Terminal for providing information about geographic object and method using same

Abstract

A method for providing information about a geographic object, according to an embodiment, is performed by the steps of: classifying each of a plurality of lines connected to define an area of the ground, by using type and shape information of geographic objects adjoining each other across a corresponding line; grouping the plurality of lines by using a result of the classification and the shape information of the plurality of lines; deriving, for each group that is a result of the grouping, a candidate footprint area for a building to have a maximum area; and determining a footprint area for the building from among the derived one or more candidate footprint areas by considering a predetermined constraint. Here, in the step of deriving the candidate footprint area, in a group including one line from among groups that are a result of the grouping, the one line becomes a reference line and is used to derive a candidate footprint area, and in a group including two or more lines, one line obtained by merging the two or more lines becomes a reference line and is used to derive a candidate footprint area.

Description

Terminal for providing information on geographical objects and method using the same

[0001] The present invention relates to a terminal for providing information on a geographic object and a method using the same.

[0002] For reference, this application claims priority to Korean Patent Application No. 10-2024-0063431, filed May 14, 2024, Korean Patent Application No. 10-2024-0087864, filed July 3, 2024, Korean Patent Application No. 10-2024-0087865, filed July 3, 2024, and Korean Patent Application No. 10-2024-0087866, filed July 3, 2024. The entire contents of these priority applications are incorporated herein by reference.

[0003] Unless otherwise indicated herein, the materials described in this section are not prior art to the claims of this application, and their inclusion in this section is not intended to be admitted as prior art.

[0004] A Geographic Information System (GIS) is an information system that converts geographic information essential to human life into computer data for efficient utilization. A GIS integrates and manages spatial, attribute, and temporal data about objects with geographic locations, enabling the provision of diverse information, such as maps, charts, and graphics. In a broad sense, GIS refers to a system for the entire process of manipulating geographic information—from observation and collection to preservation, analysis, and output—necessary for supporting human decision-making.

[0005] These GIS systems can provide useful information during the design process for a building on a specific site. For example, they can provide the addresses and areas of potential sites for a building.

[0006] However, the type and quality of information available in existing GIS is limited. For example, 3D modeling of the target site, reflecting its elevation, or geographic objects such as buildings, parcel numbers, green spaces, or roads located on the target site are often not available. Furthermore, legal requirements for the target site or the aforementioned geographic objects, or surrounding infrastructure information, are often insufficient. Furthermore, because accurate information on the target site is difficult to obtain, accurate information on the area within the site where buildings can be located, or the footprint area, is also difficult to obtain.

[0007] The problem to be solved, according to one embodiment, is to provide a terminal for providing information on geographic objects and a method using the same. The terminal for providing information on geographic objects may be a server or a user terminal. However, the following description assumes that the terminal for providing information on geographic objects is a server. However, it should be understood that the technical features described for such a server can also be implemented in a user terminal.

[0008] Additionally, implementing a technology for determining the footprint area of ​​a building within a land area may be included in the aforementioned tasks.

[0009] Additionally, the aforementioned tasks may include implementing technologies capable of providing geographic information about the land. For example, this task may include providing a 3D model of the target land, reflecting its elevation above sea level.

[0010] Additionally, the aforementioned tasks may include ensuring that sufficient information on the target land's regulations and surrounding infrastructure is provided.

[0011] However, the problems to be solved according to one embodiment are not limited to those mentioned above.

[0012] A method for providing information on a geographic object according to a first embodiment, the method comprises the steps of: classifying each of a plurality of lines connected to define an area of ​​a land using information on the type and shape of a geographic object interposed between the lines; grouping the plurality of lines using the classification results and the shape information of the plurality of lines; deriving, for each group resulting from the grouping, a candidate footprint area for a building to have the maximum area; and determining a footprint area for the building by considering a predetermined constraint among the derived one or more candidate footprint areas. In this case, in the step of deriving the candidate footprint area, in a group resulting from the grouping, one line is used as a reference line to derive the candidate footprint area, and in a group resulting from the grouping, one line is merged from the two or more lines and serves as a reference line to derive the candidate footprint area.

[0013] Additionally, the original lines defining the area of ​​the land may include curves, each of the plurality of lines may be straight lines, and some of the plurality of straight lines may be curves converted based on curvature.

[0014] Additionally, in the above classification, whether the geographic object is a road and, if so, the width of the road may be considered.

[0015] Additionally, in the grouping, the minimum distance between any one point of each of the plurality of lines and any one point of each of the other lines and the intersection angle when the virtual extension lines of each of the plurality of lines intersect can be used.

[0016] In addition, each group resulting from the grouping is first sorted using the classified results, and then secondarily sorted using the length of the baseline of each group, and the candidate footprint area can be sequentially derived for each group according to the sorting results obtained from the first sorting to the second sorting.

[0017] In addition, the sides defining the candidate footprint area include a first side and a second side parallel to the reference line, wherein the minimum value for the first side is a preset value, and the minimum and maximum values ​​for the second side are determined for each group according to the length of the reference line in each group, and in the deriving step, the length of the first side is changed in consideration of the minimum value for the first side, and the length of the second side is changed in consideration of the minimum and maximum values ​​for the second side, and the area defined according to the lengths of the first side and the second side can be derived according to whether the area invades the reference line for each of the plurality of groups and the area of ​​the defined area.

[0018] In addition, the sides defining the candidate footprint area include a first side and a second side parallel to the reference line, and in the derivation step, the results of the grouping are provided to a pre-learned candidate footprint area derivation model, so that the length and position of each of the first side and the second side can be obtained.

[0019] Additionally, the above constraints may include at least one of the building coverage ratio and the floor area ratio.

[0020] Additionally, a plurality of lines connected to define the area of ​​the land may be derived using three-dimensional modeling information generated for the land and information acquired for the geographic object.

[0021] In addition, the three-dimensional modeling information generated for the land may be generated by performing the steps of: obtaining three-dimensional modeling information for the target area; converting the three-dimensional mesh into a two-dimensional mesh, wherein the target area is divided into a plurality of tiles in which the elevation above sea level is reflected in the form of a three-dimensional mesh in the three-dimensional modeling information; obtaining two-dimensional shape information for the land; obtaining two or more intersection points where the shape of the land intersects the two-dimensional mesh using the two-dimensional shape information acquired for the land and the two-dimensional mesh; and connecting the two or more intersection points.

[0022] Additionally, in the conversion to the above two-dimensional mesh, the elevation above sea level in the above three-dimensional mesh can be converted to 0.

[0023] A computer-readable recording medium according to a second embodiment includes a computer program, the computer program comprising: a step of classifying each of a plurality of lines connected to define an area of ​​a land using information on the type and shape of a geographic object interposed between the lines; a step of grouping the plurality of lines using the classified results and the shape information of the plurality of lines; a step of deriving, for each group resulting from the grouping, a candidate footprint area for the building to have the maximum area; and a step of determining a footprint area for the building by considering a predetermined constraint among the derived one or more candidate footprint areas. In this case, in the step of deriving the candidate footprint area, in a group resulting from the grouping, one line is used as a reference line to derive the candidate footprint area, and in a group resulting from the grouping, one line is merged from the two or more lines and serves as a reference line to derive the candidate footprint area.

[0024] In one embodiment, the footprint area of ​​a building on a site can be determined. Therefore, users involved in building design can decide where to place the building without having to visit the site. Furthermore, this footprint area can be visualized and provided for building design.

[0025] Furthermore, according to another embodiment, 3D modeling of geographic objects can be provided. Specifically, the 3D modeling can be provided by reflecting the elevation above sea level of the ground surface where these geographic objects are located. Therefore, users involved in building design can easily and precisely obtain accurate height information and 3D modeling information for these geographic objects without having to go to the site.

[0026] Furthermore, according to another embodiment, a 3D model of a target area including a site may be provided differently depending on the zoom level. Specifically, at a low zoom level, the height is 0, but as the zoom level increases, a 3D model reflecting the height in a more realistic manner may be provided. Accordingly, users involved in building design can easily and precisely obtain height information for the site in question as well as the target site including such site without having to go to the site.

[0027] Furthermore, as previously mentioned, information on buildings, as well as their respective address numbers, green spaces, and roads, can be obtained and provided from public data. In some cases, information on surrounding infrastructure can also be obtained and provided. Furthermore, as will be described later, this information can be visualized and provided for reviewing the site feasibility of a building, allowing users to closely review the site feasibility of a building based on the geographic information provided in one embodiment.

[0028] FIG. 1 conceptually illustrates a server providing information on a geographic object according to one embodiment connected to a network.

[0029] FIG. 2 is a block diagram of a server providing information on a geographic object in one embodiment.

[0030] Figure 3 conceptually illustrates the architecture for deep learning.

[0031] Figures 4 and 5 illustrate the process performed in pre-learning during transfer learning.

[0032] FIG. 6 illustrates an exemplary flowchart of a technology for providing land information for a building according to one embodiment.

[0033] Figure 7 illustrates, as an example, a flowchart of the specific process of designing the data schema and creating a database for the objects illustrated in Figure 6.

[0034] FIG. 8 illustrates an exemplary flowchart of a method for providing information on a geographic object according to one embodiment.

