JP7219143B2 - Filming planning device and method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a technique for making a photographing plan for photographing an aerial photographed image with high added value.

Satellite images have the characteristic of being able to capture a wide area in a single shot, so it is possible to detect changes in land cover conditions in the same area and the appearance of features that were not there in the previous image in the new image. With the increase in large-scale disasters, there is a demand for an analysis platform that can quickly grasp the situation in a wide area (extraction of objects and changes) as a technology using satellite images. To monitor changes over a wide area over a long period of time and continue to obtain meaningful information, many satellite images are required. Conventional satellites that provide free downloadable imagery include Landsat and MODIS®, but with low resolutions of 30 m and 250 m in multispectral bands.

In recent years, due to the advent of ultra-small satellites equipped with high-resolution sensors, satellite image acquisition costs are becoming cheaper. In addition, subscription-type archive utilization services for high-resolution satellite images are at the dawn, such as Digital Globe, which provides high-resolution satellite images for the private sector, has released the GBDX platform. Consideration of such a solution has begun.

The subscription-type archive use service is a service that allows satellite images taken in the past to be used a predetermined number of times at a low cost. This is because it is considered that the conditions for which someone has requested photography in the past, such as AOI (Area of Interest) and the time of photography, are almost never completely compatible with what the user wants to request photography.

If the archive user does not have a satellite image of the AOI taken at the desired time, he or she must request the shooting himself or wait for the AOI to be taken by accident. In many cases, if there is no AOI image of the desired period, the user himself/herself requests the photographing of the AOI. At this time, the problem is the cost of taking new satellite images. With some exceptions, there are a limited number of satellites for photographing the ground. In order to make the best use of the limited shooting opportunities, there is a minimum shooting area requirement for new shoots. It is important to make effective use of shooting opportunities even when there is no request for shooting. should be done.

Japanese Patent Application Laid-Open No. 2002-200002 discloses a technique for designating imaging conditions for an artificial satellite and visualizing the progress of imaging. Further, Japanese Patent Laid-Open No. 2002-200003 discloses a technique for automatically planning a new photographing plan and notifying it to a satellite for an area in which a certain amount of change has been recognized compared to past photographing.

JP-A-2009-9436 JP 2015-28759 A

How to collect useful data when there is no photography request is a common problem faced by satellite imagery companies. It is desirable for satellite imagery operators to capture and store (archive) images of high-demand areas where users may appear at a later date. When a disaster occurs, it is clear that there is a high demand for photography in the disaster area even if no one requests photography. With satellite images, it is necessary to estimate with high accuracy the coordinates that are likely to be used in later analyses, and the appropriate timing.

However, although Patent Document 1 displays the progress of satellite images, it does not mention creating a shooting plan.

On the other hand, as a technique related to the photographing plan, Patent Document 2 discloses a technique of newly photographing an area in which a certain amount of change has been recognized from images taken in the past. It is unknown whether it will become a valuable image in the future.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to propose a technique for creating a photographing plan for photographing potentially valuable images.

One aspect of the imaging plan planning apparatus of the present invention for solving the above problems is a storage device for storing a user management table for managing user profiles including user names for a plurality of users, and a device for requesting planning of a satellite imaging plan. an input unit for inputting a user name and a keyword; a name identification unit for identifying a user's area of interest from the keyword and storing it in the user management table corresponding to the user; Based on the number of past shots in the area of interest of each user, the degree of interest is calculated for each grid on the map. and a photographing candidate management unit that uses grids with a high degree of interest shown in the heat map as photographing request areas.

According to the present invention, it is possible to determine potentially valuable shooting positions and create an efficient shooting plan.

It is a figure explaining the hardware constitutions of an Example. 4 is a diagram illustrating functional blocks of the first embodiment; FIG. 2 is a hardware block diagram of the imaging plan drafting device of the first embodiment; FIG. It is the figure which showed the user management table of 1st Example. FIG. 10 is a diagram showing an interest level map representing the interest level of user category A; FIG. 10 is a diagram showing an interest level map representing the interest level of user category D; FIG. 10 is a diagram showing a heat map representing interest levels of all users; It is the figure which showed the work table of the 1st Example. It is a figure explaining the imaging|photography time slot of a satellite image. FIG. 4 is a diagram for explaining the operation of the scheduler of the first embodiment; FIG. It is a figure explaining the functional block of a 2nd Example. It is a figure explaining operation|movement of the scheduler of a 3rd Example. FIG. 10 is a diagram illustrating a procedure for creating a user interest degree map managed for each shooting angle; It is a figure explaining the functional block of a 4th Example.

A mode for carrying out the present invention will be described by citing several embodiments. In the following description, the components with the same reference numerals in the drawings operate in the same manner, so the description is omitted. Derivative forms with different actions are indicated by adding a hyphen and alphabet at the end of the same code number.

Embodiments will be described with reference to the drawings. It should be noted that the embodiments described below do not limit the invention according to the claims, and that all of the elements described in the embodiments and their combinations are essential to the solution of the invention. is not limited.

