Learning model, signal processing device, flying body, and teacher data generation method

Abstract

The present invention makes it possible to detect a target, such as a ground object, by using RAW data. This learning model is trained by using teacher data, where the input is a first synthetic aperture radar reception signal based on a reflected electromagnetic wave which is reflection of an electromagnetic wave irradiated to a learning region, and the output is first target information of a target in the learning region corresponding to a characteristic amount of the first synthetic aperture radar reception signal. The learning model causes a computer to function so as to output, in response to an input of a second synthetic aperture radar reception signal based on a reflected electromagnetic wave which is reflection of an electromagnetic wave irradiated to a detection region, second target information of the target in the detection region corresponding to a characteristic amount of the second synthetic aperture radar reception signal.

Description

Learning model, signal processing device, flying object, and training data generation method

[0001] The present invention relates to a learning model, a signal processing device, a flying object, and a program.

[0002] Observation of the state of the Earth's surface, including land and sea, is widely carried out using flying objects such as artificial satellites, aircraft, and drones. Satellite observation methods include an observation method that acquires radar images, so-called SAR images, obtained using synthetic aperture radar (SAR) technology, and an observation method that acquires optical images and SAR images and combines the two images. Patent Document 1 describes the use of raw data obtained by synthetic aperture radar to detect objects without undergoing SAR imaging processing.

[0003] Japanese Patent Application Laid-Open No. 2023-000897

[0004] There are various possible methods for detecting objects such as features using RAW data. Therefore, an object of the present invention is to provide a learning model, a signal processing device, a flying object, and a training data generation method that enable detection of objects such as features using RAW data.

[0005] A learning model according to one aspect of the present invention is trained using teacher data that takes as input the features of a first synthetic aperture radar received signal based on the reflected electromagnetic waves of electromagnetic waves irradiated onto a learning area and outputs first object information of an object in the learning area, and causes a computer to function such that, in response to an input of the features of a second synthetic aperture radar received signal based on the reflected electromagnetic waves of electromagnetic waves irradiated onto a detection area, it outputs second object information of an object in the detection area.

[0006] According to this aspect, since object information can be output based on the feature amount of the synthetic aperture radar received signal, it becomes possible to detect objects such as features on the ground using RAW data.

[0007] In the above aspect, the feature of the first synthetic aperture radar reception signal may include a frequency spectrum of the first synthetic aperture radar reception signal, and the feature of the second synthetic aperture radar reception signal may include a frequency spectrum of the second synthetic aperture radar reception signal.

[0008] In the above aspect, the feature of the first synthetic aperture radar received signal may include a signal strength of the first synthetic aperture radar received signal, and the feature of the second synthetic aperture radar received signal may include a signal strength of the second synthetic aperture radar received signal.

[0009] In the above aspect, the feature of the first synthetic aperture radar reception signal may include a phase of the first synthetic aperture radar reception signal, and the feature of the second synthetic aperture radar reception signal may include a phase of the second synthetic aperture radar reception signal.

[0010] According to these aspects, target information based on the frequency spectrum, signal strength, or phase can be output as a feature of the synthetic aperture radar received signal, making it possible to detect targets such as ground features using RAW data.

[0011] In the above aspect, the first synthetic aperture radar received signal is a signal corresponding to a part of the learning area whose feature quantity satisfies a predetermined condition, and the feature quantity of the first synthetic aperture radar received signal may include a feature quantity based on an object in the learning area.

[0012] According to this aspect, the received signal that is the source of the feature quantities of the received signal that are used as training data for the learning model is a signal corresponding to an area where the feature quantities satisfy a predetermined condition, so the amount of data required for learning is reduced. Furthermore, the feature quantities of the received signal include feature quantities based on objects in the learning area, so that objects such as ground features can be detected using RAW data while reducing the amount of data.

[0013] In the above aspect, the second synthetic aperture radar received signal is a signal corresponding to a part of the detection area whose feature quantity satisfies a predetermined condition, and the feature quantity of the second synthetic aperture radar received signal may include a feature quantity based on an object in the detection area.