[0035] Figure 9 conceptually illustrates an example of a tiling scheme.

[0036] FIG. 10 illustrates an example of a 3D mesh generated at a given zoom level according to one embodiment.

[0037] FIG. 11 illustrates an example of a 3D mesh generated at a different zoom level than that in FIG. 10, according to one embodiment.

[0038] FIG. 12 illustrates an exemplary flowchart of a method for providing information on a geographic object according to one embodiment.

[0039] Figure 13 conceptually illustrates an example of a three-dimensionally modeled building according to one embodiment.

[0040] Figure 14 conceptually illustrates an example of a three-dimensional modeled land lot number according to one embodiment.

[0041] FIG. 15 illustrates an exemplary flowchart for a process of defining an area in which a building can be located within a land according to one embodiment.

[0042] FIG. 16 illustrates an example concept for a tile map for objects according to one embodiment.

[0043] Figure 17 illustrates an example of a concept in which a tile map for an object according to one embodiment is composed of a mesh.

[0044] FIG. 18 illustrates an example of a concept where a mesh of a tile map for an object intersects an area representing a specific object according to one embodiment.

[0045] Figure 19 illustrates an example of a concept of a specific land number and a geographic object adjacent to such land number.

[0046] Figure 20 illustrates an example of a concept for obtaining the shape of a road among geographic objects and the width of such a road.

[0047] Figure 21 conceptually illustrates an adjacent land boundary line according to one embodiment.

[0048] FIG. 22 illustrates an exemplary flowchart of a process for determining a footprint area for a building according to one embodiment.

[0049] FIG. 23 illustrates an example of classification of each of a plurality of lines forming the adjacent land boundary line obtained according to one embodiment.

[0050] Figure 24 illustrates an example of a result in which multiple lines shown in Figure 23 are grouped.

[0051] Figure 25 illustrates an example of a concept for obtaining a baseline.

[0052] Figure 26 illustrates an example of the result of sorting the groups as a result of the grouping illustrated in Figure 24.

[0053] FIG. 27 illustrates an example of a footprint area derived for a building according to one embodiment.

[0054] Hereinafter, the embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Regardless of the drawing numbers, identical or similar components will be given the same reference numbers, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably only for the convenience of writing the specification, and do not in themselves have distinct meanings or roles. In addition, when describing the embodiments disclosed in this specification, if it is determined that a specific description of a related known technology may obscure the gist of the embodiments disclosed in this specification, a detailed description thereof will be omitted. In addition, the attached drawings are only intended to facilitate easy understanding of the embodiments disclosed in this specification, and the technical ideas disclosed in this specification are not limited by the attached drawings, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and technical scope of the present invention.

[0055] Terms that include ordinal numbers, such as first, second, etc., may be used to describe various components, but the components are not limited by these terms. These terms are used solely to distinguish one component from another.

[0056] When a component is referred to as being "connected" or "connected" to another component, it should be understood that it may be directly connected or connected to that other component, but that there may be other components intervening. Conversely, when a component is referred to as being "directly connected" or "connected" to another component, it should be understood that there are no other components intervening.

[0057] In this application, terms such as “include” or “have” are intended to specify the presence of a feature, number, step, operation, component, part or combination thereof described in the specification, but should be understood not to exclude in advance the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts or combinations thereof.

[0058] In this specification, the term "unit" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Furthermore, a single unit may be realized using two or more pieces of hardware, and two or more units may be realized by a single piece of hardware.

[0059] Some of the operations or functions described herein as being performed by a terminal, apparatus, or device may instead be performed by a server connected to the terminal, apparatus, or device. Similarly, some of the operations or functions described herein as being performed by a server may also be performed by a terminal, apparatus, or device connected to the server.

[0060] Hereinafter, the present invention will be described in detail with reference to the attached drawings.

[0061] FIG. 1 conceptually illustrates a server providing information on geographical objects connected to a network. Referring to FIG. 1, a server (100) providing information on geographical objects according to one embodiment may be connected to at least one of a user terminal (200) and an external server (300) via a network (400). Note that FIG. 1 is merely exemplary, and the scope of the present invention is not limited to what is depicted in FIG. 1.

[0062] Here, the network (400) refers to a wireless or wired network. Among these, in the case of a wireless network, for example, at least one of LTE (long-term evolution), LTE-A (LTE Advance), CDMA (code division multiple access), WCDMA (wideband CDMA), UMTS (universal mobile telecommunications system), WiBro (Wireless Broadband), WiFi (wireless fidelity), Bluetooth, NFC (near field communication), and GNSS (global navigation satellite system) may be included. In addition, in the case of a wired network, for example, at least one of USB (universal serial bus), HDMI (high definition multimedia interface), RS-232 (recommended standard 232), LAN (local area network), WAN (wide area network), the Internet, and a telephone network may be included.

[0063] Next, the user terminal (200) is a terminal of a user who wishes to receive various types of information. Such a user terminal (200) may include a smartphone, tablet PC, desktop PC, or server, as illustrated in FIG. 2. Through such a user terminal (200), a user can provide information on a desired building as well as information on a site considered as a candidate to the geographic object information providing server (100), and in response, receive information on various types of geographic objects desired by the user.

[0064] Next, the external server (300) may be implemented to provide public data on at least one of various types of objects, such as buildings, land lots (PNU), green spaces, and roads, but the data or information that can be provided is not limited thereto. Specifically, data such as a DEM file in which the elevation above sea level for each of two or more points on the land is described, an SHP file indicating the shape of the aforementioned object, as well as the address, area, floor area ratio, building coverage ratio, or altitude for each land or building can be obtained from the external server (300), but the data is not limited thereto.

[0065] Next, let's look at a server (100) that provides information on geographical objects according to one embodiment. This server (100) that provides information on geographical objects may be configured to provide geographical information using data or information previously constructed from an external server (300) in response to a request from a user terminal (200).

[0066] Specifically, information regarding the site on which a building will be constructed may be provided. For example, if information regarding a structure to be constructed is provided, the information provision server (100) for such geographic objects may provide information regarding laws and regulations that must be considered in the construction of the structure, as well as information necessary to determine the location of the site on which the building will be constructed, and information regarding the surrounding infrastructure of the site.

[0067] In addition, the information provided in this manner can be shared with those who have access rights and can access the information providing server (100) for the geographic object.

[0068] Therefore, when utilizing the information provision server (100) for such geographical objects, various people involved in the design or construction of a building can easily obtain the desired information without having to visit the construction site or land in person, and the information thus obtained can also be easily shared.

[0069] Additionally, 3D modeling of various types of geographic objects, such as buildings, land lots, green spaces, or roads, can be provided by a geographic object information providing server (100). Here, such 3D modeling can be performed by reflecting the elevation above sea level of the ground or land surface on which such geographic objects are located. Therefore, more accurate 3D modeling information can be acquired and utilized.

[0070] Below, let us look specifically at the information providing server (100) for a geographical object according to one embodiment.

[0071] FIG. 2 is a block diagram showing a server providing information on a geographical object in one embodiment, but the spirit of the present invention is not limited to what is shown in FIG. 2.

[0072] First, the server (100) providing information on geographical objects can be implemented on a computer, such as a server. Based on this, let us examine each configuration of the server (100) providing information on geographical objects.

[0073] First, the communication unit (110) can be configured regardless of the communication mode, such as wired or wireless, and can be configured with various communication networks, such as a personal area network (PAN) and a wide area network (WAN). In addition, the communication unit (110) can operate based on the well-known World Wide Web (WWW), and can also utilize a wireless transmission technology used for short-distance communication, such as infrared (IrDA: Infrared Data Association) or Bluetooth. For example, the communication unit (110) can be responsible for transmitting and receiving data required to perform a technique according to an embodiment of the present disclosure.

[0074] The memory (120) may refer to any type of storage medium. For example, the memory (120) may include at least one type of storage medium among a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (e.g., SD or XD memory, etc.), a RAM (Random Access Memory), a SRAM (Static Random Access Memory), a ROM (Read-Only Memory), an EEPROM (Electrically Erasable Programmable Read-Only Memory), a PROM (Programmable Read-Only Memory), a magnetic memory, a magnetic disk, and an optical disk. A database may also be implemented in the memory (120).

[0075] The memory (120) can store at least one instruction that can be executed by the processor (130). In addition, the memory (120) can store any type of information generated or determined by the processor (130) and various types of information provided from an external server (200). In addition, the memory (120) can store various types of modules, instruction sets, or models.

[0076] The processor (130) may perform technical features according to embodiments of the present disclosure, which will be described later, by executing at least one instruction stored in the memory (120). In one embodiment, the processor (130) may be configured with at least one core and may include a processor for data analysis and / or processing, such as a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), or a tensor processing unit (TPU) of a computer device.

[0077] This processor (130) can train a neural network or model designed using machine learning or deep learning methods. To this end, the processor (130) can perform calculations for neural network training, such as processing input data for training, extracting features from the input data, calculating errors, and updating the weights of the neural network using backpropagation.

[0078] Additionally, the processor (130) may perform inference for a predetermined purpose using a model implemented in an artificial neural network manner.