In the following description, the information may be expressed in terms of [AAA table], but the information may be expressed in any data structure. That is, the [AAA table] can be described as [AAA information] because the information does not depend on the data structure.

Also, in the following description, [CPU] is a Central Processing Unit including one or more processors. A processor may include hardware circuitry that performs some or all of the processing.

Further, in the following explanation, the processing may be explained with the [program] as the main body of the operation, but the program is executed by the CPU, and the specified processing is appropriately performed by using storage resources (for example, memory) etc. , the actual processing is performed by the CPU. Therefore, the processing described with the program as the subject of the operation may be described with the processor as the processing subject. A part or all of the processing performed by the processor may be performed by a hardware circuit such as an Application Specific Integrated Circuit (ASIC) or a Field-Programmable Gate Array (FPGA).

A computer program may be installed on the device from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium.

In drafting a photographing plan for photographing an aerial image with high added value, it is necessary to predict photographing positions that will have a high value in the future. In order to determine a photographing position that will be of high value in the future, it is effective to bring the photographing plan closer to the content of the photographing request from the user who has a record in the past. In other words, it is effective to formalize the method of determining photographing conditions based on empirical rules. This is because a user who regularly requests photography for the same photography position often makes similar requests in the future.

In addition, if an operating company that takes aerial photographs using satellites publishes a photographing plan, users do not have to make unnecessary requests for photographing their own areas of interest (regions of interest). On the other hand, if the operating company also asks users to disclose their interest areas, they will be able to take pictures of potential users' interest areas in their free time, and as a result, they will be able to provide services at low cost. As a result, demand from users who have refrained from shooting due to the cost bottleneck can be discovered.

The imaging planning (scheduling) of the first embodiment will be described. One of the important points in the first embodiment is how to capture an image of high value to the user at a timing when a management company that uses a satellite for aerial photography does not receive a photography request.

FIG. 1A shows the hardware configuration of the first embodiment. A group of satellites 50 orbiting in space, a satellite communication device 10 that communicates with the satellite group 50 by radio, and a shooting planning device 30 that transmits a shooting command to the satellite 50 via the satellite communication device 10. Configured. The satellite communication device 10 and the imaging planning device 30 constitute a imaging planning system. The imaging plan formulation device 30 plays a role of creating an imaging schedule for adjusting the satellite imaging attitude and timing, and the created imaging schedule is transmitted to the satellite 50 via the satellite communication device 10 . Calculation resources for creating a schedule in the shooting plan drafting device 30 may be a locally arranged computer, or may be a computer installed at an external site, such as a resource on a cloud computer. .

FIG. 1B shows functional blocks of the imaging planning system. In the first embodiment, the trigger for creating a schedule is input of a user name and a keyword input to the keyword analysis unit 300, which is an input unit, from a user requesting planning of a satellite imaging plan. The user name does not necessarily have to be the user name as long as it is user identification information that can identify the user. A keyword analysis part 300 analyzes the input keyword and user name, determines whether the input keyword includes a proper noun indicating a place name, and returns a response to include the place name if not. In many cases, at this point, it is assumed that the place name is not given in the form of an address or AOI that can uniquely identify the shooting location. For example, assuming that the user has input "Kanagawa Prefecture, Minato", the shooting candidates are broad and ambiguous.

The Web crawler 301 uses a search engine to collect information on homepages related to "Kanagawa Prefecture, Minato" in order to specify the address of the AOI from the ambiguous information such as "Kanagawa Prefecture, Minato" input by the user. and sends the collected information to the address search unit 302 . The address search unit 302 extracts address information such as “Yokohama port” and “Yokosuka port” from the input keywords and the collected text information of the home page, and outputs it to the name identification unit 303 together with the user name. The web crawler 301 robotically collects text information from home pages using a technique such as web scraping. For example, crawl the access information and history information pages from the official homepages of Yokohama Port and Yokosuka Port, extract addresses and list them. For example, when searching for the port of Yokohama, the home page of the Yokohama City Port and Harbor Bureau is hit, and the address of the Yokohama City Port and Harbor Bureau appears at the bottom of the home page. The URL of the homepage, address information, the name of the homepage owner, the keyword used by the used web crawler, etc. are notified to the name identification unit 303 .

The name identification unit 303 performs so-called name identification, in which the address information extracted by the address search unit 302, such as "Yokohama port" and "Yokosuka port", is grouped into the same ports with different names. The name identification unit 303 associates the user name and the keyword input to the keyword analysis unit 300 with the address identified by the name identification unit 303 and outputs them. The address output from the name identification unit 303 is the region of interest identified from the keyword.

In the storage device 320, information other than the interest area T15 in the user management table T1 shown in FIG. 2 is stored in advance for each user as a user profile. The name identification unit 303 stores address information corresponding to the user name as a key in the region of interest T15 of the user management table T1.

