Flying Body

Abstract

To provide a missile and a program capable of making the amount of data to be transmitted smaller than the amount of observation data in transmitting data to a ground station side and notifying a user of detection of a desired detection object faster.SOLUTION: A missile comprises an observation data generation unit which generates observation data on the basis of a radio wave received by a radar, an image generation unit which generates an image representing a monitoring space on the basis of the observation data generated by the observation data generation unit, and a detection unit which detects an object to be detected on the basis of the image generated by the image generation unit.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a flying object and a program. This application claims priority based on Japanese Patent Application No. 2017-034122, filed on February 24, 2017, the contents of which are incorporated herein by reference. [Background technology]

[0002] Research and development is being conducted on technologies related to artificial satellites that transmit and receive radio waves between the observed object and the satellite. For example, such artificial satellites generate observation data representing the observed object based on the radio waves received from the observed object, and transmit the generated observation data to a receiving device on the ground. The receiving device performs processing based on the received observation data.

[0003] In this regard, a satellite is known that frequency-multiplexes observation data into a plurality of channels and performs hardware compression for each of the plurality of channels (see Patent Document 1).Also, a technology is known that generates an image representing an observation target based on observation data received from the satellite by a receiving device.Also, software is known that detects ships in an ocean area included in the observation target based on the image (see Non-Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 2014-510913 [Non-patent literature]

[0005] [Non-Patent Document 1] VESSELFINDER powered by HuygensWorks: http: / / www.mss.co.jp / information / 20141205 / VesselFinder2.pdf [Non-Patent Document 2] MARITIME PRODUCTS USING TERRASAR-X and SENTINEL-1 IMAGERY: S.Lehner, B.Tings, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-7 / W3, 2015 Summary of the Invention [Problem to be solved by the invention]

[0006] Although conventional satellites can compress the amount of observation data, they also transmit observation data other than the observation target to the ground station. This can result in a long time required for data transmission (including the time for downlinking from the satellite, the waiting time in the ground station buffer storage, and the transmission time from the buffer storage to the data processing center). As a result, conventional satellites may not be able to quickly find the observation target. In addition, conventional software detects ships in the sea area included in the observation target based on an image representing the observation target, so the difference between the time when the ship is observed by the satellite and the time when the ship is detected by the software may be about 10 hours to several days. As a result, it may not be possible to detect ships in the desired sea area that the user wants to monitor at the desired time when the user wants to detect them.

[0007] The present invention provides a flying object and a program that, when transmitting data to a ground station, can transmit a smaller amount of data than the amount of observation data, and can more quickly notify a user that a desired detection target has been detected. [Means for solving the problem]

[0008] One aspect of the present invention is a flying object comprising an observation data generation unit that generates observation data based on radio waves received by a radar, an image generation unit that generates an image representing a monitored space based on the observation data generated by the observation data generation unit, and a detection unit that detects a detection target based on the image generated by the image generation unit.

[0009] Another aspect of the present invention is a flying object comprising an observation data generation unit that generates observation data based on radio waves received by a radar, a processing unit that range-compresses the observation data generated by the observation data generation unit, and a detection unit that detects a detection target based on the signal range-compressed by the processing unit.

[0010] Another aspect of the present invention may be a configuration in which, in the flying object, the monitored space includes an ocean area, and the detection target includes a ship in the ocean area.

[0011] In another aspect of the present invention, in the flying object, the detection unit may be configured to detect an object that is estimated to be the detection target from among candidates for the detection target in the monitored space based on a plurality of pre-stored parameters.

[0012] In another aspect of the present invention, in a flying object, a configuration may be used in which the detection unit detects the detection target by comparing a base map with an image generated by the image generation unit.

[0013] In another aspect of the present invention, in the flying object, a configuration may be used in which the detection target includes crustal movement in the monitored space.

[0014] In another aspect of the present invention, a configuration may be used in which the flying object further includes a position calculation unit that calculates a position of the detection target detected by the detection unit and generates position information indicating the calculated position.

[0015] In another aspect of the present invention, a configuration may be used in which the flying object further includes a feature extraction unit that extracts features of the detection target detected by the detection unit.

[0016] Another aspect of the present invention is a program that causes a computer equipped in an airborne vehicle to generate observation data based on radio waves received by a radar, generate an image representing a monitored space based on the generated observation data, and detect a detection target in the generated image.

[0017] Another aspect of the present invention is a program that causes a computer included in an airborne vehicle to generate observation data based on radio waves received by a radar, range-compress the generated observation data, and detect a detection target based on the range-compressed signal. Effect of the Invention

[0018] According to the above-mentioned flying object and program, when transmitting data to the ground station, the amount of data to be transmitted can be made smaller than the amount of observation data, so that the user can be notified more quickly that the desired detection target has been detected. [Brief description of the drawings]

[0019] [Figure 1] 1 is a diagram showing an example of the configuration of a satellite observation system 1. FIG. [Diagram 2] FIG. 2 is a diagram illustrating an example of a hardware configuration of a control device 3. [Diagram 3] 2 is a diagram illustrating an example of a functional configuration of a control device 3. FIG. [Figure 4] 10 is a flowchart showing an example of a process flow in which the control device 3 detects ships included in the area D based on observation data. [Diagram 5] 5 is a diagram showing an example of a binarized image. Image P1 shown in Fig. 5 is an example of a binarized image. [Figure 6] 13 is a flowchart showing another example of the processing flow in which the control device 3 detects ships included in the area D2 based on observation data. [Figure 7] FIG. 13 is a diagram showing an example of a compressed data image in which one ship is included in area D2. [Figure 8] 13 is a flowchart showing yet another example of the process flow in which the control device 3 detects a detection target included in the area D2 based on observation data. [Figure 9] 13 is a flowchart showing an example of a process flow in which the control device 3 detects crustal deformation occurring in at least a part of the land area included in the area D based on observation data. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] <Embodiment> Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0021] <Outline of the satellite observation system> First, an overview of a satellite observation system 1 according to an embodiment will be described.

[0022] The satellite observation system 1 includes an airborne vehicle 2 and a receiving device 4 . In this embodiment, the flying object 2 is an artificial satellite that orbits above the surface of the Earth ET along a predetermined orbit. The flying object 2 may be an artificial satellite that orbits above the surface of another celestial body or object along the orbit instead of above the surface of the Earth ET. The celestial body includes planets other than the Earth ET, such as Mars and Venus, satellites such as the Moon and Titan, and asteroids such as Itokawa. The object includes rocks and the like. The flying object 2 may be another flying object, such as an airplane or a drone, instead of an artificial satellite.

[0023] In this embodiment, the flying object 2 observes an area D, which is a part of the area included in the surface of the Earth ET, as an observation target, which is a desired object to be observed. That is, the flying object 2 irradiates (transmits) radio waves toward the area D. In other words, the observation target is a monitored space monitored by the flying object 2. The monitored space (i.e., the observation target) may be a two-dimensional plane like the area D in this example, or may be a three-dimensional space instead of a two-dimensional plane. Note that the flying object 2 may be configured to observe another object present on the Earth ET as an observation target, instead of a configuration in which a part of the region is observed as an observation target. When the flying object 2 passes through an area above the surface of the earth where radio waves can be irradiated to the region D, the attitude of the flying object 2 is controlled so that the radio waves irradiated from the flying object 2 are irradiated to the region D. In the following, the method of controlling the attitude of the flying object 2 may be a known control method or a method to be developed in the future, so a description thereof will be omitted. The flying object 2 observes the region D by irradiating the region D with radio waves and then receiving the radio waves reflected by the surface of the region D.

