Program, information processing method, and information processing device

Abstract

To provide a program and the like capable of grasping a detailed outline of a building existing around a receiver.SOLUTION: A program according to one embodiment of the present disclosure causes a computer to execute processing of acquiring observation amount information based on an observation amount of a signal received by a receiver from an artificial satellite, and outputting information regarding the presence or absence of a building by inputting the acquired observation amount information to a first learning model which has been learned so as to output the information about the presence or absence of the building between the receiver and the artificial satellite when the observation amount information is input.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present technology relates to a program, an information processing method, and an information processing device. [Background technology]

[0002] Conventionally, a technology has been proposed for grasping the surrounding environment of a receiver that receives a signal based on a signal received from an artificial satellite. For example, a search support device described in Patent Document 1 grasps the elevation angle and azimuth The position of each satellite is grasped for each section divided by the corners, and for each satellite, based on a first indicator corresponding to the elevation angle and a second indicator corresponding to the radio wave strength, the reception status of the satellite radio waves from the grasped position of each satellite is compared with the ideal reception status of the satellite radio waves, and surrounding objects are displayed as the comparison result. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2005-18450 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the exploration support device described in Patent Document 1 displays the object (building) for each section, and therefore cannot grasp the detailed outline of the building.

[0005] The present disclosure has been made in consideration of the above circumstances, and aims to provide a program etc. that can grasp the detailed outlines of buildings existing around a receiver. [Means for solving the problem]

[0006] A program according to one embodiment of the present disclosure causes a computer to execute a process of acquiring observation quantity information based on the observation quantity of a signal received by a receiver from a satellite, and outputting information regarding the presence or absence of a structure by inputting the acquired observation quantity information into a first learning model that has been trained to output information regarding the presence or absence of a structure between the receiver and the satellite when the observation quantity information is input.

[0007] An information processing method according to one embodiment of the present disclosure acquires observation quantity information based on the observation quantity of a signal received by a receiver from a satellite, and outputs information regarding the presence or absence of a structure by inputting the acquired observation quantity information into a first learning model that is trained to output information regarding the presence or absence of a structure between the receiver and the satellite when observation quantity information is input.

[0008] An information processing device according to one embodiment of the present disclosure includes a processing unit that acquires observation quantity information based on observation quantities of a signal received by a receiver from a satellite, and outputs information regarding the presence or absence of a structure by inputting the acquired observation quantity information into a first learning model that is trained to output information regarding the presence or absence of a structure between the receiver and the satellite when observation quantity information is input. Effect of the Invention

[0009] A program according to an embodiment of the present disclosure makes it possible to grasp the detailed outlines of buildings existing around the receiver. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic diagram of an information processing system. [Diagram 2] 1 is a block diagram showing an example of the configuration of an information processing device; [Diagram 3] FIG. 11 is an explanatory diagram illustrating an example of a receiver table. [Figure 4] FIG. 4 is an explanatory diagram illustrating an example of an observation quantity table. [Diagram 5] FIG. 2 is an explanatory diagram showing an example of a first learning model. [Figure 6] FIG. 11 is an explanatory diagram showing an example of a second learning model. [Figure 7] FIG. 2 is an explanatory diagram showing an example of a 3D map generation module. [Figure 8] FIG. 2 is an explanatory diagram showing an example of a 3D map generation module. [Figure 9] FIG. 11 is an explanatory diagram for explaining a method for identifying a structure position. [Figure 10] FIG. 13 is an explanatory diagram for explaining a method for identifying a building height. [Figure 11] FIG. 2 is an explanatory diagram illustrating an example of processing by an information processing device. [Figure 12] FIG. 11 is a schematic diagram of an information processing device system according to a second embodiment. [Figure 13] FIG. 11 is an explanatory diagram showing an example of a receiver table according to the second embodiment. [Figure 14] FIG. 11 is an explanatory diagram showing an example of an observation quantity table according to the second embodiment. [Figure 15] FIG. 11 is an explanatory diagram showing an example of processing by an information processing device according to a second embodiment; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] (Embodiment 1) Hereinafter, the present invention according to the first embodiment will be described with reference to the drawings. FIG. 1 is a schematic diagram of an information processing system S. The information processing system S includes an information processing device 1, a plurality of receivers 2(1)-2(n), and a plurality of artificial satellites 3(1)-3(n). The information processing device 1 is, for example, a server computer, and can communicate with the receiver 2 via a network N using, for example, wide-area wireless communication. The information processing device 1 generates a three-dimensional map around the receiver 2 based on information acquired from the receiver 2. The receiver 2 transmits (outputs) information (observation quantity information) related to signals received from the artificial satellites to the information processing device 1. The artificial satellite 3 is, for example, an NSS (Navigation Satellite System) satellite, which includes a GPS (Global Positioning System) satellite, a GLONASS (Global Navigation Satellite System) satellite, a Galileo satellite, a BeiDou satellite, a QZSS (Quasi-Zenith Satellite System), or a GAGAN (GPS Aided GEO Augmented Navigation) satellite, and transmits a signal including a navigation message (latitude, longitude, altitude, speed, and acceleration of the artificial satellite 3).