[0079] Hereinafter, we will examine artificial neural networks. A model in this specification may refer to any form of computer program that operates based on a network function, an artificial neural network, and / or a neural network. Throughout this specification, the terms model, neural network, network function, and neural network may be used interchangeably. A neural network is composed of one or more nodes interconnected through one or more links, forming input and output node relationships within the neural network. The characteristics of a neural network can be determined based on the number of nodes and links within the neural network, the correlation between the nodes and links, and the weight values ​​assigned to each link. A neural network may be composed of a set of one or more nodes. A subset of the nodes constituting the neural network may constitute a layer.

[0080] Among neural networks, a deep neural network (DNN) can refer to a neural network that includes multiple hidden layers in addition to an input and output layer. As illustrated in Figure 3, a deep neural network can have one or more, and preferably two or more, hidden layers in the middle.

[0081] These deep neural networks may include convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, generative pre-trained transformers (GPTs), autoencoders, generative adversarial networks (GANs), restricted boltzmann machines (RBMs), deep belief networks (DBNs), Q-networks, U-networks, Siamese networks, generative adversarial networks (GANs), transformers, etc.

[0082] Alternatively, depending on the embodiment, the deep neural network may be a model trained using transfer learning. Here, transfer learning refers to a learning method in which a large amount of unlabeled training data is pre-trained using a semi-supervised learning or self-learning method to obtain a pre-trained model (or base part) having a first task using techniques (MLM and NSP) as illustrated in FIGS. 4 and 5, respectively, and then the pre-trained model is trained using labeled training data using a supervised learning method to fine-tune it to be suitable for a second task, thereby implementing a target model. One of the models trained using this transfer learning method includes, but is not limited to, BERT (Bidirectional Encoder Representations from Transformers).

[0083] Neural networks, including the aforementioned deep neural networks, can be trained to minimize output errors. Training a neural network involves repeatedly inputting training data into the neural network, calculating the neural network output and target error for the training data, and backpropagating the neural network error from the output layer to the input layer to update the weights of each node in the neural network to reduce the error.

[0084] Meanwhile, a model according to one embodiment may be implemented to borrow at least a portion of a transformer. Here, the transformer may be composed of an encoder that encodes embedded data and a decoder that decodes the encoded data. The transformer may have a structure that receives a series of data, performs encoding and decoding steps, and outputs a series of data of different types. In one embodiment, the series of data may be processed into a form operable by the transformer. The process of processing the series of data into a form operable by the transformer may include an embedding process. Expressions such as data tokens, embedding vectors, and embedding tokens may refer to data embedded in a form that the transformer can process.

[0085] To enable a Transformer to encode and decode a series of data, the encoders and decoders within the Transformer can utilize an attention algorithm. Here, the attention algorithm can refer to an algorithm that, for a given query, calculates the similarity for one or more keys, reflects this similarity in the values ​​corresponding to each key, and then weights and adds the values ​​to calculate an attention value.

[0086] At this point, various types of attention algorithms can be categorized depending on how the query, key, and value are set. For example, attention can be obtained by setting the query, key, and value to be identical, which may indicate a self-attention algorithm. Alternatively, attention can be obtained by reducing the dimensionality of the embedding vector and obtaining individual attention heads for each segmented embedding vector to process a series of input data in parallel, which may indicate a multi-head attention algorithm.

[0087] In one embodiment, a transformer may be composed of modules that perform multiple multi-head self-attention algorithms or multi-head encoder-decoder algorithms. In one embodiment, the transformer may also include additional components that are not attention algorithms, such as embedding, normalization, or softmax. Methods for constructing a transformer using attention algorithms may include methods disclosed in Vaswani et al., Attention Is All You Need, 2017 NIPS, which are incorporated herein by reference.

[0088] Transformers can be applied to various data domains, such as embedded natural language, segmented image data, or audio waveforms. As a result, a transformer can transform a series of input data into a series of output data. Data from various data domains can be transformed to be processed by a transformer, a process called embedding.

[0089] Additionally, the transformer may process additional data representing the relative positional relationship or phase relationship between a series of input data. Alternatively, vectors representing the relative positional relationship or phase relationship between the input data may be additionally reflected in the series of input data to embed the series of input data. In one example, the relative positional relationship between the series of input data may include, but is not limited to, word order within a natural language sentence, the relative positional relationship between each segmented image, the time order of segmented audio waveforms, etc. The process of adding information representing the relative positional relationship or phase relationship between a series of input data may be referred to as positional encoding.

[0090] Hereinafter, let us look at various operations or functions that can be performed by the information providing server (100) for geographic objects by executing at least one command stored in the memory (120) by the processor (130).

[0091] First, the processor (130) can control the communication unit (110). Through this, the server (100) providing information on geographical objects can obtain information by performing communication via a network through the communication unit (110).

[0092] Additionally, the processor (130) can read the aforementioned data or commands stored in the memory (120) and record new data or commands in the memory (120). Additionally, the processor (130) can modify or delete data or commands that have already been recorded.

[0093] Additionally, the processor (130) can execute various models or modules stored in the memory (120). These models or modules may be implemented using the aforementioned artificial neural network or rule-based methods. For example, language models can be implemented. Below, we will examine language models.

[0094] A language model refers to a model generated based on human language. Such language models can be acquired in various ways. In one embodiment, the language model can be acquired using transfer learning. In transfer learning, as previously discussed, a large amount of unlabeled training data (e.g., a corpus) is pre-trained using semi-supervised or self-learning methods to obtain a pre-trained model (or base) for a first task. Subsequently, a process of fine-tuning the pre-trained model to suit a second task is performed. In this fine-tuning, labeled training data is used for training using supervised learning. In one embodiment, models trained using this transfer learning method include, but are not limited to, BERT (Bidirectional Encoder Representations from Transformers) or GPT (Generative Pre-trained Transformer).

[0095] Here, in one embodiment, the language model may be chatGPT, based on GPT-3 or GPT-4. Specifically, GPT-3 is the third-generation language prediction model in the GPT-n series created by OpenAI. GPT-3 consists of 175 billion parameters, more than twice as large as its predecessor, GPT-2, introduced in May 2020. It is part of a pre-trained natural language processing (NLP) system.

[0096] It is known that the tasks that GPT-3 can perform include solving various language-related problems, random writing, arithmetic operations, translation, simple web coding based on given sentences, and conversation.

[0097] Let's examine the learning process of these language models. In one embodiment, the language model may be pre-trained using general language, or it may be pre-trained based on at least one of papers, patent publications, and utility model publications. Therefore, more specialized searches of papers or patent publications are possible. This is because, when pre-trained using the aforementioned Masked Language Model (MLM) or Next Sentence Prediction (NSP) methods based on papers, patent publications, or utility model publications, these pre-trained models can acquire the structure of how the aforementioned knowledge is described.

[0098] In addition, reinforcement learning by human feedback (RLHF) can be performed in the fine-tuning of these language models. RLHF refers to the use of information judged by humans for learning in fine-tuning. For example, the fine-tuning process in RLHF can involve a learning process using a set of human-generated dialogues and a learning process using rankings selected by humans for multiple outputs generated by the language model. More specifically, a supervised fine-tuned model (SFT) is generated from a set of human-generated dialogues, and then the human ranking of the outputs of this SFT model is fed back to the model (Reward Model, RM), after which fine-tuning is performed using Proximal Policy Optimization (PPO). Here, fine-tuning by PPO refers to a reinforcement learning policy algorithm that continuously adjusts the current policy based on the actions performed by the agent and the rewards received. In this PPO, the process proceeds in the order of new prompt -> PPO -> Generate output -> Calculate reward, but the Calculate reward is updated and provided again as a new prompt.

[0099] These language models can be utilized in a variety of ways. For example, when there is information in a sequence or order, they can infer and inform users of the timing or order at which a notification should be triggered based on the context of that information. For example, in the case of stocks, they can be used to suggest buy or sell points based on the current price over time.

[0100] Hereinafter, we will look at functions that can be performed by the server (100) providing information on geographical objects by executing at least one command stored in the memory (120) by the processor (130). At this time, even without a separate description below, it is assumed that the function described as being performed by the server (100) providing information on geographical objects is performed by executing at least one command stored in the memory (120) by the processor (130) as described above.

[0101] FIG. 6 illustrates an exemplary flowchart of a method for providing land information for a building according to one embodiment. However, this flowchart is merely exemplary and the scope of the present invention is not limited thereto. For example, depending on the embodiment, each step may be performed in a different order than that illustrated in FIG. 6, or at least one step not illustrated in FIG. 6 may be additionally performed, or at least one of the steps illustrated in FIG. 6 may not be performed.

[0102] First, the method illustrated in FIG. 6 can be performed by a server (100) providing information on geographic objects.

[0103] Referring to Fig. 6, a data schema design and database creation step (S10) for objects is performed.

[0104] Next, a step (S20) of generating a 3D model, e.g., a 3D mesh, for the target area (terrain) is performed.

[0105] Next, a step (S30) of generating a 3D tileset for each object is performed.

[0106] Next, a step (S40) is performed to design the footprint of the target building, i.e. the boundary for the area where the building touches the land.

