JP2019146058A - Photographing system, photographing unit and photographing device - Google Patents

Abstract

To solve the problem that in a situation where an on-vehicle camera tries to derive a photographing condition for obtaining a good-quality photographed image, the on-vehicle camera cannot derive the photographing condition on time in a point where a scene to be photographed varies rapidly, so that good-quality photographed image with high recognition rate of a photographic subject cannot be obtained.SOLUTION: The photographing system comprises: a photographing device; communication means that communicates with the photographing device; a photographing part that photographs a reference photographic subject at a specific point where the photographing device photographs and outputs a photographed image; an image recognizing part that derives a recognition rate at which the reference photographic subject is recognized in the photographed image on the basis of a result of comparison of photographed image information subjected to image processing on the basis of the output photographed image with correct-answer image information representing a prepared reference photographic subject; a condition deriving part that derives a photographing setting for improving the recognition rate of the reference photographic subject in the photographed image accompanying the derivation of the recognition rate, and derives a photographing condition which is utilized by the photographing device on the basis of the photographing setting; and a condition renewing part that transmits the photographing condition to the photographing device through the communication means to make the device to utilize the condition.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an imaging system for deriving an imaging condition corresponding to an imaging environment, an imaging unit used in the imaging system, and an imaging apparatus that uses an imaging condition derived by the imaging system.

  2. Description of the Related Art Conventionally, a photographing apparatus that photographs a subject, such as an in-vehicle camera or a surveillance camera, can recognize the subject accurately while the scene of photographing, that is, the photographing environment changes every moment, that is, the recognition rate of the subject is high. It is necessary to provide a photographed image.

  Therefore, in Patent Document 1, an in-vehicle camera that captures a subject outdoors obtains information on the imaging environment in which the vehicle on which the device is mounted travels from outside, and derives appropriate parameters related to imaging, that is, imaging conditions. There has been proposed a method for providing a high-quality photographed image by setting the own apparatus.

JP 2006-222844 A

  However, as in the prior art, even if the photographing device installed outside the vehicle photographs a subject in a photographing environment that changes from moment to moment, and the in-vehicle camera acquires information on the photographing environment photographed at this time, it is good quality. If the in-vehicle camera itself derives the imaging conditions for obtaining a captured image, for example, the in-vehicle camera derives the imaging conditions at a point where the scene photographed on the travel route of the vehicle equipped with the in-vehicle camera changes rapidly. There is a possibility that a high-quality captured image with a high recognition rate of the subject cannot be obtained in time.

  The present invention is for solving the above-described problem, and uses an imaging system for deriving an imaging condition according to an imaging environment, an imaging unit used in the imaging system, and an imaging condition derived by the imaging system. An object of the present invention is to provide a photographing apparatus that performs the above.

  The present invention includes a photographing device, a communication unit that performs communication for transmitting and receiving various information between the photographing device, and a photographing unit that photographs a reference subject and outputs a photographed image at a specific point where the photographing device photographs. And a recognition rate indicating accuracy when recognizing the reference subject in the photographed image based on a comparison between the photographed image information image-processed based on the output photographed image and correct image information representing the reference subject prepared in advance And a condition deriving unit for deriving a shooting setting for improving the recognition rate of the reference subject in the shot image and deriving a shooting condition used by the shooting apparatus based on the shooting setting. A storage unit that stores the imaging condition derived by the condition deriving unit, and a condition for transmitting the imaging condition via the communication unit so that the imaging apparatus uses the imaging condition stored in the storage unit. And Shinbu a photographing system provided with a.

  According to the present invention, the photographing system can obtain an effect of photographing a subject with a high recognition rate when the photographing device photographs at a specific point.

It is a system configuration figure showing an example of application of a photography system in Embodiment 1 of the present invention. It is a schematic diagram which shows imaging | photography vector information when an imaging | photography unit image | photographs a reference subject. It is a schematic diagram which shows establishment of the radio | wireless communication in an imaging | photography system. It is a functional block diagram which shows the structure of an imaging | photography system. It is a functional block diagram which shows the modification of a structure of an imaging | photography system. It is an information related figure which shows the relationship of the various information handled with an imaging | photography system. It is an information list figure which shows an example of the information contained in imaging condition information. It is a flowchart figure which shows the process in which an imaging | photography system transmits imaging | photography conditions with respect to an imaging device. It is a flowchart figure which shows the predefined process in which an imaging | photography unit image | photographs a reference object and derives a recognition rate. It is a schematic diagram for demonstrating the image conversion process for deriving imaging | photography conditions. 10 is a system configuration diagram illustrating an application example of an imaging system in Embodiment 2. FIG. It is a functional block diagram which shows the structure of an imaging | photography system. It is an information related figure which shows the relationship of the various information handled with an imaging | photography system. It is a flowchart figure which shows the process in which an imaging | photography system transmits imaging | photography conditions with respect to an imaging device.

Embodiment 1 FIG.
FIG. 1 is a system configuration diagram showing an application example of the imaging system 1 in Embodiment 1 of the present invention. The photographing system 1 includes at least a plurality of photographing units 4. The photographing system 1 may further include a photographing device 3.

  The photographing system 1 derives photographing conditions that are various settings when the photographing device 3 photographs a subject based on information of a photographed image photographed by the photographing unit 4. Derivation of shooting conditions will be described later.

  The mobile body 2 is, for example, a vehicle such as an automobile or a train, a ship, and an aircraft. In the present embodiment, a description will be given by taking an automobile as an example.

  A route along which the moving body 2 moves is referred to as a moving route R. The movement route R is a predetermined route. The mobile body 2 can detect its own position on the movement route R by using, for example, a satellite positioning system.

  The imaging device 3 is, for example, an in-vehicle camera that captures a scene outside the moving body 2, that is, a surrounding scene.

  The light source L indicates, for example, a light source such as the sun or a street lamp, and illuminates a shooting environment in which the shooting device 3 captures a subject. Incidentally, in addition to the light source L, the moving body 2 includes an illuminating device such as a headlight, so that the photographing environment can be illuminated using the illuminating device as a light source.

An object around the moving path R that shields or reflects light from the light source is referred to as an object E. The object E affects the image quality of a captured image obtained by capturing the subject by the image capturing apparatus 3.
In FIG. 1, as an example of the object E, the object E1 indicates a tunnel, the object E2 indicates a forest, and the object E3 indicates a sea. Since the objects E1 and E2 block light and the object E3 reflects light, the captured image of the subject is affected by the type and arrangement of the object E that is a part of the shooting environment.

  By the way, the imaging system 1 has a communication unit and can communicate with the moving body 2 or the imaging apparatus 3. The imaging system 1 transmits imaging conditions to the moving body 2 or the imaging device 3 using communication means. The photographing apparatus 3 can use the photographing conditions transmitted by the photographing system 1.

  For communication between the imaging system 1 and the moving body 2 or the imaging device 3, for example, medium-range wireless communication, long-range wireless communication, or wide-area wireless communication can be used. The imaging system 1 transmits imaging conditions in a state where communication with the moving body 2 or the imaging device 3 is established. The establishment of communication between the imaging system 1 and the moving body 2 or the imaging device 3 will be described later.

  The photographing units 4a, 4b,..., 4e are photographing devices for simulating photographing with the photographing device 3 mounted on the moving body 2 arranged along the moving route R. The photographing units 4a, 4b,..., 4e are collectively referred to as a photographing unit 4.

Reference subjects 5a, 5b,..., 5e are shooting targets of the shooting unit 4 arranged along the movement route R. In FIG. 1, the photographing unit 4a photographs the reference subject 5a, the photographing unit 4b photographs the reference subject 5b, the photographing unit 4c photographs the reference subject 5c, the photographing unit 4d photographs the reference subject 5d, and the photographing unit. 4e photographs the reference subject 5e. Reference subjects 5a, 5b,..., 5e are collectively referred to as reference subject 5.
The reference subject 5 simulates a subject photographed by the photographing device 3, such as a car, a person, a road marking, or an animal. The reference subject 5 visually represents a subject to be simulated in two dimensions or three dimensions, such as a model, a photograph, or a display.

  The photographing unit 4 is installed on the moving route R at a point where the scene taken by the photographing device 3 is likely to change, that is, the photographing environment is not gradually changed. Such a point is defined as a specific point A. In FIG. 1, specific points A corresponding to the photographing units 4a, 4b,..., 4e are specified as specific points A1, A2,.

  The photographing unit 4 always photographs the reference subject 5 at the specific point A. “Continuous photographing” means that photographing is repeated periodically or periodically, and an object is to obtain a photographed image corresponding to a change in photographing environment.

  Since the shooting environment changes depending on the date and time, for example, there is no change in the weather factor because there is no cloudiness, so the frequency of shooting is reduced during the period when the shooting environment changes slowly, and vice versa. However, the frequency of shooting may be increased during a period in which there are many changes in the shooting environment.

  In FIG. 1, a captured image when the photographing unit 4a photographs the reference subject 5a is a photographed image P1, a photographed image when the photographing unit 4b photographs the reference subject 5b is a photographed image P2, and the photographing unit 4c is the reference subject. A photographed image when the photographing unit 4d photographed the reference subject 5e is taken as a photographed image P3, a photographed image when the photographing unit 4d photographed the reference subject 5d, and a photographed image P4. The image is P5.

  Here, the accuracy of recognizing a subject in a photographed image obtained by photographing the subject by the photographing device 3 is defined as a recognition rate. Similarly, the accuracy of recognizing the reference subject 5 in the photographed image obtained by photographing the reference subject 5 by the photographing unit 4 can also be expressed using the recognition rate.

  FIG. 2 is a schematic diagram showing shooting vector information when the shooting unit 4 captures the reference subject 5. Consider the relationship between the position of the photographing unit 4 and the reference subject 5 and the recognition rate of the photographed image in the photographing unit 4.

  The object E1 shown in FIG. 2 is a tunnel. FIG. 2A shows a case where the reference subject 5a1 is installed outside the tunnel. FIG. 2B shows a case where the reference subject 5a2 is installed inside the tunnel. Thus, the reference subject 5a1 in FIG. 2A is not blocked by sunlight, while the reference subject 5a2 in FIG. 2B is blocked by sunlight. As a result, a difference occurs in the recognition rate between the photographed image P11 obtained by photographing the reference subject 5a1 by the photographing unit 4a and the photographed image P12 obtained by photographing the reference subject 5a2.

  That is, depending on the difference in the position of the reference subject 5 with respect to the photographing unit 4a, there is almost no difference in the position of the reference subject 5 in the photographed image taken by the photographing unit 4a, but there is a difference in the recognition rate due to the difference in focus in photographing. It is possible that this will occur. Here, “the difference in the position of the reference subject 5 in the photographed image taken by the photographing unit 4a hardly occurs depending on the difference in the position of the reference subject 5 with respect to the photographing unit 4a” means that the photographed reference subject 5 This means that even if there is a difference in image size, there is almost no difference in the center position of the image.