[0012] 2 is a block diagram showing a configuration example of the information processing device 1. The information processing device 1 includes a processing unit 11, a storage unit 12, and a communication unit 13. The processing unit 11 is configured with a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), and performs various control processes, arithmetic processes, and the like by reading and executing a program P (program product) and a database stored in advance in the storage unit 12. Note that a database server or the like may be provided outside the information processing device 1, and the database may be read from the database server or the like. Furthermore, the functions of the information processing device 1 may be realized by a plurality of server devices or computers.

[0013] The storage unit 12 of the information processing device 1 is, for example, a volatile memory and a non-volatile memory. The storage unit 12 stores a program P, a receiver table 121, an observation amount table 122, a first learning model M1, a second learning model M2, and a three-dimensional map generation module M3. The program P may be provided to the information processing device 1 using a storage medium 12a in which the program P is stored so as to be readable by a computer. The storage medium 12a is, for example, a portable memory. Examples of the portable memory include a CD-ROM, a USB (Universal Serial Bus) memory, an SD card, a micro SD card, and a compact flash memory (registered trademark). When the storage medium 12a is a portable memory, the processing element of the processing unit 11 may read the program P from the storage medium 12a using a reading device (not shown). The read program P is written to the storage unit 12. Furthermore, the program P may be provided to the information processing device 1 by the communication unit 13 communicating with an external device. The receiver table 121, the observation quantity table 122, the first learning model M1, the second learning model M2, and the 3D map generation module M3 will be described in detail later.

[0014] The communication unit 13 is a communication module or a communication interface for communicating with the receiver 2 by wire or wirelessly, and is, for example, a wide-area wireless communication module such as LTE (registered trademark), 4G, or 5G. The processing unit 11 communicates with the receiver 2 via the communication unit 13 through an external network N such as the Internet.

[0015] FIG. 3 is an explanatory diagram showing an example of the receiver table 121. The receiver table 121 stores the latitude and longitude of the location where each receiver 2 is installed. The management items (fields) of the receiver table 121 include, for example, a receiver number field, a latitude field, and a longitude field. The receiver number field stores a number (receiver number) assigned to each receiver 2. The receiver 2(n) is assigned the receiver number Rn. That is, the receiver number assigned to the receiver 2(1) is R1. The latitude field stores the latitude of the location where the receiver 2 is installed. The longitude field stores the longitude of the location where the receiver 2 is installed.

[0016] FIG. 4 is an explanatory diagram showing an example of the observation quantity table 122. The observation quantity table 122 stores information (observations) for each receiver 2 regarding the signal received by the receiver 2 from the artificial satellite 3. The storage unit 12 of the information processing device 1 stores an observation quantity table 122 for each combination of the receiver 2 and the artificial satellite 3. The observation quantity table 122 is provided with attribute information for distinguishing between the multiple observation quantity tables 122. The attribute information includes the receiver number assigned to each receiver 2 and the artificial satellite number assigned to each artificial satellite 3. The artificial satellite 3(n) is assigned the artificial satellite number Sn. In other words, the artificial satellite number assigned to the artificial satellite 3(1) is S1.