[0107] Here, the results obtained when steps S10 to S30 are performed, i.e., the results of database-izing the information acquired about the geographic object, as well as the results of organizing this information and generating a 3D tileset, can be referred to as CYLO or CY(CY)ILO. In other words, a user can access this CYLO and obtain various GIS information about the area he or she has targeted.

[0108] Additionally, the tool or platform that can implement and visualize the results obtained by performing step S40, i.e., the results in CYLO using a 3D engine, may be referred to as ADEK or ADEK. That is, users can access ADEK and receive results of their targeted region implemented and visualized using a 3D engine, etc.

[0109] Among these, let us look at step S20 in more detail with reference to Fig. 7.

[0110] FIG. 7 exemplarily illustrates a flowchart illustrating a specific process for designing the data schema and creating a database for the objects illustrated in FIG. 6. However, this flowchart is merely exemplary, and the scope of the present invention is not limited thereto. For example, depending on the embodiment, each step may be performed in a different order than that illustrated in FIG. 7, or at least one step not illustrated in FIG. 7 may be additionally performed, or at least one of the steps illustrated in FIG. 7 may not be performed.

[0111] Referring to Fig. 7, a step (S11) of designing a data schema is performed. The data schema may be for various types of objects, such as buildings (or structures), land lots (PNU), green spaces, and roads, but the types are not limited thereto.

[0112] Each schema may include parameters whose values ​​can be retrieved from public data, as well as parameters input by the user or obtained from external software, such as open source-based SW modules. For example, parameters indicating the ID of each object, legal building code, building use code, building structure code, building structure name, building integrated building area, use approval date, total floor area, land area, height, building coverage ratio, floor area ratio, building ID, whether the building is illegal, and whether it is a fire hazard facility may be included. In addition, parameters that can describe the centroid, maximum and minimum height, and each multi-polygon may also be included.

[0113] Meanwhile, a data schema can include parameters common to each object, as well as parameters specific to each object. In other words, the types of parameters included in a data schema can vary for each object.

[0114] Next, a step (S12) is performed to acquire data, i.e., values ​​to be entered into the parameters of the aforementioned data schema. This data may be public data, as described above, or user-entered data. Examples include, but are not limited to, the following.

[0115]

[0116] Si-Do Code"00" : "Nationwide""11" : "Seoul Metropolitan City""26" : "Busan Metropolitan City""27" : "Daegu Metropolitan City"...Beop Jeong Dong Code11000000000 Seoul Metropolitan City 11111000000 Jongno-gu, Seoul 11111010100 Cheongun-dong, Jongno-gu, Seoul 11111010200 Singyo-dong, Jongno-gu, Seoul 11111010300 Gungjeong-dong, Jongno-gu, Seoul 11111010400 Hyoja-dong, Jongno-gu, Seoul 1...Building Structure Code10 Masonry structure11 Brick structure12 Block structure13 Stone structure14 Steel house structure...Building Use Code01000 Single-family house01001 Single-family house01002 Multi-family house01003 Multi-family house01004 Official residence...Cultural Heritages1 Gongse-ri 5-story stone pagoda 0 11 32 9999 9999 Individual review 11m Processed in accordance with Yongin City Urban Planning Ordinance and other related laws (However, buildings or facilities with a maximum height of 32m or more are subject to individual review)...Land Category Code 01 Field 02 Rice Paddy 03 Orchard 04 Pasture...Land District Code 11 Cultural Resources Conservation District 12 Important Facility Conservation District 13 Ecosystem Conservation District 14 Natural Landscape District...Land Use Code 100 Residential 110 Single-family 120 Townhouse 130 Multi-family...Land Zone Code 11 Type 1 Exclusive Residential Zone 12 Type 2 Exclusive Residential Zone 13 Type 1 General Residential Zone 14 Type 2 General Residential Zone...

[0117] Additionally, SHP files representing the shapes of each building, PNU, green space, and road are obtained.

[0118] Next, a data organization step (S13) is performed to align the data schema of each object. Specifically, data is selected and written according to the data schemas for buildings, land lots, green spaces, and roads. This written data is then sorted according to the trial code. Then, SHP files for each sorted data are loaded.

[0119] Next, the loaded SHP file is converted into a multi-polygon (step S14). At this point, an open-source SW module can be used. The converted multi-polygon is assigned values ​​to the parameters of the aforementioned data schema.

[0120] Next, a step (S15) of converting the coordinate system for each object is performed. For example, the projected coordinate system is converted to a geographic coordinate system, where the projected coordinate system may be EPSG 5174 and the geographic coordinate system may be EPSG 4326.

[0121] Next, a step (S15) of obtaining the centroid value for each object is performed.

[0122] Next, a step (S16) is performed in which the data schema to which values ​​are assigned is sorted by attempt. The sorted results are stored in the aforementioned memory (120) in the form of a database.

[0123] Meanwhile, the above-described steps S11 to S16 may be performed one or more times for each building, lot number, green space, and road.

[0124] Referring back to FIG. 6, after step S10, step S20, i.e., a 3D modeling step for the target area (terrain), i.e., a step of generating a 3D mesh, is performed. This step S20 will be examined in more detail with reference to FIG. 8.

[0125] FIG. 8 illustrates a flowchart of a three-dimensional modeling process for a target area illustrated in FIG. 6, specifically, a method for providing information on a geographic object according to one embodiment. However, this flowchart is merely exemplary, and the spirit of the present invention is not limited thereto. For example, depending on the embodiment, each step may be performed in a different order than that illustrated in FIG. 8, or at least one step not illustrated in FIG. 8 may be additionally performed, or at least one of the steps illustrated in FIG. 8 may not be performed.

[0126] First, the method for providing information on a geographic object illustrated in FIG. 8 can be performed by the information providing server (100) for a geographic object illustrated in FIG. 1.

[0127] Referring to Figure 8, a step (S100) is performed to acquire an elevation map for a target area (terrain) including a land parcel. Here, the land parcel may be a parcel number or may refer to a land parcel on which a specific structure is to be built. Furthermore, the target area (terrain) refers to a specific area that includes such a parcel. In other words, the target area refers to an area larger than the land parcel itself, encompassing the land parcel itself.

[0128] The elevation map acquired in step S100 records the elevation of the land in the target area. However, the recorded elevation may only be for one or more points within the target area, for example, one or more points specified by a given latitude and longitude. In contrast, in one embodiment, the required elevation is for points with a finer granularity than the aforementioned latitude and longitude. This elevation map may be, for example, a DEM (digital elevation model) file provided by NASA, but is not limited thereto.

[0129] Here, such an elevation map may be defined based on a predetermined first coordinate system. Furthermore, such an elevation map may be divided into multiple regions based on latitude and longitude according to such first coordinate system.

[0130] Next, a step (S120) is performed to divide the target area into different numbers of tiles, each from a minimum zoom level to a maximum zoom level. Specifically, first, a scheme for tiling the Earth's flat surface, i.e., a tiling scheme, is selected. For example, an XYZ scheme may be selected, and the tile size, i.e., the number of pixels included in each tile, may be 64*64, but is not limited thereto. In addition, the zoom level may be 0 (minimum zoom level) to 13 (maximum zoom level), but is not limited thereto.

[0131] Figure 9 shows examples of tiling when the zoom level is 0 and when the zoom level is 1, respectively. When the zoom level is 0, X (row) is 2 to the power of 0 and Y (column) is 2 to the power of 1, and when the zoom level is 1, X is 2 to the power of 1 and Y is 2 to the power of 2. Similarly, when the zoom level is 13, X is 2 to the power of 13 and Y is 2 to the power of 14.

[0132] Examining the tiling at each zoom level reveals that the number of tiles varies for each zoom level. Not only is the number of pixels within each tile the same for the same zoom level, but the number of pixels within each tile can also be the same for different zoom levels. This allows users to experience smoother screen transitions as they zoom in and out, as the number of pixels remains constant across different zoom levels.

[0133] The tiling of the aforementioned target area may be performed according to a second coordinate system that differs in at least one of a reference point and a scale from the first coordinate system that serves as the basis for the elevation map. In this case, the boundary line of the tiles according to the second coordinate system at the maximum zoom level may not be aligned or coincide with the boundary line of the region of the elevation map divided into multiple sections according to the first coordinate system. This is because the reference point and scale of the second coordinate system and the first coordinate system are different from each other.

[0134] In step S120, a process is performed to convert an elevation map according to a first coordinate system, more specifically, the boundaries of a plurality of zones constituting the elevation map and the elevation values ​​in these zones, based on a second coordinate system. More specifically, in the above-described conversion, the boundaries of a plurality of zones and the elevation values ​​in these zones are converted based on a plurality of tiles partitioned at the maximum zoom level and also based on the first coordinate system.

[0135] The result of this transformation is a changed elevation map, so it can be referred to as a virtual elevation map or Virtual DEM file.

[0136] In addition, the conversion process is as follows, but is not limited thereto. First, the elevation map is divided into multiple regions according to the first coordinate system as described above. If each region is a rectangular shape, a single point where four adjacent squares meet exists in the first coordinate system. This single point is matched with a single point where four adjacent tiles meet at the maximum zoom level in the second coordinate system, and then conversion or interpolation is performed using bilinear interpolation. There may be various methods for matching here, and for example, this may include, but is not limited to, dividing or dividing the elevation map divided into multiple regions into multiple tiles generated according to the maximum zoom level in the second coordinate system.