The name identification unit 303 not only performs name identification of address information, but also decodes user profiles, identifies users of the same company and similar users, and manages them as shown in the user management table T1 in FIG. This is because it is more efficient to capture similar users as the same user as satellite images. For this reason, similar users are identified by the name identification unit 303 not only because they have similar names, but also because they have similar analysis work and industries.

FIG. 2 shows a user management table T1 for managing user profiles by the name identification unit 303. As shown in FIG. The storage device 320 stores the user management table T1 and updates and adds information sequentially under the control of the name identification unit 303 . For example, when a new user is input from the keyword interpretation unit 300, the name identification unit 303 adds an entry to the user management table T1.

The ID T11 of the user management table T1 stores, for example, "0" as identification information for uniquely identifying each user. User name Corporate name T12 stores, for example, "XX company" as the name of the user. The business category T13 stores, for example, “GIS (Geogprahical Information System)” as the business category of the user identified by the identification information. Also, the GIS user category T14 stores, for example, "A" as the GIS category of the user identified by the identification information. This GIS user category T14 will be detailed later.

The region of interest T15 stores, for example, "somewhere in XX city" as a user's region of interest identified by identification information. This region of interest T15 is input corresponding to the user name T12 whose address information is extracted by the address search unit 302 based on the keyword input by the user to the keyword analysis unit 300 and matches the input user name. .

The analysis task T16 stores, for example, "difference detection" as the analysis task of the user identified by the identification information. The number of times of photographing in the past T17 stores the number of times of photographing the region of interest T15 among the number of times of photographing of the user identified by the identification information or the user name, for example, "10". When the same user inputs different keywords and different regions of interest T15 are stored in the user management table T1, the number of entries is increased so that the number of times T17 of past photography is managed for each region of interest. Similarly, for similar users identified by the name identification unit 303, the number of times T17 of past photography is managed for each region of interest.

In the user management table T1, as described above, except for the area of interest T15, the user identification information ID T11, user name company name T12, industry type T13, analysis work T16, and past photographing times T17 are user profiles. are registered in the storage device 320 in advance and updated sequentially.

In the GIS user category T14, the user name T12 who is in charge of the same category of industry as "Company XX" updating the map information and whose industry T13 is managed as "GIS" is registered as "Company XX". Users who are associated store "A" together in order to treat them as similar users. For example, GIS user category "A" is a group of GIS professional companies, and organizations such as ministries and agencies are registered as GIS user category "B", and each user is divided into certain categories and managed. In addition, although not directly dealing with GIS, insurance companies with large-scale address information are registered as GIS user category "C", agricultural management associations are registered as GIS user category "D", infrastructure etc. For example, public institutions are assigned a GIS user category "Z", and users with similar business types and tasks are included in the same category. For each GIS user category, the number of times of past photography may be aggregated, and processing of users belonging to a category with a large number of times of photography in the past may be preferentially handled. This is because a user belonging to a category with a large number of times of photographing T17 in the past is likely to make many photographing requests in the future as well.

Here, the number of times of photographing in the past T17 is the number of times the region of interest T15 was actually photographed, and the number of times the scene was photographed when photographing including most or all of the region of interest was executed even if no photographing order was received. , the number of shots may be counted up. However, since it is not an area photographed on order, it may be managed as the number of times by multiplying by a coefficient of 1 or less. In addition, the number of incidental shootings may be separately managed and counted up. It is also possible to predict the tendency of what kind of work users are doing for each industry from the information of users who have already ordered, and to infer the work contents of potential users by analogy.

The user name includes the place name information of the location, and the location is often the area of interest. The region of interest T15 referred to here includes not only the information entered by the user at the time of user registration, but also the place name information included in the user name, the address information of the home page obtained when the user name is searched on the Internet, and the like. OK. Since the user management table T1 reflects the past results, it is preferable to sequentially add the user name and the area of interest included in the photographing request each time a photographing request is received.

Returning to FIG. 1B, when an area of interest is input from the name identification unit 303, the AOI candidate creation unit 306 performs geocoding for conversion into latitude and longitude, and determines the center point of the area of interest T15 (AOI) to be photographed. .

If the address information is insufficient at this point and the search location is not narrowed down, the image analysis unit 311 may detect AOI candidates using satellite images captured in the past. For example, to detect a port, the AOI candidate creation unit 306 sends information “port” to the image analysis unit 311, and the image analysis unit 311 searches along the coast from the archive image to detect piers and anchored ships. When searching for large-scale facilities such as factories and power plants, and commercial facilities, the image analysis unit 311 compares time-series images from archive images, detects vast vacant lots, and gradually increases the number of buildings. to detect Since this search requires a sufficient number of images, it is convenient to use low-resolution images to handle time-series images.

When the satellite image including the AOI candidate is extracted by the image analysis unit 311, the satellite image usually has coordinate information about the upper right point of the image. Coordinate information (latitude, longitude) can be obtained for each pixel in the image. The obtained coordinate information is sent to the imaging candidate management unit 307 via the AOI candidate creation unit 306 .