  Therefore, shooting vector information is defined by combining information on the direction in which the shooting unit 4 shoots the reference subject 5 and information on the shooting distance.

The photographing system 1 outputs the photographing vector information corresponding to the reference subject 5 together with the recognition rate together with the recognition rate derived when the photographing unit 4 has photographed the reference subject 5, thereby allowing the photographing system 1 to the photographing device 3. On the other hand, it is possible to derive and provide shooting conditions including shooting vector information.
The shooting vector information corresponding to the reference subject 5 may be held in advance by the shooting unit 4 or may be derived with adjustment of the direction and distance in shooting.

  Thereby, the imaging device 3 can set the imaging condition in association with the imaging vector information, and the difference between the imaging direction and the imaging distance of the subject to be imaged from the imaging direction and the imaging distance included in the imaging vector information. Based on this, the photographing apparatus 3 can further adjust the photographing conditions for photographing the subject. As a result, it is possible to further improve the recognition rate in the photographed image by the photographing apparatus 3.

  In FIG. 1, the photographing unit 4 c photographs the reference subject 5 c with the movement route R in between. By adding the shooting vector information to the shooting conditions at this time, the shooting unit 4c can include the intention of simulating the shooting device 3 for shooting the other moving body 2 coming from the opposite lane in the shooting conditions. . Further, the photographing unit 4d photographs the reference subject 5d without sandwiching the movement route R. Similarly, by adding the shooting vector information to the shooting conditions at this time, the shooting unit 4d can include in the shooting conditions the intention of simulating the shooting device 3 for shooting the preceding moving body 2 in the same lane. I can do it.

As described above, the relative position of the reference subject 5 with respect to the photographing unit 4 affects the photographing of the photographing device 3 that receives and sets the photographing conditions, so that the photographing unit 4 has a plurality of reference subjects having different photographing vector information. It is also possible to provide the photographing apparatus 3 with photographing conditions in which information is distinguished in association with a plurality of pieces of photographing vector information by deriving a recognition rate in association with photographing vector information each time the image is taken. is there.
For example, a plurality of reference subjects 5 are arranged at a specific point A, and the photographing unit 4 at the specific point A images the plurality of reference subjects 5 to derive respective recognition rates and correspond to the recognition rates. The shooting vector information may be associated.
At this time, the photographing unit 4 may be provided with a mechanism for changing the photographing direction autonomously or by an external operation in order to photograph a plurality of reference subjects 5.

By the way, in order that the imaging device 3 can immediately use the imaging conditions received from the imaging system 1 at the specific point A, the imaging system 1 moves before the moving body 2 reaches the specific point A. The imaging conditions are transmitted to the body 2 or the imaging device 3. Therefore, in the photographing system 1, a position where the moving body 2 moving in the traveling direction D on the moving route R arrives before the specific point A is determined as the specific position B in advance. The imaging system 1 can provide imaging conditions prior to the mobile body 2 reaching the specific point A by transmitting the imaging conditions when the mobile body 2 reaches the specific position B.
In FIG. 1, specific positions B corresponding to specific points A1, A2,..., A5 are set as specific positions B1, B2,.

That is, since the photographing apparatus 3 has already received the photographing condition at the specific position B, the photographing apparatus 3 can immediately perform photographing using the photographing condition at the specific point A.
In FIG. 1, a state in which the photographing apparatus 3 uses the photographing condition derived by photographing by the photographing unit 4b by switching to the photographing condition derived by photographing by the photographing unit 4c is indicated as a condition switching Renew.

For this reason, the position of the specific position B needs to be a position that can be used at the specific point A after the photographing apparatus 3 receives the photographing condition. Such a specific position B can be determined by the range of the moving speed of the moving body 2.
Further, the specific position B may be determined as a range in which wireless communication with the imaging system 1 is possible. In this case, within the wireless communication range of the imaging system 1, for example, the imaging condition is transmitted based on the position information of the moving body 2 equipped with the imaging device 3 or in response to a request from the imaging device 3. Anyway.

The communicator 6 is a part of communication means of the photographing system 1 and is arranged along the movement route R, and relays communication between the photographing unit 4 and the moving body 2 or the photographing device 3. The communication device 6 may be installed at a high place, not on the ground.
In FIG. 1, among the communication devices 6, the communication device 6c relays the communication of the photographing unit 4c, and the communication device 6e relays the communication of the photographing unit 4e.

  A plurality of photographing units 4 may share the communication device 6 by distinguishing the communication relayed by the communication device 6. Further, the photographing unit 4 may directly communicate with the moving body 2 or the photographing device 3 without using the relay of the communication device 6.

The imaging system 1 uses communication of at least one of the imaging unit 4 and the communication device 6 as a communication unit for transmitting the derived imaging condition to the moving body 2 or the imaging device 3.
In FIG. 1, communication from the imaging system 1 to the moving body 2 or the imaging device 3 is indicated by an arrow, and communication from the moving body 2 or the imaging device 3 to the imaging system 1 is indicated as a communication Receive.

  The communication device 6 relays communication between the imaging system 1 and the moving body 2 or the imaging apparatus 3, so that the distance between the specific point A and the specific position B is wireless between the imaging unit 4 and the mobile body 2 or the imaging apparatus 3. When the communication range is exceeded, or when there is an object E that shields wireless communication between the specific point A and the specific position B, the communication device 6 is within the wireless communication range from the specific position B. By arranging, it is possible to reliably communicate between the photographing unit 4 arranged at the specific point A and the moving body 2 or the photographing device 3 that has reached the specific position B.

  Next, establishment of wireless communication in the imaging system 1 will be described.

  FIG. 3 is a schematic diagram for explaining establishment of wireless communication in the imaging system 1. FIG. 3A is a schematic diagram illustrating a process of establishing communication with the moving body 2 or the imaging apparatus 3 from the imaging system 1 side. FIG. 3B is a schematic diagram for explaining a process of establishing communication with the photographing system 1 from the moving body 2 or the photographing apparatus 3 side.

  In FIG. 3A, the radio wave notification signal Wb <b> 1 or Wb <b> 2 is intermittently transmitted from the imaging unit 4 or the communication device 6 of the imaging system 1 to the moving body 2 or the imaging device 3. The mobile body 2 or the imaging device 3 that has received the notification signal Wb1 or Wb2 transmits a radio wave response signal Wr1 to the imaging unit 4 or the communication device 6 that is the transmission source of the notification signal Wb1 or Wb2. Thereby, communication between the imaging system 1 and the moving body 2 or the imaging device 3 is established.

In FIG. 3 (b), the moving body 2 or the imaging device 3 has the imaging unit 4 or the communication device within the radio wave communication range in the traveling direction D based on the position information PosInfo obtained by using the satellite positioning system. When 6 is detected, a radio wave notification signal Wb3 is transmitted to the photographing unit 4 or the communication device 6 of the photographing system 1. The imaging unit 4 or the communication device 6 that has received the notification signal Wb3 transmits a radio wave response signal Wr2 or Wr3 to the mobile body 2 or the imaging device 3 that is the transmission source of the notification signal Wb3. Thereby, communication between the imaging system 1 and the moving body 2 or the imaging device 3 is established.
Even when communication is established from the mobile unit 2 or the imaging device 3 side, the position information PosInfo is not used, and intermittently in the same manner as the notification signals Wb1 and Wb2 in FIG. The notification signal Wb3 may be transmitted.

  3A and 3B, when communication between the imaging system 1 and the moving body 2 or the imaging device 3 is established, the imaging system 1 communicates with the imaging system 1 from the mobile body 2 or the imaging device 3. By receiving, the position information PosInfo of the moving body 2 can be acquired periodically or at an arbitrary timing, so that it is possible to detect that the moving body 2 has reached the specific position B. When the imaging system 1 detects that the moving body 2 has reached the specific position B, the imaging system 1 transmits imaging conditions by communication Send from the imaging system 1 to the moving body 2 or the imaging device 3.

Next, the configuration of the photographing system 1 will be described.
FIG. 4 is a functional block diagram illustrating the configuration of the imaging system 1. The imaging system 1 includes a plurality of imaging units 4 and a communication device 6, and an electronic computer 500. The photographing system 1, the photographing unit 4, and the electronic computer 500 are in the real space Sr.

  The external network 20 is a communication line such as a wide area network, the Internet network, or a dedicated line network. The electronic computer 500 of the photographing system 1 is connected to the external network 20 and transmits / receives information to / from other devices.

  The external server 30 is a server device on the cloud, for example, and is connected to the external network 20 to provide various information to the electronic computer 500 of the imaging system 1.

Next, the configuration of the photographing unit 4 will be described.
The shooting unit 4 includes a shooting unit 411, an image adjustment unit 412, an image recognition unit 413, and a first communication unit 414.
Further, the photographing unit 4 holds correct image information Dc in a storage medium (not shown) as information indicating a correct image of the reference subject 5 in a photographed image obtained by photographing the reference subject 5. It is assumed that a correct image representing the reference subject 5 is prepared in advance.

  The imaging unit 411 is an imaging device that images the reference subject 5 and outputs a captured image D1 that is a captured image.

  The image adjustment unit 412 outputs captured image information D2 obtained by performing image processing for adjusting the image quality so that the reflected image of the reference subject 5 is clearly displayed on the captured image D1 output by the imaging unit 411.

  The image recognition unit 413 performs image recognition of the reference subject 5 based on a comparison between the captured image information D2 output from the image adjustment unit 412 and the correct image information Dc as image processing. Image recognition here is extracting image information similar to the correct image information Dc from the captured image information D2.

  Then, the image recognition unit 413 derives a recognition rate that quantifies the degree to which the image information extracted from the captured image information D2 matches the correct image information Dc. The recognition rate when the image information extracted from the captured image information D2 completely matches the correct image information Dc is 100%.

  The image recognition unit 413 outputs a signal Sig1 including the derived recognition rate to the electronic computer 500. The image recognition unit 413 may further include shooting vector information in the signal Sig1 output at this time.

  The first communication unit 414 communicates with the moving body 2 or the imaging device 3, the communication device 6, and the electronic computer 500. Communication between the first communication unit 414 and the moving body 2 or the imaging device 3 is wireless communication, but communication between the first communication unit 414 and the communication device 6 and the electronic computer 500 is wireless communication even if it is wired communication. It may be.

  The electronic computer 500 is a computer that can execute high-speed arithmetic processing and communication processing, and is, for example, a supercomputer that constitutes cloud computing. The electronic computer 500 communicates with the first communication unit 414 of the photographing unit 4 or the communication device 6 in the real space Sr.

  The electronic computer 500 derives the shooting setting of the shooting unit 4 that improves the recognition rate in the shot image based on the shot image obtained by the shooting unit 4 shooting the reference subject 5. Then, the electronic computer 500 derives an imaging condition that improves the recognition rate when the imaging device 3 captures the subject at the specific point A based on the derived imaging setting.