[0024] More specifically, the flying object 2 includes a synthetic aperture radar unit 21, a communication antenna unit 22, and a control device 3. The flying object 2 is an example of a flying object.

[0025] The synthetic aperture radar unit 21 is provided with a plurality of antenna elements arranged in a first direction as a phased array antenna A1. In the following description, the phased array antenna A1 has both transmitting and receiving functions. However, the phased array antenna A1 may have a configuration in which the transmitting antenna and the receiving antenna are separate. The phased array antenna A1 is provided at a predetermined position of the synthetic aperture radar unit 21. This position is a position where the phased array antenna A1 can irradiate (transmit) radio waves in a second direction. The radio waves are radio waves corresponding to a signal acquired by the synthetic aperture radar unit 21 from the control device 3. The second direction is a direction perpendicular to the first direction. When the flying object 2 passes through an area above the surface of the earth ET where radio waves can be irradiated to an area D, the second direction coincides with the direction toward the area D. In the following, a case where the first direction is the azimuth direction will be described as an example. The azimuth direction is the traveling direction of the flying object 2. That is, the second direction in this example is the range (slant range in the side-looking method) direction. The phased array antenna A1 receives radio waves arriving toward the phased array antenna A1. The first direction may be another direction instead of the azimuth direction. That is, the second direction may be another direction instead of the range direction. The synthetic aperture radar unit 21 may be configured to include another antenna such as a slot antenna or a parabolic antenna instead of the phased array antenna A1. The antenna included in the synthetic aperture radar unit 21 may be a separate transmitting antenna and receiving antenna, similar to the phased array antenna A1 in this embodiment. In this case, the flying object 2 may include only the receiving antenna, and may form a tandem satellite with the satellite including the transmitting antenna.

[0026] The communication antenna unit 22 includes an antenna for transmitting and receiving radio waves corresponding to various information to and from the receiving device 4. The antenna is not particularly limited, and may be a parabolic antenna or a phased array antenna.

[0027] The control device 3 controls the entire flying object 2. In this embodiment, the control device 3 is built into the flying object 2. The control device 3 may be separate from the flying object 2. In this case, for example, the control device 3 is provided in another artificial satellite, and the flying object 2 is controlled from that artificial satellite via wireless communication.

[0028] The control device 3 outputs a transmission pulse signal to the synthetic aperture radar unit 21 at PRI (Pulse Repetition Frequency) intervals within a synthetic aperture time (i.e., one cycle). As a result, the control device 3 causes the phased array antenna A1 to irradiate (transmit) radio waves corresponding to the transmission pulse signal as irradiated radio waves toward the area D. The transmission pulse signal is a chirp signal, and is a signal whose frequency changes over time. In this embodiment, the frequency band of the transmission pulse signal is the microwave frequency band. Note that the frequency band of the transmission pulse signal may be a frequency band lower than the microwave frequency band or a frequency band higher than the microwave frequency band, instead of the microwave frequency band.

[0029] Here, some of the radio waves emitted from the phased array antenna A1 are reflected toward the phased array antenna A1 by virtual point-like backscatterers at each position on the surface of the region D. The control device 3 receives the radio waves reflected in this manner by the phased array antenna A1. For convenience of explanation, therefore, in the following description, the point-like backscatterers at each position will be referred to as backscatterers.

[0030] The control device 3 generates observation data based on the strength of radio waves received by the phased array antenna A1 and the time at which the radio waves were received. The observation data is two-dimensional data having cells representing the time at which the phased array antenna A1 received the radio waves. Each cell of the observation data is associated with the strength of the radio waves received by the phased array antenna A1 at the time represented by each cell. The method by which the control device 3 generates the observation data may be a known method or a method to be developed in the future. For this reason, further detailed explanation of the method (e.g., explanation of the process of removing noise, etc.) will be omitted below.

[0031] Here, a control device X (for example, a conventional control device) different from the control device 3 outputs the generated observation data to the communication antenna unit 22, and causes the communication antenna unit 22 to transmit radio waves corresponding to the observation data to a receiving device installed on the surface of the Earth ET. This allows the receiving device to transmit the observation data received from the control device X to a device corresponding to the request at a timing corresponding to the request in response to a request from the user. However, the timing is often not determined by the convenience of a single user. Also, the time required to transmit (transmit) the observation data from the control device X to the receiving device becomes longer as the amount of observation data increases. Due to these circumstances, there are cases where the satellite observation system equipped with the control device X cannot detect the detection target in the area D at the timing desired by the user. Here, the user is, for example, a person who operates the device on the ground. Note that the user may be another person instead of the person.

[0032] Therefore, the control device 3 detects the detection target in the area D (i.e., the observation target) based on the generated observation data. That is, the flying object 2 itself detects the detection target in the area D based on the observation data. As a result, the flying object 2 can make the amount of data transmitted to the target (in this example, the receiving device 4) to which the data is transmitted smaller than the amount of observation data. Also, the flying object 2 can reduce the memory capacity required to store the data. As a result, the flying object 2 can notify the user that the desired detection target has been detected more quickly. Also, for observations in areas where the data can be directly transmitted to the ground station, the transmission data may not be stored in the memory unit and may be sent directly to the communication unit. After detecting the detection target, the control device 3 generates transmission data including information indicating the result of detecting the detection target. The control device 3 outputs the generated transmission data to the communication antenna unit 22 and causes the communication antenna unit 22 to transmit the transmission data toward the receiving device 4. The control device 3 receives radio waves corresponding to various control signals from the receiving device 4 via the communication antenna unit 22. The control device 3 performs processing corresponding to the received radio waves.

[0033] The receiving device 4 is equipped with an antenna capable of transmitting and receiving various information via radio waves to and from the flying object 2. The receiving device 4 is, for example, a dedicated or general-purpose computer connected to the antenna. The receiving device 4 is installed on the surface of the Earth ET at a location desired by the user. The receiving device 4 receives the transmission data transmitted from the flying object 2 to the receiving device 4 as received data. The receiving device 4 stores the received data. This allows the user to perform an operation according to the information indicating the detection target contained in the received data stored by the receiving device 4.

[0034] Hereinafter, the functional configuration of the control device 3 and the process of the control device 3 detecting a detection target in the area D based on observation data will be described in detail.

[0035] <Control device hardware configuration> The hardware configuration of the control device 3 will be described below with reference to Fig. 2. Fig. 2 is a diagram showing an example of the hardware configuration of the control device 3.

[0036] The control device 3 includes, for example, an FPGA (Field Programmable Gate Array) 31, a storage unit 32, and a communication unit 34. These components are connected to each other via a bus so that they can communicate with each other. The control device 3 also communicates with each of the synthetic aperture radar unit 21 and the communication antenna unit 22 via the communication unit 34. The control device 3 may be configured to include a CPU (Central Processing Unit) instead of the FPGA 31.

[0037] The FPGA 31 realizes the functional configuration of the control device 3, which will be described later, by using a hardware function unit. The storage unit 32 includes, for example, an Electrically Erasable Programmable Read-Only Memory (EEPROM), a Read-Only Memory (ROM), a Random Access Memory (RAM), a flash memory, etc. The storage unit 32 stores various types of information and various images processed by the control device 3. The communication unit 34 includes, for example, analog or digital input / output ports and the like that comply with various communication standards.