If all of the coordinate information from the AOI candidate creation unit 306 is input to the scheduler 308, the few satellite image capturing opportunities will be wasted. to decide. For example, only the coordinate information of the area whose degree of interest exceeds a certain threshold is output to the scheduler 308 as the imaging request area. As a result, satellite images, which are likely to be in demand in the future, can be captured on limited opportunities.
The degree of interest for each user is calculated for each grid obtained by dividing the area on the map and stored in the storage device 320 .

FIG. 3A shows a degree-of-interest map of the degree of interest of the GIS user category A in the industry T13. FIG. 3A shows grids that divide an area on the map into 3 in the horizontal direction (X direction) and 3 in the vertical direction (Y direction), and each grid is represented by (X, Y). A in grid A11 indicates a user category, and 11 corresponds to coordinate information (X, Y). That is, the grid A11 indicates that the interest level indicated by the grid for the GIS user category A is "1".

Grid A22 is "3", indicating the highest degree of interest. The interest level of the grid A33 is "0", indicating that the user of the GIS user category A has no interest in the satellite image of this area, that is, it is unlikely that he will receive a request for photographing in the future. Since the business of GIS user category A is GIS, there is a high possibility that the user will be requested to take a satellite image of a wide area.

Similarly, FIG. 3B is a user of the GIS user category D, showing an interest degree map of the interest degree of the user whose industry T13 is an agricultural cooperative. Since users of agricultural cooperatives belonging to user category D are interested only in the grids corresponding to the areas to which the agricultural cooperatives belong, only the grid D33 has a numerical value indicating the degree of interest "2".

The degree of interest for each grid shown in FIGS. 3A and 3B quantifies future imaging values in the corresponding grid. The degree of interest is calculated based on the number of times the grid has been photographed in the past and the user category. As a simple method of calculating the degree of interest, the number of times of photographing in the past for each grid is multiplied by a coefficient for each user category. The number of times of photographing for each grid can be obtained from the past number of times of photographing T17 managed corresponding to the region of interest T15 in FIG. Among user categories A to Z, the coefficient for each user category is set to be the largest for user category A and the smallest for user category Z.

It is assumed that the more professional the GIS, the greater the range of interest, or the stronger the demand for 3D rendering. may also be taken into account to calculate the degree of interest. In addition, when the shooting period is limited, such as when used for yield prediction, the score of the degree of interest may be cleared if the timing is off.

FIG. 3C shows a heat map of interest levels of all users by summing the interest level maps for each user shown in FIGS. 3A and 3B for all registered users stored in the user management table T1. .

Enlarging FIG. 3C shows detailed interest degrees for each shooting direction (for each off-nadir angle) on the grid. This is because the degree of interest for each off-nadir angle is important in AOI where three-dimensional imaging is important. This is because the MVS (Multi View Stereo) technology used for 3D rendering requires viewpoints from multiple directions. , the sum of these determines the interest level of the region.

The shooting directions are eight directions that divide the horizontal plane by 45 degrees. FIG. 3C shows the degree of interest in three of the eight directions. The numerical value obtained by summing the interest degrees in all eight directions of the relevant grid is "15". Since information from oblique directions is also necessary for 3D rendering, it is better to manage 16 directions by adding two types of pitch angles (how much it is tilted) to each of the 8 directions and solid angles in the vertical direction. . The 16 directions are a guideline and may be managed in detail. Even if a photograph is taken once, it is treated as photographed only in a specific direction out of the 16 directions. Therefore, the photographing flag for the area in question is not completed until it is recognized that the photograph was taken from all directions. prevent it from happening.

The imaging candidate management unit 307 updates the degree of interest for each grid of imaging candidate points as needed, as shown in FIG. 3C. There should be some bias in the timing when the user wants to shoot. Therefore, a mechanism for managing the foot map of FIG. 3C in chronological order may be provided.

If a heat map of interest levels as shown in FIG. 3C is displayed in the photographing candidate display section 312 and made available to a third party who will be a future user, even a user who has never requested satellite image photographing can realize the possibility of future photographing. can be predicted, and it is possible to prevent a new photographing request for overlapping areas. In addition, the heat map shows that if the user wants to obtain a satellite image of an area with a low degree of interest, a new request for photographing must be made. The degree of interest heat map may be displayed on the imaging candidate display unit 312, or may be transmitted to a user terminal such as a user's smartphone or PC via a network.

The photographing condition management unit 304 receives the analysis work T16 of the user management table T1 from the name identification unit 303 . Based on this information, typical imaging conditions for each job stored in the job table T2 shown in FIG. 4 are extracted.

The storage device 320 stores the work table T2, and under the control of the imaging condition management unit 304, updates and adds information to the work table T2 sequentially.