Next, the configuration of the electronic computer 500 will be described.
The electronic computer 500 includes a storage unit 501, a condition derivation unit 502, a setting change unit 503, a condition update unit 504, and a second communication unit 505.

  The storage unit 501 is a storage medium for storing information related to the photographing unit 4 and information related to the moving body 2 and the photographing device 3. Various information stored in the storage unit 501 will be described later.

  The condition deriving unit 502 receives the signal Sig1 including the recognition rate from the image recognition unit 413 of the shooting unit 4, and derives shooting settings in which various setting contents are adjusted so as to improve the recognition rate of shooting by the shooting unit 4. To do. The condition deriving unit 502 stores the obtained shooting setting in the storage unit 501 or obtains it from the storage unit 501 by the signal Sig2.

  The condition deriving unit 502 adjusts the setting contents here through a learning process using a learning algorithm. As the learning algorithm, for example, a known method such as reinforcement learning (Re-inforce Learning) or a neural network is used.

  Further, the condition deriving unit 502 derives a photographing condition for photographing by the photographing device 3 at the specific point A where the photographing unit 4 is arranged based on the derived photographing setting. The condition deriving unit 502 stores the acquired photographing condition in the storage unit 501 or obtains it from the storage unit 501 by the signal Sig2.

  In addition, the condition deriving unit 502 communicates with the external server 30 via the second communication unit 505 and the external network 20, and collects information related to various shooting environments at the specific point A from the external server 30. And stored in the storage unit 501. Communication when the condition deriving unit 502 collects a shooting environment from the external server 30 is indicated by a signal ExInfo.

The setting change unit 503 sets the shooting setting derived by the condition deriving unit 502 for the shooting unit 411, the image adjustment unit 412, and the image recognition unit 413.
In FIG. 4, the shooting setting for the shooting unit 411 is a signal Sig4, the shooting setting for the image adjustment unit 412 is a signal Sig5, and the shooting setting for the image recognition unit 413 is a signal Sig6.

  The condition update unit 504 communicates with the moving body 2 or the imaging device 3 via the second communication unit 505 of the electronic computer 500 and the first communication unit 414 of the imaging unit 4. Then, the condition update unit 504 causes the storage unit 501 to store various information received from the moving body 2 or the imaging device 3.

  When the condition updating unit 504 detects that the moving body 2 has reached the specific position B based on the position information PosInfo received from the moving body 2 or the imaging device 3, the condition updating unit 504 stores the moving body stored in the storage unit 501 by the condition deriving unit 502. 2 is acquired and transmitted to the moving body 2 or the imaging device 3.

It is assumed that the moving body 2 can exchange its position information PosInfo with the satellite positioning system 50 and detect its own position on the moving route R. Note that the photographing apparatus 3 may acquire the position information PosInfo handled by the moving body 2 and transmit the position information PosInfo to the electronic computer 500.
For the satellite positioning system 50, for example, a GPS (Global Positioning System) can be used.

  The second communication unit 505 communicates with the imaging unit 4, the communication device 6, the moving body 2 or the imaging device 3, and the external server 30 via the external network 20. Communication between the second communication unit 505 and the moving body 2 or the imaging device 3 is wireless communication, but communication between the second communication unit 505 and the imaging unit 4, the communication device 6, and the external server 30 via the external network 20. May be wired communication or wireless communication.

  The imaging system 1 communicates with the moving body 2 or the imaging device 3 via communication between the second communication unit 505 of the electronic computer 500 and at least one of the first communication unit 414 of the imaging unit 4 and the communication device 6. A means for performing communication between them is referred to as a communication means 10. That is, the imaging system 1 includes a communication unit 10 for communicating with the moving body 2 or the imaging device 3.

  Here, the relationship between the imaging settings of the imaging unit 4 and the imaging conditions of the imaging device 3 derived by the electronic computer 500 in the imaging system 1 will be described.

  When the recognition rate of the reference subject 5 is high in the photographed image taken by the photographing unit 4 at the specific point A, the situation where the photographing unit 4 photographs the reference subject 5 and the situation where the photographing device 3 photographs the subject at the same specific point A It can be said that the recognition rate for recognizing the subject tends to increase in the photographed image obtained by photographing the subject by the photographing apparatus 3 using the photographing setting adjusted in the photographing unit 4.

As described above, in the photographing system 1, it is desirable that the photographing environment in which the photographing unit 4 photographs the reference subject 5 is as close as possible to the photographing environment in which the photographing device 3 photographs the subject.
For this purpose, the photographing unit 4 and the reference subject 5 are installed so that the photographing environment that is a scene photographed by the photographing unit 4 is approximated to the photographing environment of the photographing device 3 as much as possible.

  For example, the direction in which the photographing unit 4 shoots the reference subject 5 is adjusted to a range of a photographing direction with a high photographing frequency in the photographing device 3 or a range of a photographing direction in which a specific subject to be photographed with a high recognition rate exists. Install. The range of the photographing direction in the photographing device 3 is that the subject photographed by the photographing device 3 is not necessarily the center of the photographed image, and the subject reflected in the corner of the photographed image may be subjected to image processing. This is because it is not necessary to match the shooting direction of the shooting unit 4 with the shooting direction of the shooting device 3.

  Moreover, it can be said that the degree of approximation of the shooting environment has a positive correlation with the degree of approximation of the positional relationship between the image sensor, the subject and the light source, and the amount of light from the light source, for example. Therefore, the photographing unit 4 may include an illumination device that simulates the headlight of the moving body 2 so that the photographing environment of the photographing unit 4 is approximated to the photographing environment of the photographing device 3. The photographing unit 4 can photograph in the photographing environment equivalent to the photographing device 3 at night by controlling the lighting device.

  In the configuration of the imaging system 1 illustrated in FIG. 4, the electronic computer 500 can perform learning processing for improving the recognition rate by using learning results of a plurality of specific points A. As a result, learning at a plurality of specific points A where the shooting environment is similar or a high positive correlation is seen in the change of the shooting environment, rather than individually performing learning processing using information at the specific point A By using the result, it is possible to expect an effect of obtaining a shooting setting that can be recognized more accurately by executing learning processing with a higher degree of convergence and with a higher recall rate.

  Note that “shows a high positive correlation” means that there is a positive correlation between certain information or a set of data obtained by digitizing certain information, and the standardized correlation coefficient r for covariance is reliable. Say that you are in the leg. As the confidence interval of the correlation coefficient r at this time, for example, a 95% confidence interval using the standard error of the correlation coefficient is used. As the correlation coefficient, Pearson's product moment correlation coefficient can be used.

  When the computer 500 collects information corresponding to a plurality of specific points A from the external server 30, the external server 30 provides information by associating with the position information of each specific point A, or specifying information It is possible to compare and select the degree of change at the point A to provide information, and to efficiently provide information to the electronic computer 500. As a result, the electronic computer 500 efficiently collects information from the external server 30 and processes it efficiently, or processes the associated information together to improve immediacy in deriving shooting settings and shooting conditions. The improvement effect can be expected.

  Further, as compared with the case where the photographing conditions are derived by the photographing unit 4, the photographing unit 4 can be configured with a smaller number of implementations, so that the cost of installing many photographing units 4 can be reduced.

Next, a modified example of the configuration of the photographing system 1 will be described.
FIG. 5 is a functional block diagram illustrating a modified example of the configuration of the imaging system 1. As shown in FIG. 5, each photographing unit 4 of the photographing system 1 has the function of the electronic computer 500 shown in FIG. That is, each photographing unit 4 shown in FIG. 5 operates individually to derive the photographing condition of the photographing device 3.

The photographing unit 4 includes a photographing unit 411, an image adjusting unit 412, an image recognizing unit 413, a storage unit 501, a condition deriving unit 502, a setting changing unit 503, a condition updating unit 504, and a first communication unit 414.
Further, the photographing unit 4 holds correct image information Dc in a storage medium (not shown) as information indicating a correct image of the reference subject 5 in a photographed image obtained by photographing the reference subject 5.
The description of the same reference numerals as in FIG. 4 will not be repeated.

  In the configuration of the imaging system 1 shown in FIG. 5, each imaging unit 4 individually derives imaging settings for improving the recognition rate and imaging conditions to be transmitted to the imaging device 3 at the arranged specific point A, and the moving body 2. It is possible to transmit the shooting conditions individually while detecting the position. As a result, the effect of providing the imaging conditions to the imaging device 3 can be obtained without constructing the large-scale imaging system 1 including the electronic computer 500 shown in FIG.

  In addition, even if the installation number or installation location of the imaging units 4 is changed in accordance with the system expansion or maintenance work in the imaging system 1, the imaging units 4 that are not subject to change can continue to operate during that time. It is. This leads to an increase in expandability and maintainability of the imaging system 1 compared to the configuration shown in FIG.

Next, various information stored in the storage unit 501 of the imaging system 1 will be described.
FIG. 6 is an information relation diagram showing the relation between various information handled by the photographing system 1. 6, the case of the imaging system 1 shown in FIG. 4 will be described.

  The storage unit 501 stores shooting unit information 30a that is information about the shooting unit 4, and moving body and shooting device information 30b that is information about the moving body 2 and the shooting device 3.

  Examples of information included in the photographing unit information 30a include arrangement information 31, photographing environment information 32, photographing information 33, photographing setting information 34, and learning information 35.

  The arrangement information 31 is information prepared in advance, and includes specific point information 31a indicating the position of each specific point A and specific position information 31b indicating the position of each specific position B.

The shooting environment information 32 includes date and time information 32a, weather information 32b, and shooting field of view information 32c.
The date information 32a is information indicating the date.
The weather information 32b is information indicating the weather and temperature at each specific point A.
The imaging field of view information 32c is information on an object that affects imaging at each specific point A. For example, the imaging field information 32c is information on the object E shown in FIG. Or
The date / time information 32a, the weather information 32b, and the photographing visual field information 32c are associated with each other.

The condition deriving unit 502 collects each piece of information included in the shooting environment information 32 from the external server 30 periodically, periodically, or at a predetermined timing. The predetermined timing is, for example, when information is provided from the external server 30 side.
The condition deriving unit 502 acquires each piece of information included in the shooting environment information 32 in association with each piece of information included in the arrangement information 31 and stores it in the storage unit 501. Information that associates the arrangement information 31 and the shooting environment information 32 is referred to as association information Rel1.

  The shooting information 33 includes recognition rate information 33a that is a recognition rate derived by each shooting unit 4, and shooting vector information 33b that is shooting vector information associated with the recognition rate information 33a.

  The shooting setting information 34 is information on setting contents related to shooting processing and image processing in each shooting unit 4, and includes shooting control information 34a, illumination control information 34b, and image processing control information 34c.

  The shooting control information 34 a is information on setting contents related to shooting processing in the shooting unit 411 of each shooting unit 4. In a case where the photographing unit 4 includes a mechanism for controlling its own device such as a mechanism for changing the photographing direction, the photographing control information 34a may further include information related to mechanism control of the photographing unit 4.

