Program, information processing apparatus, information processing method, and learning model

Abstract

To generate data at high frequency which is obtained by observing an observation area by using a plurality of flying objects.SOLUTION: A program causes a computer to execute, on the basis of first observation data that is information on an observation area at first timing and that is observed by first observation means, and second observation data that is information on the observation area at second timing different from the first timing and that is observed by second observation means different from the first observation means, generating integrated data including first information based on the first observation data and second information based on the second observation data.SELECTED DRAWING: Figure 2

Description

[Technical field]

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

[0002] Using flying objects such as satellites, aircraft, and drones, both on land and at sea Observation of the surface condition of the sphere is widely used. Synthetic Aperture Radar (SAR) is an observation method that uses a The system acquires radar images, or SAR images, obtained using SAR (Signal Acquisition and Reconstruction) technology. There are various observation methods.

[0003] In Patent Document 1, in order to increase the frequency of grasping the situation on the Earth's surface, A new image obtained by photographing a predetermined observation point on the Earth's surface by a flying first flying object; and A second vehicle flying on a second orbit different from the first orbit, the same as a predetermined observation point By aligning and comparing the image with a previous image taken at the same location, changes can be detected. A method for detecting regions of Earth's surface situational awareness is described. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2022-167343 A Summary of the Invention [Problem to be solved by the invention]

[0005] In the method described in Patent Document 1, a new image observed from a first orbit is obtained by The new image is treated as a different image from the previously observed image. New images observed by the first vehicle from one orbit and not the second vehicle's observations. The data does not take into account the observation of the observation point. The data itself is not generated frequently, only the frequency of change detection has been improved. When conducting observations using satellites, the observation data is used to obtain a more detailed understanding of the state of the observed area. It is preferable that the production of is carried out at a high frequency.

[0006] Therefore, the present invention is a method for collecting data obtained by observing an observation area using multiple flying objects at high frequency. The purpose is to generate [Means for solving the problem]

[0007] A program according to an aspect of the present invention causes a computer to The first observation data is information on a region, the first observation data being observed by a first observation means, and a first timing. is information on the observation area at a different second timing, and is a second observation means different from the first observation means. Based on the second observation data observed by the measurement means, first information based on the first observation data is obtained. generating integrated data including the first information based on the first observation data and the second information based on the second observation data. .

[0008] According to this aspect, the integrated data includes first information based on the first observation data and second observation data. and second information based on the first observation data or the second observation data. When either one is updated or acquired, the integrated data is also updated. The computer is the observation area at either the first timing or the second timing. By generating integrated data based on regional information, it is possible to generate integrated data frequently. It becomes Noh.

[0009] In the above aspect, the first observation data is data observed under a first observation condition, The computer receives information on the observation area at a third timing different from the first timing. and data observed under third observation conditions by a third observation means different from the first observation means. and obtaining third observation data observed based on a common observation method to the first observation data. The learning observation data, which is information in the learning observation domain, and the learning perspective corresponding to the learning observation data are The measurement conditions and the learning observation timing at which the learning observation data was observed are input, and the learning observation data is converted into converted observation data, and the learning observation region is the first The transformation model outputs transformed observation data that represents the data observed under the observation conditions. The third observation data, the third observation condition, and the third timing are input, and the third observation data is converted. Obtaining the observation data and further executing the process to generate the integrated data is the first information and updating the information based on the transformed observation data to generate integrated data. .

[0010] According to this aspect, when there is third observation data in a common observation method with the first observation data, In addition, the third observation data can be converted into data observed under the first observation conditions. For example, when performing an analysis using integrated data, it is necessary to compare data observed under different observation conditions. The data can be treated as first information, since it was observed under the first observation conditions. As a result, the first information in the integrated data is updated frequently, and as a result, the integrated data is also updated frequently. It is possible to generate

[0011] In the above aspect, the first observation condition is first aircraft information of the first flying object which is the first observation means; and a first environmental information indicating an observation environment by the first flying object. The third observation condition includes: and third environment information indicating an observation environment by the third flying object. may include:

[0012] According to this aspect, data conversion is possible taking into account the aircraft information and the observation environment. This makes it possible to convert the third observation data into observation data under the first observation conditions with high accuracy. It becomes.

[0013] In the above aspect, the first observation data is data observed under a first observation condition. and acquiring the first observation data and the first observation condition in the computer, and The information is information on the observation area at a third timing different from the first observation means. Data observed under the third observation condition by the third observation means, and common to the first observation data and acquiring third observation data observed based on the observation method of the learning observation area. The learning observation data is, the learning observation conditions corresponding to the learning observation data, and the learning observation data is The observed learning observation timing and the converted observation data are input. The data is obtained by observing the observation area under standard observation conditions at the learning observation timing. The first observation data, the first observation condition, and the second observation condition are input to a conversion model that outputs converted observation data that indicates the first observation data. The first observation data is converted into converted observation data, and the converted observation data is input as the first converted observation data. The third observation data, the third observation condition, and the third timing are added to the conversion model. The third observation data at the third timing is converted into converted observation data. and obtaining second converted observation data. That is, the first information based on the first converted observation data is converted into the first information based on the second converted observation data. to generate integrated data.

[0014] According to this aspect, the first observation data and the third observation data observed under a common observation method are Each of the data is converted to converted observation data observed under the common reference observation conditions. By generating integrated data based on the converted observation data, This reduces the differences between each piece of data, and improves information processing during the analysis of integrated data. In addition, the first information in the integrated data is updated frequently, resulting in It will also be possible to generate integrated data at high frequency.

[0015] In the above aspect, the first observation condition is first aircraft information of the first flying object which is the first observation means; and a first environmental information indicating an observation environment by the first flying object. The third observation condition includes: Second aircraft information of a third flying object, which is a stage of the observation, and second environmental information indicating an observation environment by the third flying object. may include:

[0016] According to this aspect, data conversion is possible taking into account the aircraft information and the observation environment. This makes it possible to convert the second observation data into observation data under the first observation conditions with high accuracy. It becomes.

[0017] In the above aspect, the first observation data is data observed based on a first observation method. The second observation data is data observed based on a second observation method that is different from the first observation method. It may also be a data.

[0018] According to this aspect, data observed based on different observation methods are generated as integrated data. For example, the first observation method is SAR observation, and the second observation method is optical imaging. This allows the observation of multiple observation methods. Data can be integrated into one data set, allowing for more information to be obtained through analysis or for higher accuracy. This makes it possible to analyze the degree of

[0019] In the above aspect, the computer is configured to execute the program at a third timing different from the first timing. The information is obtained by a third observation method different from the first observation method under a third observation condition. The second observation method is different from the first observation method. 3. Obtaining observation data, learning observation data, and the learning perspective in which the learning observation data was observed The observation timing is input, and the observation area is determined based on the first observation method at the learning observation timing. The estimation model outputs the estimated observation data that shows the observed data in the region, and the third observation data and , and a third timing are input to obtain estimated observation data at the third timing. and generating integrated data by further executing the first information based on the inferred observation data. and updating the data to generate the consolidated data.