[0014] According to this aspect, the feature quantities of the second synthetic aperture radar received signal used in inference using the learning model can also be based on signals corresponding to a portion of the area where the feature quantities satisfy predetermined conditions, thereby reducing the amount of data required for inference. Furthermore, because the feature quantities of the second synthetic aperture radar received signal include feature quantities based on objects in the learning area, it is possible to detect objects such as ground features using RAW data while reducing the amount of data.

[0015] In the above aspect, the first object information may include information indicating attributes of the object in the learning area, and the second object information may include information indicating attributes of the object in the detection area.

[0016] According to this aspect, it is possible to obtain information indicating the attributes of the object, and therefore it is possible to detect the object while obtaining more detailed information about the object.

[0017] An information processing device according to another aspect of the present invention includes a memory unit in which a learning model is stored, a signal acquisition unit that acquires a second synthetic aperture radar received signal, and an estimation unit that inputs the second synthetic aperture radar received signal to the learning model and infers second object information.

[0018] A flying object according to another aspect of the present invention includes a memory unit in which a learning model is stored, a signal acquisition unit that acquires a second synthetic aperture radar received signal, an estimation unit that inputs the second synthetic aperture radar received signal into the learning model and infers second object information, and a signal output unit that outputs an output signal based on the second object information to the outside.

[0019] A teacher data generation method according to another aspect of the present invention includes a computer acquiring a first synthetic aperture radar reception signal based on electromagnetic waves reflected from electromagnetic waves irradiated onto a learning area, extracting from the first synthetic aperture radar reception signal features of a learning signal corresponding to an area of ​​the learning area in which an object is present, acquiring object information of the object, and storing a pair of the features of the learning signal and the object information as teacher data used for training a learning model that causes the computer to function so as to output second object information of the object in the detection area corresponding to the features of the second synthetic aperture radar reception signal in response to input of features of a second synthetic aperture radar reception signal based on electromagnetic waves reflected from electromagnetic waves irradiated onto the detection area.

[0020] According to the present invention, it is possible to detect objects such as features using RAW data.

[0021] FIG. 1 is a block diagram of an observation system according to the present embodiment. FIG. 2 is a diagram illustrating a learning model according to the present embodiment. FIG. 3 is a diagram illustrating learning of the learning model according to the present embodiment. FIG. 4 is a flowchart illustrating an example of processing in a flying object according to the present embodiment. FIG. 5 is a diagram illustrating an example of inference using the learning model according to the present embodiment. FIG. 6 is a flowchart illustrating another example of processing in a flying object according to the present embodiment. FIG. 7 is a diagram illustrating another example of inference using the learning model according to the present embodiment. FIG. 8 is a flowchart illustrating another example of processing in a flying object according to the present embodiment. FIG. 9 is a diagram illustrating another example of the learning model according to the present embodiment. FIG. 10 is a diagram illustrating an example of a teacher data generation method used for learning the learning model according to the present embodiment. FIG. 11 is a flowchart illustrating another example of processing in a flying object according to the present embodiment. FIG. 11 is a diagram illustrating another example of inference using the learning model according to the present embodiment. FIG. 12 is a block diagram illustrating another aspect of the observation system.

[0022] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A preferred embodiment of the present invention will be described with reference to the accompanying drawings. In the drawings, components with the same reference numerals have the same or similar configurations.

[0023] FIG. 1 shows a block diagram of an observation system 10 according to this embodiment. The observation system 10 includes an air vehicle 100 and an observation device 200. The air vehicle 100 is located in space above the Earth, and the observation device 200 is located on the Earth. In this embodiment, the air vehicle 100 observes a detection area D on the Earth's surface using radar, and an observation signal O processed by the air vehicle 100 is transmitted to the observation device 200. The observation signal O is, for example, a received signal acquired by the air vehicle 100, or a signal corresponding to the received signal, which indicates object information, such as information about an object in the detection area D, such as a feature on the ground or a ship at sea, as described below.