  The illumination control information 34b is information relating to parameters such as the presence / absence of illumination, the amount of illumination, and the illumination direction for controlling the illumination device serving as the light source when the photographing unit 4 includes a light source such as an illumination device. The illumination control information 34b is information used when the moving body 2 controls the headlight.

  The image processing control information 34c is information on an image processing algorithm used when the image adjusting unit 412 and the image recognizing unit 413 of each photographing unit 4 process a photographed image, and parameters related to the image processing. The image processing algorithm includes an image extraction algorithm, an image recognition algorithm, and an image inference algorithm.

Here, it is assumed that the photographing process and the image processing performed by each photographing unit 4 are equivalent to the photographing process and the image processing performed by the photographing apparatus 3. Here, “equivalent” means that a photographed image having the same image quality can be obtained by adjusting the setting even if the manufacturer and model differ from device to device.
Regarding the fact that the manufacturer and model differ depending on the device, the electronic computer 500 of the photographing system 1 stores and uses information in which the specifications and setting contents of each device are associated with each other in advance so that the photographing setting of the photographing unit 4 can be made. Based on this, it is possible to derive the photographing conditions of each photographing device 3.

  The learning information 35 includes learning control information 35a and learning data 35b.

  The learning control information 35a includes learning algorithm information used for learning processing executed by the condition derivation unit 502 of the imaging system 1, information on parameters related to learning processing, weighting information on learning data, and bias information of learning processing. It is information to show.

  The learning data 35b is a learning result of learning processing executed by the condition deriving unit 502 of the imaging system 1.

  The arrangement information 31, the shooting environment information 32, the shooting information 33, the shooting setting information 34, and the learning information 35 are associated with each other, and information indicating this association is referred to as association information Rel2.

  Information included in the moving body and imaging device information 30b includes moving body information 36, imaging device information 37, and imaging condition information 38.

  The moving body information 36 includes moving body identification information 36 a, position information 36 b, and imaging device information 37.

  The moving body identification information 36a is information prepared in advance, and is information unique to the moving body 2 that is predetermined to identify each moving body 2. The condition update unit 504 may acquire the moving body identification information 36a together with the position information PosInfo.

The position information 36b is information indicating the position of each moving body 2. Based on the position information PosInfo of the satellite positioning system 50 acquired from each mobile unit 2, it is updated by the condition update unit 504. Note that the condition update unit 504 may acquire the position information PosInfo directly from the satellite positioning system without using the moving body 2 or the imaging device 3.
Information that associates the mobile object information 36 with the arrangement information 31 is referred to as association information Rel3.

The imaging device information 37 includes imaging device identification information 37a.
The imaging device identification information 37a is information prepared in advance, and is information unique to the imaging device 3 that is predetermined to identify each imaging device 3. The condition update unit 504 may acquire the imaging device identification information 37a together with the position information PosInfo.

  The shooting condition information 38 includes shooting control information 38a, illumination control information 38b, and image processing control information 38c.

  The shooting control information 38 a is information on setting contents related to shooting processing of each shooting apparatus 3. When the imaging device 3 includes a mechanism for controlling the own device such as a mechanism for changing the imaging direction, the imaging control information 38a may further include information related to mechanism control of the imaging device 3.

  The illumination control information 38b is information regarding parameters such as the presence / absence of illumination, the amount of illumination, and the illumination direction for controlling the headlights included in each moving body 2.

  The image processing control information 38c is information on an image processing algorithm used when each photographing apparatus 3 performs image processing on a photographed image and parameters related to the image processing. The image processing algorithm includes an image extraction algorithm, an image recognition algorithm, and an image inference algorithm.

  FIG. 7 is an information list diagram illustrating an example of information included in the shooting condition information 38.

  The contents of the shooting control information 38a in the shooting condition information 38 are shown in a list 71. The list 71 includes information on image sensor sensitivity, exposure time, exposure area, convergence brightness, white balance, convergence color space, gain control, black level control, filter coefficient, light transmission wavelength band, and mechanism control. Is included. The information on the mechanism control includes, for example, shooting direction, light shielding when a light shielding unit is provided, shake correction, and camera selection when a plurality of cameras are provided.

  The contents of the illumination control information 38b in the shooting condition information 38 are shown in a list 72. The list 72 includes information on the presence / absence of illumination (that is, whether illumination is on or off), the amount of illumination, and the illumination direction. The illumination control information 38b is information for controlling the headlight of the moving body 2. The illumination control information 38b is information that can be used by the photographing apparatus 3 for photographing with a higher recognition rate, and the headlight from the photographing apparatus 3 side. In this case, it is included in the photographing condition information 38.

  List 73 shows the contents of the image processing control information 38c in the photographing condition information 38. The list 73 includes information on an image extraction algorithm, an image recognition algorithm, and an image inference algorithm. Further, the information on the image extraction algorithm includes information on contour enhancement, hue, saturation, contrast adjustment, and color extraction as extraction parameters. The information of the image recognition algorithm includes information of recognition parameters. The information of the image inference algorithm includes information of inference parameters.

  There are a plurality of pieces of information shown in FIGS. In the case of the photographing system 1 shown in FIG. 4, the arrangement information 31 and the photographing environment information 32 exist for the number of photographing units 4 to be arranged. Further, the shooting information 33, the shooting setting information 34, the learning information 35, and the shooting condition information 38 exist for the reference subject 5 to be shot by each of the shooting units 4 arranged. Further, the moving body information 36 and the imaging device information 37 exist for the imaging device 3 that uses the imaging system 1.

  On the other hand, in the case of the photographing system 1 shown in FIG. 5, the arrangement information 31 and the photographing environment information 32 exist for the photographing unit 4 that is the own device. Further, the shooting information 33, the shooting setting information 34, the learning information 35, and the shooting condition information 38 exist for the reference subject 5 shot by the shooting unit 4 that is the device itself. Further, the moving body information 36 and the imaging device information 37 exist for the imaging device 3 that uses the imaging system 1.

  Next, operations of the electronic computer 500 and the photographing unit 4 in the photographing system 1 will be described with reference to FIGS. 4, 6, 8 and 9.

  FIG. 8 is a flowchart showing processing in which the photographing system 1 transmits photographing conditions to the photographing device 3. FIG. 9 is a flowchart showing a predefined process in which the photographing unit 4 photographs the reference subject 5 and derives the recognition rate.

First, FIG. 8 will be described.
In process S801, the electronic computer 500 performs initial settings for executing various processes in the imaging system 1 when given initial values of various information stored in the storage unit 501.
The initial setting refers to the specific spot information 31a and the specific position information 31b included in the arrangement information 31 of the photographing unit information 30a stored in the storage unit 501, and the moving object identification information included in the moving object and the photographing device information 30b. This is a predetermined information setting of 36a and imaging device identification information 37a. It is assumed that appropriate initial information is set for information that is updated every other time.
Then, it progresses to process S802.

In the process S802, the condition deriving unit 502 of the computer 500 receives the signal ExInfo from the external server 30 after the initial setting of the arrangement information 31 in the process S801 is completed, and the shooting environment at each specific point A Information 32 is collected. Then, the condition deriving unit 502 refers to the association information Rel1 between the arrangement information 31 and the shooting environment information 32, and starts an update process for updating the shooting environment information 32. This update process is executed periodically or periodically or at a predetermined timing.
Then, it progresses to process S803.

  The fact that the condition deriving unit 502 collects the shooting environment information 32 from the external server 30 triggered by the initial setting of the arrangement information 31 in process S802 is indicated by an arrow Base1 in FIG.

  In process S803, when the condition updating unit 504 of the electronic computer 500 receives the position information PosInfo from the moving body 2 or the imaging device 3, the position information 36b is updated for each moving body 2, and the arrangement information 31 and the moving body information 36 are updated. With reference to the association information Rel3, it is determined whether there is the moving body 2 that has reached the specific position B. As a result of the determination, if it exists, the process proceeds to step S813. If it does not exist, the process proceeds to step S804.

  In addition, after the processing S803, for example, each processing such as acquisition of the position information PosInfo of each moving body 2 and derivation of the photographing setting of each photographing unit 4 is executed individually.

  In process S804, the condition deriving unit 502 refers to the association information Rel2 between the arrangement information 31 and the learning information 35, and the learning information 35 associated with the specific point A where the photographing unit 4 is arranged is stored in the storage unit 501. Check whether it is. Then, it progresses to process S805.

  In process S805, if learning information 35 is stored in process S804, the process proceeds to process S807, and if not stored, the process proceeds to process S806.

  In step S806, the setting change unit 503 of the electronic computer 500 initializes the shooting unit 411, the image adjustment unit 412 and the image recognition unit 413 of the shooting unit 4 based on the information of the shooting setting information 34 stored in the storage unit 501. To do. The photographing unit 4 starts operation after initialization. Thereafter, the process proceeds to the predefined process S809.

In process S807, the condition deriving unit 502 executes learning processing while updating the shooting setting information 34 and the learning information 35 based on the association information Rel2, and sets the shooting unit 411, the image adjustment unit 412, and the image recognition unit 413. The shooting setting information 34 is derived.
More specifically, the condition deriving unit 502 acquires the recognition rate included in the signal Sig1 from the image recognition unit 413 of the shooting unit 4, and further increases the recognition rate in the shooting environment of the shooting unit 4. Setting of shooting processing performed by the shooting unit 411 of the shooting unit 4 (that is, shooting control information 34a and illumination control information 34b), and setting of image processing performed by the image adjustment unit 412 and the image recognition unit 413 of the shooting unit 4 (That is, the image processing control information 34c) is adjusted. These settings are collectively referred to as shooting settings (that is, shooting setting information 34). The condition deriving unit 502 individually adjusts the shooting settings of each shooting unit 4.
Further, the condition deriving unit 502 outputs a signal Sig3 including the shooting setting information 34 to the setting changing unit 503 in order to reflect the adjusted shooting setting on the corresponding shooting unit 4.
Then, it progresses to process S808.

  The fact that the condition deriving unit 502 derives the shooting setting information 34 by referring to the learning information 35 in the process S807 is indicated by an arrow Refer1 in FIG.

  In the learning process, the condition deriving unit 502 predicts a shooting environment that changes from moment to moment, such as the amount of light from the light source, for example, when the time advances from the time when the shooting unit 4 outputs the recognition rate. Thus, the shooting setting may be adjusted to improve the shooting recognition rate.

In process S808, the setting change unit 503 searches for the shooting unit 4 corresponding to the shooting setting information 34 included in the signal Sig3 output in process S807, and the shooting unit 411, the image adjustment unit 412 and the shooting unit 4 of the shooting unit 4 are searched. A signal Sig4, a signal Sig5, and a signal Sig6 for reflecting the shooting setting are output to the image recognition unit 413, respectively. The imaging unit 411 inputs the signal Sig4 to reflect the corresponding imaging setting, the image adjustment unit 412 inputs the signal Sig5 to reflect the corresponding imaging setting, and the image recognition unit 413 inputs the signal Sig6 to respond. Reflect shooting settings.
Thereafter, the process proceeds to the predefined process S809.