The ID T21 of the task table T2 in FIG. 4 stores, for example, "0" as identification information for uniquely identifying each task. Analysis work T22 stores, for example, "difference detection" as the analysis work of the work identified by the identification information. Analysis work T22 corresponds to analysis work T16 in FIG. The seasonality consideration T23 stores, for example, "necessary (summer)" as information indicating whether the business identified by the identification information needs to take into account the season in which satellite images are taken. The shooting direction T24 stores, for example, "directly above" as the shooting direction of the business identified by the identification information. The multiple-picture necessity T25 stores, for example, "required" as information indicating whether or not the work identified by the identification information requires multiple-picture photography for one photography area. The reflectance transform T26 stores the information when the business identified by the identification information needs to acquire an image in consideration of the reflectance.

The photographing management unit 305 reads archived images of satellite images taken in the past stored in the storage device 320, and obtains information on various photographing conditions such as photographing timing/time, photographing angle, and sun angle of the archived image. , with the imaging conditions from the imaging condition management unit 304 . Reading from the archive image may be performed only for the image containing the latitude and longitude from the imaging candidate management unit 307 . For example, when the seasonality consideration T23 of the business management table T2 is "required (summer)", the photographing management unit 305 determines that new photographing is unnecessary according to the archived image if the photographing time of the archived image is summer. It makes a decision and outputs the result to the scheduler 308 . If the archival image is to be captured in winter, it is determined that a new image must be captured, and the result is output to the scheduler 308 .

Similarly, the photographing management unit 305 determines whether the time, photographing angle, and sun angle of the archived image satisfy the photographing direction T24 and the reflectance T26 of the work table T2. When an archive image cannot be used, the scheduler 308 is requested to shoot a satellite image according to the shooting conditions read by the shooting condition management unit 304. This enables efficient use of shooting opportunities and reduces shooting costs. can.

The imaging management unit 305 receives the imaging conditions stored in the work table T2 from the imaging condition management unit 304 and the grid information indicating grids of high interest shown in the heat map from the imaging candidate management unit 307. An archival image corresponding to the grid information is read from storage device 320 . In other words, an archive image of the area specified by the grid information is read.

The archive images to be processed are narrowed down from the heat map information from the imaging candidate management unit 307 based on grid coordinate information equal to or greater than a certain threshold. The photographing management unit 305 uses the archived image when the narrowed down archive image satisfies the photographing conditions from the photographing condition management unit 304 . If not available, the coordinate information from the imaging candidate management unit 307 and the imaging conditions from the imaging condition management unit 304 are output to the scheduler 308 as a new imaging request.

FIG. 9 illustrates how interest level heat maps, such as those shown in FIGS. 3A, 3B, and 3C, are updated from user profile analysis. The name identification unit 303 reads the user's profile from the user management table T1 stored in the storage device 320 (S902). The name identification unit 303 estimates the analysis task T16 from the read user profile (S903). When the analysis work is grasped, the imaging condition management unit 304 specifies the imaging conditions such as the imaging angle from the work table T2 (S904). Next, the scheduler 308 distributes the degree of interest of each grid to a plurality of shooting angles based on the shooting method direction T24 and the like included in the shooting conditions (S905). This makes it possible to manage the degree of interest according to the shooting direction as shown in FIG. 3C. The interest level for each grid and the interest level distributed to the shooting angles are input to the shooting candidate display section 312 and displayed.

The scheduler 308 uses the past photography statistics received from the photography management unit 305, the heat map information of the photography candidates received from the photography candidate management unit 307, and the weather forecast information to schedule satellite image photography. create. Among the operations of the scheduler 308, the creation of the satellite image capturing schedule itself is performed when an image capturing request is actually received. This can be performed in the same manner as when a satellite image capturing schedule is performed based on the above. In the first embodiment, instead of the photographing conditions included in the actual photographing request, the photographing conditions that are highly likely to be requested in the future are input to the scheduler, thereby preliminarily photographing images that will be in high demand in the future. Thereby, a satellite image can be captured efficiently.

Even if the satellite imaging angle input to the scheduler 308 is controlled too finely, there will be a lead time for re-setting and wear of parts. The angle formed by the satellite's shooting direction and the vertical direction on the ground, which is generally called the off-nadir angle, is how much the satellite is tilted in elevation (pitch angle) and how much it tilts left and right (yaw angle) with respect to the satellite's traveling direction (azimuth). Therefore, the attitude information of these three points of azimuth, pitch angle, and yaw angle, and information on the timing of change may be transmitted as information for creating a schedule.

FIG. 6 shows an outline of the operation of the scheduler 308 of the present invention.

In order to explain the outline of the scheduler operation, FIG. shown in Slots for capturing grids of high interest are shown in gray. The scheduler 308 creates a schedule for capturing satellite images at the locations shown in gray.

First, since the scheduler 308 determines the attitude of the satellite and the time slot for each day shown in FIG. 5, the operation flow is started for each day (step S601).

Next, a loop process for performing the following process is started for each yaw angle and pitch angle (step S602).