[0020] According to this aspect, the observation data at the third timing based on the first observation method is It can be estimated from the third observation data at the third timing based on the observation method. This allows, for example, the first observation data to be obtained when the first observation data cannot be obtained due to factors such as bad weather. However, the data observed based on the first observation method is supplemented with the third observation data. It is possible to generate integrated data including the above at high frequency.

[0021] In the above aspect, the computer is configured to acquire first observation data and learn observation data. The timing at which the learning observation data was observed differs from the timing at which the learning observation data was observed. The timing at which the observation area is observed is input, and the observation area is obtained based on the second observation method at the different timing. The first observation data is input to an inference model that outputs inferred observation data indicating the observed data. Input the first timing and the second timing, and estimate the observation data at the second timing. and further executing to generate integrated data at a second timing. generating integrated data including second information based on the inferred observational data at stomach.

[0022] According to this aspect, the observation data at the second timing based on the second observation method is It can be estimated from the first observation data at the first timing based on the observation method. This allows, for example, the second observation data to be obtained when the second observation data cannot be obtained due to factors such as bad weather. However, the first observation data can be used to complement the first observation data to generate integrated data.

[0023] In the above aspect, the computer is configured to detect the observation area at a third timing before the first timing. Obtaining the third observation data when the area is observed, and the learning observation data and the learning observation data The learning observation timing at which the data was observed and the prediction timing after the learning observation timing are Data that is input and observed in the observation area based on the first observation method at the predicted timing The estimation model outputs the estimated observation data showing the third observation data, the third timing, and the first inputting a timing and acquiring estimated observation data at the first timing; The further execution of the simulation and generation of integrated data will be in accordance with the estimated observation data at the first timing. generating integrated data including the first information based on the first information.

[0024] According to this aspect, the observation data at the first timing based on the first observation method is It can be estimated from the third observation data at the third timing based on the observation method. This allows, for example, the first observation data to be obtained when the first observation data cannot be obtained due to factors such as bad weather. However, the data was obtained from the third observation data, which was observed at the third timing before the first timing. They can be complemented to generate integrated data.

[0025] In the above aspect, the computer is configured to execute a learning observation process based on the learning observation data, which is information on the learning observation area. The learning integration data is input to the meta-information output model, which outputs meta-information of the learning observation area. , inputting the integrated data and obtaining meta information of the observation area. .

[0026] According to this aspect, the integrated data generated at high frequency based on the observation of the observation area is used. , meta-information of the observation area can be obtained. This improves the accuracy of the meta-information of the observation area. can be increased.

[0027] According to another aspect of the present invention, there is provided an information processing device for detecting an object of an observation region at a first timing. The first observation data observed by the first observation means is different from the first timing. This is information on the observation area at two different times, and is obtained by a second observation method different from the first observation method. and second observation data observed by the first observation data. and an integrated data generation unit that generates integrated data including the measurement data and second information based on the measurement data.

[0028] According to another aspect of the present invention, there is provided an information processing method, comprising: Information on an observation area, including first observation data observed by a first observation means and a first timing The second timing is different from the first observation method. based on the first observation data and second observation data observed by the second observation means generating integrated data including the first information and second information based on the second observation data. .

[0029] A learning model according to another aspect of the present invention includes learning observation data, which is information on a learning observation domain, and The training observation conditions corresponding to the training observation data are input, and the time when the training observation data was obtained is In the learning observation data, the learning observation area is collected as the learning observation data according to the standard observation conditions. The first transformed observation data is used as the output for training data, and the second transformed observation data is used as the output for training data. The observation data, which is the information of the observation area, and the observation conditions corresponding to the observation data are input, and the observation The data is obtained when the target observation area is observed under standard observation conditions. and causing the computer to output second transformed observation data obtained by transforming the observation data. Make it work.

[0030] In a learning model according to another aspect of the present invention, a learning observation domain is observed based on a first observation method. The learning observation data when the learning observation data was observed, the observation timing, and the observation time The input is a prediction timing different from the actual timing, and the learning observation area is A first guess observation indicates data observed based on a second observation method that is different from the first observation method. The training data is used to output the observation data, which is information on the target observation area. The timing at which the observation data was acquired in response to the input of the observation conditions corresponding to the observation data and A second estimated observation data showing data observed in the target observation area based on the second observation method in The computer is then caused to output the data. Effect of the Invention

[0031] According to the present invention, data obtained by observing an observation area using multiple flying objects is collected at high frequency. It becomes possible to generate. [Brief description of the drawings]

[0032] [Figure 1] FIG. 1 is a schematic diagram of an observation system 10. [Diagram 2] FIG. 2 is a diagram illustrating generation of integrated data according to the first embodiment. [Diagram 3] FIG. 1 is a block diagram of an observation data processing device 101 according to a first embodiment. [Figure 4] FIG. 2 is a diagram illustrating a meta information output model according to the first embodiment. [Diagram 5] 11 is a flowchart of an integrated data generation process according to the first embodiment. [Figure 6] 11 is a flowchart of a meta information acquisition process according to the first embodiment. [Figure 7] FIG. 11 is a block diagram of an observation data processing device 101A according to a second embodiment. [Figure 8] FIG. 11 is a diagram illustrating a first conversion model according to the second embodiment. [Figure 9] 13 is an example of a flowchart of an integrated data generation process according to the second embodiment. [Figure 10]FIG. 11 is a diagram illustrating an example of generation of integrated data according to the second embodiment. [Figure 11] FIG. 11 is a diagram illustrating a second conversion model according to the second embodiment. [Figure 12] 13 is another example of a flowchart of the integrated data generation process according to the second embodiment. [Figure 13] FIG. 11 is a diagram illustrating another example of generation of integrated data according to the second embodiment. [Figure 14] FIG. 11 is a block diagram of an observation data processing device 101B according to a third embodiment. [Figure 15] FIG. 13 is a diagram illustrating a first guess model according to the third embodiment. [Figure 16] 13 is an example of a flowchart of an integrated data generation process according to the third embodiment. [Figure 17] FIG. 13 is a diagram illustrating generation of integrated data according to the third embodiment. [Figure 18] FIG. 13 is a diagram illustrating a second estimation model according to the third embodiment. [Figure 19] 13 is another example of a flowchart of the integrated data generation process according to the third embodiment. [Figure 20] FIG. 13 is a diagram illustrating a third estimation model according to the third embodiment. [Figure 21] 13 is another example of a flowchart of the integrated data generation process according to the third embodiment. [Figure 22] FIG. 13 is a diagram illustrating generation of integrated data according to the third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0033] A preferred embodiment of the present invention will be described with reference to the accompanying drawings. , components with the same reference numerals have the same or similar configurations.

[0034] (First embodiment) FIG. 1 shows an observation system 10 according to a first embodiment. The same applies to the first and second embodiments.