[0017] The management items (fields) of the observation quantity table 122 include, for example, a time field, an elevation angle field, an azimuth angle field, a signal strength field, a pseudorange residual field, a Doppler frequency residual field, and a structure presence probability field. The time field stores the time when the receiver 2 receives a signal from the satellite 3. The elevation angle field stores the elevation angle of the satellite 3 relative to the receiver 2 when the receiver 2 receives the signal. The azimuth angle field stores the azimuth angle of the satellite 3 relative to the receiver 2 when the receiver 2 receives the signal. The elevation angle and azimuth angle of the satellite 3 relative to the receiver 2 are calculated based on the navigation message (latitude, longitude, and altitude) included in the signal and the latitude and longitude of the receiver. The elevation angle and azimuth angle define coordinates with respect to the receiver 2.

[0018] The signal strength field of the observation amount table 122 stores the strength of the signal received by the receiver 2 from the satellite 3. The pseudorange residual field stores the residual obtained by subtracting the theoretical distance between the receiver 2 and the satellite 3, which is calculated based on the navigation message (latitude, longitude, and altitude) and the latitude and longitude of the receiver 2, from the actual distance between the receiver 2 and the satellite 3, which is calculated based on the time required for the signal transmitted from the satellite 3 to be received by the receiver 2. For example, if there is an obstruction such as a building on the straight line connecting the receiver 2 and the satellite 3, the signal received by the receiver 2 is a signal that has been reflected by another building or a mountain slope and reached the receiver 2, so the actual distance will be long and the pseudorange residual will be a positive value. Also, for example, if the signal reaches the receiver 2 after hitting the corner of a building and being diffracted, the actual distance will be short and the pseudorange residual will be a negative value.

[0019] The Doppler frequency residual field of the observation quantity table 122 stores the residual obtained by subtracting the theoretical frequency calculated based on the navigation message (speed and acceleration) contained in the signal from the actual measured frequency of the signal that has arrived at the receiver 2.

[0020] The structure existence probability field stores the probability that a structure exists at coordinates specified by the elevation angle and azimuth angle from the receiver 2, output by the first learning model M1 described below. That is, the structure existence probability field stores information regarding the presence or absence of a structure, output by inputting observation amount information into the first learning model M1. Note that the observation amount table 122 may include a field for storing the determination result of whether or not a structure exists between the receiver 2 and the artificial satellite 3, instead of the structure existence probability field.

[0021] 5 is an explanatory diagram showing an example of the first learning model M1. When an observation amount (signal strength, pseudorange residual, and Doppler frequency residual) at a coordinate stored in the observation amount table 122 is input, the first learning model M1 outputs the probability that a structure exists at the coordinate. The first learning model M1 according to this embodiment is configured by, for example, an RNN (Recurrent Neural Network). Note that the first learning model M1 may be other neural networks such as, for example, a Transformer, a BERT (Bidirectional Encoder Representations from Transformers), a GPT (Generative Pre-trained Transformer), a CNN (Convolution Neural Network), a DNN (Deep Neural Network), or a LSTM (Long Short Term Memory), or may be other learning algorithms such as a decision tree, a support vector machine, a logistics regression, a random forest, or an XGBoost (eXtreme Gradient Boosting).

[0022] The input layer of the first learning model M1 accepts input of observables in the order of time when the signal was received. The input layer may accept, for example, input of the most recent 10 observations in the order of time. The intermediate layer of the first learning model M1 has a plurality of neurons that extract features of the input data, and passes the extracted features to the output layer. The neurons of the intermediate layer also receive (recurrently input) features extracted by the intermediate layer from the observations input at the previous time. The output layer has neurons that output the probability that a structure exists at the coordinates where the observations are input, and outputs the probability that a structure exists based on the features output from the intermediate layer. The processing unit 11 of the information processing device 1 stores the output probability of the structure existence in the structure existence probability field of the observation table 122, in association with the coordinates where the observations are input.

[0023] The first learning model M1 is trained using training data that associates an observation amount at each coordinate with a correct answer value label of the probability that a structure exists. The correct answer value label is based on, for example, an omnidirectional photograph taken at the location where the receiver 2 is installed. When it is recognized that a structure exists at each coordinate, a probability of 1 is associated with the observation amount as a correct answer value label. When it is recognized that a structure does not exist at each coordinate, a probability of 0 is associated with the observation amount as a correct answer value label. The first learning model M1 is trained by correcting the weights and biases by backpropagation so that when an observation amount is input, it outputs the same value as the correct answer value label. The first learning model M1 is trained (constructed) by performing the same process on a large number of training data.