[0038] <Control device functional configuration> The functional configuration of the control device 3 will be described below with reference to Fig. 3. Fig. 3 is a diagram showing an example of the functional configuration of the control device 3.

[0039] The control device 3 includes a storage unit 32, a communication unit , and a control unit .

[0040] The control unit 36 ​​controls the entire control device 3. The control unit 36 ​​includes a communication control unit 361, a radar control unit 363, an observation data generation unit 364, a processing unit 365, an image generation unit 367, a detection target detection unit 369, a position calculation unit 371, a feature extraction unit 373, and a transmission data generation unit 375. Some or all of these functional units included in the control unit 36 ​​may be hardware functional units such as a large scale integration (LSI) or an application specific integrated circuit (ASIC).

[0041] The communication control unit 361 transmits and receives radio waves corresponding to various information to and from the receiving device 4 via the communication antenna unit 22. Specifically, for example, the communication control unit 261 receives radio waves corresponding to various control signals from the receiving device 4 via the communication antenna unit 22. In addition, the communication control unit 261 causes the communication antenna unit 22 to transmit, to the receiving device 4, radio waves corresponding to the transmission data generated by the transmission data generation unit 275.

[0042] The radar control unit 363 outputs a transmission pulse signal to the synthetic aperture radar unit 21 at PRI intervals within a synthetic aperture time (i.e., one cycle). Then, the radar control unit 363 causes the phased array antenna A1 of the synthetic aperture radar unit 21 to irradiate (transmit) an irradiated radio wave corresponding to the transmission pulse signal toward the area D. The radar control unit 363 also causes the phased array antenna A1 to receive, by the phased array antenna A1, a part of the irradiated radio waves irradiated from the phased array antenna A1, which are reflected by each backscatterer.

[0043] The observation data generation unit 364 generates the above-mentioned observation data based on the radio waves received by the phased array antenna A1 of the synthetic aperture radar unit 21.

[0044] The processing unit 365 performs various processes on the observation data generated by the observation data generating unit 364 .

[0045] The image generating unit 367 generates an image representing the area D (i.e., the observation target) based on the observation data generated by the processing unit 365.

[0046] The detection target detection unit 369 detects the detection target in the area D based on the observation data generated by the observation data generation unit 364.

[0047] Position calculation section 271 calculates the position of the detection target detected by detection target detection section 369 based on the observation data generated by observation data generation section 364 .

[0048] Feature extraction section 373 extracts features of the detection target detected by detection target detection section 369 based on the observation data generated by observation data generation section 364 .

[0049] The transmission data generation unit 375 generates information including some or all of the image generated by the image generation unit 367, the position information indicating the position calculated by the position calculation unit 271, and the feature information indicating the features detected by the feature extraction unit 373 as transmission data.

[0050] <Processing by which the control device detects a detection target in an observation target based on observation data> Hereinafter, with reference to Fig. 4, a process in which the control device 3 detects a detection target in area D based on observation data will be described. As an example, a process in which the control device 3 detects a ship in area D as a detection target based on observation data will be described below. In this case, area D includes at least an ocean area among areas included in the surface of the Earth ET. Fig. 4 is a flowchart showing an example of a process flow in which the control device 3 detects a ship in area D (i.e., a ship in the ocean area) based on observation data. Note that the detection target may be configured to include other objects and phenomena in addition to a ship.

[0051] When the flying object 2 is located at a position where the synthetic aperture radar unit 21 can irradiate radio waves to at least a part of the area D, the radar control unit 363 controls the synthetic aperture radar unit 21 to observe the area D (step S110). Specifically, in this case, the radar control unit 363 outputs a transmission pulse signal to the synthetic aperture radar unit 21 at PRI intervals within the synthetic aperture time (i.e., one cycle). Then, the radar control unit 363 causes the phased array antenna A1 of the synthetic aperture radar unit 21 to irradiate (transmit) radio waves corresponding to the transmission pulse signal toward the area D. When the flying object 2 moves to a position where the synthetic aperture radar unit 21 cannot irradiate radio waves to at least a part of the area D, the radar control unit 363 stops outputting the transmission pulse signal to the synthetic aperture radar unit 21. In addition, the radar control unit 363 receives, by the phased array antenna A1, a part of the irradiated radio waves irradiated from the phased array antenna A1, which are reflected by each backscattering object. Then, the radar control unit 363 performs processing such as A / D conversion on the received radio wave information indicating the received radio waves. Note that in this embodiment, a description of this processing is omitted. In this manner, the radar control unit 363 observes the area D using the synthetic aperture radar unit 21.

[0052] Next, the observation data generating unit 364 generates observation data based on the received radio wave information indicating the radio waves received in step S110 (step S120). Here, the observation data generating unit 364 may be configured to store the generated observation data in the storage unit 32 after generating the observation data in step S120, or may be configured not to store the observation data in the storage unit 32.

[0053] Next, the image generating unit 367 generates an image representing the region D based on the observation data generated in step S120 (step S150). Specifically, the image generating unit 367 compresses the observation data in the range direction and in the azimuth direction to generate an image representing the region D. Here, the position of each pixel of the image generated by the image generating unit 367 represents the position of the backscatterer on the image, and the pixel value of each pixel represents the luminance value of the radio wave reflected by the backscatterer. In addition, each pixel of the image is associated with phase information indicating the phase of the radio wave arriving from the backscatterer represented by each pixel toward the phased array antenna A1. The image generating unit 367 generates the image representing the region D by, for example, a range Doppler method, a chirp scaling method, an omega-kappa method, or the like. At this time, the image generating unit 367 may improve the calculation accuracy of the phase by performing a high-precision orbit determination process and an ionospheric delay correction. The image generating unit 367 may be configured to perform a process of removing phase components due to the altitude on the surface of the region D using a digital elevation model previously stored in the storage unit 32. In this case, the image generating unit 367 reads out the digital elevation model from the storage unit 32. Here, the digital elevation model is a three-dimensional model that represents the shape of at least a part of the surface of the Earth ET. The image generating unit 367 may be configured to generate the image from the observation data by a known method different from these methods, or may be configured to generate the image from the observation data by a method to be developed in the future. For this reason, the method of generating the image by the image generating unit 367 will not be described in further detail.

[0054] Next, based on the image generated by the image generating unit 367 in step S150, the detection target detection unit 369 detects an object that is considered (estimated) to be a ship included in the image as a ship (step S160). Note that the object that is considered (estimated) to be a ship included in the image may include an object such as a structure on the sea, a small island, etc., in addition to a real ship. Here, the processing of step S160 will be described. For example, the detection target detection unit 369 reads out land area information indicating a land area that is an area different from a sea area among the areas included in the area D from the storage unit 32. The land area is a land area among the areas included in the area D. The detection target detection unit 369 applies a land area filter to the image based on the read land area information. Specifically, the detection target detection unit 369 changes the luminance value of a pixel included in the land area portion indicated by the land area information in the image to a predetermined first luminance value. The first luminance value is, for example, 0. Note that the first luminance value may be any luminance value, instead of 0, as long as the detection target detection unit 369 can distinguish the difference between the first luminance value and a third luminance value described later. For ease of explanation, the image obtained by applying the land area filter to the image will be referred to as a land area removed image below.