[0035] The observation system 10 includes an observation data processing device 101 and artificial satellites 201, 202, and 203. , and an observation data receiving device R.

[0036] In the observation system 10, a plurality of satellites including artificial satellites 201, 202, and 203 are used to observe the earth. The observation data obtained by observing the spherical surface O is transmitted to the observation data receiving device R. Observation conditions, which indicate the environment in which the data was observed, are associated with or attached to the observation data. The observation data may be transmitted to the observation data receiving device R in such a manner that the observation data is included in the observation data processing device 10. 1 receives observation data from the observation data receiving device R and performs information processing on the observation data. The observation system 10 is configured to observe the observation data from the satellites 201 and 202. As shown in Figure 02,203, it will be deployed in space as a satellite capable of acquiring observation data. The satellites that collect the data are geostationary satellites. The flying object may be an aircraft, a helicopter, a drone, or the like. Any device capable of being positioned above the Earth's atmosphere may be used.

[0037] In the observation system 10, the observation data observed by the artificial satellites 201, 202, and 203 are The data may be observed based on different observation methods, or may be observed based on a common observation method. The data may be data observed at a given time.

[0038] The observation method is, for example, a method using SAR signals (SAR method), and the observation data is Microwaves (electromagnetic waves) are irradiated to the object of observation from a satellite equipped with a radar device. The signal corresponds to the electromagnetic waves reflected by the object being measured. The SAR data is, for example, processed by the observation data processing device 101 to become a SAR image. The following visualization process is performed on the observed SAR data. There is also a second level of SAR data that has been range-compressed and multi-look azimuth-compressed. The second level of SAR data provides a geometrically corrected visualization of the SAR image. Other SAR data can be obtained by applying range pressure to the observed SAR data. There is also a third level of SAR data that is compressed, single-look azimuth compressed and orthorectified. By adding orthorectification, the SAR image can be overlaid with the optical image described below. can be obtained.

[0039] Other observation methods include those that use optical data obtained using optical sensors (optical methods). Optical data is obtained by sensors that observe light with wavelengths in the visible light range or near infrared range. Optical data is based on the light produced when sunlight is reflected off the Earth's surface. This is simple data.

[0040] Another observation method is to use meteorological data obtained using an infrared sensor. Meteorological data is obtained by observing infrared radiation emitted by clouds, the earth's surface, and the atmosphere. This is used to generate an infrared image or a water vapor image. The optical data and meteorological data are explained separately according to the difference in the wavelength ranges used. There are.

[0041] The generation of integrated data by the observation data processing device 101 will be described with reference to FIG. In FIG. 2, each of the artificial satellites 201, 202, and 203 has a different observation method. The following describes an example in which observation is performed based on the method 1, method 2, or method 3. For example, method 1 uses SAR data, and method 2 uses optical data. Method 1 is a method using meteorological data, and method 2 is a method using meteorological data.

[0042] The artificial satellite 201 observes an observation area at a certain time T1 on the time axis T and collects observation data. The observation data D11 is transmitted to the observation data processing device 101, and The data processing device 101 generates integrated data ID1 based on the observation data D11.

[0043] After that, the artificial satellite 202 observes the observation data D22 at a certain time T2 on the time axis T. The observation data D22 is transmitted to the observation data processing device 101. The device 101 is integrated data based on previously observed observation data D11 and observation data D22. Generate ID2.

[0044] The integrated data ID2 is the observation data based on the method 1 at time T1 and the The data is based on the observation data based on the method 2 and the most recent observation data in the observation area at time T2. This data contains valuable information.

[0045] The artificial satellite 203 observes observation data D33 at a certain time T3 on the time axis T. The observation data D33 is transmitted to the observation data processing device 101. is an integrated data set based on the previously observed data D11, D22, and D33. The integrated data ID3 is the observation data based on the method 1 at time T1. , the observation data based on method 2 at time T2, and the observation data based on method 3 at time T3 The data is based on observation data and contains the most recent information on the observation area at time T3. It is.

[0046] After that, at time T4, the artificial satellite 201 observes the same area as that observed at time T1. The satellite 201 observes the observation area and obtains observation data D14. The time from one observation to the next is determined, for example, by the orbit of the artificial satellite 201. The observation data D14 is transmitted to the observation data processing device 101. 1 generates integrated data ID4 based on observation data D14. Observation data processing device 101 The observation data D11 obtained by the observation by the artificial satellite 201 at the time T1 is newly The obtained observation data D14 is replaced with the integrated data ID 14. 4 is data containing the most recent information of the observation area at time T4.

[0047] Similarly, at time T5, the satellite 202 observes the same area as it observed at time T2. The observation data D25 is obtained by the observation data processing device 10. 1, and the observation data processing device 101 processes the integrated data ID5 based on the observation data D25. The observation data processing device 101 generates the observation The observation data D22 obtained in the previous step was replaced with the newly obtained observation data D25 to create the integrated data. The integrated data ID5 is the most recent information of the observation area at time T5. By integrating the data at each time, the observation data processing device 10 1 is a method for comparing data obtained by observing an observation area using multiple aircraft with that obtained by a single aircraft. It can be generated more frequently than observed.

[0048] The information included in the integrated data is the time when the observation data on which each piece of information is based was observed. The information in the consolidated data is related and can be sorted based on time. For example, in integrated data ID5, the observation observed at time T3 is Data D33, observation data D14 observed at time T4, and observation data D14 observed at time T5 It is possible to order the data D25. Note that the observation data does not include time information. The integrated data may be sorted in the order of acquisition. The observation data is generated in a way that allows for sorting in chronological order. By using the integrated data, it is possible to detect changes in the observation area more accurately. In addition, the integrated data does not necessarily mean that the observation data can be rearranged in chronological order. The format of the integrated data may be determined appropriately depending on the purpose of the integrated data. Can be set.

[0049] Here, the data to be integrated may be the signals themselves from each satellite, or may be the data based on each signal. The image data may be generated based on the image data. In this case, each image data is input to the meta-information output model as a channel. The data to be integrated is obtained from signals or image data from each satellite. In addition, the signals, image data, or meta information from each satellite may be, for example, For example, the simulation process is carried out by an information processing device based on observation data. In other words, the data included in the integrated data may be information generated by observation data. Any information based on the above will suffice.

[0050] The observation data processing device 101 generates integrated data each time it acquires observation data from each satellite. This will update the integrated data to include the latest observations for the observation area. For example, if a change occurs in the observation area between time T2 and time T5, the satellite 20 If only 2 is used, the change will not appear in the observation data until time T5. However, Observation data D33 from satellite 203 at time T3 and observation data D34 from satellite 201 at time T4 From the observation data D14, the integrated data is generated as integrated data ID3, ID4. The integrated data ID3 and ID4 contain the data that occurred in the observation area between time T2 and time T5. The changes occurring in the observation area are also reflected in the integrated data ID3 and ID4. By analyzing integrated data based on observation data from multiple satellites, It will be possible to grasp changes that occur within the observation interval of a single satellite in a timely manner. An event that prevents the satellite 202 from observing the observation area at time T5, such as bad weather Even if an event occurs, the data will be collected using the SAR method, which is not affected by weather. By analyzing the generated integrated data, it becomes possible to understand changes in the observation area.