[0024] FIG. 6 is an explanatory diagram showing an example of the second learning model M2. When a plurality of two-dimensional array data (array data group) indicating the probability of a structure existing at each coordinate for the receiver 2 (information regarding the presence or absence of a structure) is input, the second learning model M2 outputs a silhouette image in which the outlines of structures existing around the receiver 2 are projected. The second learning model M2 is configured by, for example, a U-net. Note that the second learning model M2 may be a learning algorithm capable of generating images, such as CNN, GAN (Generative Adversarial Network), or VAE (Variational AutoEncoder).

[0025] The two-dimensional array data input to the second learning model M2 is data indicating the probability of the existence of a structure at the coordinates for a specific receiver 2. The following describes a case where a silhouette diagram related to the receiver 2(1) is generated. The processing unit 11 of the information processing device 1 performs the same process for the receivers 2(2) to 2(n). The processing unit 11 reads out the observation quantity table 122 whose receiver number included in the attribute information is R1 from among the multiple observation quantity tables 122 stored in the storage unit 12. The processing unit 11 generates two-dimensional array data k[x][y]=p (x: elevation angle, y: azimuth angle, p: probability) based on the elevation angle, azimuth angle, and probability stored in the read observation quantity table 122. For example, the two-dimensional array data generated based on the top record of the observation quantity table 122 shown in FIG. 4 is k[x][y]=p. The processing unit 11 generates two-dimensional array data for all records of the multiple observation quantity tables 122 read out, and generates an array data group.

[0026] The input layer of the second learning model M2 accepts the input of a group of sequence data. The intermediate layer of the second learning model M2 has a plurality of neurons that extract features of the input data, and passes the extracted features to the output layer. The output layer has neurons that output a silhouette diagram on which the contours of buildings existing around the receiver 2 are projected, and outputs a silhouette diagram on which the contours of buildings existing around the receiver 2(1) are projected based on the features output from the intermediate layer. The silhouette diagram is output with the elevation angle from the receiver 2(1) as the first axis and the azimuth angle as the second axis. The elevation angle indicates the angle from the ground at the location of the receiver 2(1). The azimuth angle indicates the direction from the receiver 2(1), with north being 0° (360°), east being 90°, south being 180°, and west being 270°.

[0027] The second learning model M2 is trained using training data that links the array data group in each receiver 2 with a correct silhouette image. The correct silhouette image is, for example, created by visual inspection or computer based on an omnidirectional photograph taken at the location where the receiver 2 is installed. The second learning model M2 is trained by correcting the weights and biases using error backpropagation so that when an array data group is input, the second learning model M2 outputs the correct silhouette image. The second learning model M2 is trained (constructed) by performing the same process on a large amount of training data.

[0028] 7 and 8 are explanatory diagrams showing an example of the 3D map generation module M3. The 3D map generation module M3 is composed of a third learning model M31 and a stereo vision module M32. The third learning model M31 detects areas showing the same building in silhouette diagrams related to the inputted multiple receivers 2, and outputs the areas showing the same building in each silhouette diagram by linking them. The stereo vision module M32 generates and outputs a 3D map based on the latitude and longitude of each receiver 2 and the silhouette diagram showing the areas showing the same building. The 3D map generation module M3 may be a learning algorithm capable of generating a 3D image, such as CNN, GAN, VAE, U-net, NV (Neural Volumes), SRN (Scene Representation Networks), LLFF (Local Light Field Fusion), or NeRF (Neural Radiance Fields).

[0029] 7 shows a third learning model M31 of the 3D map generation module M3. The third learning model M31 is configured by a Neural Network (NN) having a segmentation function, such as RegionsCNN (RCNN). The third learning model M31 may be a learning algorithm such as FastRCNN, FasterRCNN, YOLO (You Look Only Once), or SSD (Single Shot Multibox Detector).