Since the shooting candidates should naturally be determined from the point with the highest degree of interest, the strip is predicted for each yaw angle and pitch angle in the trajectory to be scheduled (S603). Then, the degree of interest that can be collected from the strip is calculated (step S604). The retrievable interest degree means, for example, the total interest degree of the grid covered by the strip. It is evaluated in comparison with the degree of interest of the strip calculated in other yaw angle and pitch angle loops (step S604). When all the yaw angles and pitch angles are processed for the required number of days, the yaw angle and pitch angle that involve the most areas with high points of interest are determined (step S607).

If the posture is changed, that is, the number of slots is set to several times at most, the score of the degree of interest can be evaluated in a finite number of trials. In order to collect the points of the grid for which three-dimensionalization is managed, a constraint is added so that the degree of interest is not added up unless the image is taken from a direction in which the image is not taken. In order to collect the points of the grid in which the photographing period is specified, it is necessary to satisfy the conditions.

In addition, weather forecast information is important because the ground cannot be photographed when it is covered with clouds, which is the biggest problem with optical satellite images. The score may be re-evaluated by adding, as an expected value, an expectation regarding cloud coverage at the scheduled shooting time. In that case, since a satellite image covered by clouds over a certain ratio is useless, the score that can be recovered in the area is adjusted to be almost zero. Clear the interest level of the grid once the shooting was done.

Attitude information and slot information determined in this manner are sent to the satellite communication transmitting/receiving unit 10b as a result of creating a schedule, converted into radio signals, and then radio waves are transmitted from the antenna 10a and come into orbit. input to the satellite. The timing at which an orbiting satellite comes within the range of the antenna is limited, and a shooting command can be issued to the satellite about once a day per site.

In addition, as shown in FIG. 5, information on the expected trajectory of the photographing strip, the photographing time, and the photographing posture is added to the photographing candidate display section 312 to determine what kind of schedule the photographing is actually scheduled to be performed. may be displayed and disclosed as schedule information. In addition to such schedule information, an interest level map such as that shown in FIG. 3 may be published as reference information indicating how much other users are interested. These publications are performed by displaying them in the shooting candidate display section 312 and by transmitting schedule information and heat map information to the user's PC or smartphone.

Considering the accuracy of weather forecasts and the necessity of update frequency, it is sufficient to create a schedule for about three days. It can be fine-tuned on a daily basis as long as there is communication between the antenna on the ground and the satellite platform.

FIG. 1C is an example of the hardware configuration of the imaging plan drafting apparatus. In FIG. 1C, a communication device 353, an input device (not shown) such as a keyboard and a mouse that constitute the keyword analysis unit 300, an output device (not shown) such as a display that constitutes the imaging candidate display unit 312, and a control unit It includes a CPU (Central Processing Unit) 351 , a memory 352 and a storage device 320 .

The CPU 351 implements various functions according to programs stored in the memory 352 . Programs stored in the memory are keyword interpretation, web crawler, address search, name identification, shooting condition management, shooting management, AOI candidate creation, shooting candidate management, and scheduler programs, and the processing unit 351 executes each program. 1B, the keyword analysis unit 300, the web crawler 301, the address search unit 302, the name identification unit 303, the shooting condition management unit 304, the shooting management unit 305, the AOI candidate creation unit 306, the shooting candidate management unit 307, and the scheduler Each functional block such as 308 is realized.

The storage device 320 includes a magnetic disk drive, a magneto-optical drive, an optical drive, a raid (RAID), a solid state memory device, a solid state drive, and the like. An archive image 310 and an interest degree map 330 indicating the degree of interest for each user are stored.

As described above, according to the first embodiment, it is possible to create a photographing plan for photographing potentially valuable images.

In addition, if an operating company that takes aerial photographs using satellites publishes a photographing plan, users do not have to make unnecessary requests for photographing their own areas of interest. On the other hand, if the operating company also asks users to disclose their interest areas, they will be able to take pictures of potential users' interest areas in their free time, and as a result, they will be able to provide services at low cost.
Furthermore, it is possible to unearth the demand of users who have refrained from shooting due to the cost bottleneck.

In the second embodiment, an operation will be described in which the AOI is directly input by the user without inferring the AOI from the keyword as in the first embodiment.

FIG. 7 is a diagram for explaining functional blocks of the second embodiment. An AOI reception unit 700 is added as a new function to the input unit of the first embodiment. The AOI reception unit 700 of FIG. 7 provides an interface for users to upload information about AOIs. The AOI may be specified in the form of an address book, in the form of vector data such as ESRI Shapefile, or in the form of raster data such as mask images. In the second embodiment, latitude and longitude coordinate information output from the AOI reception unit 700 is input to the region of interest T15 of the user management table T1 shown in FIG.

The Web crawler 301-A crawls and acquires related home pages using names as keys in order to register the user's area of interest that is inferred from the registered user's profile information.

As in the first embodiment, the address search unit 302 may extract address information from the collected home pages and add it to the interest area of the user in the user management table. However, in the second embodiment, since there is an AOI directly specified by the user, it is treated as the main area of interest.

When the AOI candidate creation unit 306-A receives AOIs in a format such as an address book, it converts them into latitude and longitude information by geocoding. The converted latitude and longitude information is sent to the imaging candidate management unit 307 .