[0051] The observation data processing device 101 updates data on the observation area at shorter time intervals. This allows the analysis to be applied to tasks that require real-time performance. One example of the task is to measure NDVI, a vegetation index, using satellites, etc. (Normalized Difference Vegetation Index) is measured. While VI measurements can be made accurately, optical data are limited to daytime observations and are subject to cloud cover. There are restrictions on time and weather, such as not being able to observe due to weather conditions. By measuring NDVI using integrated data including R data, the number of measurements can be increased, This can improve real-time performance.

[0052] In addition, since the integrated data is data that integrates information from each satellite, for example, When using a satellite to obtain meta-information and analyze the observation area, the information from each satellite is learned individually. This enables more frequent and multifaceted analysis than when inputting the data into a learning model.

[0053] In addition, observation systems that use data from a single satellite may be converted to systems that use data from other satellites. It is also possible to use the system to generate integrated data and enable analysis. become.

[0054] Each unit of the observation data processing device 101 will be described. A block diagram of a data processing device 101 is shown. The observation data processing device 101 includes a communication unit 1 011, a memory unit 1012, an observation data acquisition unit 1013, an integrated data generation unit 1014, and The observation data processing device 101 includes a meta information acquisition unit 1015. Each part of the device 101 is stored in a storage device in an information processing device such as a personal computer. This can be realized by a processor executing the program.

[0055] The communication unit 1011 communicates with the observation data processing device 101 and external devices including the observation data receiving device R. The observation data processing device 101 controls communication between the observation data processing device and the information processing device. The communication unit 1011 may acquire the observation data from an information processing device such as a server that records the , and controls communications with other information processing terminals.

[0056] The memory unit 1012 stores various information used in processing by the observation data processing device 101. The storage unit 1012 stores a meta information output model 10121.

[0057] The meta information output model 10121 will be described with reference to Fig. 4. 10121 is a system that inputs the integrated data obtained from observations of the observation area, and then converts the meta information of the observation area into the It is a learning model that outputs information.

[0058] The meta information output model 10121 is trained using the training data LD1 as training data. The learning data LD1 is based on multiple learning observation data obtained by observing the observation area. It includes a set of integrated learning data and meta-information MD0 corresponding to the integrated learning data. The learning model 10212 is trained using the integrated learning data as input and the meta-information MD0 as output. do.

[0059] Meta information MD0 has various items depending on the object of observation. The meta information for the moving object includes information on the moving path of the moving object. The change in SAR data and the shift in optical images due to the change in the interference of reflected electromagnetic waves caused by the It is obtained from the change in position of a moving object.

[0060] If flooded areas during a disaster are the subject of observation, the meta-information may be information on the extent of the flooded areas. In addition, as meta-information for managing agricultural crops, NDVI, which indicates the growth rate of crops, etc. The degree of crop growth can be calculated based on the backscatter coefficient of SAR data and the actual observation data. Correlations calculated based on the growth of the crops and color changes in optical data The crop temperature is estimated based on the spectral reflectance characteristics. In this case, the meta-information includes the actual measured crop temperature. Height information may also be included.

[0061] In the detection of buildings, information about new buildings may be used as meta information. Information on new buildings is based on the backscatter coefficients of SAR data and the actual observed new buildings. Based on correlations calculated based on information about the building and color changes in optical data, In addition, information on the type of building is obtained based on maps, etc., and is used as meta-information. This may be reported as information.

[0062] The integrated data generated by the observation data processing device 101 is input to the trained meta-information output model 1 When input to 0121, the meta information output model 10121 corresponds to the integrated data. The meta-information output model 10121 outputs the meta-information MD1 obtained by processing the observation data. It does not have to be stored in the device 101, but may be stored in another information processing device.

[0063] Returning to FIG. 2, the observation data acquisition unit 1013 acquires the observation data ( Obtain the first observation data and the observation conditions (first observation conditions) for the observation of the artificial satellite 201. The observation data acquisition unit 1013 receives, for example, observation data and observation Observation conditions may be acquired, and actual observation data or observation data generated by simulation may be used. Observation data and observation conditions may be obtained from the server where the data is recorded. When generating the data, the observation data processing device 101 does not acquire the observation conditions, but The integrated data may be generated based on the observation data from the artificial satellite 202 (second observation data). , observation conditions (second observation conditions) for observing the artificial satellite 202, and the artificial satellite 203 The same applies to the observation data and observation conditions from the satellite.

[0064] The observation conditions are information based on the characteristics of the aircraft, and the observation environment of the aircraft. The aircraft information includes the frequency of the light used by the aircraft for observation, the polarization of the light, and the environmental information. The information includes waves, pulse information, and beam patterns. If a constellation is formed, the constellation identifier and the constellation number of each vehicle The aircraft information may include the aircraft number or identifier within the aircraft.

[0065] The environmental information includes the orbit direction at the time of observation (northbound ascending or southbound descending), Angle, observation direction (left or right), aircraft speed, aircraft attitude information, and receiving system parameters. If the observation data is image data, the environmental information includes information on the resolution of the image. If the observation data is SAR image data, the environmental information may include the Information on the angle of incidence of the illumination signal may be included for each pixel. When removing noise, the environment information includes noise information, the type of noise filter, and the noise removal method. The following parameters may be included:

[0066] The integrated data generating unit 1014 performs a process of generating integrated data. 014 is, for example, information on the observation area at time T1, observed by the artificial satellite 201. The observation data D11 is information on the observation area at time T2, and is sent to the artificial satellite 202. Based on the observation data D22 thus observed, first information based on the observation data D11 The integrated data ID1 is generated based on the second information based on the observation data D22.

[0067] The meta information acquisition unit 1015 inputs the integrated data to the meta information output model 10121, Obtain meta information of the observation area. The obtained meta information is used by the observation data processing device 101. The observation data may be presented to a user of the observation data processing device 101 by a display unit that displays the observation data. The information may be transmitted from the processing device 101 to another information processing device and presented to the user.

[0068] The integrated data generation process will be described with reference to FIG. 5. In step S501, The observation data acquisition unit 1013 acquires first observation data at a first timing. For example, observed data acquisition unit 1013 acquires observed data D11 at time T1.

[0069] In step S502, the observation data acquisition unit 1013 acquires the first For example, the observation data acquisition unit 1013 acquires the observation data at time T2. Get data D22.

[0070] In step S503, the integrated data generating unit 1014 generates a first integrated data based on the first observation data. Here, the first information is, for example, information generated based on the first observation data. and meta information obtained from the image data or signals from the artificial satellite 201. In other words, the data included in the integrated data should be information based on observation data. The first information may be, for example, the observation data from the artificial satellite 201 itself.