[0030] The input layer of the third learning model M31 accepts input of a plurality of silhouette diagrams linked with the latitude and longitude of the receiver 2. The intermediate layer of the third learning model M31 has a plurality of neurons that extract features of the input silhouette diagrams, and transfers the extracted features to the output layer. The output layer has neurons that detect the area in which the building is shown in the silhouette diagram, and outputs the area in which the building exists based on the features output from the intermediate layer. In addition, at this time, when it is inferred that the same building is included in the plurality of silhouette diagrams, the output layer outputs the area by surrounding the same building with the same type of bounding box. In the example shown in FIG. 7, bounding boxes A to H are output to each silhouette diagram. For example, the building included in the bounding box A output in the silhouette diagram related to the receiver 2(1) and the building included in the bounding box A output in the silhouette diagram related to the receiver 2(2) are inferred to be the same building. The buildings included in the bounding boxes A to H, respectively, are b(A) to b(H).

[0031] The third learning model M31 is trained using multiple silhouette images and training data in which areas containing the same building in the multiple silhouette images are linked. When multiple silhouette images are input, the third learning model M31 is trained by correcting weights and biases using backpropagation so that, in each silhouette image, areas that are the same as areas linked in the training data are output by linking areas containing the same building. The third learning model M31 is trained (constructed) by performing the same process on a large amount of training data.

[0032] FIG. 8 shows the stereo vision module M32 of the three-dimensional map generation module M3. The stereo vision module M32 is a program module that generates a three-dimensional map showing the shape, position, and height of a building by plotting the three-dimensional contour of the building using a stereo vision method. The stereo vision module M32 generates a three-dimensional map based on a silhouette diagram showing an area including the building, output by the third learning model M31, and the latitude and longitude of the receiver 2 related to each silhouette diagram. Specifically, the stereo vision module M32 calculates the disparity between multiple receivers 2 based on the latitude and longitude of the multiple receivers 2, and sets the buildings included in the same type of bounding box shown in the multiple silhouette diagrams as landmarks. The stereo vision module M32 plots the location where the building exists in a three-dimensional space based on the calculated disparity and the landmark. The stereo vision module M32 also calculates the position and height of the building, and can generate a three-dimensional map showing the shape, position, and height of the building by adding the calculated position and height information to the plotted building. In addition, the latitude and longitude of each building shown in the 3D map of FIG. 8 are written only in minutes and seconds, with the degrees omitted. In the example shown in FIG. 8, the latitude north of the position of each building is 34 degrees, and the longitude east of the position is 135 degrees. In addition, in the 3D map of FIG. 8, the positions of the receivers 2(1) to 2(3) are shown, but the display of the position of the receiver 2 may be omitted. Details of the calculation method of the position and height of the building will be described later. In addition, the stereo vision module M32 may calculate the position and height of the building, and plot the location where the building exists in the 3D space based on the calculated position and height information.

[0033] Fig. 9 is an explanatory diagram explaining a method for identifying a structure position. This diagram shows an example of calculating the position of structure b(A) based on the latitude and longitude of receiver 2(1) and receiver 2(2). Note that structure b(A) is a structure included in bounding box A shown in Fig. 7. Processing unit 11 of information processing device 1 calculates distance da between receiver 2(1) and receiver 2(2) based on the latitude and longitude of receiver 2(1) and receiver 2(2).

[0034] The processing unit 11 of the information processing device 1 identifies the azimuth angle at which the building b(A) (bounding box A) is located in the silhouette diagram of the receiver 2(1) shown in Fig. 7. The processing unit 11 also calculates the azimuth angle of the receiver 2(2) relative to the receiver 2(1) based on the latitude and longitude of the receiver 2(1) and the receiver 2(2). The processing unit 11 calculates the angle θ1 between the receiver 2(2) and the building b(A) relative to the receiver 2(1) based on the azimuth angle of the building b(A) relative to the receiver 2(1) and the azimuth angle of the receiver 2(2).

[0035] The processing unit 11 of the information processing device 1 identifies the azimuth angle at which the building b(A) (bounding box A) is located in the silhouette diagram of the receiver 2(2) shown in Fig. 7. The processing unit 11 also calculates the azimuth angle of the receiver 2(1) relative to the receiver 2(2) based on the latitude and longitude of the receiver 2(1) and the receiver 2(2). The processing unit 11 calculates the angle θ2 between the receiver 2(1) and the building b(A) relative to the receiver 2(2) based on the azimuth angle of the building b(A) relative to the receiver 2(2) and the azimuth angle of the receiver 2(1).