[0024] The flying object 100 includes a communication antenna 101, a radar device 102, and a signal processing device 103. The flying object 100 is an artificial satellite capable of acquiring and processing received signals, and is placed in space and orbits the Earth. Note that the flying object 100 may be a geostationary satellite. Alternatively, the flying object 100 may be any device capable of being positioned above the Earth, such as an airplane, a helicopter, or a drone.

[0025] The communication antenna 101 is an antenna for the flying object 100 to communicate with an external device located on the Earth or in outer space.

[0026] The radar device 102 is a device that irradiates a detection area D on the Earth's surface with electromagnetic waves EM1, such as microwaves, and acquires reflected electromagnetic waves EM2 that are the electromagnetic waves EM1 reflected by an object to be observed in the detection area D. The radar device 102 is, for example, a synthetic aperture radar (SAR). The reflected electromagnetic waves EM2 are processed and recorded by the radar device 102 as received signals (RAW data) based on fluctuations in the electromagnetic waves that can be handled by the flying object 100. The received signals are recorded, for example, as complex number signals corresponding to each coordinate of the detection area D. The radar device 102 transmits an observation signal O to the observation device 200 via the communication antenna 101.

[0027] The radar device 102 includes a processor for controlling the acquisition process of the received signal and a storage device for storing programs required for this control.

[0028] The signal processing device 103 is an information processing device that processes the received signal acquired by the radar device 102. The signal processing device 103 is a computer that has a storage area such as a memory, and performs predetermined processing by having a processor execute a program stored in the storage area.

[0029] The signal processing device 103 includes a storage unit 104 and a control unit 105. The storage unit 104 is, for example, a semiconductor memory such as a RAM or an optical disk. The storage unit 104 stores various types of information used in processing by the signal processing device 103.

[0030] The storage unit 104 stores a learning model 1041. The learning model 1041 is a program that has been trained to receive a received signal as input and output object information corresponding to the feature quantities of the received signal. Details of the feature quantities and the learning model 1041 in this embodiment will be described later.

[0031] The control unit 105 performs signal processing in the signal processing device 103. The control unit 105 also controls transmission of the processing results of the received signal through the flying object 100. The control unit 105 has a signal acquisition unit 1051, an estimation unit 1052, and a signal output unit 1053.

[0032] The signal acquisition unit 1051 acquires a received signal from the radar device 102 .

[0033] The estimation unit 1052 inputs the feature amount of the received signal acquired by the signal acquisition unit 1051 into the learning model 1041 and acquires object information from the learning model 1041.

[0034] The signal output unit 1053 outputs the object information acquired by the estimation unit 1052 to the observation device 200 as an observation signal O via the communication antenna 101. The signal output unit 1053 may also output a received signal corresponding to the object information as the observation signal O together with the object information.

[0035] The observation device 200 is a device that transmits a control signal to the flying object 100 to control observation of the detection area D by the flying object 100, and acquires an observation signal O from the flying object 100. The observation device 200 has a communication unit 201 that includes an antenna and a control unit that controls communication via the antenna. Information is transmitted and received with the flying object 100 via the communication unit 201.

[0036] The signal processing unit 202 processes the observation signal O from the flying object 100. Based on the observation signal O acquired from the flying object 100, the signal processing unit 202 performs processing to visualize the observation results in the detection area D, for example, using an image.

[0037] The learning of the learning model 1041 according to this embodiment will be described with reference to FIGS.

[0038] FIG. 2 is a diagram illustrating the learning and inference of the learning model 1041. The learning model 1041 is trained using training data LD1 as training data. The training data LD includes a set of feature quantities of a received signal R0 (first synthetic aperture radar received signal) acquired by irradiating a certain area (learning area) with electromagnetic waves and object information TI0 (first object information) corresponding to the feature quantities of the received signal R0. The learning model 1041 is trained using the feature quantities of the received signal R0 as input and the object information TI0 as output. Here, the feature quantities in this embodiment refer to information including the frequency spectrum, signal intensity, or phase of the received signal R0.