[0071] In step S504, the integrated data generating unit 1014 generates a second Two pieces of information are acquired. The second pieces of information are the same as the first pieces of information.

[0072] In step S505, the integrated data generating unit 1014 generates a first information and a second information including the first information and the second information. At this time, the integrated data generating unit 1014 generates the integrated data. The integrated data may be stored in the storage unit 1012. The data may be transmitted to an external information processing device.

[0073] The integrated data generating unit 1014 may generate integrated data each time observation data is acquired. For example, the integrated data generating unit 1014 may determine that the satellite 201 has an observation area at time T4. Observation data D14 obtained by observation is an observation data obtained at time T1 before time T4. Integrated data may be generated to replace data D11.

[0074] The meta information acquisition process based on the integrated data will be described with reference to FIG. In 601, the meta information acquisition unit 1015 acquires the integrated data generated by the integrated data generation unit 1014. The meta information acquisition unit 1015 acquires data, for example, from the integrated data stored in the storage unit 1012. In step S602, the meta information is acquired. The unit 1015 inputs the integrated data to the meta information output model 10121. In FIG. 3, the meta-information acquisition unit 1015 acquires the meta-information of the observation area from the meta-information output model 10121. Obtain data information.

[0075] By the above processing, the observation data processing device 101 can obtain the observation area using multiple flying objects. By generating data obtained by observing the area at high frequency, meta-information of the observation area can be obtained. do.

[0076] Second embodiment A second embodiment will be described. The second and subsequent embodiments are common to the first embodiment. The second embodiment is a conversion method. The observation data is transformed using the model, and the integrated data is generated based on the transformed data. This embodiment differs from the first embodiment in that the image is generated.

[0077] FIG. 7 shows a block diagram of an observation data processing device 101A according to the second embodiment. The observation data processing device 101A includes a first conversion module in addition to the components of the observation data processing device 101. A storage unit 1012A having a second transformation model 10122, a second transformation model 10123, and a transformation observation data The data acquisition unit 1016 is also included.

[0078] The first conversion model 10122 will be described with reference to FIG. 2 is the third observation data obtained by observing the observation area, the third observation conditions, and the third observation timing. This is a learning model that outputs transformed observation data for input of a training set.

[0079] The first conversion model 10122 is trained using the training data LD2 as training data. The learning data LD2 includes the learning observation data obtained by observing the learning observation area, The conditions, the learning observation timing, and the corresponding converted observation data are included. The converted observation data is data observed under the first observation condition and the observation timing is the observation timing for learning. The first conversion model 10122 is The observation conditions and the learning observation timing are input, and the converted observation data is output. .

[0080] The first conversion model 10122 converts observation data at a certain timing into This is a model that generates observation data obtained by changing the first transformation model 10122. Observation data under different observation conditions is classified into observation data under a certain observation condition (e.g., first observation condition). The data is then converted into integrated data.

[0081] The converted observation data acquisition unit 1016 acquires a first conversion model 10122 and a second conversion model 10123. Using 10123, the converted observation data obtained by converting the observation data is obtained.

[0082] 9 and 10, the first conversion model 1 in the observation data processing device 101A is The process of generating integrated data using the satellite 201 and the satellite 0122 will be described. 202 and 203 are used to acquire observation data based on a common observation method (e.g., SAR method). In addition, the observation system 10 is different from the artificial satellites 201, 202, and 203. An artificial satellite 204 that acquires observation data based on an observation method (for example, an optical method) is further included. include.

[0083] In step S901, the observation data acquisition unit 1013 acquires the first For example, the observation data acquisition unit 1013 acquires observation data at time T1. Get data D11.

[0084] In step S902, the observation data acquisition unit 1013 acquires the first For example, the observation data acquisition unit 1013 acquires the observation data at time T2. Get data D42.

[0085] In step S903, the observation data acquisition unit 1013 acquires the third For example, the observation data acquisition unit 1013 acquires the observation data at time T3. Get data D23.

[0086] In step S904, the converted observation data acquisition unit 1016 performs For example, the converted observation data acquisition unit 1016 acquires a third observation condition for the observation data D2 More specifically, the observation conditions for the satellite 202 and the observation data are acquired. The environmental information of the observation environment of the data D23 is acquired as the observation conditions.

[0087] In step S905, the converted observation data acquisition unit 1016 converts the first converted model 101 22, the third observation data, the third observation condition, and the third timing are input. For example, The observation data acquisition unit 1016 inputs the observation data D23 and the observation data The observation conditions for D23 and time T3 are input.

[0088] In step S906, the converted observation data acquisition unit 1016 converts the third observation data For example, the converted observation data acquisition unit 1016 acquires the converted observation data. The converted observation data D13 obtained by converting the observation data D23 is obtained from the first conversion model 10122.

[0089] In step S907, the integrated data generation unit 1014 converts the first information into converted observation data and generating integrated data including the first information and the second information. For example, the integrated data generating unit 1014 may generate an integrated data set of observed data D11 and observed data D42. Next, the integrated data generating unit 1014 generates the integrated data ID2 based on the observation data D 11 generates integrated data ID3 updated by the converted observation data D13.

[0090] Similarly, the integrated data generation unit 1014 converts the observation data D34 into converted observation data The integrated data generating unit 1014 generates integrated data ID4 based on the integrated data D14. In the data, the converted observation data D14 is the integrated data updated by the observation data D15. Generate data ID5.

[0091] As a result, for example, between the artificial satellite 201 and the artificial satellites 202 and 203 Used to generate integrated data even when there are differences in aircraft characteristics or observation environments Observation data can be collected under the observation conditions of satellite 201. This allows for the generation of integrated data with reduced differences between observation methods. It will be possible to generate data observed under the observation conditions of satellite 201 at high frequency. As a result, it becomes possible to update the integrated data frequently. By using this, meta-information of the observation area can be obtained more accurately.

[0092] The second conversion model 10123 will be described with reference to FIG. 123 is a first observation data obtained by observing an observation area, a first observation condition, and a first observation time. This is a learning model that outputs transformed observation data for timing input.

[0093] The second conversion model 10123 is trained using the training data LD3 as training data. The training data LD3 includes training observation data obtained by observing the training observation area, The conditions, the learning observation timing, and the corresponding converted observation data are included. The converted observation data is data observed under standard observation conditions and the observation timing is the learning observation timing. The learning model 10212 is The observation conditions and the learning observation timing are input, and the converted observation data is output. Here, the reference observation conditions are different from the observation conditions of each of the artificial satellites 201, 202, and 203. The reference observation conditions are conditions that are determined in advance by the user.

[0094] The second conversion model 10123 converts observation data at a certain timing into This is a model that generates observation data obtained by converting the Observation data with different observation conditions are converted to observation data based on the standard observation conditions, and the integrated data is The data is generated.

