step of constructing a target partitioning vertex set based on (N+M) first projection points and (N+M) second projection points includes: A first dividing line perpendicular to the X-axis is generated through each of the first projection points, resulting in (N+M) first dividing lines. A second dividing line perpendicular to the Y-axis is generated through each of the second projection points, resulting in (N+M) second dividing lines. The intersection point of any first dividing line and any second dividing line is determined as... One initial partitioned vertex; from The target partitioning vertices that fall within the target region are selected from the initial partitioning vertices to obtain the target partitioning vertex set.

4. The method according to claim 2, characterized in that, The step of generating Q partitioning strategies that satisfy the target partitioning conditions based on the target partitioning vertex set yields the partitioning strategy set, including: The number of target functional areas in the target area is determined based on the area of ​​the target area. Construct a rectangular region as a functional area within the target region by taking any two target partitioning vertices from the target partitioning vertex set as diagonal vertices, thus obtaining Q partitioning strategies as the partitioning strategy set, wherein each partitioning strategy includes the number of functional areas specified in the target functional area set.

5. The method according to claim 2, characterized in that, Before obtaining the target vertex information and associated vertex information corresponding to the target region, the method further includes: The vertices of each region other than the target region in the multiple regions included in the target scene are obtained as candidate vertices, thus obtaining a candidate vertex set; The candidate vertices in the candidate vertex set whose distance from the boundary of the target region is less than the target distance are identified as the associated vertices.

6. The method according to claim 2, characterized in that, Before projecting the N target vertices and the M associated vertices onto the X and Y axes of the target Cartesian coordinate system, the method further includes: Obtain the minimum bounding rectangle of the target region; A target Cartesian coordinate system is generated based on the minimum bounding rectangle, wherein the X-axis and Y-axis of the target Cartesian coordinate system are perpendicular or parallel to the sides of the minimum bounding rectangle.

7. The method according to claim 1, characterized in that, The step of inputting the target region into the target partitioning model and obtaining the partitioning information output by the target partitioning model includes: The target region is input into the target feature extraction layer to obtain the target geometric features and target attribute features output by the target feature extraction layer. The target geometric features are used to characterize the boundary, area and distance of the target region from the target scene exit, and the target attribute features are used to characterize the scene function of the target scene. The target geometric features and the target attribute features are input into the target region segmentation layer to obtain the segmentation information output by the target region segmentation layer. The target segmentation model includes the target feature extraction layer and the target region segmentation layer connected in sequence. The target segmentation model is obtained by training the model parameters of the initial feature extraction layer and the initial region segmentation layer using multiple sets of region samples labeled with segmentation information.

8. The method according to claim 1, characterized in that, In each functional area division method, the area ratio is the ratio between the sum of the areas of the functional areas included in the target area and the area of ​​the target area; the union-intersection ratio is used to indicate the ratio between the union and intersection of the areas covered by each functional area; and the number of overlapping points is used to indicate the number of vertices among the N target vertices that overlap with the vertices of each functional area.

9. A functional area division device for a scene region, characterized in that, include: The input module is used to input the target region into the target partitioning model and obtain the partitioning information output by the target partitioning model. The target region is the region of the functional area to be partitioned in the target scene, and the partitioning information is used to indicate the location of the functional area included in the target region. The matching module is used to match the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy. The partitioning strategy set includes Q partitioning strategies that satisfy the target partitioning conditions, where Q is a positive integer greater than or equal to 1. Each partitioning strategy is used to indicate a way of dividing the target region into functional areas. In the partitioning method indicated by the partitioning strategy that satisfies the target partitioning conditions, the number of target vertices located on the boundaries of each functional area is greater than a threshold for the number of target vertices. The target vertices are the vertices of the target region. A segmentation module is used to segment the target region using the target segmentation strategy; The device further includes: a first acquisition module, configured to acquire target vertex information and associated vertex information corresponding to the target region before matching the partitioning strategy corresponding to the partitioning information from the partitioning strategy set corresponding to the target region as the target partitioning strategy, wherein the target vertex information is used to indicate the position of each vertex among the N target vertices of the target region, where N is a positive integer greater than or equal to 4, and the associated vertex information is used to indicate the position of each vertex among the M associated vertices in the target scene associated with the target region, where M is a positive integer greater than or equal to 1; and a first generation module, configured to generate Q partitioning strategies that satisfy the target partitioning conditions based on the target vertex information and associated vertex information corresponding to the target region, thereby obtaining the partitioning strategy set. Wherein, the partitioning strategy that satisfies the target partitioning condition means that the boundary of the functional area obtained by partitioning the target region according to the partitioning strategy passes through the target vertex or the associated vertex; The matching module includes: a search unit, configured to search for P partitioning strategies consistent with the partitioning information from the partitioning strategy set as reference partitioning strategies, wherein P is a positive integer greater than or equal to 1; a first determination unit, configured to determine the reference partitioning strategy as the target partitioning strategy when P equals 1; and a second determination unit, configured to obtain the partitioning parameters of each reference partitioning strategy when P is greater than 1, wherein the partitioning parameters include area ratio, intersection ratio, and number of overlapping points; and to determine the reference partitioning strategy with the largest partitioning parameters as the target partitioning strategy.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method of any one of claims 1 to 8.

11. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method of any one of claims 1 to 8 through the computer program.

Images