[0039] FIG. 3 shows the received signal R0 and the feature quantities of the received signal when a learning area D0 in which two objects, an object O1 and an object O2, exist is observed.

[0040] When the learning region D0 is observed, a received signal R0 is obtained. Based on the received signal R0, a feature quantity (frequency spectrum, signal intensity, or phase) of the received signal R0 is calculated. In order to associate the feature quantity of the received signal R0 with the object information TI0, information processing is performed to enable a user to grasp the object information TI0 corresponding to the received signal R0.

[0041] For example, a graph may be generated based on the received signal R0, and a user may associate the target information TI0 with the graph. Alternatively, a computer may associate the target information with the feature of the received signal R0 using a learning model that associates the received signal R0 with the target information TI0. The received signal R0 may be converted into user-understandable information through a predetermined conversion process for SAR imaging. Alternatively, the target information TI0 associated with the feature may be information acquired from another device observing the learning area D0 without undergoing the above-described imaging process. For example, when detecting a ship, the target information TI0 may be information acquired using an Automatic Identification System (AIS) in addition to information obtained from processing the received signal R0.

[0042] The object information TI0 is information associated with the feature quantity (frequency spectrum, signal intensity, or phase) of the received signal R0. The received signal R0 is a signal obtained in the form of a complex number, for example, I(t)+jQ(t) (j is the imaginary unit, t is time). When the received signal R0 is written in polar form, the received signal R0 is expressed as A(t)e ｊθ（ｔ） In this case, the signal strength of the received signal R0 is A(t), the phase of the received signal R0 is θ(t), and the signal strength and phase are obtained for each coordinate in the detection area D0. Furthermore, by performing a frequency conversion on the received signal R0 in the time domain, the frequency spectrum X(ω) of the received signal R0 can be obtained.

[0043] The learning model 1041 is trained to output object information in response to input of these feature quantities calculated from the received signal R0. For example, the control unit 105 calculates feature quantities based on the received signal R0, and the learning model 1041 is trained to output object information using the feature quantities as input. Note that the learning model 1041 may also be trained to receive input of the received signal R0 itself and output object information, in which case the learning model 1041 may be configured to calculate feature quantities based on the received signal R0.

[0044] The object information T10 includes information on the probability distribution of the probability that an object exists in each pixel of the learning area. Furthermore, the object information TI0 includes, for example, information on the number of objects present in the learning area and the attributes of the objects. The attributes of the objects include various information related to the objects, such as the type of object (building, moving object, topography, natural object, etc.), the size of the object, the moving speed of the object, etc. The information on the attributes of the objects may be the probability that the object is classified into a certain attribute category.

[0045] The learning model 1041 is trained by a general machine learning method such as a method using a neural network, using the training data LD1 prepared as described above. Note that the learning model 1041 may be configured as a single learning model, or may be configured as a learning model combining multiple learning models.

[0046] When the features of the received signal R1 (second synthetic aperture radar received signal) acquired by the flying object 100 based on the electromagnetic waves irradiated to the detection area by the flying object 100 are input to the trained learning model 1041, the learning model 1041 outputs object information TI1 (second object information).

[0047] The processing by the flying object 100 will be described with reference to Figures 4 and 5. In the processing in Figures 4 and 5, the target object information output by the learning model 1041 is assumed to be information corresponding to the frequency spectrum of the received signal R1.

[0048] 4, the radar device 102 irradiates the detection area D1 with electromagnetic waves EM1. The irradiation timing may be controlled by the observation device 200 or may be a timing designated in advance by the flying object 100. As shown in FIG. 5, objects O3, O4, and O5 are present in the detection area D1.

[0049] In step S402, the signal acquisition unit 1051 acquires, from the radar device 102, a reception signal R1 based on the reflected electromagnetic wave EM2 detected by the radar device 102.

[0050] In step S403, the estimation unit 1052 inputs the feature amount of the received signal R1 into the learning model 1041. At this time, the estimation unit 1052 performs a calculation to calculate the feature amount (for example, frequency spectrum) of the received signal R1.