[0137] Next, a step (S130) is performed to obtain a 3D model of the target area for each zoom level by calculating the height for each tile at each zoom level using the result converted in S120.

[0138] At this time, the tile height for each zoom level can be calculated so that the height calculation method for each tile from the minimum zoom level to the zoom level immediately before the predetermined zoom level is different from the height calculation method for each tile from the predetermined zoom level to the maximum zoom level.

[0139] Specifically, the height of each tile can be calculated assuming that the height of each tile is 0 from zoom level 0 to zoom level 7, which is the previous zoom level. In this case, each tile is modeled as a 3D mesh (3D, but effectively 2D because the height is 0) with a height of 0, and tiles modeled in this way are exemplarily illustrated in Fig. 10.

[0140] In contrast, from the given zoom level 8 to the maximum zoom level 13, the height for each tile can be calculated using the result converted in S120. Specifically, at zoom level 8, since X is 2^8 and Y is 2^9, 256*512 tiles are generated. These generated tiles have the same number of pixels, for example, 64*64. That is, each tile can be divided into 64*64 pixels. The center value of each pixel has latitude and longitude, and the result converted in S120 also includes latitude and longitude as well as the elevation value for the latitude and longitude. Accordingly, the elevation value for the corresponding latitude and longitude among the results converted in S120 is assigned to the center value in each pixel. In the same manner, the elevation value is assigned to the center value in the pixel of each tile repeatedly from zoom level 9 to the maximum zoom level.

[0141] Next, a 3D mesh is generated for each tile using the elevation value for the pixel of each tile. When generating the 3D mesh, a method is used to divide the tile into four or more triangles by considering the elevation value of each pixel. At this time, the triangles may be divided into four or more triangles so that the centers of other pixels in the tile are not included in the triangles connecting the centers of the pixels in the tile. This method may include, but is not limited to, Delaunay Triangulation. Fig. 11 illustrates 3D meshes for two tiles generated using this method as an example. Unlike Fig. 10, the 3D mesh illustrated in Fig. 11 has a height value.

[0142] That is, according to one embodiment, a 3D modeling of a target area including a site may be provided differently depending on the zoom level. Specifically, at a low zoom level, the height is 0, but as the zoom level increases, a 3D modeling reflecting the height in a more realistic manner may be provided. Accordingly, users involved in building design can easily and in detail obtain height information about the site in question as well as the target site including such site without having to go to the site.

[0143] Furthermore, as previously mentioned, information on buildings, as well as their respective address numbers, green spaces, and roads, can be obtained and provided from public data. In some cases, information on surrounding infrastructure can also be obtained and provided. Furthermore, as will be described later, this information can be visualized and provided for reviewing the site feasibility of a building, allowing users to closely review the site feasibility of a building based on the geographic information provided in one embodiment.

[0144] Additionally, the aforementioned zoom level at which height is provided, the maximum zoom level that provides the highest resolution information to the user, or the pixel count for each tile may vary depending on the intended use or situation. Therefore, users can receive geographic information optimized for them. Let's examine this further.

[0145] First, the predetermined zoom level mentioned above, that is, the predetermined zoom level at which the height starts to be provided, can be determined by considering the ratio between the width of the tile according to the predetermined zoom level and the longest straight-line distance that can be drawn in the target area. For example, if the longest straight-line distance that can be drawn in the target area is 10 km, the zoom level can be determined as the smallest zoom level among the zoom levels that have tiles with a width greater than 120% of 10 km. In this case, at the point in time when the height is provided to the user (the point in time of the predetermined zoom level), the entire target area can be modeled and provided to the user within a single tile, so that the user can obtain information about the target area more intuitively.

[0146] Next, the maximum zoom level is not always 13 and may vary depending on the situation. Specifically, the maximum zoom level may be determined by considering the maximum deviation in elevation in the target area and the pixel width of the tile according to the maximum zoom level. For example, let's say the maximum deviation in elevation in the target area is 30 m. Furthermore, let's say the pixel width of the tile at zoom level 13 is 40 m and the pixel width of the tile at zoom level 14 is 20 m. Then, the maximum deviation in elevation in the target area of ​​30 m is smaller than the pixel width of 40 m at zoom level 13, but larger than the pixel width of 20 m at zoom level 14. In this case, the maximum zoom level may be determined as 14. In other words, among the zoom levels whose pixel width is smaller than the maximum deviation, the zoom level with the smallest number may be determined as the maximum zoom level. In this case, the maximum deviation in elevation can be sufficiently expressed at the user's desired resolution.

[0147] Meanwhile, the 3D modeling of the aforementioned target area at each zoom level can be shared among users utilizing the information provision server (100) for such geographic objects. Therefore, users involved in the design and construction of buildings within the target area can more easily and conveniently share diverse information and engage in discussions with one another.

[0148] As described above, according to one embodiment, a 3D model of a target area including a site can be provided differently depending on the zoom level. Specifically, at a low zoom level, the height is 0, but as the zoom level increases, a 3D model reflecting the height can be provided to provide a more realistic sense. Accordingly, users involved in building design can easily and precisely obtain height information for the site in question as well as the target site including such site without having to go to the site.

[0149] Furthermore, as previously mentioned, information on buildings, as well as their respective address numbers, green spaces, and roads, can be obtained and provided from public data. In some cases, information on surrounding infrastructure can also be obtained and provided. Furthermore, as will be described later, this information can be visualized and provided for reviewing the site feasibility of a building, allowing users to closely review the site feasibility of a building based on the geographic information provided in one embodiment.

[0150] Let us now examine step S30, which is a step of generating a three-dimensional (3D) tileset or tile map for each object, i.e., each type of geographic object, as illustrated in FIG. 6. Step S30 is specifically illustrated in FIG. 12.

[0151] Figure 12 is a flowchart illustrating a method for providing information on a geographic object according to one embodiment. However, this flowchart is merely exemplary and the scope of the present invention is not limited thereto. For example, depending on the embodiment, each step may be performed in a different order than that illustrated in Figure 12, or at least one step not illustrated in Figure 12 may be additionally performed, or at least one of the steps illustrated in Figure 12 may not be performed.

[0152] First, the method for providing information on a geographic object illustrated in FIG. 12 can be performed by the information providing server (100) for a geographic object illustrated in FIG. 1.

[0153] Referring to Fig. 12, a step (S200) of obtaining a tile map for a target area is performed.

[0154] Here, the tile map for the target area refers to the 3D modeling information for the target area generated in step S20 illustrated in FIG. 6, i.e., the elevation for the target area is modeled by generating it as a 3D mesh. However, the tile map is not limited to being generated in this manner. For example, the aforementioned tile map, i.e., the modeling reflecting the elevation for the target area, can be generated in the form of a 3D mesh by a method not described herein, and the 3D modeling generated in this way can be acquired in step (S200).

[0155] Meanwhile, as previously discussed, these target area tile maps contain elevation information for the corresponding areas of each area tile when the target area is divided into multiple area tiles. As previously mentioned, these area tiles differ depending on the zoom level. As the zoom level increases, the number of area tiles may increase. However, the number of pixels constituting each tile may remain the same at each zoom level.

[0156] Next, a step (S210) of obtaining a tile map for a geographic object, i.e., a tile map for an object, is performed. Specifically, the tile map for an object has the same tile map scheme as the tile map for a target area. This is because, since the geographic object is located on the target area (terrain), it must be modeled on the same coordinate system as the target area to enable alignment. For example, the tile map for an object can be tiled in an XYZ scheme, just like the tile map for the target area.

[0157] Here, the target area tile map is generated for each zoom level corresponding to the minimum zoom level and the maximum zoom level, while the object tile map is generated only for a certain zoom level. For example, the object tile map can be generated only for the zoom level corresponding to the maximum zoom level of the target area tile map. In this case, the target area tile map and the object tile map have the same size and number of tiles at the maximum zoom level.

[0158] However, depending on the embodiment, the number of pixels in the target area tile map and the object tile map may be different at the maximum zoom level. For example, the number of pixels in each tile in the target area tile map may be 64*64, but the number of pixels in each tile in the object tile map may be 256*256. In other words, the number of pixels in each tile in the target area tile map may be less than the number of pixels in each tile in the object tile map. Due to this difference in pixels, as will be discussed later, an interpolation method may be used to obtain the elevation above sea level for corners (vertices) of buildings, etc.

[0159] Next, a step (S220) of acquiring shape information and location information for a geographic object is performed. Here, the shape information and location information refer to geometry information for the geographic object. For example, a 2D multi-polygon representing a geographic object and the centroid of the geographic object may be included in such geometry information, but the present invention is not limited thereto. Here, such shape information and location information may be acquired from a database created for each object described in step S10 of FIG. 6, and specifically, the process is exemplarily illustrated in FIG. 7. Since FIG. 7 has already been examined, the relevant portion will be cited.