The shooting candidate management unit 307 creates a heat map of interest levels, and uses different interest level evaluation methods for users who have input AOIs and users who have not input AOIs, for example, users who have input keywords. You can change it. For example, when an AOI is directly specified by the user, the interest level to be set may be treated as N times the estimated AOI area of the user who did not specify it. However, N is a number of 1 or more. That is, the degree of interest of the region of interest input from the AOI reception unit is calculated to be higher than that of the region of interest based on the keyword in the first embodiment. This is because users who are searching for shooting locations at the keyword level and users who have determined the area they want to see and are highly motivated to take satellite images, the latter is more likely to want highly accurate information. This is a measure to take into account the high As for the processing of other parts, the same operations as those shown in the first embodiment are performed.

As described above, according to the second embodiment, in addition to the first embodiment, the AOI is directly input by the user, so that a more accurate AOI can be used as a basis.

In Example 3, a method for creating a schedule for multiple satellites is described.
The basic operation of the scheduler for multiple satellites is the same as that for a single satellite. However, since multiple satellites are used, flexibility regarding shooting has increased, and more detailed shooting angles can be specified. For example, the typical shooting conditions for business are shown in the business table T2 in FIG. It may be set so that the degree of interest is not cleared by photographing from any direction.

Normally, shooting from an oblique angle increases the impact of collapsing buildings, and satellite images using DEM (digital representation of three-dimensional coordinates using plane positions and elevation values measured by various surveys). Correction becomes difficult, and later processing of satellite images becomes difficult. In particular, for jobs that require the use of a plurality of images, the photographing direction conditions may be stricter. For example, it has been explained that the degree of interest of the user should be determined so that the order of priority can be determined for each user category, but as shown in FIG. is used to provide a method of managing the degree of interest, including the satellite's shooting attitude. This makes it possible to manage satellite resources, including shooting angles when creating schedules.

As shown in FIG. 8, availability of photographing of a plurality of satellites is checked, and a schedule is created for satellites for which photographing requests have not been received for the scheduled schedule. Similar to scheduling for a single satellite, the orbit and strip are estimated, and cloud cover conditions are predicted for each satellite from weather forecasts. The processing of FIG. 8 is basically the same as that of FIG. 6 except that processing loops corresponding to the number of satellites are added to the first step.

As described above, according to the third embodiment, in addition to the first and second embodiments, it is possible to efficiently create a photographing plan using a plurality of satellites.

The fourth embodiment provides a mechanism that takes pictures when news is detected without any input from the user. For example, news is detected about the region of interest that has been photographed in the past.

As shown in FIG. 10, the Web crawler 301-B constantly collects news articles published on the Internet. A news interpretation unit 800 is added as a new function to the input unit of the first embodiment. The news interpreter 800 tags collected articles and collects address information that can be read from the news. In the collected address information, as in the case where a specific user exists as in Examples 1 to 3, the tagged content is regarded as a user, a user category is assigned, and the collected address information is assigned to an area of interest. store the information of
That is, the tag attached to the news is registered instead of the user name T12 in FIG.

Articles for which address information could not be obtained are not dealt with here. The Web crawler 301-B may collect data not only from news sites but also from resources on the Internet that are invisible to search engines, such as SNS sites.

In general, it is rare for a news article to include complete address information. I need to narrow it down. Here, the intended image processing is not the image obtained from the satellite for which the schedule is created, but the low-resolution image provided free of charge, and the result of image processing such as difference detection, It is intended to be the process of extracting where a difference is found.

The image collection unit 801 provides a mechanism for collecting archived images by requesting downloads from sites that publish free satellite images. For example, there are optical satellites such as MODIS, Landsat, Sentinel, etc. that are open to the public free of charge. I submitted a download request for images containing the place name of the news to these sites, and downloaded multiple images. After that, appropriate preprocessing such as brightness correction, cloud cover removal, and orthocorrection are performed, and change extraction is performed. In recent years, US platform companies such as Amazon and Google have provided a service of collecting these free satellite images to some extent and distributing them as catalog archives as image data sets. If the robot downloads the original site every day, it will increase the access load, so using such a catalog archive service, although it is not necessarily the latest, use a service with a relatively small access load to collect images. , the change extraction process should always be prepared.

The news interpreting unit 800 infers by analogy the work performed by a user who has taken pictures in the past for an area requested to be photographed, and searches for and tags news articles reported on the area in question at the time of photographing. management. This makes it possible to cue the tag information as a news article that induces shooting in the future.

Information on the time and area of searched news and captured satellite images is stored as a combination, and in the future, when the number of accesses to articles with the same tag suddenly increases, the target area can be selected from the articles. By extracting and combining with change extraction by the image analysis unit 311, it is determined whether or not to actually photograph a high-resolution satellite.

As a simple calculation method for the degree of interest in this case, the number of hit news items for each grid can be multiplied by a coefficient.