[0036] The processing unit 11 of the information processing device 1 calculates the angle θ3 between the receiver 2(1) and the receiver 2(2) relative to the building b(A) based on the angle θ1 and the angle θ2. The processing unit 11 calculates the distance db between the receiver 2(1) and the building b(A) and the distance dc between the receiver 2(2) and the building b(A) based on the distance da, the angle θ1, the angle θ2, and the angle θ3 by the sine theorem. The processing unit 11 specifies the latitude and longitude of the building b(A) based on the latitude and longitude of the receiver 2(1), the azimuth angle of the building b(A) relative to the receiver 2(1), and the distance db, and calculates the position of the building b(A). The processing unit 11 may correct the latitude and longitude of the building b(A) based on the latitude and longitude of the receiver 2(2), the azimuth angle of the building b(A) relative to the receiver 2(2), and the distance db. In this case, it is possible to specify the latitude and longitude of the building b(A) with higher accuracy.

[0037] Fig. 10 is an explanatory diagram explaining a method for identifying the height of a structure. This diagram shows an example of calculating the height of a structure based on the elevation angle from the receiver 2(1) to the highest point of the structure b(A) and the distance db explained in Fig. 7. If the elevation angle from the receiver 2(1) to the highest point of the structure b(A) is angle θ4, the height h of the structure b(A) is calculated by h=db*tan θ4.

[0038] FIG. 11 is an explanatory diagram showing an example of processing by the information processing device 1. The processing unit 11 of the information processing device 1 executes the following processing at a predetermined timing (for example, a predetermined time on a predetermined day of the week). The processing unit 11 may start the processing when information instructing the execution of the following processing is input. The processing unit 11 acquires an observation amount related to the artificial satellite 3 from the receiver 2 (S1), and stores the acquired observation amount in the observation amount table 122 (S2). The processing unit 11 reads out the observation amount of each coordinate (elevation angle and azimuth angle) from the observation amount table 122 (S3). The processing unit 11 inputs the read observation amount to the first learning model M1 (S4), and outputs the probability that a structure exists at each coordinate (S5). The processing unit 11 stores the probability output by the first learning model M1 in the probability field of the observation amount table 122 (S6). The processing unit 11 executes S1 to S6 for multiple artificial satellites 3 for one receiver 2.

[0039] The processing unit 11 of the information processing device 1 reads out coordinates (elevation angle and azimuth angle) and the probability at the coordinates from the observation quantity table 122 in which the receiver number of the specific receiver 2 is stored in the attribute information (S7). The processing unit 11 creates multiple two-dimensional array data (array data group) based on the read out coordinates and probability (S8). The processing unit 11 inputs the created array data group to the second learning model M2 (S9), and outputs a silhouette diagram related to the specific receiver 2 (S10). The processing unit 11 executes S1 to S10 for multiple receivers 2.

[0040] The processing unit 11 of the information processing device 1 inputs silhouette diagrams of the multiple receivers 2 to the third learning model M31 of the 3D map generation module M3 (S11) and outputs the area including the structure (S12). The processing unit 11 inputs the silhouette diagrams of the multiple receivers 2 to which the area including the structure is output to the stereo vision module M32 of the 3D map generation module M3 (S13) and plots the location where the structure exists in the 3D space (S14). The processing unit 11 also calculates the position and height of the structure by the stereo vision module M32 (S15). The processing unit 11 assigns the calculated position and height to the plotted structure (S16) and generates a 3D map (S17). The processing unit 11 can display the generated 3D map on a display device connected to the information processing device 1, for example, or transmit it to a terminal device and display it on the terminal device.

[0041] According to the above configuration and processing, the receiver 2 can grasp the detailed outlines of buildings existing around the receiver based on the signal acquired from the artificial satellite 3. Also, a highly accurate 3D map can be generated based on the outlines of the buildings. The processing unit 11 of the information processing device 1 may include the elevation angle and azimuth angle of the artificial satellite 3 relative to the receiver 2 in the input data to be input to the first learning model M1. The processing unit 11 may also output a silhouette diagram by inputting the observation amount at each coordinate to the second learning model M2.