[0095] 12 and 13, in the observation data processing device 101A, the second conversion model The process of generating integrated data using the satellite 201 will be described. ,202,203 acquire observation data based on a common observation method (e.g., SAR method) In addition, the observation system 10 is different from the artificial satellites 201, 202, and 203. and an artificial satellite 204 that acquires observation data based on an observation method (e.g., an optical method) according to the present invention. Included.

[0096] In step S1201, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the observation data at time T1. The measurement data D11 is acquired.

[0097] In step S1202, the converted observation data acquisition unit 1016 converts the first observation data For example, the converted observation data acquisition unit 1016 acquires a first observation condition for the observation data D More specifically, the satellite 201's aircraft information and observation Environmental information of the environment is obtained as the observation condition.

[0098] In step S1203, the converted observation data acquisition unit 1016 acquires the second converted model 10 The first observation data, the first observation condition, and the first timing are input to 123. For example, The observation data acquisition unit 1016 inputs the observation data D11 and the observation data D12 into the second conversion model 10123. The observation conditions for data D11 and time T1 are input.

[0099] In step S1204, the converted observation data acquisition unit 1016 converts the first observation data For example, the converted observation data acquisition unit 1016 acquires the first converted observation data. Obtain the converted observation data D51, which is the result of converting the observation data D11, from the first conversion model 10122. do.

[0100] In step S1205, the observation data acquisition unit 1013 For example, the observation data acquisition unit 1013 acquires the second observation data at time T2. The measurement data D42 is acquired.

[0101] In step S1206, the integrated data generating unit 1014 generates a first converted observation data based on the first converted observation data. and generating integrated data based on the first information based on the second observation data. For example, the integrated data generation unit 1014 generates an integrated data based on the converted observation data D51 and the observation data D42. The combined data ID2 is generated.

[0102] In step S1207, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the third observation data at time T3. The measurement data D23 is acquired.

[0103] In step S1208, the converted observation data acquisition unit 1016 converts the third observation data For example, the converted observation data acquisition unit 1016 acquires a third observation condition for the observation data D More specifically, the satellite 201's aircraft information and observation Environmental information of the environment is obtained as the observation condition.

[0104] In step S1209, the converted observation data acquisition unit 1016 converts the second converted model 10 The third observation data, the third observation condition, and the third timing are input to 123. For example, The observation data acquisition unit 1016 inputs the observation data D23 and the observation data The observation conditions for data D23 and time T3 are input.

[0105] In step S1210, the converted observation data acquisition unit 1016 converts the third observation data For example, the converted observation data acquisition unit 1016 acquires the converted third converted observation data. Obtain the converted observation data D53, which is the result of converting the observation data D23, from the first conversion model 10122. do.

[0106] In step S1211, the integrated data generating unit 1014 generates a first converted observation data based on the first converted observation data. The first information based on the second converted observation data is updated by the first information based on the second converted observation data, and the first information and the second information are For example, the integrated data generating unit 1014 generates integrated data including observation data, D51 generates integrated data ID3 updated by the converted observation data D53.

[0107] Similarly, the integrated data generation unit 1014 converts the observation data D34 into converted observation data The integrated data generator 1014 generates integrated data ID4 based on the observation Integrated data ID5 is generated based on converted observation data D55 obtained by converting data D15.

[0108] As a result, for example, between the artificial satellite 201 and the artificial satellites 202 and 203 Used to generate integrated data even when there are differences in aircraft characteristics or observation environments It is possible to collect observation data as data observed under specified observation conditions. This produces integrated data with reduced differences between observation methods. By using this data, meta-information of the observation area can be obtained more accurately.

[0109] Third embodiment A third embodiment will be described. The third embodiment is a method for detecting the presence or absence of an observed data by using an estimation model. The first embodiment and the second embodiment are similar in that an inference is made and integrated data is generated based on the inferred data. Here, the estimation of observed data is different from the first embodiment in that the observed data is It means inferring observed data from other observed data.

[0110] FIG. 14 shows a block diagram of an observation data processing device 101B according to the third embodiment. The observation data processing device 101B includes, in addition to the components of the observation data processing device 101, a first estimation model 10124, second guess model 10125, and third guess model 10126. The estimated observation data acquisition unit 1017 includes a memory unit 1012B and an estimated observation data acquisition unit 1017. 017 is a first guess model 10124, a second guess model 10125, or a third guess model 10126, which will be described later. Using the model 10126, the observed data is estimated to obtain inferred observed data.

[0111] The first guess model 10124 will be described with reference to FIG. 24 is for inputting the first observation data and the second observation timing obtained by observing the observation area. The input first observation data is the first observation The data is based on a satellite observation method (e.g., SAR method) and the output is an inferred observation. The data is obtained based on a second observation method (e.g., optical method) that is different from the first observation method. This is the data collected.

[0112] The first guess model 10124 is trained using the training data LD4 as training data. The learning data LD4 includes the learning observation data and learning observation data obtained by observing the learning observation area. The set of measurement timing A and the corresponding estimated observation data is included. , the data was observed based on the first observation method, and the observation timing was timing A The first guess model 10124 inputs the learning observation data and the learning observation timing. The model is trained using the input and output of inferred observation data.

[0113] The first guess model 10124 is a model of observation data at a certain time, This is a model that generates observation data obtained by changing the observation method to another one. First estimation model 1 0124, observation data with different observation methods is estimated based on certain observation data, Consolidated data is generated.

[0114] 16 and 17, in the observation data processing device 101B, the first estimation model The process of generating integrated data using the satellite 201 will be described. The satellite 202 acquires observation data based on a first observation method (for example, a SAR method). , 204 acquires observation data based on a second observation method (e.g., an optical method), and transmits the observation data to the satellite. 203 is a method for obtaining observation data based on a third observation method (e.g., an observation method using meteorological data). shall obtain and

[0115] In step S1601, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the observation data at time T1. The measurement data D11 is acquired.

[0116] In step S1602, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the second observation data at time T2. The measurement data D42 is acquired.

[0117] In step S1603, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the third observation data at time T3. The measurement data D23 is acquired.

[0118] In step S1604, the converted observation data acquisition unit 1016 converts the third observation data For example, the converted observation data acquisition unit 1016 acquires a third observation condition for the observation data D More specifically, the observation conditions for the satellite 202 are acquired. Environmental information on the observation environment of the data D23 is acquired as the observation conditions.

[0119] In step S1605, the converted observation data acquisition unit 1016 converts the first guess model 10 The third observation data, the third observation condition, and the third timing are input to 124. For example, The observation data acquisition unit 1016 inputs the observation data D23 and the observation data D24 to the first conversion model 10122. The observation conditions for data D23 and time T3 are input.

[0120] In step S1606, the converted observation data acquisition unit 1016 converts the third observation data For example, the converted observation data acquisition unit 1016 acquires the converted observation data. The data D23 is converted into the estimated observation data D13, which is obtained from the first guess model 10124. .