  The predefined process S809 is a predefined process shown in FIG. In the predefined process S809, the photographing unit 4 derives a recognition rate when the reference subject 5 is photographed. Then, it progresses to process S810.

  In process S810, the condition deriving unit 502 acquires the shooting information 33 included in the signal Sig1 output from the image recognition unit 413, and updates the information in the storage unit 501. Then, it progresses to process S811.

  In process S811, the condition deriving unit 502 determines whether the recognition rate information 33a included in the acquired shooting information 33 has been improved or decreased as compared with the information already stored. Then, it progresses to process S812.

In the process S812, the condition deriving unit 502 updates the learning data 35b of the learning information 35 in association with each piece of information associated with the association information Rel2 based on the determination result in the process S811.
More specifically, the condition deriving unit 502, for example, in the shooting environment indicated by the shooting environment information 32 at the specific point A in the specific point information 31a when the recognition rate is improved in the determination in the process S811, the shooting unit 4 It is reflected in the learning data 35b that it is appropriate to set the shooting setting to the contents of the shooting setting information 34.
In addition, for example, when the recognition rate decreases in the determination in step S811, the condition deriving unit 502 sets the shooting setting of the shooting unit 4 in the shooting environment indicated by the shooting environment information 32 at the specific point A in the specific point information 31a. It is reflected in the learning data 35b that it is not suitable for the contents of the shooting setting information 34.
Then, it progresses to process S803.

  The fact that the condition deriving unit 502 updates the learning information 35 with reference to the association information Rel2 in the process S812 is indicated by an arrow Refer2 in FIG.

  As described above, each photographing unit 4 repeatedly shoots the reference subject 5 based on the photographing setting adjusted by the condition deriving unit 502 and derives the recognition rate, and the condition deriving unit 502 is based on the recognition rate derived by the photographing unit 4. By repeatedly deriving shooting settings that improve the recognition rate of the shooting unit 4 through learning processing, each shooting unit 4 performs shooting while maintaining a good recognition rate even if the shooting environment changes from moment to moment. I can do it.

In the process S813, the condition update unit 504 learns the learning result associated with the specific point A (that is, the specific point information 31a) corresponding to the specific position B (that is, the specific position information 31b) reached by the mobile body 2 in the process S803. It is determined whether the content of (i.e., learning data 35b) is sufficient. The criterion for determining whether or not the learning data 35b is sufficient is, for example, a case where the recognition rate (that is, the recognition rate information 33a) that is the result of learning converges within a predetermined fluctuation range, or is determined in advance. It is conceivable that the recognition rate has been reached or the learning process in the condition deriving unit 502 has passed a predetermined time.
By performing the determination here, it is possible to avoid a situation in which learning is not sufficiently advanced and a photographing condition that is unlikely to improve the recognition rate in the photographing device 3 is provided to the photographing device 3. I can do it.
As a result of the determination, if it is determined that it is sufficient, the process proceeds to step S814. When it determines with it not being enough, it progresses to process S816.

  In the process S814, the condition deriving unit 502 derives the photographing condition (that is, the photographing condition information 38) to be transmitted to the photographing apparatus 3 based on the derived photographing setting (that is, the photographing setting information 34). Then, it progresses to process S815.

  The fact that the condition deriving unit 502 derives the shooting condition information 38 with reference to the shooting setting information 34 in the process S814 is indicated by an arrow Refer3 in FIG.

  In the process S815, the condition update unit 504, based on the association information Rel3 between the arrangement information 31 and the moving body information 36, the moving body 2 where the position information 36b has reached the specific position B (that is, the specific position information 31b) or the movement The derived imaging conditions are transmitted to the imaging device 3 of the body 2. Thereafter, the process ends.

  In the process S816, the condition updating unit 504, based on the association information Rel3 between the arrangement information 31 and the moving body information 36, the moving body 2 where the position information 36b has reached the specific position B (that is, the specific position information 31b) or the movement The derived imaging conditions are not transmitted to the imaging device 3 of the body 2. Thereafter, the process ends.

  In steps S815 and S816, the condition update unit 504 refers to the association information Rel3 and transmits the shooting condition information 38, as indicated by an arrow Base2 in FIG.

  Note that the derivation of the shooting condition information 38 performed in step S814 in the flowchart of FIG. 8 may be performed every time the shooting setting information 34 is derived in step S807.

Next, FIG. 9 will be described.
In process S901, the photographing unit 411 of the photographing unit 4 photographs the reference subject 5. The photographing unit 411 outputs a photographed image D1 that is a photographed photographed image. Then, it progresses to process S902.

  In process S902, the image adjustment unit 412 of the photographing unit 4 adjusts the image quality of the photographed image D1 so that the reference subject 5 becomes clearer in the photographed image D1 output in process S901. Then, the image adjustment unit 412 outputs captured image information D2, which is image information whose image quality has been adjusted. Then, it progresses to process S903.

  In step S903, the image recognition unit 413 of the photographing unit 4 extracts image information similar to the correct image information Dc from the photographed image information D2 output in step S902. Then, it progresses to process S904.

  In process S904, the image recognition unit 413 derives a recognition rate based on the comparison between the image information extracted in process S903 and the correct image information Dc. Thereafter, the process ends.

  By the way, it has been described that it is desirable to approximate the photographing environment of the photographing unit 4 to the photographing environment of the photographing device 3 in deriving the photographing condition of the photographing device 3 based on the photographing setting of the photographing unit 4.

  Here, an example of a process in which the condition deriving unit 502 derives the photographing condition of the photographing apparatus 3 based on the photographing setting of the photographing unit 4 in the process S814 of FIG.

The electronic computer 500 converts an image captured by the image capturing unit 4 into an image captured as if the image capturing device 3 captured an object, based on information regarding the difference in image capturing environment between the image capturing unit 4 and the image capturing device 3. Consider processing. The image processing at this time is referred to as image conversion processing Trans. In addition, an image generated by the image conversion process Trans is referred to as a converted captured image Pt.
The electronic computer 500 derives the photographing condition of the photographing device 3 based on the information regarding the difference for generating the converted photographed image Pt and the photographing setting of the photographing unit 4.

  FIG. 10 is a schematic diagram for explaining an image conversion process Trans for deriving a shooting condition.

The condition deriving unit 502 of the electronic computer 500 acquires a captured image obtained by capturing the reference subject 5 from the photographing unit 4. The captured image acquired at this time is referred to as an original captured image Po.
The condition deriving unit 502 then compares the photographing position between the photographing unit 4 and the photographing position of the specific photographing device 3X, the photographing direction and photographing distance of the reference subject 5 with the photographing unit 4 as a base point, and the specific photographing device 3X. The image conversion processing Trans is performed based on the information about the difference between the shooting direction and the shooting distance of the specific subject 2X with reference to the difference between the shooting environments of the shooting unit 4 and the specific shooting device 3X, The original photographed image Po is converted.

Note that the difference between the shooting positions of the shooting unit 4 and the shooting position of the specific shooting device 3X, the shooting direction and shooting distance of the reference subject 5 with the shooting unit 4 as a base point, and a specific point with the specific shooting device 3X as a base point It is assumed that the electronic computer 500 holds in advance information regarding the difference between the shooting direction and shooting distance of the subject 2X and the difference in shooting environment between the shooting unit 4 and the specific shooting device 3X.
Further, the specific photographing device 3X and the specific subject 2X at this time are given in advance to the electronic computer 500 as a basic model case.

  The condition deriving unit 502 performs a coordinate conversion process Tr1 and a light amount estimation process Tr2 in the image conversion process Trans.

The coordinate conversion process Tr1 converts the original photographed image Po of the reference subject 5 acquired from the photographing unit 4 based on information on the photographing position, photographing direction, and photographing distance difference between the photographing unit 4 and the specific photographing device 3X. This is a process for deriving a captured image when a specific subject 2X is captured from a specific capturing device 3X.
As an example of the derivation process, the difference in shooting position between the shooting unit 4 and the specific shooting device 3X is a position where the specific shooting device 3X has moved by 1 meter northward from the shooting unit 4. Further, the difference between the photographing direction and the photographing distance between the photographing unit 4 and the specific photographing device 3X is a position where the specific subject 2X has moved by 2 meters toward the north and west from the reference subject 5, respectively. Thereby, the shooting direction of the specific subject 2X viewed from the specific shooting device 3X is 1 meter in the north direction and 2 meters in the west direction relative to the shooting direction of the reference subject 5 viewed from the shooting unit 4. It can be seen that a difference of Based on the information on the difference between these shooting positions, shooting directions, and shooting distances, it is possible to estimate and derive a shot image when a specific subject 2X is shot from a specific shooting device 3X.
In this way, in the coordinate conversion processing Tr1, based on the photographed image obtained when the photographing unit 4 photographed the reference subject 5, a photographed image when different subjects are photographed at different photographing positions, photographing directions, and photographing distances is derived. To do.

  The light amount estimation process Tr2 is based on the difference in the shooting environment depending on the shooting position between the shooting unit 4 and the specific shooting device 3X and the difference in shooting environment based on the installation position of the reference subject 5 and the specific subject 2X. This is a process of estimating and deriving the amount of light from the specific subject 2X and the shooting environment incident on the specific shooting device 3X when the specific subject 2X is shot from the device 3X. The estimated value of the amount of light at this time can be derived by estimating the amount by which the amount of light incident on the specific photographing apparatus 3X increases or decreases with respect to the photographing unit 4 from the difference in the photographing environment described above.

  As described above, the condition deriving unit 502 performs the image conversion process Trans, and based on the original photographed image Po photographed by the photographing unit 4, the reference subject 5 and the reference subject 5 from the photographing device 3 at a photographing position different from the photographing unit 4. A converted photographed image Pt obtained by converting the original photographed image Po can be generated as if a different subject was photographed.

Then, the condition deriving unit 502 converts the original captured image Po into the converted captured image Pt in the image conversion process Trans, and converts the captured setting corresponding to the original captured image Po into the captured condition corresponding to the converted captured image Pt. To do.
The conversion process from the shooting setting to the shooting condition is performed so as to maintain the recognition rate in the shot image before and after the image conversion process Trans. That is, along with the conversion from the original photographed image Po to the converted photographed image Pt in the image conversion process Trans, the value of the recognition rate in the photographed image during or after conversion is converted into the original photographed image Po before the image conversion process Trans is performed. The adjustment for maintaining the recognition rate value in is performed on the shooting setting. In other words, correction is performed while adjusting the content of the shooting settings so as to cancel the change in image quality caused by the conversion process to the shot image shot by the shooting unit 4. By doing in this way, the imaging conditions of the specific imaging device 3X can be derived.

  This is a case where it is difficult to install the photographing unit 4 at a position where the photographing device 3 is frequently photographed, or to install the reference subject 5 at a position where the photographing device 3 photographs frequently. However, the electronic computer 500 can derive and provide shooting conditions for shooting a subject with a high recognition rate to the shooting apparatus 3.