[0042] (Embodiment 2) Hereinafter, the present invention according to the first embodiment will be described with reference to the drawings. Among the configurations of the second embodiment, the same configurations as those of the first embodiment are given the same reference numerals, and detailed description thereof will be omitted. The receiver 2 according to the second embodiment is mounted on a moving body such as a vehicle or a ship. The receiver 2 may be realized by a location information acquisition function mounted on a smartphone or tablet terminal owned by a person. The information processing device 1 outputs a silhouette diagram of a specific point based on the observation amount of a signal received from an artificial satellite 3 by the receiver 2 that has passed the specific point.

[0043] Fig. 12 is a schematic diagram of an information processing system S according to a second embodiment. A receiver 2 according to this embodiment is mounted on a moving body. In this embodiment, an example will be described in which the moving body is a vehicle C. A receiver 2(n) is mounted on the vehicle C(n). The vehicle number of the vehicle C on which the receiver 2(n) is mounted is denoted by Cn.

[0044] 13 is an explanatory diagram showing an example of the receiver table 121 according to the second embodiment. Management items (fields) of the receiver table 121 according to the second embodiment include a receiver number field and a vehicle number field. The receiver number field stores the receiver number of the receiver 2. The vehicle number field stores the vehicle number of the vehicle C on which the receiver 2 is mounted.

[0045] 14 is an explanatory diagram showing an example of an observation quantity table 122 according to the second embodiment. The management items (fields) of the observation quantity table 122 according to the second embodiment include a receiver latitude field and a receiver longitude field. The receiver latitude field stores the latitude of the point where the receiver 2 was located when the receiver 2 received a signal from a satellite 3. The longitude field stores the longitude of the point where the receiver 2 was located when the signal was received. The receiver 2 is able to acquire the latitude and longitude of the point where the receiver 2 was located based on signals received from multiple satellites 3.

[0046] FIG. 15 is an explanatory diagram showing an example of processing by the information processing device 1 according to the second embodiment. The processing unit 11 of the information processing device 1 acquires from the receiver 2 the observation amount of the signal received from the artificial satellite 3, and the latitude and longitude of the location where the receiver 2 was located when it received the signal from the artificial satellite 3 (S21). The processing unit 11 stores the acquired observation amount and the latitude and longitude of the location where the receiver 2 was located in the observation amount table 122 (S22). The processing unit 11 reads out the observation amount of each coordinate (elevation angle and azimuth angle) from the observation amount table 122 (S23). The processing unit 11 inputs the read observation amount to the first learning model M1 (S24), and outputs the probability that a structure exists at each coordinate (S25). The processing unit 11 stores the probability output by the first learning model M1 in the probability field of the observation amount table 122 (S26). The processing unit 11 executes S21 to S26 for a plurality of satellites 3 for one receiver 2, and also executes S21 to S26 for a plurality of receivers 2.

[0047] The processing unit 11 of the information processing device 1 reads out the coordinates (elevation angle and azimuth angle) of the records in which the location (latitude and longitude) of the receiver 2 is the same from the multiple observation amount tables 122 and the probability at the coordinates (S27). The processing unit 11 may read out the coordinates and the probability at the coordinates by regarding the latitude and longitude within a predetermined error range as the same location. The processing unit 11 creates multiple two-dimensional array data (arrangement data group) for the specific location based on the read out coordinates and probability (S28). The processing unit 11 inputs the created array data group to the second learning model M2 (S29) and outputs a silhouette diagram related to the specific location (S30). The processing unit 11 executes S27 to S30 for multiple locations.

[0048] The processing unit 11 of the information processing device 1 inputs silhouette diagrams of multiple locations to the third learning model M31 of the 3D map generation module M3 (S31) and outputs the area including the structure (S32). The processing unit 11 inputs the silhouette diagrams of multiple receivers 2, from which the area including the structure has been output, to the stereo vision module M32 of the 3D map generation module M3 (S33) and plots the location where the structure exists in a three-dimensional space (S34). The processing unit 11 also calculates the position and height of the structure by the stereo vision module M32 (S35). The processing unit 11 assigns the calculated position and height to the plotted structure (S36) and generates a three-dimensional map (S37).