[0055] After generating the land area removed image, the detection target detection unit 369 applies a binarization filter to the land area removed image. Specifically, the detection target detection unit 369 performs binarization of the land area removed image by setting the luminance values ​​of pixels that make up the land area removed image and have luminance values ​​less than a predetermined luminance value to a second luminance value and setting the luminance values ​​of pixels that make up the land area removed image and have luminance values ​​equal to or greater than a first predetermined luminance value to a third luminance value that is determined in advance. The second luminance value is, for example, 0. Note that the second luminance value may be any luminance value instead of 0 as long as the detection target detection unit 369 can distinguish the difference in luminance value from a third luminance value described later. The third luminance value is, for example, 255. Note that the third luminance value may be any luminance value instead of 255 as long as the detection target detection unit 369 can distinguish the difference in luminance value from both the first luminance value and the second luminance value. The first predetermined luminance value may be any luminance value that is greater than both the first luminance value and the second luminance value and less than the third luminance value. For ease of explanation, the image obtained after applying the binarization filter to the land area removed image will be referred to as a binarized image below. FIG. 5 is a diagram showing an example of a binarized image. Image P1 shown in FIG. 5 is an example of a binarized image. In image P1, region SC1 represents an area made up of pixels whose luminance value is the third luminance value. Furthermore, hatched region SC2 represents an area made up of pixels whose luminance value is the second luminance value. Furthermore, detection target detection unit 369 may be configured to binarize the land area removed image using a standard deviation filter instead of the binarization filter.

[0056] When the binarized image contains one or more regions constituted by a predetermined number or more pixels, each of which has a luminance value of the third luminance value, the detection target detection unit 369 detects each of the regions (i.e., the regions corresponding to the object considered to be a ship as described above) as a ship. The predetermined number is, for example, 10. Note that the predetermined number may be a number smaller than 10 or larger than 10. Here, the region SC1 shown in FIG. 5 is an example of a region constituted by a predetermined number of pixels, each of which has a luminance value of the third luminance value. That is, in the example shown in FIG. 5, the detection target detection unit 369 detects the region SC1 as a ship.

[0057] The detection target detection unit 369 in step S160 may be configured to detect a ship from the image generated in step S150 using a machine learning algorithm without applying at least a part of the land area filter, the binarization filter, and the standard deviation filter to the image. In this case, the detection target detection unit 369 pre-stores (learns) a plurality of images including a ship and information associating the position and shape of the ship in each of the images, and a plurality of images not including a ship and information indicating that the ship is not included in each of the images, as a plurality of parameters. Here, each of the images is a ship image cut out from an image of one scene. Then, the detection target detection unit 369 detects a likely candidate as a ship from among the ship candidates included in the image generated in step S150 based on the plurality of parameters stored in advance. In this case, the detection target detection unit 369 may be configured to detect the likely candidate as a ship using the algorithm in step S160 and to perform the process of step S180 described later (specifically, extraction of the characteristics of the ship detected in step S160). Here, the algorithm may be any known algorithm (including deep learning) or an algorithm to be developed in the future. For this reason, further detailed explanation of the machine learning algorithm will be omitted.

[0058] After the process of step S160 is performed, the detection target detection unit 369 determines whether or not a ship has been detected in step S160 (step S165). If the detection target detection unit 369 determines that a ship has not been detected in step S160 (step S165-NO), the control unit 36 ​​ends the process. On the other hand, if the detection target detection unit 369 determines that a ship has been detected in step S160 (step S165-YES), the image generation unit 367, the position calculation unit 271, the feature extraction unit 373, the communication control unit 361, and the transmission data generation unit 375 each repeat the processes of steps S170 to S210 for one or more ships detected in step S160 (step S167). In the following, for convenience of explanation, the ship selected in step S167 will be referred to as the target ship.

[0059] After the target ship is selected in step S167, the position calculation unit 271 calculates the position of the target ship (step S170). Specifically, the position calculation unit 271 calculates the latitude and longitude represented by a predetermined position of the area detected as the target ship among the areas included in the binarized image in step S160 as the position of the target ship. At this time, the position calculation unit 271 acquires flying object position information (e.g., GPS (Global Positioning System) information) indicating the position of the flying object 2 at each time and flying object attitude information (e.g., ACS (Attitude Control System) information) indicating the attitude of the flying object 2 at each time, and calculates the latitude and longitude represented by the predetermined position as the position of the target ship based on the acquired flying object position information and flying object attitude information. The predetermined position is, for example, the position of the centroid (or center of gravity) of the area. Note that the predetermined position may be another position based on the area instead of the position of the centroid of the area.

[0060] Next, the image generating unit 367 trims the image generated in step S150 based on the position of the target ship calculated in step S170 (step S175). Specifically, the image generating unit 367 trims (cuts out) a partial image representing an area of ​​a predetermined shape centered on the position of the target ship calculated in step S170 among the areas included in area D from the image. The predetermined shape is, for example, a rectangle with a predetermined distance on each side. The predetermined shape is a rectangular area having sides parallel to the latitude direction and sides parallel to the longitude direction in the image. The predetermined distance is, for example, 500 meters. The predetermined distance may be shorter than 500 meters or longer than 500 meters. The predetermined shape may be another shape such as a circle or an ellipse instead of a rectangle. The image generating unit 367 generates the trimmed partial image as a transmission image.

[0061] Next, the feature extraction unit 373 extracts the features of the target ship (step S180). In this example, the features include the overall length of the target ship, the type of the target ship, the course of the target ship, the speed of the target ship, and the navigation state of the target ship. The navigation state of the target ship is either a state in which the target ship is stopped or a state in which the target ship is moving. Note that the features may be configured to include other information representing the features of the target ship instead of some or all of these, or may be configured to include other information representing the features of the target ship in addition to some or all of these. Here, the processing of step S180 will be explained.

[0062] The feature extraction unit 373 uses a machine learning algorithm to extract features of the target ship based on the binary image generated in step S160 and the transmission image generated by the image generation unit 367. In this case, the feature extraction unit 373 pre-stores (learns) information in which the features of the area representing the target ship and each of the features of the target ship are associated with each other as multiple parameters. In this example, the features of the area include the longitudinal length of the area, the lateral length of the area, the direction in which the longitudinal direction of the area faces, the shape of the area, and the area of ​​the area. Note that the features may be configured to include other information representing the area instead of some or all of these features, or may be configured to include other information representing the area in addition to some or all of these features.

[0063] That is, the feature extraction unit 373 pre-stores information as a plurality of parameters that associate a combination of the longitudinal length of the area representing the target ship, the transverse length of the area, the longitudinal direction of the area, the shape of the area, and the area of ​​the area with the overall length of the ship.The feature extraction unit 373 then extracts a plausible candidate for the overall length of the target ship from among the candidates for the overall length of the target ship based on the pre-stored parameters and the area representing the target ship included in the binarized image.

[0064] The feature extraction unit 373 also stores in advance, as a plurality of parameters, information associating a combination of the longitudinal length of the area representing the target ship, the transverse length of the area, the longitudinal direction of the area, the shape of the area, and the area of ​​the area with the type of ship.The feature extraction unit 373 then extracts a plausible candidate as the type (category) of the target ship from among the candidates based on the prestored parameters and the area representing the target ship included in the binarized image.

[0065] The feature extraction unit 373 also stores information in advance as a plurality of parameters that associate a combination of the longitudinal length of the area representing the target ship, the transverse length of the area, the longitudinal direction of the area, the shape of the area, and the area of ​​the area with the ship's course.The feature extraction unit 373 then extracts a plausible candidate as the target ship's course from among the candidates for the course of the target ship based on the prestored parameters and the area representing the target ship included in the binary image.