[0121] In step S1607, the integrated data generating unit 1014 converts the first information into inferred observation data. Based on the data, the first information and the second information are updated to generate integrated data including the first information and the second information. The combined data generating unit 1014 generates an integrated data in which the observation data D11 and the observation data D42 are integrated. Next, the integrated data generation unit 1014 generates the observation data D11 as the estimated observation data ID2. The integrated data ID3 is generated based on the measurement data D13.

[0122] Similarly, the integrated data generation unit 1014 converts the observation data D34 into converted observation data The integrated data generating unit 1014 generates integrated data ID4 based on the integrated data D14. In the data, the converted observation data D14 is the integrated data updated by the observation data D15. Generate data ID5.

[0123] This allows, for example, observation between the artificial satellite 201, the artificial satellite 202, and the artificial satellite 203. Even if there are differences in the observation methods, the observation data used to generate the integrated data should be Observation method, for example, to compile data observed under the observation method of artificial satellite 201 In addition, it is possible to generate data observed under the observation method of artificial satellite 201 at high frequency. This makes it possible to harmonize the differences between observation methods and to provide integrated data at high frequency. By using such integrated data, meta-information of the observation area can be obtained more accurately. It will be possible to obtain it.

[0124] The second guess model 10125 will be described with reference to FIG. 25 is the first observation data obtained by observing the observation area, the first observation timing, and the second observation This is a learning model that outputs estimated observation data for the input of observed timing. The first observation data is data observed based on the first observation method (e.g., SAR method). The output inferred observation data is generated using a second observation method (e.g., optical) different from the first observation method. The data was observed based on the method.

[0125] The second guess model 10125 is trained using the training data LD5 as training data. The training data LD5 includes training observation data obtained by observing the training observation area, Timing A, and timing B different from the learning observation timing A and the corresponding The estimated observation data is a set of the estimated observation data and the estimated observation data. The data was collected at time B. Second guess model 10125 Input the learning observation data, learning observation timing, and timing B, and estimate the observation data. The data is used as the output for learning.

[0126] The second guess model 10125 is a method of observing data at a certain time. A model that changes to a different observation method and generates observation data obtained at a different timing. The second guess model 10125 uses observation data with different observation methods and timing. The data is inferred based on certain observational data to generate integrated data.

[0127] 19 and 22, in the observation data processing device 101B, the second estimation model The process of generating integrated data using the satellite 201 will be described. ,203 acquires observation data based on a first observation method (e.g., SAR method) and converts it into artificial satellite data. The star 202 is assumed to acquire observation data based on a second observation method (e.g., an optical method). do.

[0128] In step S1901, the observation data acquisition unit 1013 performs For example, the observation data acquisition unit 1013 acquires the first observation data at time T2. The measurement data D22 is acquired.

[0129] In step S1902, the estimated observation data acquisition unit 1017 assigns a first estimated model to 1. Input the observation data, the first timing, and the second timing. For example, as shown in FIG. As shown in FIG. 1, the estimated observation data acquisition unit 1017 receives the observation data D22, the time T2, and the time The time T3 is input to the second guess model 10125.

[0130] In step S1903, the estimated observation data acquisition unit 1017 acquires the second estimated model 10 125, the estimated observation data at the second timing is obtained. For example, As shown in FIG. 1, the estimated observation data acquisition unit 1017 acquires the estimated observation data D63 at time T3. is obtained from the second guess model 10125.

[0131] In step S1904, the integrated data generating unit 1014 generates an integrated data based on the first observation data. Generate integrated data based on the first information and the second information based on the inferred observation data. For example, As shown in FIG. 22, the integrated data generation unit 1014 generates the observation data D11 and the estimated observation data D12. Generate integrated data ID3 based on data D63.

[0132] As a result, for example, between the artificial satellite 202 and the artificial satellites 201 and 203 Even if there are differences in the observation methods, the observation data used to generate the integrated data are The data collected under the observation methods of the artificial satellite 201 and artificial satellite 203 will be collected. This allows for the generation of integrated data with reduced differences between observation methods. By using such integrated data, meta-information of the observation area can be obtained with greater accuracy.

[0133] The third guess model 10126 will be described with reference to FIG. 26 is the first observation data obtained by observing the observation area, the third observation data before the first observation timing, Learning to output estimated observation data for the input of the observation timing and the first observation timing The input first observation data and the output inferred observation data are common observations. The third guess model 101 is data observed based on a method (e.g., SAR method). 26 is a model that performs time series regression prediction and generates inferred observation data.

[0134] The third guess model 10126 is trained using the training data LD6 as training data. The learning data LD6 includes the learning observation data obtained by observing the learning observation area, Timing A, and timing C before the learning observation timing A and the corresponding estimation The training observation data and the estimated observation data are paired with the first observation method. The data was observed based on the observation timing of timing C. The observation model 10126 uses the training observation data, the training observation timing, and the timing C. Learning is done by using inferred observation data as input and output as output.

[0135] The third guess model 10126 is an observation method based on observation data at a certain timing. Without changing the formula, the observation data obtained after that timing The third guess model 10126 generates observations with different observation timing. Data is inferred based on certain observed data, and integrated data is generated.

[0136] 20 and 21, in the observation data processing device 101B, a third estimation model The process of generating integrated data using the satellite 201 will be described. ,203 acquires observation data based on a first observation method (e.g., SAR method) and converts it into artificial satellite data. The star 202 is assumed to acquire observation data based on a second observation method (e.g., an optical method). do.

[0137] In step S2101, the observation data acquisition unit 1013 acquires the observation data before the first timing. Obtain third observation data when the observation area is observed at the third timing. For example, The data acquisition unit 1013 acquires the observation data when the observation area is observed at time T2, which is before time T5. Get data D22.

[0138] In step S2102, the estimation observation data acquisition unit 1017 adds a third observation to the estimation model. The measurement data, the third timing, and the first timing are input. For example, as shown in FIG. Thus, the estimated observation data acquisition unit 1017 receives the observation data D22, the time T2, and the time T5. is input to the third guess model 10126.

[0139] In step S2103, the estimated observation data acquisition unit 1017 acquires the third estimated model 10 126, the estimated observation data at the first timing is obtained. For example, As shown in FIG. 1, the estimated observation data acquisition unit 1017 acquires the estimated observation data D75 at time T5. is obtained from the third guess model 10126.

[0140] In step S2104, the integrated data generating unit 1014 performs Integrated data based on first information based on estimated observation data and second information based on second observation data For example, as shown in FIG. 22, the integrated data generator 1014 generates a speculative The integrated data ID5 is generated based on the measurement data D53 and the observation data D14.

[0141] This allows, for example, if the satellite 202 is unable to make an observation at a certain point in time However, the observation data used to generate the integrated data may be estimated and used as the integrated data. This makes it possible to obtain the time interval even if some of the observation data is not available. Short integrated data can be generated. Such integrated data can be used to measure the area of ​​interest. This makes it possible to obtain meta-information with greater accuracy.