  The condition deriving unit 502 may derive the shooting conditions of the shooting device 3 based only on the shooting settings of the shooting unit 4 without performing the image conversion process Trans, In addition, the imaging condition of the imaging device 3 may be derived based on at least one of the information regarding the difference between the coordinate conversion processing Tr1 and the light amount estimation processing Tr2.

  That is, in the first embodiment, by using the shooting setting derived based on the shooting by the shooting unit 4, the shooting condition that allows the shooting device 3 to take a higher recognition rate than before can be set. One of the purposes is to do.

  For example, depending on the specifications of the photographing apparatus 3, it is possible to reflect only the illumination control information 38b among the photographing conditions received from the photographing system 1 and reflect the information in the setting of the own apparatus.

  In addition, the photographing apparatus 3 includes a storage device (not shown) so that the initial photographing conditions prepared in advance and the photographing conditions received from the photographing system 1 can be switched or partially set and photographed. I can do it. For example, when the moving body 2 starts moving, the photographing device 3 sets initial photographing conditions prepared and stored in the storage device of the own device, receives the photographing conditions from the photographing system 1, sets the photographing conditions, and then The initial photographing conditions stored in the storage device can be set again. By doing in this way, the imaging device 3 operates to return to the initial imaging conditions held by itself when the received imaging conditions cannot be properly captured or when the received imaging conditions are no longer necessary. It becomes possible to do.

As described above, according to the first embodiment, the imaging system 1 includes the imaging unit 411, and the imaging unit 411 captures the reference subject 5 at the specific point A to derive the recognition rate in the captured image. The imaging system 1 further includes a condition deriving unit 502. The condition deriving unit 502 derives an imaging setting for improving the recognition rate of the imaging in the repeated imaging unit 411. The condition deriving unit 502 is based on the imaging setting. The imaging condition is derived, and the imaging system 1 further includes a condition updating unit 504, which transmits the condition updating unit 504 to the imaging apparatus 3 using the communication unit 10 before the imaging apparatus 3 captures the image at the specific point A. Since it comprised, the imaging device 3 can utilize the imaging conditions which can image | photograph at a specific point A with high recognition rate.
As a result, an effect is obtained that enables the subject to be photographed with a high recognition rate when the photographing device 3 takes a picture at the specific point A or starts taking a picture at the specific point A. This enables the photographing apparatus 3 to photograph the subject with good image quality regardless of the change of the photographing environment at the specific point A.

Embodiment 2. FIG.
In the first embodiment, the photographing conditions provided to the photographing device 3 are derived based on the photographed image photographed by the photographing unit 4. In the second embodiment, the photographing system represents in a virtual space the photographing environment in which the photographing device 3 photographs in the real space and the photographing of the subject performed by the photographing device 3, and the photographing device 3 represents the subject in the virtual space. Based on the photographed image to be photographed, a photographing condition for providing to the photographing device 3 in the real space is derived.
The description of the same reference numerals as in Embodiment 1 will not be repeated.

  FIG. 11 is a system configuration diagram illustrating an application example of the imaging system 100 in the second embodiment. The imaging system 100 includes an electronic computer 511 and a communication device 6.

  The electronic computer 511 is an arithmetic device that can execute high-speed arithmetic processing, and is, for example, a supercomputer that constitutes cloud computing. The electronic computer 511 communicates with the communication device 6 in the real space Sr. In FIG. 11, the electronic computer 511 communicates with the communication devices 6a, 6b,.

  The electronic computer 511 constructs a virtual space Sv in which the real space Sr is modeled in three dimensions by executing simulation analysis.

  A reference subject 150 for deriving a recognition rate is arranged in the virtual space Sv. In FIG. 11, reference subjects 150a, 150b,..., 150e are schematically shown as reference subjects 150.

  In the virtual space Sv, a photographing unit 140 that photographs the reference subject 150 and derives a recognition rate is arranged. FIG. 11 schematically shows that as the photographing unit 140, the photographing units 140a, 140b,..., 140e each photograph the reference subjects 150a, 150b,. ing.

  The electronic computer 511 estimates a captured image of the reference subject 150 in the imaging unit 140 in the virtual space Sv. In FIG. 11, captured images estimated in each of the capturing units 140a, 140b,..., 140e are schematically illustrated as captured images P110, P120,.

In the virtual space Sv constructed by the electronic computer 511, for example, a movement path Rv, a light source Lv, and an object Ev are arranged as models of the surrounding environment of the movement path R along which the moving body 2 on which the photographing apparatus 3 is mounted moves. . In FIG. 11, as an example of the object Ev in the virtual space Sv, the object Ev1 indicates a tunnel, the object Ev2 indicates a forest, and the object Ev3 indicates a sea.
In the virtual space Sv, the objects Ev1 and Ev2 shield light and the object Ev3 reflects light, so that the captured image of the subject is affected by the type and arrangement of the object Ev that is a part of the shooting environment.

  In the virtual space Sv, the shooting unit 140 performs shooting processing of the reference subject 150.

The electronic computer 511 derives a recognition rate while performing image processing on the photographed image of the photographing unit 140, and also performs a learning process to derive a photographing condition that improves the recognition rate.
Then, the electronic computer 511 transmits the derived imaging condition to the moving body 2 or the imaging device 3 in the real space Sr.

  The electronic computer 511 uses the communication of the communication device 6 as a communication unit for transmitting the derived imaging condition to the moving body 2 or the imaging device 3.

  Further, the electronic computer 511 transmits / receives various information to / from the moving body 2 or the imaging device 3 via the communication device 6. The establishment of communication between the communication device 6 and the moving body 2 or the imaging device 3 is the same as in the first embodiment.

  The specific points A1, A2,..., A5 are shooting points in the real space Sr, and can be arbitrarily set for the electronic computer 511. The electronic computer 511 arranges the specific points A110, A120,..., A150 so that the specific points A1, A2,..., A5 of the set real space Sr correspond to the virtual space Sv. In FIG. 11, specific points A110, A120,..., A150 that the computer 511 places in the virtual space Sv are schematically shown.

  The specific position B is a position where the imaging conditions are transmitted to the moving body 2 or the imaging device 3 in the real space Sr, and is arranged in a range where communication between the communication device 6 and the moving body 2 or the imaging device 3 is possible. . In FIG. 11, specific positions B110, B120,..., B150 are schematically shown as specific positions B that the electronic computer 511 places in the real space Sr.

  In addition, the information regarding the arrangement of the specific point A and the specific position B in the real space Sv is associated with information about the arrangement of the communication device 6 and the communication range.

  FIG. 12 is a functional block diagram illustrating the configuration of the imaging system 100. The description of the same reference numerals as those in FIGS. 4 and 5 of the first embodiment will not be repeated.

  The electronic computer 511 includes an image adjustment unit 412, an image recognition unit 413, a storage unit 501, a condition derivation unit 502, a setting change unit 503, a condition update unit 504, a second communication unit 505, and an analysis unit 1201.

  The second communication unit 505 of the electronic computer 511 communicates with the external server 30 via the external network 20 in the real space Sr, and for example, specifies the moving route R on which the moving body 2 on which the imaging device 3 is mounted moves. Various types of information handled by the photographing system 100 such as information on the photographing environment at the point A are collected. Communication between the second communication unit 505, the communication device 6, and the external server 30 may be wired communication or wireless communication.

  The second communication unit 505 communicates with the moving body 2 or the imaging device 3 via the communication device 6.

  A means for the photographing system 100 to communicate with the mobile body 2 or the photographing apparatus 3 via the second communication unit 505 and the communication device 6 of the electronic computer 511 in the real space Sr is referred to as a communication means 10. That is, the imaging system 100 includes a communication unit 10 for communicating with the moving body 2 or the imaging device 3.

  The analysis unit 1201 constructs a virtual space Sv for expressing a shooting environment at a specific point A in the real space Sr that changes every moment. In this way, the simulation performed by the analysis unit 1201 to construct the virtual space Sv in which the shooting environment of the real space Sr is modeled in three dimensions is referred to as a construction simulation SimA.

In the construction simulation SimA, the analysis unit 1201 performs a simulation analysis using various information collected from the external server 30 to express a shooting environment for the shooting unit 140 to shoot in the virtual space Sv.
More specifically, the analysis unit 1201, for example, in a shooting environment expressed by reflecting information such as the arrangement of various objects Ev including the moving body 2 and information on the density of dust and water in the atmosphere, For example, a simulation analysis is performed on a situation in which light is reflected, refracted, and absorbed based on the light irradiation direction, light intensity, and illuminance from a light source such as the light source Lv and the headlight of the moving body 2. As an example of the simulation analysis at this time, a ray tracing method (ray tracing) can be used.

Further, the analysis unit 1201 arranges a photographing unit 140 that models the photographing device 3 in the real space Sr in the virtual space Sv.
The modeled photographing unit 140 includes a photographing unit 1202 that represents the photographing function of the photographing device 3. The photographing unit 1202 is modeled using information such as an internal structure, a lens shape, a filter coefficient, an exposure area, and sensitivity as information regarding the photographing function of the photographing device 3.

  Then, the analysis unit 1201 derives a photographed image D1 when the photographing unit 140 photographs the reference subject 150 by simulation analysis.

  In this way, the analysis unit 1201 arranges the imaging unit 140 modeled in the virtual space Sv, and a simulation performed for deriving information of the captured image in the imaging unit 140 is an imaging simulation SimB.

  The image adjustment unit 412 and the image recognition unit 413 perform image processing on the photographed image D1 derived by the photographing unit 1202 of the photographing unit 140 in the same manner as in the first embodiment, and a signal Sig1 including the recognition rate and the photographing vector information. Is output.

  Here, the electronic computer 511 holds the correct image information Dc as information on the correct image of the reference subject 150 in the captured image obtained by capturing the reference subject 150 in the virtual space Sv.

  When the condition deriving unit 502 receives the signal Sig1 from the image recognizing unit 413 in the same manner as in the first embodiment, the condition deriving unit 502 executes various learning processes while performing learning processing so as to improve the recognition rate in shooting of the shooting unit 140. Adjust the settings and derive the shooting conditions.

  In the first embodiment, the shooting conditions of the shooting device 3 are derived after the shooting settings of the shooting unit 4 are derived. However, in this embodiment, the shooting unit 140 in the virtual space Sv directly models the shooting device 3. Therefore, the shooting setting for improving the recognition rate of shooting by the shooting unit 140 can be set as the shooting condition of the shooting apparatus 3 as it is.

  FIG. 13 is an information relation diagram showing the relation between various types of information handled by the imaging system 100. Various information shown in FIG. 13 is stored in the storage unit 501 of the computer 511 in FIG.

  The virtual space information 130a includes arrangement information 131, shooting environment information 132, shooting information 133, shooting condition information 134, and learning information 135 as information handled by the construction simulation SimA and the shooting simulation SimB. However, the arrangement information 131 is also handled as information related to the real space Sr.