[0160] Next, using the location information obtained in step S220, the tile for the object in which the object is located is selected (S230) from among the multiple object tiles obtained in step S210. Since the object includes a centroid, it is possible to select a file for the object that includes this centroid.

[0161] Next, a step (S230) is performed to select a regional tile corresponding to the object tile selected in S230 from among the multiple regional tiles acquired in step S200. As described above, the target regional tile map acquired in S200 and the object tile map acquired in S210 have the same size and number of tiles, so the regional tiles and the object tiles can be matched 1:1 with each other. Therefore, the above-described selection is possible.

[0162] Next, a step (S250) is performed to generate 3D modeling information for the geographic object. Specifically, the regional tiles selected in step S230 include elevation information for the region. Furthermore, shape information for the object is acquired in step S220. The aforementioned 3D modeling is then performed using the elevation information for the region and the shape information for the object. This will be examined in more detail with reference to FIG. 13.

[0163] FIG. 13 conceptually illustrates an example of a three-dimensionally modeled building according to one embodiment. However, FIG. 13 is merely exemplary, and the spirit of the present invention is not limited to what is illustrated in FIG. 13.

[0164] Referring to Fig. 13, the two 3D tiles illustrated in Fig. 11 are depicted in the form of a 3D mesh. The tile on the right, where a building, one of the geographic objects, is located, is the tile on the left and right. In addition, a 3D mesh, i.e., an elevation above sea level, is reflected in each tile. At this time, this elevation above sea level is assigned to each pixel of the tile. The tile illustrated in Fig. 13 is at the maximum zoom level of 13, and this tile is divided into 64*64 pixels. Therefore, each of the left and right tiles illustrated in Fig. 13 has an elevation above sea level assigned to each pixel at a resolution of 64*64.

[0165] Here, a building is located on the tile on the right, and the point where the building meets the ground surface, i.e., the boundary, can be derived from the geometry information or 2D multi-polygon information of the building mentioned above. Here, this 2D multi-polygon information is derived under the assumption that the ground surface is flat, i.e., the elevation above sea level is zero. Therefore, the boundary obtained according to this 2D multi-polygon information can be projected onto each triangular face of the 3D mesh from the bottom to the top of the tile, and as a result, as illustrated in FIG. 13, the boundary is depicted as a dotted line on each triangular face.

[0166] Here, since the height information is written by pixels on the face of each triangle, these boundaries also have height information, and therefore, as shown in Fig. 13, it can be seen that each has its own unique height compared to the tiles.

[0167] Next, these boundaries may be polygons of a given shape. In this case, the polygon has vertices or corners, and a process is performed to obtain height information for these corners.

[0168] If this corner meets the vertex or side of each triangle, the height information of the vertex of each triangle is known, so it is possible to obtain the height of this corner using interpolation.

[0169] In contrast, if this corner does not intersect with any vertex or edge of each triangle, but exists on a specific triangle face, then the Z value can be estimated using barycentric coordinates, but the estimation method itself is not limited to this. In this estimation method, in order to know the height of the corner existing on the triangle face, the height information of the three vertices existing in the triangle is used.

[0170] Next, if the geographic object is a building, there is a number of floors. For example, if each floor is 2 meters high, a three-story building would have a height of 6 meters, and a five-story building would have a height of 10 meters.

[0171] Here, the corner with the highest elevation above sea level among the boundary corners is selected. Then, from this selected corner, the height reflecting the number of floors of the building is determined as the maximum height of the building.

[0172] Next, lines are drawn from the remaining corners to the maximum height of the building, and 3D modeling of the building is performed using these lines. Since the remaining corners are lower in elevation than the corner with the highest elevation, the distance from each of these remaining corners to the maximum height of the building can be calculated by reflecting the height difference between the maximum elevation corner and each of the remaining corners, as well as the number of floors in the building. In other words, in 3D modeling of a building, the top surface of the building can be modeled as flat.

[0173] Meanwhile, unlike buildings, land lots, green spaces, and roads are simply located on the ground surface and do not have their own elevations. Therefore, as depicted on the face of a specific triangle in the left tile in the shape of an ellipse in Figure 14, land lots, green spaces, or roads can be modeled in three dimensions. Of course, if these land lots, green spaces, or roads exist as polygons spanning multiple triangles, the elevations corresponding to the corners of each polygon can be derived in the same or similar manner as the elevations at the corners of the aforementioned buildings.

[0174] Let us now examine step S40, illustrated again in Figure 6, the process of designing or determining the footprint area of ​​a building. Step S40 is specifically illustrated in Figures 15 and 22.

[0175] FIG. 15 is a flowchart for a process of deriving or defining an area in which a building within a land can be located in one embodiment, and FIG. 22 is a flowchart for a process of determining a footprint area for a building within a land according to one embodiment.

[0176] However, the flowcharts illustrated in FIGS. 15 and 22 are merely exemplary, and the spirit of the present invention is not limited thereto. For example, depending on the embodiment, each step may be performed in a different order than that illustrated in FIG. 15 or FIG. 22, or at least one step not illustrated in FIG. 15 or FIG. 22 may be additionally performed, or at least one of the steps illustrated in FIG. 15 or FIG. 22 may not be performed.

[0177] First, each of the flowcharts or methods illustrated in FIG. 15 and FIG. 22 can be performed by the information provision server (100) for the geographic object illustrated in FIG. 1.

[0178] Referring to FIG. 15, a step (S300) of acquiring 3D modeling information for a target area (terrain) where the land is located is performed. Here, this 3D modeling information may be acquired by performing the method illustrated in FIG. 8, but is not limited thereto.

[0179] Here, in the aforementioned 3D modeling information, the target area is divided into a plurality of tiles in which the elevation above sea level is reflected in the form of a 3D mesh, as exemplarily illustrated in Fig. 16. That is, each of the plurality of tiles constituting the target area is divided into at least 4 triangles, and the vertices where these triangles meet each other are implemented in the form of a 3D mesh to reflect the elevation above sea level of the corresponding vertex, as exemplarily illustrated in Fig. 17.

[0180] In addition, in the 3D modeling information, the region may include a tile map divided into different numbers of tiles for each zoom level from the minimum zoom level to the maximum zoom level. Since the tile map for each zoom level is the same as described above, a description of the overlapping portion will be omitted.

[0181] Next, a step (S320) of converting the 3D mesh obtained in step S300 into a 2D mesh is performed. In the process of converting into a 2D mesh, a 3D mesh for each of a plurality of tiles included in the tile map at the maximum zoom level among the tile maps each of which is divided into a different number of tiles for each of the minimum zoom level to the maximum zoom level described above may be used. That is, the reason for converting into a 3D mesh, as will be explained later, is to model a building on a site in 3D, and in order to model such a building at the highest resolution, the highest zoom level among the multiple zoom levels is required, and therefore, a 3D mesh for a tile included in the tile map at the maximum zoom level is used.

[0182] At this time, the 3D mesh at the aforementioned maximum zoom level can be converted into a 2D mesh in various ways. For example, in the 3D mesh, as discussed above, the elevation above sea level is assigned as a value to the pixels at the points where triangles meet or near those points, and the conversion into a 2D mesh can be performed by converting the elevation above sea level in the 3D mesh to a specific value, for example, 0. Here, the elevation being converted to 0 is merely exemplary, and it is of course possible to convert it to a specific value depending on the embodiment.

[0183] Next, a step (S320) is performed to acquire two-dimensional shape information about the land. The two-dimensional shape information about the land here refers to information indicating the shape of the land on a two-dimensional plane. This two-dimensional information may be acquired in step S220 of FIG. 12.

[0184] Next, a step (S330) may be performed to acquire two or more intersection points where the shape of the land intersects with the two-dimensional mesh, using the two-dimensional shape information acquired for the land and the two-dimensional mesh. Since the shape of the land is two-dimensional and the two-dimensional mesh is also two-dimensional, their intersection points may be acquired, as exemplarily illustrated in Fig. 18.

[0185] Next, a step (S340) of generating three-dimensional modeling information for the land connecting two or more intersections is performed. This step S340 may be performed including a step of connecting the two or more obtained intersections as lines, a step of converting the result of connecting the lines into a two-dimensional mesh, and a step of deriving the elevation of the result converted into the two-dimensional mesh using the elevation in the three-dimensional modeling information acquired for the target area. At this time, the elevation derived for the result may be used to generate the three-dimensional modeling information for the land.

[0186] Next, a step (S350) of acquiring information on a geographic object in contact with the land is performed. Here, the geographic object in contact with the land can be selected using a median value that has been previously acquired for each geographic object, for example, the median value acquired in step S15 of FIG. 7. Specifically, the geographic object in contact with the land can be selected from among geographic objects having a median value within a predetermined distance based on the median value of the land.

[0187] Next, a step (S36) is performed to derive an area in which the building can be located within the land by using the information acquired about the geographic object and the three-dimensional modeling information generated about the land. Specifically, the information acquired about the geographic object includes information on the type and shape of the geographic object in contact with the land, the boundary surface of the land is composed of a plurality of straight lines, and the area is defined by each of the plurality of straight lines constituting the boundary surface of the land moving toward the center of the land by a predetermined distance, and the distance by which each of the plurality of straight lines moves can be determined dependently (e.g., 3 m or 1 m) depending on the type and shape information of the geographic object in contact with the land based on each straight line, and examples thereof are exemplarily illustrated in FIGS. 19 to 21.