[0051] In step S404, the estimation unit 1052 acquires object information MD1 corresponding to the feature quantities of the received signal R1 from the learning model 1041. The object information includes, as an example, object information TI1a and object information TI1b shown in FIG. 5. The object information TI1a is a probability distribution of the probability that an object exists in the detection area. The object information TI1a indicates areas A1, A2, and A3 as areas where there is a high probability that an object exists. The object information TI1b is the number of objects in the detection area and the attributes of the objects. In this example, three objects are detected.

[0052] In step S405, the signal output unit 1053 outputs an output signal based on the object information TI1 to the observation device 200. The output signal based on the object information TI1 is a signal that transmits all or part of the object information TI1. Alternatively, the output signal may be a signal that transmits information resulting from information processing performed on the object information.

[0053] 6 and 7, a case will be described in which the object information output by the learning model 1041 is information based on the signal strength of the received signal R1. As shown in Fig. 7, it is assumed that objects O3, O4, and O5 exist in the detection area D1, as in Fig. 5.

[0054] 6, the radar device 102 irradiates the detection area D1 with electromagnetic waves EM1. In step S602, the signal acquisition unit 1051 acquires from the radar device 102 a received signal R1 based on the reflected electromagnetic waves EM2 detected by the radar device 102.

[0055] In step S603, the estimation unit 1052 inputs the signal strength of the received signal R1 to the learning model 1041 as a feature of the received signal R1. The received signal R1 is, for example, a signal capable of generating the image IG1 shown in FIG. 7. The image IG1 can take various forms depending on the signal. As an example, the image IG1 may plot the signal strength obtained by detecting each object, and the changes in signal strength may be displayed as a pattern. The estimation unit 1052 acquires object information MD1 corresponding to the signal strength of the received signal R1 from the learning model 1041. As shown in FIG. 7, the object information includes, for example, object information TI1a and object information TI1b, similar to the case shown in FIG. 5.

[0056] In step S605, the signal output unit 1053 outputs an output signal based on the object information TI1 to the observation device 200.

[0057] 8, a case will be described in which the object information output by the learning model 1041 is information based on the phase of the received signal R1. In step S801 of FIG. 8, the radar device 102 irradiates the detection area D1 with an electromagnetic wave EM1. In step S602, the signal acquisition unit 1051 acquires from the radar device 102 a received signal R1 based on a reflected electromagnetic wave EM2 detected by the radar device 102. In step S803, the estimation unit 1052 inputs the phase of the received signal R1 to the learning model 1041. In step S804, the estimation unit 1052 acquires object information MD1 corresponding to the signal strength of the received signal R1 from the learning model 1041. In step S605, the signal output unit 1053 outputs an output signal based on the object information TI1 to the observation device 200.

[0058] As described above, when inferring object information using the learning model 1041, the learning model 1041 is trained to input the frequency spectrum, intensity, or phase as the feature of the received signal and output object information, and inference is performed using the learning model 1041. In this case, it is possible to combine the frequency spectrum, intensity, or phase as the feature of the received signal.

[0059] Another learning method for the learning model 1041 according to this embodiment will be described with reference to FIGS. 9 and 10. FIG.

[0060] 9 is a diagram illustrating the learning and inference of the learning model 1041. The learning model 1041 is trained using the training data LD2 as training data. The training data LD2 includes a pair of a reception signal R0a, the feature of which satisfies a predetermined condition, among reception signals R0 (first synthetic aperture radar reception signals) acquired by irradiating a certain area (learning area) with electromagnetic waves, and object information TI0 (first object information) corresponding to the feature of the reception signal R0a.