  The arrangement information 131, the specific spot information 131a, and the specific position information 131b correspond to the arrangement information 31, the specific spot information 31a, and the specific position information 31b in FIG. 6, respectively.

  The shooting environment information 132, date / time information 132a, weather information 132b, and shooting field information 132c correspond to the shooting environment information 32, date / time information 32a, weather information 32b, and shooting field information 32c in FIG. 6, respectively.

  The analysis unit 1201 executes the construction simulation SimA using the information on the arrangement information 131 and the shooting environment information 132.

  The shooting information 133, the recognition rate information 133a, and the shooting vector information 133b correspond to the shooting information 33, the recognition rate information 33a, and the shooting vector information 33b of FIG. 6, respectively.

  The shooting condition information 134, the illumination control information 134b, the image processing control information 134c, and the shooting control information 134a correspond to the shooting setting information 34, the illumination control information 34b, the image processing control information 34c, and the shooting control information 34a of FIG. 6, respectively. The imaging condition information 38 of FIG. 6 does not exist because in this embodiment, the imaging settings of the imaging unit 140 are treated as they are as the imaging conditions of the imaging device 3.

  The learning information 135, the learning data 135b, and the learning control information 135b correspond to the learning information 35, the learning data 35b, and the learning control information 35b in FIG. 6, respectively.

  The real space information 130b includes moving body information 36 and imaging device information 37 as information regarding the real space Sr.

  Further, each of the association information Rel11, Rel12, and Rel13 corresponds to the association information Rel1, Rel2, and Rel3 in FIG. 6, and each of the arrows Base11 and Based12 corresponds to each of the arrows Based1 and Based2 in FIG. 6, and the arrows Refer11 and Refer12. Respectively correspond to the arrows Refer1 and Refer2 in FIG.

  As shown in FIG. 13, the external server 30 is, for example, an astronomical information system 141, a weather forecast system 142, a map information system 143, and a road information system 144 as information sources for providing various information to the computer 511. For example, a system for managing information related to the shooting environment may be used.

  Astronomical information system 141 can use, for example, an astrophysical data system (ASTROPHYSICS DATA SYSTEM, ADS).

  The weather forecasting system 142 can use, for example, a World Weather Watch (WWW).

  The map information system 143 can use, for example, a geographic information system (GIS).

  The road information system 144 can use, for example, images from a navigation satellite system (NSS) and a wide-area monitoring system for traffic use in each region.

  In addition to the system described above, the external server 30 may collect and use information related to the imaging environment from the imaging device 3 that captures the periphery of the movement route R.

  The external server 30 can be connected to these information sources via a cloud network or a dedicated line. The electronic computer 511 can access these information sources via the external server 30 or directly.

  Next, operations of the electronic computer 511 and the photographing unit 140 in the photographing system 100 will be described with reference to FIGS.

  FIG. 14 is a flowchart illustrating processing in which the photographing system 100 transmits photographing conditions to the photographing device 3.

In the process S1401, the condition updating unit 504 of the electronic computer 511, the position information 36b of the moving body 2 in the real space Sr, the moving body identification information 36a for identifying the moving body 2 and the photographing apparatus 3, and the photographing apparatus identification information 37a. Are stored in the moving body information 36 and the photographing apparatus information 37 of the storage unit 501.
Here, the condition update unit 504 requests the imaging system 100 to transmit the imaging conditions in advance as the imaging apparatus 3 to be acquired information and the mobile body 2 on which the imaging apparatus 3 is mounted, or imaging in advance. The processing load of simulation analysis in the electronic computer 511 can be reduced by narrowing down the imaging apparatus 3 and the mobile body 2 equipped with the imaging apparatus 3 that have undergone registration procedures to the system 100.
Then, it progresses to process S1402a.

  The process S1402a is a loop process indicating repetition with the process S1402b.

  During the loop processing of the processing S1402a and the processing S1402b, the analysis unit 1201 of the electronic computer 511 is mounted with the imaging device 3 in order to derive imaging conditions to be transmitted to the imaging device 3 that acquired information in the processing S1401. The modeling of the shooting environment at the specific point A through which the moving body 2 passes and the derivation of the shooting conditions at the specific point A are repeated.

Specifically, the analysis unit 1201 executes a construction simulation SimA for constructing the virtual space Sv for the specific point A, and a photographed image obtained by photographing the reference subject 150 at the specific point A of the constructed virtual space Sv. The imaging simulation SimB for deriving is performed.
After the transition to the loop process of process S1402a, the process proceeds to process S1403.

  Note that, after the processing S1402a, the electronic computer 511 executes the processing individually for each of the imaging device 3 that has been requested and the mobile body 2 in which the imaging device 3 is mounted.

In process S1403, the condition deriving unit 502 of the electronic computer 511 receives the signal ExInfo from the external server 30 in order to construct a virtual space Sv for each of the imaging apparatuses 3 requested in process S1401, and the imaging apparatus 3 Collects the location information 131 of the specific point A passing through and the shooting environment information 132 of the specific point A, and refers to the association information Rel11 between the location information 131 and the shooting environment information 132, and the shooting environment information An update process for updating 132 is started. This update process is executed periodically or periodically or at a predetermined timing. It is assumed that appropriate initial information is set for information that is updated every other time.
Then, it progresses to process S1404.

  In process S <b> 1404, when the condition update unit 504 receives the position information PosInfo from the moving body 2 or the imaging device 3, the position updating unit 504 updates the position information 36 b for each moving body 2 and also associates the arrangement information 131 with the moving body information 36. Referring to Rel12, it is determined whether or not there is a moving body 2 that has reached the specific position B. As a result of the determination, if it exists, the process proceeds to step S1414. If it does not exist, the process advances to step S1405.

  In process S1405, the condition deriving unit 502 refers to the association information Rel13 between the arrangement information 131 and the learning information 135, and the learning information 135 associated with the specific point A where the photographing unit 140 is arranged in the virtual space Sv is stored in the storage unit. It is confirmed whether or not it is stored in 501. Then, it progresses to process S1406.

  In process S1406, if learning information 135 is stored in process S1405, the process proceeds to process S1408, and if not stored, the process proceeds to process S1407.

  In process S 1407, the setting change unit 503 of the electronic computer 511 initializes the image processing settings of the image adjustment unit 412 and the image recognition unit 413 of the electronic computer 511 based on the information of the shooting condition 134 stored in the storage unit 501. To do. Then, it progresses to process S1410.

In process S1408, the condition deriving unit 502 executes learning processing while updating the shooting condition information 134 and the learning information 135 based on the association information Rel13, and sets the shooting unit 1202, the image adjustment unit 412 and the image recognition unit 413. The imaging condition information 134 to be derived is derived.
In addition, the condition deriving unit 502 outputs a signal Sig3 including the shooting condition information 134 to the setting changing unit 503 in order to reflect the adjusted shooting condition in the corresponding shooting unit 140.
Then, it progresses to process S1409.

  The fact that the condition deriving unit 502 derives the shooting setting information 134 with reference to the learning information 135 in the process S1408 is indicated by an arrow Refer11 in FIG.

In processing S1409, the setting change unit 503 searches for the imaging unit 140 corresponding to the imaging condition information 134 included in the signal Sig3 output in processing S1408, and the imaging unit 1202 of the imaging unit 140 and the electronic computer 511. The signal Sig4, the signal Sig5, and the signal Sig6 for reflecting the shooting conditions are output to the image adjustment unit 412 and the image recognition unit 413, respectively. The imaging unit 1202 inputs the signal Sig4 to reflect the corresponding imaging condition, the image adjustment unit 412 inputs the signal Sig5 to reflect the corresponding imaging condition, and the image recognition unit 413 inputs the signal Sig6 to respond. Reflect the shooting conditions.
Then, it progresses to process S1410.

  In process S1410, the photographic unit 140 outputs a photographic image D1 obtained by photographing the reference subject 150 in the virtual space Sv. Then, it progresses to process S1411.

In process S1411, the image adjustment unit 412 outputs captured image information D2 obtained by adjusting the image quality of the captured image D1 so that the reference subject 150 becomes clearer in the captured image D1 output in process S1410. .
The image recognition unit 413 extracts image information similar to the correct image information Dc from the output captured image information D2, derives a recognition rate based on the comparison with the correct image information Dc, and outputs a signal Sig1.
The condition deriving unit 502 acquires the shooting information 133 included in the signal Sig1 output from the image recognition unit 413, and updates the information in the storage unit 501.
Then, it progresses to process S1412.

  In process S1412, the condition deriving unit 502 determines whether the recognition rate information 133a included in the acquired shooting information 133 has improved or decreased compared to the previous time. Then, it progresses to process S1413.

In process S1413, the condition deriving unit 502 updates the learning data 135b of the learning information 135 in association with each piece of information associated with the association information Rel13 based on the determination result in the process S1412.
More specifically, the condition deriving unit 502, for example, in the shooting environment indicated by the shooting environment information 132 at the specific spot A in the specific spot information 131a when the recognition rate is improved in the determination in step S1412, the shooting unit 140. It is reflected in the learning data 135b that it is appropriate to set the shooting condition of the shooting condition information 134 as the content of the shooting condition information 134.
For example, when the recognition rate decreases in the determination in step S1412, the condition deriving unit 502 sets the shooting condition of the shooting unit 140 in the shooting environment indicated by the shooting environment information 132 at the specific point A in the specific point information 131a. The fact that it is not suitable for the contents of the shooting condition information 34 is reflected in the learning data 135b.
Thereafter, the process proceeds to process S1402b, and the loop process is continued.

  In step S1413, the condition derivation unit 502 refers to the association information Rel12 and updates the learning information 135 with an arrow Refer12 in FIG.

  As described above, the imaging unit 140 in the virtual space Sv repeatedly captures the reference subject 150 according to the imaging conditions adjusted by the condition deriving unit 502 and outputs the captured image. The condition deriving unit 502 captures the imaging output by the imaging unit 140. By repeatedly deriving shooting conditions that improve the recognition rate of the shooting unit 140 based on the image through learning processing, the shooting unit 140 can capture images while maintaining a good recognition rate even if the shooting environment changes from moment to moment. Can be done.

In the process S1414, the condition update unit 504 learns the learning result associated with the specific point A (that is, the specific point information 131a) corresponding to the specific position B (that is, the specific position information 131b) reached by the moving body 2 in the process S1404. Is determined as to whether the content (i.e., learning data 135b) is sufficient. The criterion for determining whether or not the learning data 135b is sufficient is, for example, a case where the recognition rate (that is, the recognition rate information 133a) that is the result of learning converges within a predetermined fluctuation range, or is determined in advance. It is conceivable that the recognition rate has been reached or the learning process in the condition deriving unit 502 has passed a predetermined time.
By performing the determination here, it is possible to avoid a situation in which learning is not sufficiently advanced and a photographing condition that is unlikely to improve the recognition rate in the photographing device 3 is provided to the photographing device 3. I can do it.
As a result of the determination, if it is determined that it is sufficient, the process proceeds to step S1415. If it is determined that it is not sufficient, the process advances to step S1416.