[0142] The above-described embodiment is intended to facilitate understanding of the present invention and does not limit the present invention. The elements of the embodiment and their arrangement, conditions, and shapes are not intended to be interpreted as being related to the present invention. The dimensions and size are not limited to those shown in the example and can be changed as appropriate. In addition, configurations shown in different embodiments may be partially substituted or combined. do. [Explanation of symbols]

[0143] 10...observation system, 101, 101A, 101B...observation data processing device, 201, 20 2,203…Satellite, 1013…Observation data acquisition unit, 1014…Integrated data generation unit, 1 015: Meta information acquisition unit, 1016: Converted observation data acquisition unit, 1017: Estimated observation data Acquisition department

Claims

1. On the computer, The first observation means observes the first region at the first timing. Observation data and information of the observation area at a second timing different from the first timing. second observation data observed by a second observation means different from the first observation means; Based on the first observation data, first information based on the first observation data and second information based on the second observation data are generated. generating integrated data including the information; A program that executes the following.

2. The program according to claim 1, the first observation data is data observed under a first observation condition, The computer includes: information on the observation area at a third timing different from the first timing, Data observed under third observation conditions by a third observation means different from the first observation means. and acquiring third observation data observed based on a common observation method to the first observation data. To do, Learning observation data, which is information on the learning observation area, and learning observation conditions corresponding to the learning observation data The learning observation data and the learning observation timing at which the learning observation data were observed are input. The learning observation timing is a learning observation timing. A transformation that outputs transformed observation data representing data observed under the first observation condition in the observation area. The third observation data, the third observation condition, and the third timing are input to the model, obtaining converted observation data obtained by converting the third observation data; Then, generating the integrated data, The first information is updated based on the converted observation data to generate the integrated data. , including , a program.

3. The program according to claim 2, The first observation condition includes first aircraft information of a first flying object which is the first observation means, and First environmental information indicating an observation environment by the flying object; The third observation condition includes third aircraft information of a third flying object which is the third observation means, and and second environmental information indicating an observation environment by the flying object.

4. The program according to claim 1, the first observation data is data observed under a first observation condition, The computer includes: acquiring the first observation data and the first observation condition; information on the observation area at a third timing different from the first timing, Data observed under third observation conditions by a third observation means different from the first observation means. and acquiring third observation data observed based on a common observation method to the first observation data. To do, Learning observation data, which is information on a learning observation area, and a learning observation corresponding to the learning observation data The conditions and the learning observation timing at which the learning observation data was observed are input, and the learning The observation data is converted into converted observation data, A conversion model that outputs converted observation data that indicates data observed under standard observation conditions in the observation area. the first observation data, the first observation conditions, and the first timing are input to the Obtaining first converted observation data as the converted observation data obtained by converting the first observation data. And, The third observation data, the third observation condition, and the third timing are input to the conversion model. the converted observation data obtained by converting the third observation data at the third timing; acquiring second converted observation data as the second converted observation data; Then, generating the integrated data, The first information based on the first converted observation data is converted to the previous information based on the second converted observation data. updating the first information to generate the integrated data.

5. The program according to claim 4, The first observation condition includes first aircraft information of a first flying object which is the first observation means, and First environmental information indicating an observation environment by the flying object; The third observation condition includes third aircraft information of a third flying object which is the third observation means, and and second environmental information indicating an observation environment by the flying object.

6. The program according to claim 1, the first observation data is data observed based on a first observation method, The second observation data is observed based on a second observation method different from the first observation method. A program is a piece of data.

7. The program according to claim 6, The computer includes: information on the observation area at a third timing different from the first timing, Data observed under third observation conditions by a third observation means different from the first observation means. and acquiring third observation data observed based on an observation method different from the first observation method. To do, The learning observation data and the learning observation timing at which the learning observation data was observed are input. The observation area is observed based on the first observation method at the learning observation timing. The third observation data and the and acquiring the estimated observation data at the third timing. And, Then, generating the integrated data, updating the first information based on the inferred observation data to generate the integrated data; The program includes:

8. The program according to claim 6, The computer includes: acquiring the first observation data; A learning observation data, a learning observation timing at which the learning observation data was observed, and A timing different from the observation timing is input, and at the different timing, outputting estimated observation data indicative of data observed in the observation area based on a second observation method; The estimation model includes the first observation data, the first timing, and the second timing. and acquiring the estimated observation data at the second timing; Then, generating the integrated data, The integrated information including the second information based on the estimated observation data at the second timing. generating data.

9. The program according to claim 6, a third observation in a case where the observation area is observed at a third timing prior to the first timing; obtaining measurement data; The learning observation data, the learning observation timing at which the learning observation data was observed, and the learning perspective A predicted timing after the measured timing is input, and the predicted timing is Outputting estimated observation data indicating data observed in the observation area based on one observation method The third observation data, the third timing, and the first timing are input to an estimation model. acquiring the estimated observation data at the first timing; Then, generating the integrated data, The integrated information including the first information based on the estimated observation data at the first timing generating data.

10. The program according to claim 1, The computer includes: Learning integration data based on learning observation data, which is information on the learning observation domain, is input, and The integrated data is input to a meta-information output model that outputs meta-information of the observation area. Obtaining meta information of a region; Further execute the program.

11. The first observation means observes the first region at the first timing. Observation data and information of the observation area at a second timing different from the first timing. second observation data observed by a second observation means different from the first observation means; Based on the first observation data, first information based on the first observation data and second information based on the second observation data are generated. an integrated data generating unit for generating integrated data including the information; An information processing device comprising:

12. The computer The first observation means observes the first region at the first timing. Observation data and information of the observation area at a second timing different from the first timing. second observation data observed by a second observation means different from the first observation means; Based on the first observation data, first information based on the first observation data and second information based on the second observation data are generated. generating integrated data including the information; An information processing method comprising:

13. Learning observation data, which is information on the learning observation area, and learning observation conditions corresponding to the learning observation data The learning observation area is input at the timing when the learning observation data is acquired. The first transformation is obtained by converting the learning observation data into data observed under the reference observation conditions. It is trained using teacher data that outputs transformed observation data, Input of observation data, which is information on the target observation area, and observation conditions corresponding to the observation data At the time when the observation data is acquired, the target observation area is the reference observation area. second converted observation data obtained by converting the observed data, which indicates data observed under the observation conditions; A learning model that causes a computer to output the following:

14. learning observation data when the learning observation area is observed based on a first observation method; The timing at which the observation data was observed and the predicted timing that differs from the observation timing and the learning observation domain is different from the first observation method at the prediction timing. The output is first guess observation data indicating data observed based on a different second observation method. It is trained using training data, Input of observation data, which is information on the target observation area, and observation conditions corresponding to the observation data On the other hand, at the timing when the observation data is acquired, the target observation area is the second observation area. The computer is adapted to output second guess observation data indicative of the data observed based on the observation method. A learning model that puts data to work.

Images