[0066] The feature extraction unit 373 also stores in advance, as a plurality of parameters, information associating the combination of the longitudinal length of the area representing the target ship, the transverse length of the area, the longitudinal direction of the area, the shape of the area, the area of ​​the area, and the wake (if there is a wake outside the area), with the ship's speed.The feature extraction unit 373 then extracts a plausible candidate as the target ship's speed from among the candidates for the speed, based on the prestored parameters and the area representing the target ship included in the binarized image.

[0067] The feature extraction unit 373 also stores in advance, as a plurality of parameters, information associating a combination of the longitudinal length of the area representing the target ship, the transverse length of the area, the longitudinal direction of the area, the shape of the area, the area of ​​the area, and the wake (if there is a wake outside the area), with the ship's navigation state. The feature extraction unit 373 then extracts, from among the candidates for the navigation state of the target ship, a candidate that is most likely to be the navigation state, based on the prestored parameters and the area representing the target ship included in the binary image.

[0068] Here, the machine learning algorithm used by the feature extraction unit 373 in step S180 may be any known algorithm (including deep learning) or an algorithm to be developed in the future. For this reason, further detailed description of the machine learning algorithm will be omitted. Note that the feature extraction unit 373 may be configured to use the machine learning algorithm to extract features of the target ship from the image generated in step S150 (i.e., the image before the binarization filter is applied in step S160 or the image to which the standard deviation filter is applied).

[0069] Next, the transmission data generation unit 375 generates transmission data (step S200). Specifically, when an AIS signal is received in step S190, the transmission data generation unit 375 generates information including ship identification information of the target ship, ship position information of the target ship, a transmission image, ship characteristic information of the target ship, and AIS information as transmission data. The ship identification information is information that identifies the target ship. Note that the ship identification information may be any information that can identify each of the one or more ships detected in step S160. The ship position information is information indicating the position of the target ship, which is the position calculated in step S170. The transmission image is the transmission image generated in step S175. The ship characteristic information is information indicating each of the characteristics of the target ship, which is the characteristics detected in step S180. The AIS information is the AIS information stored in the memory unit 32 in step S190. In addition, if an AIS signal is not received in step S190, the transmission data generation unit 375 generates information as transmission data including the ship identification information, the ship position information, the transmission image, the ship characteristic information, and information indicating that an AIS signal was not received in step S190.

[0070] In addition, when an AIS signal is received in step S190, the transmission data generation unit 375 may be configured to compare (match) the ship characteristic information of the target ship with the AIS information indicated by the AIS signal. In this case, the transmission data generation unit 375 identifies information among the multiple pieces of information indicated by the ship characteristic information that matches any of the multiple pieces of information indicated by the AIS information. In addition, the transmission data generation unit 375 identifies information among the multiple pieces of information indicated by the ship characteristic information that does not match any of the multiple pieces of information indicated by the AIS information. In step S200, the transmission data generation unit 375 generates information including the identified information, ship identification information of the target ship, ship position information of the target ship, the transmission image, the ship characteristic information of the target ship, and the AIS information as transmission data.

[0071] Next, the transmission data generating unit 375 stores the transmission data generated in step S200 in the storage unit 32 (step S210).

[0072] In this way, the flying object 2 can generate transmission data for each of the one or more ships detected in step S160 and store the generated transmission data in the memory unit 32 by repeatedly performing the processes of steps S167 to S210.

[0073] After steps S167 to S210 are repeated, the communication control unit 361 outputs each of the transmission data stored in the storage unit 32 in step S210 to the communication antenna unit 22, causes the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4 (step S220), and ends the process. This allows the flying object 2 to, for example, reduce the amount of data to be transmitted to the receiving device 4 compared to the amount of observation data, and shortens the time required to provide the user with information indicating that a ship, which is an example of a detection target, has been detected.

[0074] In step S220, the communication control unit 361 may be configured to output a part of the transmission data stored in the storage unit 32 to the communication antenna unit 22 and cause the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4. In addition, in step S200, the communication control unit 361 may be configured to output the transmission data generated by the transmission data generation unit 375 to the communication antenna unit 22 and cause the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4.

[0075] Furthermore, some or all of the detection target detection unit 369, position calculation unit 271, and feature extraction unit 373 described above may be configured as an integrated functional unit. When the detection target detection unit 369 and the position calculation unit 271 are an integrated functional unit, the detection target detection unit 369 configured integrally with the position calculation unit 271 detects one or more ships in step S160 and calculates (detects) the position of each of the detected ships. When the detection target detection unit 369 and the feature extraction unit 373 are an integrated functional unit, the detection target detection unit 369 configured integrally with the feature extraction unit 373 detects one or more ships in step S160 and extracts the features of each of the detected ships. Furthermore, if the detection target detection unit 369, the position calculation unit 271, and the feature extraction unit 373 are an integrated functional unit, the detection target detection unit 369 configured integrally with the position calculation unit 271 and the feature extraction unit 373 detects one or more ships in step S160, calculates (detects) the position of each detected ship, and further extracts the features of each detected ship. Furthermore, in the flowchart shown in Figure 4, the processes of steps S175 to S190 may be performed in parallel with each other, or may be performed in an order different from that shown in Figure 4.

[0076] In the above embodiment, the detection process is shown to perform feature extraction (S180) after ship detection (S160), but is not limited to the above, and for example, ship detection may be performed as part of feature extraction (ship detection and feature detection may be performed at the same timing). In other words, image generation (S150) in Fig. 4 may be performed, and then features (whether it is a ship, length (total length of the ship), direction of travel (direction in which the ship is sailing), type (ship type), etc.) may be extracted.

[0077] As described above, the flying object 2 according to this embodiment generates observation data based on radio waves received by a radar (in this example, the synthetic aperture radar unit 21), generates an image showing a monitored space (in this example, the observation target) based on the generated observation data, and detects a detection target (in this example, a ship) based on the generated image. This allows the flying object 2 to make the amount of data transmitted to a target to which data is to be transmitted (in this example, the receiving device 4) smaller than the amount of observation data, and allows the user to be notified more quickly that the desired detection target has been detected.

[0078] In addition, the flying object 2 detects a likely candidate as a detection target from among candidates of detection targets in the monitored space based on a plurality of pre-stored parameters, which allows the flying object 2 to quickly notify the user that a desired detection target has been detected based on a plurality of pre-stored parameters.

[0079] In addition, the flying object 2 calculates the position of the detected detection target and generates position information indicating the calculated position, which enables the flying object 2 to more quickly notify the user that the desired detection target has been detected and also notify the user of the target's position.

[0080] In addition, the flying object 2 extracts the characteristics of the detected detection target, which allows the flying object 2 to quickly notify the user that the desired detection target has been detected, as well as the characteristics of the detection target.

[0081] <First Modification of the Embodiment> Hereinafter, a first modified example of the embodiment will be described with reference to Figs. 6 and 7. In the first modified example of the embodiment, the same components as those in the embodiment are denoted by the same reference numerals, and the description thereof will be omitted. In the first modified example, the control device 3 detects one or more ships from the observation data without generating an image representing the area D. Specifically, the control device 3 executes the process of the flowchart shown in Fig. 6 instead of the process of the flowchart shown in Fig. 4. In the first modified example, as in the embodiment, a case where the detection target is a ship will be described as an example. In the first modified example, instead of the configuration for observing the area D, the flying object 2 will be described as observing the area D2. The area D2 is an area that includes only the sea area (i.e., does not include the land area described above) among the areas included in the surface of the Earth ET.