  In the process S1415, the condition update unit 504, based on the association information Rel12 between the arrangement information 131 and the moving body information 36, the moving body 2 where the position information 36b has reached the specific position B (that is, the specific position information 131b) or the movement The derived imaging conditions are transmitted to the imaging device 3 of the body 2. Thereafter, the process ends.

  In the process S1416, the condition update unit 504, based on the association information Rel12 between the arrangement information 131 and the moving body information 36, the moving body 2 where the position information 36b has reached the specific position B (that is, the specific position information 131b) or the movement The derived imaging conditions are not transmitted to the imaging device 3 of the body 2. Thereafter, the process ends.

As described above, according to the second embodiment, the imaging system 100 includes, for example, the external server 30, the astronomical information system 141, the weather forecast system 142, the map information system 143, the road information system 144, and the travel route. Based on information collected from an information source such as the photographing device 3 that photographs the periphery of R, the photographing environment in the real space Sr photographed by the photographing device 3 is represented by a virtual space Sv, and the subject is recognized with respect to the photographing device 3 It is possible to derive and provide shooting conditions for high shooting.
As a result, it is possible to obtain an effect of shooting a subject with good image quality in a shooting environment in which the shooting apparatus 3 will take a shot.

  Further, according to the second embodiment, the photographing system 100 directly sets the photographing condition of the photographing device 3 without distinguishing the photographing setting of the photographing unit 4 and the photographing condition of the photographing device 3 as in the first embodiment. Can be derived. As a result, a shooting condition that improves the recognition rate of the subject is derived based on the shooting environment of the shooting device 3 without taking into account differences in shooting position, shooting direction, shooting distance, and the like between the shooting unit 140 and the shooting device 3. Effect is obtained.

Further, according to the second embodiment, the imaging system 100 can arbitrarily set the specific point A where the imaging device 3 captures an image and derive the imaging conditions. This enables the photographing system 100 to derive and provide the photographing conditions of the photographing device 3 in advance in accordance with the timing when any photographing device 3 starts photographing.
That is, since the photographing system 100 can provide photographing conditions to the photographing device 3 with high immediacy, the photographing device 3 can obtain a good photographed image even when the photographing environment is not gradually changed.

Embodiment 3 FIG.
In the third embodiment, the imaging system 1 of the first embodiment or the imaging system 100 of the second embodiment accumulates the derived imaging conditions of the imaging device 3 in association with various information stored in the storage unit 501. .

  The imaging device 3 has a high subject recognition rate even if the shooting conditions used on a sunny day at a specific point A are used on another day when the sunlight angle and time zone are the same and the weather is sunny. A photographed image can be obtained.

  Moreover, even if the imaging apparatus 3 uses the imaging conditions used on a bad day at another time when the imaging environment is equivalent, for example, the adjustment of the imaging control and the control of the illumination device are performed using the recognition rate of the subject. It becomes possible to carry out immediately so that becomes high.

The condition deriving unit 502 of the imaging system 1 or 100 includes an information group used when the imaging setting is derived at a certain time of a specific point A, the derived imaging setting, and the imaging condition derived based on the imaging setting. Are stored in the storage unit 501, and the correlation is derived and managed for each stored specific point A and each information group at each time point.
By doing in this way, the condition update unit 504 at a certain time point of another specific point A that has already been accumulated without deriving a new photographing condition at a certain time point T0 at a certain point A0, if not shown. From the information group, the information group having a high positive correlation with the information group at the specific point A0 and the time point T0 is selected, and the photographing condition associated with the selected information group is selected, and at the specific point A0 and the time point T0. The selected derived imaging conditions can be transmitted to the imaging device 3 that performs imaging. Here, “showing a high positive correlation” is as described above.

  At this time, the imaging device 3 feeds back the recognition rate in the captured image when the image is captured using the already derived imaging conditions to the imaging system 1 or 100, and the imaging system 1 or 100 stores each piece of information stored. By performing the review on each correlation of the group, it is possible to transmit to the photographing apparatus 3 photographing conditions having a similar photographing environment.

  As described above, according to the third embodiment, the photographing system 1000 selects the photographing condition for photographing the subject with a high recognition rate according to the change of the photographing environment at the specific point A, and selects the photographing device 3. Shooting conditions can be provided immediately.

  The second and third embodiments can be applied to, for example, a surveillance camera installed outdoors, in addition to the in-vehicle camera in which the photographing apparatus 3 is mounted on a vehicle.

When the shooting device 3 that is an outdoor monitoring camera starts shooting, the shooting conditions for shooting the subject with high recognition rate cannot be derived smoothly, and important shooting images are recorded. There is a risk of missing.
In such a case, the imaging system 100 according to the second embodiment derives imaging conditions that allow the imaging device 3 arranged at the specific point A to image the subject with a high recognition rate in the imaging environment when the imaging is started. Can be provided to the photographing apparatus 3.
As a result, the photographing apparatus 3 can start photographing under photographing conditions for photographing a subject with a high recognition rate even when the weather is bad.

In addition, the photographing device 3 that is an outdoor monitoring camera can smoothly derive photographing conditions for photographing a subject with a high recognition rate when the weather suddenly changes during photographing and heavy rain or heavy fog occurs. There is a possibility that a good quality photographed image cannot be obtained during this period.
In such a case, the imaging system 1000 according to the third embodiment is associated with the imaging apparatus 3 arranged at the specific point A, which has a high positive correlation among the information groups stored in the storage unit 501. The photographing conditions can be selected and provided to the photographing device 3.
As a result, the photographing apparatus 3 can continue photographing according to photographing conditions for photographing a subject with a high recognition rate even when the weather is bad.

DESCRIPTION OF SYMBOLS 1,100,1000 Shooting system 2 Mobile body 3 Shooting device 4,140 Shooting unit 411 Shooting unit 412 Image adjustment unit 413 Image recognition unit 414 First communication unit 5,150 Reference subject 6 Communication device 500,511 Computer 501 Storage Unit 502 condition deriving unit 503 setting changing unit 504 condition updating unit 505 second communication unit 1201 analyzing unit

Claims (12)

A photographing device;
Communication means for performing communication for transmitting and receiving various types of information to and from the imaging device;
A photographing unit that photographs a reference subject and outputs a photographed image at a specific point where the photographing apparatus photographs;
Based on the comparison between the captured image information image-processed based on the output captured image and the correct image information representing the reference subject prepared in advance, the accuracy when the reference subject is recognized in the captured image is shown. An image recognition unit for deriving a recognition rate;
A condition deriving unit for deriving a shooting setting for improving the recognition rate of the reference subject in the captured image with deriving the recognition rate, and deriving a shooting condition used by the shooting device based on the shooting setting;
A storage unit for storing the photographing conditions derived by the condition deriving unit;
A condition updating unit for transmitting the imaging condition via the communication unit in order to cause the imaging apparatus to use the imaging condition stored in the storage unit;
Shooting system equipped with. Shooting by the shooting unit, derivation of the shooting condition by the condition deriving unit, and transmission of the shooting condition by the condition updating unit are executed prior to shooting by the shooting device at the specific point. The
The imaging system according to claim 1. It further comprises an analysis unit for executing a construction simulation and a shooting simulation,
The analysis unit
By executing the construction simulation, constructing a virtual space expressing the photographing unit, a photographing environment photographed by the photographing unit at a specific point, and the reference subject,
By executing the shooting simulation, the shooting unit outputs a shot image in which the reference subject is shot in the virtual space;
The image recognition unit derives the recognition rate based on a captured image output by the analysis unit;
The imaging system according to claim 1 or 2. The condition derivation unit includes:
The storage unit associates the information group related to the shooting environment when the shooting unit is shooting at the specific point, the shooting setting of the shooting unit shooting at the specific point, and the shooting condition derived based on the shooting setting. Accumulate in
The condition update unit
Among the information groups stored in the storage unit, an information group that has a high positive correlation with an information group related to a shooting environment that the shooting apparatus is shooting is selected, and the shooting conditions associated with the selected information group Send,
The imaging system according to any one of claims 1 to 3. The imaging device is mounted on a moving body,
The condition update unit detects the position of the moving body via the communication means, and transmits the imaging condition when detecting that the moving body has reached a specific position prior to the specific point.
The imaging system according to any one of claims 1 to 4. The imaging device is a surveillance camera installed outdoors,
The condition update unit transmits the shooting conditions prior to the start of shooting of the shooting device.
The imaging system according to any one of claims 1 to 4. The imaging device is a surveillance camera installed outdoors,
The condition update unit extracts and transmits the shooting condition according to a change in a shooting environment in which the shooting apparatus is shooting,
The imaging system according to claim 3 or 4. The condition derivation unit includes:
An original photographed image obtained by photographing the reference subject is acquired from the photographing unit,
A difference in shooting position between the shooting unit and a specific shooting device;
A difference between a shooting direction and a shooting distance of the reference subject with the shooting unit as a base point, and a shooting direction and a shooting distance of a specific subject with the specific shooting device as a base point;
In addition, based on at least one of the information related to the difference in shooting environment between the shooting unit and the specific shooting device, an image conversion process for converting the original shot image into a converted shot image is performed.
Along with the conversion from the original photographed image to the converted photographed image in the image conversion process, the value of the recognition rate in the photographed image during or after the conversion is converted into the original photographed image before the image conversion process. Deriving the photographing condition of the specific photographing device by performing adjustment for maintaining the value of the recognition rate on the photographing setting;
The imaging system according to any one of claims 1 to 7. A first communication unit that performs communication for transmitting and receiving various types of information to and from the imaging device;
A photographing unit that photographs a reference subject and outputs a photographed image at a specific point where the photographing apparatus photographs;
Based on the comparison between the captured image information image-processed based on the output captured image and the correct image information representing the reference subject prepared in advance, the accuracy when the reference subject is recognized in the captured image is shown. An image recognition unit for deriving a recognition rate;
A condition deriving unit for deriving a shooting setting for improving the recognition rate of the reference subject in the captured image with deriving the recognition rate, and deriving a shooting condition used by the shooting device based on the shooting setting;
A storage unit for storing the photographing conditions derived by the condition deriving unit;
A condition updating unit for transmitting the imaging condition via the first communication unit in order to cause the imaging device to use the imaging condition stored in the storage unit;
Shooting unit equipped with. Shooting by the shooting unit, derivation of the shooting condition by the condition deriving unit, and transmission of the shooting condition by the condition updating unit are executed prior to shooting by the shooting device at the specific point. The
The imaging unit according to claim 9.   The imaging device which utilizes at least any one information among the information included in the imaging conditions of any one of Claims 1-10 for the imaging | photography of the own apparatus in the said specific point. A storage device for storing initial photographing conditions prepared in advance in the own device is provided.
The photographing condition according to any one of claims 1 to 10 is held in the storage device, and the initial photographing condition or the photographing condition is switched and used.
The imaging device according to claim 11.