[0049] According to the above configuration and processing, it is possible to grasp the detailed outline of a structure using a receiver or receiving function of a signal from the artificial satellite 3 mounted on a moving object such as a vehicle or a smartphone. In addition, it is possible to generate a highly accurate 3D map based on the outline of the structure. Note that the processing unit 11 of the information processing device 1 may include the latitude and longitude of the point where the receiver 2 is located, or the elevation angle and azimuth angle of the artificial satellite 3 relative to the receiver 2, in the input data to be input to the first learning model M1.

[0050] (Modification) In each of the above-described embodiments, the processing unit 11 of the information processing device 1 outputs a silhouette diagram by inputting a group of array data to the second learning model M2, but is not limited to this. The processing unit 11 may plot coordinates where the probability of the existence of a structure is equal to or higher than a predetermined probability, and output a silhouette diagram by performing linear interpolation processing based on the plotted coordinates.

[0051] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The technical features described in each embodiment can be combined with each other, and the scope of the present invention is intended to include all modifications within the scope of the claims and equivalents to the scope of the claims. In addition, the independent claims and dependent claims described in the claims can be combined with each other in all combinations regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. A format in which a multiple claim (multi-multi claim) that references at least one multiple claim may be used. [Explanation of symbols]

[0052] 1. Information processing device 11 Processing section 12 Storage section 12a Storage medium 13. Communications Department 121 Receiver Table 122 Observation Table M1 First learning model M2 Second Learning Model M3 3D map generation module M31 Third Learning Model M32 Stereo Vision Module 2 Receiver 3 satellite 11 Processing section 12 Storage section 12a Storage medium 13. Communications Department 121 Receiver Table 122 Observation Table M1 First learning model M2 Second Learning Model M3 3D map generation module M31 Third Learning Model M32 Stereo Vision Module P Program S Information Processing System

Claims

1. The receiver acquires observation quantity information based on the observation quantity of the signal received from the artificial satellite, The acquired observation quantity information is input to a first learning model that is trained to output information regarding the presence or absence of a structure between the receiver and the artificial satellite when the observation quantity information is input, thereby outputting information regarding the presence or absence of a structure. A program that causes a computer to carry out processing.

2. The observable information includes the strength of the signal, the pseudorange residual between the receiver and the satellite, or the Doppler frequency residual of the signal. The program according to claim 1.

3. The information regarding the presence or absence of a structure is the probability that a structure exists at coordinates specified by the elevation angle and azimuth angle from the receiver.

3. The program according to claim 1 or 2.

4. A second learning model outputs a silhouette diagram on which the contours of the buildings are projected, with the elevation angle from the receiver as a first axis and the azimuth angle as a second axis, when information on the presence or absence of buildings at a plurality of coordinates specified by the elevation angle and azimuth angle from the receiver is input, and outputs a silhouette diagram on which the contours of the buildings existing around the receiver are projected.

3. The program according to claim 1 or 2.

5. A three-dimensional map is generated based on the silhouette image of the plurality of receivers, the three-dimensional map showing the three-dimensional shape of the building, the position of the building, and the height of the building. The program according to claim 4.

6. The location of the structure is identified based on the locations of the plurality of receivers and the azimuth angles from the plurality of receivers to the outline of the structure. The program according to claim 5.

7. The height of the structure is determined based on the position of the structure and the elevation angles from the multiple receivers to the outline of the structure. The program according to claim 5.

8. The receiver is mounted on a moving object.

3. The program according to claim 1 or 2.

9. The receiver acquires observation quantity information based on the observation quantity of the signal received from the artificial satellite, The acquired observation quantity information is input to a first learning model that is trained to output information regarding the presence or absence of a structure between the receiver and the artificial satellite when the observation quantity information is input, thereby outputting information regarding the presence or absence of a structure. Information processing methods.

10. The receiver acquires observation quantity information based on the observation quantity of the signal received from the artificial satellite, The acquired observation quantity information is input to a first learning model that is trained to output information regarding the presence or absence of a structure between the receiver and the artificial satellite when the observation quantity information is input, thereby outputting information regarding the presence or absence of a structure. Processing section An information processing device comprising:

Images