[0082] Fig. 6 is a flowchart showing another example of the process flow in which the control device 3 detects ships in the area D2 based on observation data. Note that the processes of steps S110 to S120, steps S165 to S167, steps S180 to S210, and step S220 shown in Fig. 6 are similar to the processes of steps S110 to S120, steps S165 to S167, steps S180 to S210, and step S220 shown in Fig. 4, respectively, and therefore descriptions thereof will be omitted.

[0083] 6, the detection target detection unit 369 detects one or more ships in the area D2 based on the observation data generated in step S120 (step S310). Here, the process of step S310 will be described.

[0084] The detection target detection unit 369 performs pulse compression in the range direction on the observation data generated in step S120 shown in Fig. 6 based on the transmitted chirp signal. For ease of explanation, the observation data that has been subjected to this pulse compression (range compression) will be referred to as compressed data below.

[0085] Here, when one or more ships are included in area D2, a compressed data image is produced in which the strength of the radio waves contained in the compressed data is plotted on a graph with the number of range cells on the horizontal axis and the number of azimuth cells on the vertical axis, and in the compressed data image composed of multiple cells, one or more ship areas appear, which are areas corresponding to the ships and are arc-shaped areas as shown in Figure 7. This arc-shaped region is a region that represents the range curvature included in the compressed data. Here, the number of range cells (written as range cells in FIG. 7) is the number of cells on the horizontal axis of the compressed data image, and is a numerical value that can be converted into a range distance. Also, the number of azimuth cells (written as azimuth cells in FIG. 7) is the number of cells on the vertical axis of the compressed data image, and is a numerical value that can be converted into a time. FIG. 7 is a diagram showing an example of a compressed data image in which one ship is included in the region D2. Image P2 shown in FIG. 7 is an example of a compressed data image. The luminance value of the pixels that make up image P2 represents the intensity. The luminance value increases as the intensity increases. Region F1 shown in FIG. 7 is an example of a ship region corresponding to the one ship. Also, in the example shown in FIG. 7, the luminance values ​​of the pixels that make up region F1 are equal to or greater than a second predetermined luminance value. The ship region has a partial region that is approximately parallel to the vertical axis in the compressed data image. In the example shown in FIG. 7, the partial region is a partial region of the region F1, which is indicated by the region W1 in the image P2. The detection target detection unit 369 can detect the ship area by detecting such a partial area from the compressed data.

[0086] Specifically, the detection target detection unit 369 selects one or more range curvatures included in the compressed data one by one. The detection target detection unit 369 calculates a total intensity, which is the sum of the intensity of radio waves in each cell that constitutes the selected range curvature (for example, calculates the total value by integrating the intensity). The detection target detection unit 369 identifies a range curvature that corresponds to a total intensity equal to or greater than a predetermined intensity among the total intensities calculated for each range curvature. The detection target detection unit 369 detects each of the identified one or more range curvatures as a ship.

[0087] If the area D2 includes a land area, the detection target detection unit 369 can perform the process of step S310 by using a method such as applying a land area filter to the compressed data. The detection target detection unit 369 may be configured to detect a ship area from the compressed data using a machine learning algorithm. In this case, the detection target detection unit 369 stores (learns) in advance, as a plurality of parameters, information associating a plurality of compressed data including a ship with the position and shape of the ship area in each of the compressed data, and a plurality of compressed data not including a ship and information indicating that a ship is not included in each of the compressed data. Then, the detection target detection unit 369 detects, as the ship area, a candidate that is likely to be a ship area from among the candidates of the ship area included in the compressed data generated in step S310 based on the plurality of parameters stored in advance. Here, the algorithm may be any known algorithm (including deep learning) or an algorithm to be developed in the future. For this reason, a further detailed description of the machine learning algorithm will be omitted.

[0088] After the target ship is selected in step S167 shown in FIG. 6, the position calculation unit 271 calculates the position of the target ship (step S320). Specifically, the position calculation unit 271 identifies the cell with the smallest range distance among the cells constituting the range curvature identified as the target ship in step S310. The position calculation unit 271 identifies one or more cells associated with radio wave intensities equal to or greater than a predetermined threshold among the intensities of radio waves associated with the identified cells. Then, the position calculation unit 271 calculates the latitude and longitude corresponding to a cell that is the midpoint between the cell with the oldest reception time and the cell with the newest reception time among the identified one or more cells as the position of the target ship. At this time, the position calculation unit 271 acquires flying object position information indicating the position of the flying object 2 at each time and flying object attitude information indicating the attitude of the flying object 2 at each time, and calculates the latitude and longitude corresponding to the cell as the position of the target ship based on the acquired flying object position information and flying object attitude information. The position calculation unit 271 may be configured to identify the position of the target ship from the compressed data using a machine learning algorithm. In this case, the position calculation unit 271 pre-stores (learns) as a plurality of parameters information associating a plurality of compressed data including a ship with the position, shape, etc. of the ship in each of the compressed data, and a plurality of compressed data not including a ship and information indicating that a ship is not included in each of the compressed data. Then, based on the plurality of pre-stored parameters, the position calculation unit 271 identifies a likely candidate as the ship's position from among the candidates for the ship's position included in the compressed data generated in step S310. Here, the algorithm may be any known algorithm (including deep learning) or an algorithm to be developed in the future. For this reason, a further detailed description of the machine learning algorithm will be omitted.

[0089] Next, based on the observation data generated in step S120 and the position of the target ship calculated in step S320, the image generation unit 367 generates an image representing an area of ​​a predetermined shape centered on that position as a transmission image (step S330). Note that in step S330, the image generation unit 367 may be configured to generate the transmission image based on the compressed data generated in step S310 instead of the observation data. The processing method for generating the transmission image based on the observation data or the compressed data in step S330 may be a known method or a method to be developed in the future. For this reason, further detailed description of the processing method will be omitted below.

[0090] As described above, the flying object 2 according to the first modified embodiment generates observation data based on radio waves received by a radar (in this example, the synthetic aperture radar unit 21), compresses the range of the generated observation data, and detects the detection target (in this example, a ship) based on the range-compressed signal. This allows the flying object 2 to reduce the time required to generate an image representing the area D2, and can more quickly notify the user that the detection target has been detected.

[0091] <Modification 2 of the embodiment> Hereinafter, a second modification of the embodiment will be described with reference to FIG. 8. In the second modification of the embodiment, the same components as those in the embodiment are denoted with the same reference numerals, and the description thereof will be omitted. In the second modification, as in the embodiment, a case where the detection target is a ship will be described as an example. In the second modification, as in the first modification of the embodiment, a case where the flying object 2 observes the area D2 will be described. In the second modification, the control device 3 detects the characteristics of each of the one or more ships in the area D2 from the observation data, without generating an image representing the area D2. Specifically, the control device 3 executes the processing of the flowchart shown in FIG. 8 instead of the processing of the flowchart shown in FIG. 4.

[0092] Fig. 8 is a flowchart showing yet another example of the process flow in which the control device 3 detects a detection target in the area D2 based on observation data. Note that the processes of steps S110 to S320, steps S330 to S210, and step S220 shown in Fig. 8 are similar to the processes of steps S110 to S320, steps S330 to S210, and step S220 shown in Fig. 6, respectively, and therefore descriptions thereof will be omitted.