[0061] The predetermined conditions for the feature quantities are conditions that enable extraction of a received signal containing feature quantities based on the object in the learning region, and can be changed depending on the purpose of observation, etc. For example, when extracting received signal R0a from received signal R0, a received signal whose signal strength satisfies a predetermined condition can be extracted. For example, the predetermined condition for the signal strength is that the signal strength is greater than a predetermined threshold. Also, for example, the predetermined condition for the frequency spectrum is that the frequency is within a predetermined frequency band. Also, for example, the predetermined condition for the phase is that the phase is within a predetermined range.

[0062] The learning model 1041 is trained using the feature amount of the received signal R0a as input and the object information TI0 as output.

[0063] The generation of the learning data LD2 will be described with reference to Fig. 10. The example of Fig. 10 shows the generation of the learning data LD2 using the observation results of a learning area D0 in which two objects, an object O1 and an object O2, exist.

[0064] When the learning region D0 is observed, a received signal R0 is obtained. When the received signal R0 is plotted as a graph or imaged by converting it into an SAR image, for example, an image IG0 is obtained. The image IG0 can be associated with the signal strength of the received signal R0. Received signals whose signal strength is greater than a predetermined threshold are extracted from the received signal R0. For example, a received signal R0a including a received signal R0a1 corresponding to a partial region in the image IG0 and a received signal R0a2 corresponding to another partial region in the image IG0 is extracted. When the received signal R0a1 is imaged, an image IG2 is obtained, and when the received signal R0a2 is imaged, an image IG3 is obtained. The images IG2 and IG3 can be associated with the signal strengths of the received signals R0a1 and R0a2, respectively.

[0065] The feature quantities of each received signal R0a, including received signal R0a1 and received signal R0a2, are associated with object information TI0 based on image IG0. As an example, the object information includes object information TI0a and object information TI0b. The object information TI0a is a probability distribution of the probability that an object exists in the detection area. In the object information TI0a, areas A4 and A5 are indicated as areas where there is a high probability that an object exists. The object information TI0b is the number of objects in the detection area and the attributes of the objects. In this example, two objects are detected. The learning data LD2 is generated as a set of received signal R0a and object information TI0.

[0066] The learning model 1041 is trained using the training data LD2 prepared as described above, by a general machine learning method such as a method using a neural network.

[0067] The processing performed by the flying object 100 will be described with reference to FIGS.

[0068] 11, the radar device 102 irradiates the detection area D1 with an electromagnetic wave EM1. As shown in FIG. 12, objects O3, O4, and O5 exist in the detection area D1.

[0069] In step S1102, the signal acquisition unit 1051 acquires, from the radar device 102, a reception signal R1 based on the reflected electromagnetic wave EM2 detected by the radar device 102.

[0070] In step S1103, the estimation unit 1052 extracts a received signal R1a from the received signal R1 that satisfies a predetermined signal strength condition. The feature amount of the extracted received signal R1a corresponds to an image IG4 obtained by cutting out a part of the image IG1 based on the received signal R1, as shown in FIG. 12 .

[0071] In step S1104, the estimation unit 1052 inputs the received signal R1a to the learning model 1041.

[0072] In step S1105, the estimation unit 1052 acquires object information MD1 based on the feature quantities of the received signal R1a from the learning model 1041. The object information includes, as an example, object information TI1a and object information TI1b shown in FIG. 12. In the object information TI1a, as in the case of FIG. 5, areas A1, A2, and A3 are indicated as areas where there is a high probability that objects exist. The object information TI1b indicates the number of objects in the detection area and the attributes of the objects. In this example, three objects are detected.

[0073] In step S1106, the signal output unit 1053 outputs an output signal based on the object information TI1 to the observation device 200.

[0074] In this way, by making it possible to output object information based on the features of the entire received signal from the features of a partial received signal, it is possible to reduce the amount of data that is input to the learning model 1041 for inference. Furthermore, since the calculation time using the learning model 1041 can be shortened, the real-time nature of observation is improved. Furthermore, since it is possible to reduce the calculation load of the learning model 1041, it is possible to acquire object information using a lighter processing device.