[0188] Meanwhile, the information obtained about the type of geographic object may be derived in response to the three-dimensional modeling information obtained about the target area being provided to a pre-trained geographic object identification model.

[0189] In addition, the distance traveled can be determined by the width of the road, and the width of the road can be confirmed by taking steps that gradually reduce the width and length of the road, as illustrated in Fig. 20, and counting the number of times this step is taken, or by using a predetermined deep learning model.

[0190] Additionally, the area in which the aforementioned building within the land can be located can be derived by a previously learned area definition model. For example, the information acquired about the geographic object includes information on the type and shape of the geographic object that contacts the land, and the area in which the building can be located can be derived in response to the 3D modeling information generated for the land and the type and shape information acquired for the geographic object that contacts the land being provided to the previously learned area definition model.

[0191] Meanwhile, the area within a land where a building can be located can be obtained in various ways. For example, various known technologies can be used to derive this area, and further explanation of these will be omitted.

[0192] Next, let us look specifically at how to determine the footprint area for the building illustrated in Figure 22.

[0193] Referring to FIG. 22, a step (S400) is performed to classify each of a plurality of lines connected to define an area of ​​the land using information on the type and shape of a geographic object that is in contact with the line.

[0194] The plurality of lines connected here to define the area of ​​land may be those illustrated in FIG. 21 or derived from the method according to FIG. 15, or may be 'adjacent land boundary lines' obtained in any other known manner.

[0195] Additionally, the area of ​​the land here may be defined by straight lines alone, or by a combination of curves and straight lines, or by curves alone, and in either case, the multiple lines mentioned in step S400 are described to encompass them.

[0196] The features used in classification, i.e., features, may include information on the type and shape of geographic objects bordering the line. For example, if the geographic object bordering the line is a road with a width of 4m or more, the line may be classified as a "big" group. On the other hand, if it is not a road or is a road with a width of less than 4m, it may be classified as a "small" group. The classification criteria themselves are merely exemplary, and the classification into either "big" or "small" groups is also merely exemplary. Figure 23 illustrates classification results for multiple lines as an example.

[0197] In addition, this classification can be utilized in a deep learning model. For example, as input, each of a plurality of lines defining an area of ​​land can be input into a deep learning model along with information indicating the shape of the lines and the positional relationship between them, and also information on the type and shape of the geographic object bordering each line can be input into the deep learning model. Then, as a result, the classification for each line can be classified as small or big. To this end, the deep learning model can have the architecture of a classification model, and the training input data can include information on the shape of the lines, the positional relationship between the plurality of lines, and the type and shape of the geographic object bordering each line, and the training answer data can include the classification result for each line as the answer.

[0198] Here, the original line defining the area of ​​the land may include a curve, while each of the plurality of lines mentioned in step S400 may be a straight line. In this case, the curves included in the original line may be converted into straight lines, and the curvature of the curves may be utilized for this conversion. For example, a curved portion that is not a straight line among the original lines may be selected. Thereafter, the selected curved portion may be divided, and the division may be based on a point where the curvature of any two points in the curved portion reaches a predetermined critical curvature. After this, the two ends of the curves resulting from the division may be connected to each other, thereby being converted into straight lines.

[0199] Here, the process of converting a curve into a straight line itself may employ a technique called smoothing, which may not be implemented in any one embodiment.

[0200] Next, a step (S410) is performed to group the plurality of lines into one of two or more groups using the classified results from step S400 and shape information for each of the plurality of lines.

[0201] There may be various criteria used for grouping. For example, the minimum separation distance between all points or a single point of each of a plurality of lines and any or all points of each of the other lines, as well as the intersection angle when the imaginary extensions of each of the plurality of lines intersect, may be used. For example, if there are lines with a minimum separation distance of less than 10 cm and an intersection angle (acute angle) of less than 10 degrees, these lines may be grouped into the same group. Conversely, lines with a minimum separation distance of 10 cm or more or an intersection angle of 10 degrees or more may be grouped into a different group rather than the same group. Figure 24 illustrates an example of the grouping results. Big group 1 contains two lines classified as big at step S400, Big groups 2 and 3 each contain one line classified as big at step S400, Small group 1 contains one line classified as small at step S400, and Small group 2 contains three lines classified as small at step S400.

[0202] As a result, each group can include at least one line. In a group including one line among these groups, that one line becomes the reference line. However, in a group including two or more lines, a single line created by merging those lines becomes the reference line. In the merging, the aforementioned reference line can be created through a process of connecting the two points with the longest distance between the two or more lines, but is not limited thereto. Figure 25 illustrates an example of a result created by merging two short lines to create one reference line (the longest line).

[0203] Next, for each group resulting from the grouping process, a step (S420) is performed to derive candidate footprint areas for the building with the maximum area. For example, if there are two groups, one candidate footprint area with the maximum area is derived for each group. The process of deriving candidate footprint areas will be exemplified below.

[0204] First, each group can be sorted or not. When sorting, the sorted results from step S400 can be used as the primary criteria, and the length of the baseline can be used as the secondary criteria. Figure 26 illustrates an example of such sorted results. Specifically, referring to Figure 26, after the primary sorting is performed by whether it is big or small, the secondary sorting can be performed by the length of the baseline within big and by the length of the baseline within small. Of course, the sorting criteria are not limited to these.

[0205] By deriving the candidate footprint area with the maximum area for each group in this sorted order, the first group may have a candidate footprint area with a larger maximum area than the subsequent groups, and thus, there is an advantage in that the candidate footprint area with the maximum area can be provided to the user more quickly and with fewer resources.

[0206] Of course, even if not sorted, the candidate footprint area with the maximum area for each group is derived.

[0207] Meanwhile, the candidate footprint area indicates where a building can be built within the adjacent property line of the site. For buildings, a candidate footprint area with as large an area as possible is desirable. Therefore, the candidate footprint area with the maximum area can be determined in various ways. Let's examine some examples below.

[0208] First, the candidate footprint area can be a polygon of various shapes, such as a triangle or a square, or an ellipse or a circle, or a combination of straight and curved lines. In any case, the lines defining the candidate footprint area can be divided into two or more parts (in the case of a circle, a circle can be divided into two lines based on any two points on its circumference, and the same goes for an ellipse, and in the case of a polygon such as a triangle or a square, it can be divided into two or more lines based on its vertices). Hereinafter, let us call one of these two or more lines the first side, and the other the second side.

[0209] For the candidate footprint area to have maximum area, the lengths of the first and second sides must be long. However, if these lengths are continuously increased, the candidate footprint area defined by these sides may encroach on one or more of the aforementioned adjacent property boundaries, i.e., the multiple lines connected to define the area of ​​the land. Since the building must be located within the adjacent property boundaries, the lengths of the first and second sides can be extended without encroaching on the boundaries.

[0210] In addition, the building coverage ratio, that is, the ratio of the building area to the land area, is also a constraint or restriction, so the area of ​​the candidate footprint area calculated by extending the length of the first and second sides must not exceed the building coverage ratio, and this can also be applied to the floor area ratio.

[0211] Under these conditions, let us first examine the following, assuming that the aforementioned candidate footprint area is a rectangle, with the first side being the width of the rectangle and the second side being the height of the rectangle.

[0212] First, the minimum value for the first variable may be assigned a preset value. For example, the minimum value may be 25 m, but is not limited thereto.

[0213] Next, the minimum and maximum values ​​for the second variable are determined for each group based on the length of the baseline in each group. For example, if the length of the baseline is less than 73 m, the minimum value may be 80% of the length of the baseline, and the maximum value may be the same length as the baseline. Furthermore, if the length of the baseline is greater than or equal to 73 m but less than 130 m, the minimum value may be 25 m, and the maximum value may be the same length as the baseline. Furthermore, if the length of the baseline is greater than or equal to 130 m, the minimum value may be 66.6 m, and the maximum value may be the smaller of the length of the baseline and 162 m.

[0214] The method or numbers used to determine the minimum and maximum values ​​here are only examples.

[0215] Next, the position of the second side is specified. At this time, the second side can be positioned parallel to the reference line, and either end of the second side can be positioned at the midpoint of the reference line or midway between this midpoint and either end of the reference line.

[0216] In this state, the length of the first side is extended by a predetermined first amount. In the extended state, the area of ​​the candidate footprint area defined by the first side and the second side is derived, and it is examined whether such area exceeds the building-to-land ratio or the floor area ratio, and also whether the candidate footprint area defined in this way encroaches on one or more of the aforementioned multiple lines. Whether or not there is an encroachment can be examined based on whether the candidate footprint area and the multiple lines overlap when expressed on a single coordinate system, but the examination method is not limited thereto.

[0217] If the length of the first side is extended, but the area of ​​the candidate footprint area resulting from the extension does not exceed the building-to-land ratio or the floor area ratio and does not encroach on any of the plurality of lines, the length of the first side is extended again by the first amount, and the aforementioned investigation is performed until the area of ​​the candidate footprint area exceeds the building-to-land ratio or the floor area ratio or encroaches on one or more of the plurality of lines. In such case, the length of the immediately preceding first side becomes the maximum value.