[0093] After the process of step S320 shown in FIG. 8 is performed, the feature extraction unit 373 extracts the features of the target ship based on the compressed data generated in step S310 (step S410). Specifically, the feature extraction unit 373 extracts the features of the target ship from the compressed data using a machine learning algorithm. In this case, the feature extraction unit 373 pre-stores (learns) the features of the ship area representing the target ship and information in which each of the features of the target ship is associated with each other as multiple parameters. The features of the ship area include, for example, the width in the longitude direction of the ship area included in the compressed data, the shape of the ship area, and the area of ​​the ship area. Note that the features may be configured to include other information representing the ship area instead of some or all of these features, or may be configured to include other information representing the ship area in addition to some or all of them.

[0094] That is, information associating a combination of the longitudinal width of the ship area, the shape of the ship area, and the area of ​​the ship area included in the compressed data with the overall length of the ship is stored in advance as a plurality of parameters in the feature extraction unit 373. Then, the feature extraction unit 373 extracts a plausible candidate for the overall length of the target ship from among the candidates for the overall length of the target ship based on the prestored parameters and the ship area representing the target ship included in the compressed data generated in step S310.

[0095] Furthermore, information associating a combination of the longitudinal width of the ship area, the shape of the ship area, and the area of ​​the ship area included in the compressed data with the type of ship is stored as a plurality of parameters in advance in the feature extraction unit 373. Then, the feature extraction unit 373 extracts a plausible candidate as the type of the target ship from among the candidate types of the target ship based on the prestored parameters and the ship area representing the target ship included in the compressed data generated in step S310.

[0096] Furthermore, information associating a combination of the longitudinal width of the ship area, the shape of the ship area, and the area of ​​the ship area included in the compressed data with the ship's course is stored as a plurality of parameters in advance in the feature extraction unit 373. Then, the feature extraction unit 373 extracts a plausible candidate as the course of the target ship from among the candidates for the course of the target ship based on the prestored parameters and the ship area representing the target ship included in the compressed data generated in step S310.

[0097] Furthermore, information associating a combination of the longitudinal width of the ship area, the shape of the ship area, and the area of ​​the ship area included in the compressed data with the ship speed is stored in advance as a plurality of parameters in the feature extraction unit 373. Then, the feature extraction unit 373 extracts a plausible candidate as the speed of the target ship from among the candidates for the speed of the target ship based on the prestored parameters and the ship area representing the target ship included in the compressed data generated in step S310.

[0098] Furthermore, information associating a combination of the longitudinal width of the ship area included in the compressed data, the shape of the ship area, and the area of ​​the ship area with the ship's navigation state is stored in advance as a plurality of parameters in the feature extraction unit 373. Then, the feature extraction unit 373 extracts a plausible candidate as the navigation state of the target ship from among the candidates for the navigation state of the target ship based on the prestored parameters and the ship area representing the target ship included in the compressed data generated in step S310.

[0099] Here, the machine learning algorithm used by the feature extraction unit 373 in step S410 may be any known algorithm (including deep learning) or an algorithm to be developed in the future. Therefore, further detailed description of the machine learning algorithm will be omitted.

[0100] As described above, the flying object 2 according to the second modification of the embodiment extracts the characteristics of each of one or more ships from the observation data without generating an image representing the area D2. As a result, the flying object 2 can reduce the time required to generate the image, and can more quickly notify the user of ship characteristic information indicating the characteristics of the detected ship.

[0101] <Modification 3 of the embodiment> Hereinafter, a third modification of the embodiment will be described with reference to FIG. 9. In the third modification of the embodiment, the same components as those in the embodiment are denoted by the same reference numerals, and the description thereof will be omitted. In the third modification, a case where the flying object 2 observes the area D will be described, as in the first modification of the embodiment. In the third modification, a case where the detection target is crustal movement will be described. Specifically, the control device 3 detects an uplift or subsidence occurring in at least a part of the land area included in the area D as crustal movement. In order to detect the crustal movement, the control device 3 executes the process of the flowchart shown in FIG. 9 instead of the process of the flowchart shown in FIG. 4. In addition, the detection target may be configured to include other objects and phenomena in addition to crustal movement.

[0102] Fig. 9 is a flowchart showing an example of a process flow in which the control device 3 detects, based on observation data, crustal deformation occurring in at least a part of the land area included in the area D. Note that the processes of steps S110 to S150 shown in Fig. 9 are similar to the processes of steps S110 to S150 shown in Fig. 4, respectively, and therefore description thereof will be omitted.

[0103] 9, the detection target detection unit 369 reads out a base map stored in advance in the storage unit 32 from the storage unit 32 (step S520). In this example, the base map is an image representing the area D generated in step S150 previously executed by the control device 3.

[0104] Next, the detection target detection unit 369 detects crustal movements occurring in at least a part of the land area included in the area D based on the image representing the area D generated in step S150 shown in Fig. 9 and the base map read out in step S520 (step S530). Here, the processing of step S530 will be described. For convenience of explanation, each of the multiple pixels constituting the image will be referred to as a first pixel, and each of the multiple pixels constituting the base map will be referred to as a second pixel.

[0105] The detection target detection unit 369 selects a second pixel corresponding to each of a plurality of first pixels constituting the image representing the region D generated in step S150 shown in FIG. 9. The second pixel is a pixel that represents a backscattering object at the same position as the backscattering object represented by the first pixel and that constitutes the base map. The detection target detection unit 369 calculates the difference between the phase associated with a certain first pixel and the phase associated with the second pixel. The detection target detection unit 369 extracts the phase component resulting from crustal movement after removing unnecessary phase components from the calculated difference. If the extracted phase component is equal to or greater than a predetermined value, the detection target detection unit 369 specifies the first pixel as a third pixel. On the other hand, if the calculated difference is less than a predetermined value, the detection target detection unit 369 specifies the first pixel as a fourth pixel. In this case, the detection target detection unit 369 may be configured not to do anything. The detection target detection unit 369 repeats the process from the selection of the second pixel to the specification of the third pixel or the fourth pixel for each of the first pixels.

[0106] Then, the detection target detection unit 369 identifies adjacent third pixels from the identified one or more third pixels as one group. The detection target detection unit 369 excludes groups that have less than a predetermined number of third pixels from the identified one or more groups as noise. The detection target detection unit 369 detects each of the one or more groups that remain without being excluded as a crustal movement.

[0107] After the process of step S530 is performed, the detection target detection unit 369 judges whether or not a crustal movement is detected in step S530 (step S535). If the detection target detection unit 369 judges that a crustal movement is not detected in step S530 (step S535-NO), the control unit 36 ​​ends the process. On the other hand, if the detection target detection unit 369 judges that a crustal movement is detected in step S530 (step S535-YES), the position calculation unit 271, the image generation unit 367, the transmission data generation unit 375, and the communication control unit 361 each repeat the processes of steps S540 to S560 for one or more crustal movements detected in step S530 (step S357). In the following, for convenience of explanation, the crustal movement selected in step S530 will be referred to as a target crustal movement.

[0108] After the target crustal movement is selected in step S530, the position calculation unit 271 calculates the position of the target crustal movement (step S540). Specifically, the position calculation unit 271 calculates a predetermined position of the group of third pixels detected as the crustal movement in step S530 as the position of the target crustal movement. The predetermined position is, for example, the position of the centroid (or center of gravity) of the area constituted by the group. Note that the predetermined position may be another position based on the area instead of the position of the centroid of the area.