In the case of SNS, it is better to set a high coefficient because the higher the trend, the more access continues. In general, compared to the case where a request for the shooting range is received from the user, the method in which the computer spontaneously determines the shooting range from a news site or SNS is less likely to actually require shooting. It is preferable to set the coefficient to 1 or less so that the degree of interest in photographing is smaller than in Examples 1 to 3 in which the photographing area is specified.

As described above, according to Embodiment 4, in addition to Embodiments 1 to 3, it is possible to capture a satellite image of an area where future demand is expected based on information such as news without information from the user. .

The forms described in Examples 1 to 4 do not need to operate alone, and may be used in combination.

As described above, according to the present embodiment, it is possible to create a photographing plan for photographing potentially valuable images.

In addition, if an operating company that takes aerial photographs using satellites publishes a photographing plan, the user can prevent unnecessary photographing requests for his/her own area of interest. On the other hand, if the operating company also asks users to disclose their interest areas, they will be able to take pictures of potential users' interest areas in their free time, and as a result, they will be able to provide services at low cost.

Furthermore, it is possible to unearth the demand of users who have refrained from shooting due to the cost bottleneck.

In addition, since the AOI is directly input by the user, a more accurate AOI can be used as a basis.

In addition, a photographing plan can be efficiently created using a plurality of satellites.

Furthermore, even without information from the user, it is possible to take satellite images of areas where demand is expected in the future based on information such as news.

10: satellite communication device,
30: shooting planning device,
50: Satellite,
300: keyword analysis unit,
301: Web Crawler,
302: address search unit,
303: name identification unit,
304: an imaging condition management unit;
305: shooting management unit,
306: AOI candidate creation unit,
307: Imaging candidate management unit;
308: scheduler;
311: image analysis unit,
312: a shooting candidate display section;
320: Storage device.

Claims (10)

a storage device storing a user management table for managing user profiles including user names for a plurality of users;
an input unit into which a user name and a keyword are entered to request planning of a satellite imaging plan;
a name identification unit that identifies an area of interest of the user from the keyword and stores it in the user management table in correspondence with the user;
A degree of interest is calculated for each grid on a map based on the number of past shots in the region of interest for each user stored in the user management table, and the calculated degree of interest for each user is stored in the user management table. and a photographing candidate management unit that creates a heat map by summing up users who are interested, and sets grids with a high degree of interest shown in the heat map as photographing request areas. In the shooting planning device according to claim 1,
A photographing planning apparatus, comprising a scheduler for planning a satellite photographing schedule so as to collect interest levels shown in the heat map. In the shooting planning device according to claim 1,
The user management table stored in the storage device stores analysis work corresponding to each user,
The storage device further stores a work table storing imaging conditions for each analysis work and past satellite images as archive images,
a photographing condition management unit that reads photographing conditions based on the job table when the analysis job read from the user management table is input to the name identification unit based on the user name input from the input unit;
a photographing management unit that compares the photographing conditions read by the photographing condition management unit and the photographing conditions of the archive image, and requests photographing based on the photographing conditions read by the photographing condition management unit when the photographing conditions are different; A shooting planning device characterized by: In the shooting plan drafting device according to claim 3,
The archive image to be compared with the imaging conditions read by the imaging condition management unit by the imaging management unit,
A photographing plan drafting apparatus, wherein grid information with a high degree of interest shown in the heat map is input from the photographing candidate management unit, and the image is an archive image including the grid information. In the shooting plan drafting device according to claim 3,
an address search unit for extracting address information from the keyword input from the input unit;
The photographing plan drafting apparatus, wherein the name identification unit carries out name identification processing of the address information input from the address search unit. In the shooting plan drafting device according to claim 3,
The input unit is an AOI reception unit for directly inputting an area of interest from a user,
The photographing plan drafting apparatus, wherein the photographing candidate management unit calculates the degree of interest of the region of interest input from the AOI reception unit to be higher than the degree of interest of the region of interest based on the keyword. In the shooting plan drafting device according to claim 3,
The shooting plan drafting apparatus, wherein the input unit has a collecting news interpretation unit that extracts addresses from news articles. In the shooting plan drafting device according to claim 2,
An imaging plan drafting apparatus, comprising: an imaging candidate display section for displaying a heat map created by the imaging candidate management section. In the shooting planning device according to claim 8,
The imaging plan drafting apparatus, wherein the imaging candidate display unit displays a satellite imaging schedule drafted by a scheduler. A shooting plan planning method in a shooting planning device having a storage device and a control unit,
the storage device stores a user management table for managing user profiles including user names for a plurality of users;
The control unit
A user name and keyword are entered to request a satellite imaging plan,
identifying an area of interest of the user from the keyword and storing it in the user management table corresponding to the user;
A degree of interest is calculated for each grid on a map based on the number of past shots in the region of interest for each user stored in the user management table, and the calculated degree of interest for each user is stored in the user management table. A method of making a photographing plan, comprising the steps of: creating a heat map by summing up users who are interested in each other;

Images