[0075] Figure 13 shows a block diagram of an observation system 10A as another embodiment. As shown in Figure 13, the signal processing device 103 can also be provided so as to be included in the observation device 200A. A control unit 1301 of the flying object 100A transmits a received signal acquired by the radar device 102 to the observation device 200A. The above-mentioned processing can also be performed by the signal processing device 103 of the observation device 200A.

[0076] The above-described embodiments are intended to facilitate understanding of the present invention and are not intended to limit the present invention. The elements and conditions of the embodiments are not limited to those exemplified and can be changed as appropriate. Different configurations can be partially substituted or combined.

[0077] 10... observation system, 100... flying object, 101... communication antenna, 102... radar device, 103... signal processing device, 104... memory unit, 1041... learning model, 105... control unit, 1051... signal acquisition unit, 1052... estimation unit, 1053... signal output unit, 200... observation device, 201... communication unit, 202... signal processing unit

Claims

1. A learning model that is trained using teacher data with the feature amount of a first synthetic aperture radar reception signal based on the reflected electromagnetic wave obtained by irradiating a learning area as input and the first object information of an object in the learning area as output, and causes a computer to function so as to output the second object information of an object in the detection area for an input of the feature amount of a second synthetic aperture radar reception signal based on the reflected electromagnetic wave obtained by irradiating the detection area.

2. The learning model according to claim 1, wherein the feature amount of the first synthetic aperture radar reception signal includes the frequency spectrum of the first synthetic aperture radar reception signal, and the feature amount of the second synthetic aperture radar reception signal includes the frequency spectrum of the second synthetic aperture radar reception signal.

3. The learning model according to claim 1, wherein the feature amount of the first synthetic aperture radar reception signal includes the signal intensity of the first synthetic aperture radar reception signal, and the feature amount of the second synthetic aperture radar reception signal includes the signal intensity of the second synthetic aperture radar reception signal.

4. The learning model according to claim 1, wherein the feature amount of the first synthetic aperture radar reception signal includes the phase of the first synthetic aperture radar reception signal, and the feature amount of the second synthetic aperture radar reception signal includes the phase of the second synthetic aperture radar reception signal.

5. The learning model according to claim 1, wherein the first synthetic aperture radar reception signal is a signal corresponding to a part of the learning area where the feature amount satisfies a predetermined condition, and the feature amount of the first synthetic aperture radar reception signal includes the feature amount based on the object in the learning area.

6. The learning model according to claim 5, wherein the second synthetic aperture radar reception signal is a signal corresponding to a part of the detection area where the feature amount satisfies the predetermined condition, and the feature amount of the second synthetic aperture radar reception signal includes the feature amount based on the object in the detection area.

7. The learning model according to claim 1, wherein the first object information includes information indicating the attribute of the object in the learning area, and the second object information includes information indicating the attribute of the object in the detection area.

8. A signal processing apparatus comprising: a storage unit storing the learning model according to any one of claims 1 to 7; a signal acquisition unit acquiring the second synthetic aperture radar reception signal; and an estimation unit inputting the second synthetic aperture radar reception signal to the learning model and inferring the second object information.

9. An aircraft comprising: a storage unit storing the learning model according to any one of claims 1 to 7; a signal acquisition unit acquiring the second synthetic aperture radar reception signal; an estimation unit inputting the second synthetic aperture radar reception signal to the learning model and inferring the second object information; and a signal output unit outputting an output signal based on the second object information to the outside.

10. Teacher data used for learning of a learning model that causes a computer to function as follows: acquiring a first synthetic aperture radar reception signal based on a reflected electromagnetic wave obtained by reflecting an electromagnetic wave irradiated on a learning area; extracting a learning signal corresponding to an area where an object exists in the learning area from the first synthetic aperture radar reception signal; acquiring object information of the object; and storing, as teacher data, a pair of a feature amount of the learning signal and the object information for input of a feature amount of a second synthetic aperture radar reception signal based on a reflected electromagnetic wave obtained by reflecting an electromagnetic wave irradiated on a detection area, and outputting second object information of the object in the detection area corresponding to the feature amount of the second synthetic aperture radar reception signal.

Images