[0109] Next, the image generating unit 367 trims the image generated in step S150 based on the position of the target crustal movement calculated in step S540 (step S545). Specifically, the image generating unit 367 trims (cuts out) a partial image representing a region of a predetermined shape centered on the position of the target crustal movement calculated in step S540 among the regions included in the region D from the image. The predetermined shape is, for example, a rectangle with a predetermined distance on each side. The predetermined shape is a rectangular region having sides parallel to the latitude direction and sides parallel to the longitude direction in the image. The predetermined distance is, for example, 500 meters. The predetermined distance may be shorter than 500 meters or longer than 500 meters. The predetermined shape may be another shape such as a circle or an ellipse instead of a rectangle. The image generating unit 367 generates the trimmed partial image as a transmission image.

[0110] Next, the transmission data generation unit 375 generates transmission data (step S550). Specifically, the transmission data generation unit 375 generates information including crustal movement identification information, crustal movement position information, and a transmission image as transmission data. The crustal movement identification information is information for identifying the target crustal movement. Note that the crustal movement identification information may be any information as long as it is capable of identifying each of the one or more crustal movements detected in step S530. The crustal movement position information is information indicating the position of the target crustal movement, which is calculated in step S540. The transmission image is the transmission image generated in step S545.

[0111] Next, the transmission data generating unit 375 stores the transmission data generated in step S550 in the storage unit 32 (step S560).

[0112] In this way, the flying object 2 can generate transmission data for each of the one or more crustal movements detected in step S530 by repeatedly performing the processes of steps S357 to S560, and can store the generated transmission data in the storage unit 32.

[0113] After steps S357 to S560 are repeated, the communication control unit 361 outputs each of the transmission data stored in the storage unit 32 in step S550 to the communication antenna unit 22, causes the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4 (step S570), and ends the process. This allows the flying object 2 to, for example, reduce the amount of data to be transmitted to a target to which the data is to be transmitted (in this example, the receiving device 4) compared to the amount of observation data, and can shorten the time it takes to provide the user with information indicating that crustal movement, which is an example of a detection target, has been detected.

[0114] In step S570, the communication control unit 361 may be configured to output a part of the transmission data stored in the storage unit 32 to the communication antenna unit 22, and cause the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4. In addition, in step S550, the communication control unit 361 may be configured to output the transmission data generated by the transmission data generation unit 375 to the communication antenna unit 22, and cause the communication antenna unit 22 to transmit radio waves corresponding to the transmission data toward the receiving device 4.

[0115] Furthermore, instead of being configured to be applied to the detection of such crustal movements, the flying object 2 may be applied to local disaster detection, volcanic activity monitoring, infrastructure monitoring, and the like. The flying object 2 may also be configured to detect characteristics of the target crustal movement at any timing included in the period in which the processes of steps S357 to S545 shown in Fig. 9 are performed. In this case, the flying object 2 detects the characteristics using a machine learning algorithm. Then, in step S550, the flying object 2 generates transmission data including information indicating the characteristics.

[0116] As described above, the flying object 2 according to the third modification of the embodiment detects crustal movement by comparing the base map with the generated image representing the area D. This allows the flying object 2 to notify the user more quickly that crustal movement has been detected based on the base map and the generated image representing the area D.

[0117] The control device 3 described above may be configured to be mounted on another flying object such as an airplane instead of the flying object 2. In this case, the flying object is equipped with functional units corresponding to the synthetic aperture radar unit 21 and the communication antenna unit 22. Moreover, instead of a configuration that calculates a phase difference between the base map and the generated image representing the area D, the detection target detection unit 369 may be configured to obtain image similarity, coherence reduction, and the like by coherence analysis and detect local changes in the ground surface in the area D.

[0118] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and may be changed, replaced, deleted, etc. without departing from the gist of the present invention.

[0119] Furthermore, a program for implementing the functions of any of the components of the device described above (for example, the flying object 2) may be recorded on a computer-readable recording medium, and the program may be read into a computer system and executed. Note that the term "computer system" as used herein includes hardware such as an operating system (OS) and peripheral devices. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs, and compact disks (CDs)-ROMs, as well as storage devices such as hard disks built into computer systems. Furthermore, the term "computer-readable recording medium" includes devices that retain a program for a certain period of time, such as volatile memory (RAM) within a computer system that serves as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line.

[0120] The above program may be transmitted from a computer system in which the program is stored in a storage device or the like to another computer system via a transmission medium, or by transmission waves in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The above program may be for implementing some of the above functions. Furthermore, the above program may be a so-called differential file (differential program) that can implement the above functions in combination with a program already recorded in the computer system. [Industrial Applicability]

[0121] According to the above-mentioned flying object and program, when transmitting data to the ground station, the amount of data to be transmitted can be made smaller than the amount of observation data, so that the user can be notified more quickly that the desired detection target has been detected. [Explanation of symbols]

[0122] 1. Satellite Observation System 2 Projectile 3. Control device 4. Receiving device 21 Synthetic Aperture Radar Section 22 Communication antenna section 31 FPGA 32 Storage section 34 Communications Department 36 Control Unit 361 Communication Control Unit 363 Radar Control Unit 364 Observation Data Generation Unit 365 Processing Section 367 Image Generation Unit 369 Detection target detection section 371 Position calculation section 373 Feature Extraction Unit 375 Transmission Data Generation Unit

Claims

1. A synthetic aperture radar that transmits radio waves into a monitored space including a sea area and receives radio waves reflected from the monitored space; an observation data generating unit that generates observation data based on the reflected radio waves received by the synthetic aperture radar; an image generation unit that generates an image representing the monitored space based on the generated observation data; a detection unit that generates a binary image by binarizing the brightness value of each pixel of the generated image by applying a standard deviation filter to the brightness values ​​of pixels constituting each predetermined region in the generated image, and detects an area as a ship when the generated binary image includes an area constituted by a predetermined number or more pixels having a predetermined brightness value set by the binarization; a position calculation unit that acquires flying object position information indicating the position of the flying object at each time and flying object attitude information indicating the attitude of the flying object, and calculates latitude and longitude representing the position of the ship based on the acquired flying object position information and flying object attitude information and the generated image; a feature extraction unit that determines the detected course and speed of the ship based on a plurality of parameters that are set by machine learning information that associates the features of the area representing the ship in the generated binary image with the course and speed of the ship; and a transmission data generating unit that generates transmission data including information on latitude and longitude representing the calculated position of the vessel and information on the determined course and speed of the vessel; a communication antenna unit that transmits radio waves corresponding to the generated transmission data to a receiving device; A flying object equipped with the above.

2. the detection unit pre-stores, as a plurality of parameters, information associating a plurality of images including a ship with the position and shape of the ship in each of the plurality of images, and a plurality of images not including a ship and information indicating that a ship is not included in each of the plurality of images, and detects, based on the plurality of pre-stored parameters, a likely candidate as a ship from among candidate ships included in the generated images using a machine learning algorithm; The flying object according to claim 1.

3. the image generation unit crops a partial image representing an area of ​​a predetermined shape centered on the position of the ship from the generated image, and generates the cropped partial image as a transmission image; the transmission data generation unit generates the transmission data including information of latitude and longitude representing the calculated position of the ship and the transmission image; 3. The flying object according to claim 1 or 2.

4. the observation data generation unit, the image generation unit, the detection unit, the position calculation unit, the transmission data generation unit, and the feature extraction unit are realized by an FPGA (Field Programmable Gate Array); The flying object according to claim 1.

Images