[0218] Next, the length of the second side is extended by a predetermined second amount. In this state, the length of the first side is again extended by the aforementioned first amount starting from the minimum value, and the aforementioned investigation is performed to obtain the maximum value of the length of the first side before it is exceeded or invaded.

[0219] Repeating this process generates one or more pairs of the first and second sides. For each pair, the area of ​​the candidate footprint region is calculated. The candidate footprint region with the largest value becomes the one with the largest area in the group.

[0220] Meanwhile, this is a candidate footprint area obtained by expanding the lengths of the first and second sides while fixing the position of the second side at a certain location. The second side can be positioned in various ways as long as it is parallel to the reference line within the group and is contained within both ends of the reference line within the group (if this condition is satisfied). Therefore, by varying the position of the second side under conditions that satisfy the aforementioned conditions, the candidate footprint area with the maximum area can be determined for each case.

[0221] Meanwhile, specifying the location of the second side may be performed by a person, but may also be performed by software or a deep learning model configured to operate according to the aforementioned conditions. For example, the aforementioned software or deep learning model may be designed according to the following conditions: 1) parallel to the baseline, 2) not intersecting with the adjacent property boundary, and 3) when projected onto the baseline with the two ends of the baseline as a reference, neither end of the second side exceeds either end of the baseline.

[0222] Meanwhile, as previously discussed, candidate footprint regions having the maximum area can be derived for each group, and the order may vary depending on the embodiment, but may follow the sorted order for each group.

[0223] If two or more candidate footprint areas with the largest area need to be derived, the candidate footprint areas can be derived starting with the building that occupies the larger land area among the two, and then the candidate footprint areas can be derived in the order of the building that occupies the next land area.

[0224] Meanwhile, a candidate footprint area with the maximum area can be derived using a learning model for deriving footprints. This learning model for deriving footprints can receive adjacent property boundaries as input. This model may have constraints such as floor area ratio or building coverage ratio already input. Furthermore, classification or grouping results for each line forming the adjacent property boundary can be provided as input to the model. Then, the model can infer the candidate footprint area within the adjacent property boundary. To this end, the input for training the model may include image files of the adjacent property boundary, constraints, and classification or grouping results for each line forming the adjacent property boundary. The correct answer may be an image or other representation of the location of the candidate footprint area with the maximum area within the adjacent property boundary. However, the training data is not limited to these.

[0225] An example of a candidate footprint area with the maximum area discussed above is shown in Figure 27.

[0226] As discussed above, one embodiment can derive not only the adjacent property boundaries within which a building can be constructed, but also a footprint indicating the location and size of the building within these boundaries. Therefore, designers of buildings and other structures can easily, accurately, and quickly obtain the information necessary for their design, without having to visit the construction site.

[0227] Meanwhile, the method according to the various embodiments described above can be implemented in the form of a computer program stored in a computer-readable recording medium programmed to perform each step of the method, and can also be implemented in the form of a computer-readable recording medium storing a computer program programmed to perform each step of the method.

[0228] The above description is merely an illustrative illustration of the technical idea of ​​the present invention, and those skilled in the art will appreciate that various modifications and variations can be made without departing from the essential quality of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate, rather than limit, the technical idea of ​​the present invention, and the scope of the technical idea of ​​the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within a scope equivalent thereto should be interpreted as being included in the scope of the rights of the present invention.

Claims

As a method of providing information about geographic objects, A step of classifying each of a plurality of lines connected to define an area of ​​the land by using information on the type and shape of the geographic object that is adjacent to the line; A step of grouping the plurality of lines using the classified results and shape information of the plurality of lines; A step of deriving, for each group resulting from the above grouping, a candidate footprint area for a building to have the maximum area; and Including a step of determining a footprint area for the building by considering a predetermined constraint among one or more candidate footprint areas derived above, In the step of deriving the above candidate footprint area, In a group including one line among the groups resulting from the above grouping, the one line becomes a reference line and is used to derive the candidate footprint area, and in a group including two or more lines, one line obtained by merging the two or more lines becomes a reference line and is used to derive the candidate footprint area. A method for providing information about geographic objects. In the first paragraph, The original line defining the area of ​​the above land includes a curve, and each of the plurality of lines is a straight line, Some of the above multiple straight lines are curves that have been transformed based on curvature. A method for providing information about geographic objects. In the first paragraph, In the above classification, Whether the above geographic object is a road and, if so, the width of the road is taken into account. A method for providing information about geographic objects. In the first paragraph, In the above grouping, The minimum distance between any point of each of the plurality of lines and any point of each of the other lines and the intersection angle when the virtual extension lines of each of the plurality of lines intersect are used. A method for providing information about geographic objects. In the first paragraph, Each group resulting from the above grouping is: After the first sorting using the above classified results, the second sorting is performed using the length of the baseline of each group. The above candidate footprint area is, Each of the above groups is sequentially derived for each group according to the sorting results obtained from the first sort to the second sort. A method for providing information about geographic objects. In the first paragraph, The sides defining the above candidate footprint area include a first side and a second side parallel to the reference line. The minimum value for the above first variable is a preset value, The minimum and maximum values ​​for the above second variable are determined for each group according to the length of the reference line in each group, In the above derivation step, The length of the first side is changed in consideration of the minimum value for the first side, and the length of the second side is changed in consideration of the minimum and maximum values ​​for the second side, and whether the area defined by the length of the first side and the second side invades the reference line for each of the plurality of groups and is derived according to the area of ​​the defined area. A method for providing information about geographic objects. In the first paragraph, The sides defining the above candidate footprint area include a first side and a second side parallel to the reference line, In the above derivation step, The result of the grouping is provided to the learned candidate footprint area derivation model, thereby obtaining the length and location of each of the first side and the second side. A method for providing information about geographic objects. In the first paragraph, The above restrictions include: At least one of the building coverage ratio and floor area ratio is included A method for providing information about geographic objects. In the first paragraph, A plurality of lines connected to define the area of ​​the above land, It is derived using the 3D modeling information generated for the above land and the information acquired for the above geographic object. A method for providing information about geographic objects. In paragraph 9, The 3D modeling information generated for the above land is: A step of acquiring 3D modeling information for a target area including the above-mentioned land; in the 3D modeling information, the target area is divided into a plurality of tiles in which the elevation above sea level is reflected in the form of a 3D mesh, A step of converting the above three-dimensional mesh into a two-dimensional mesh; A step of obtaining two-dimensional shape information for the above land; A step of obtaining two or more intersection points where the shape of the land intersects the two-dimensional mesh using the two-dimensional shape information obtained for the land and the two-dimensional mesh; and It is generated by performing the step of connecting two or more intersections above. A method for providing information about geographic objects. In paragraph 10, In the above conversion to a two-dimensional mesh, The elevation above sea level in the above 3D mesh is converted to 0. A method for providing information about geographic objects. As a method of providing information about geographic objects, A step of obtaining a tile map for a target area (terrain); the tile map for the target area includes elevation information for an area corresponding to each area tile when the target area is divided into a plurality of area tiles, A step of obtaining an object tile map composed of object tiles of the same size and number as the tile map for the target area; A step of obtaining shape information and location information for a geographic object; A step of selecting an object tile in which the geographic object is located from among the plurality of object tiles using the acquired location information; A step of selecting a regional tile corresponding to the selected object tile among the plurality of regional tiles; and A step of generating 3D modeling information for the geographic object by using the elevation information and the obtained shape information in the region corresponding to the selected regional tile. A method for providing information about geographic objects. As a method of providing information about geographic objects, A step of obtaining an elevation map for a target area (terrain) including the ground; the elevation map can be divided into a plurality of regions defined based on a first coordinate system, A step of dividing the target area into different numbers of tiles at each zoom level from the minimum zoom level to the maximum zoom level based on a second coordinate system having different reference points and scales from the first coordinate system; A step of converting the boundary lines of the plurality of zones constituting the above-mentioned elevation map and the elevation values ​​in the plurality of zones based on the plurality of tiles partitioned at the maximum zoom level; and Using the above-mentioned converted result, a step of calculating the height for each tile at at least some zoom levels from the minimum zoom level to the maximum zoom level and obtaining a 3D modeling of the target area for each zoom level is included. A method for providing information about geographic objects. A step of classifying each of a plurality of lines connected to define an area of ​​the land by using information on the type and shape of the geographic object that is adjacent to the line; A step of grouping the plurality of lines using the classified results and shape information of the plurality of lines; For each group resulting from the above grouping, a step of deriving a candidate footprint area for the building to have the maximum area; and A computer program that is performed, including a step of determining a footprint area for the building by considering a predetermined constraint among one or more candidate footprint areas derived above, In the step of deriving the above candidate footprint area, In a group including one line among the groups resulting from the above grouping, the one line becomes a reference line and is used to derive the candidate footprint area, and in a group including two or more lines, one line obtained by merging the two or more lines becomes a reference line and is used to derive the candidate footprint area. A computer-readable recording medium that stores a computer program.

Images