JP7049858B2 - Moving object tracking device and its program - Google Patents

Description

The present invention relates to a moving object tracking device and a program thereof.

Conventionally, an automatic tracking imaging system that captures images while automatically tracking a moving object such as an aircraft has been proposed (Patent Document 1). For example, the conventional automatic tracking imaging system includes a pan head control unit that controls the pan head so that the photographing camera tracks a moving object based on the position information (sensor information) acquired by the sensor camera.

Japanese Unexamined Patent Publication No. 2016-58986

When a general robot camera pan head system is used in a conventional automatic tracking imaging system, the pan head outputs the current direction of the imaging camera in a polar coordinate system expressed in degrees. On the other hand, the pan head can be controlled only by the operation speed (speed instruction such as pan and tilt), and the operation instruction is also converted into a command command of a predetermined number of steps (for example, ± 63 steps) and is a discrete value. It may be possible to control only with. In addition, since the shooting camera has a certain mass and the drive torque of the pan head is finite, it does not reach the intended speed immediately even if it receives a command command, but it starts to move slowly and the speed increases to some extent. It becomes a constant speed over time. For this reason, in the photographing camera, the swing is delayed when the camera is stopped, and a tracking error occurs when the speed of the moving object changes. Therefore, in the conventional automatic tracking imaging system, it is necessary to use the feedback signal from the pan head, but there is a large delay and it is difficult to accurately grasp the direction of the imaging camera.

An object of the present invention is to solve the above-mentioned problems and to provide a moving body tracking device and a program thereof capable of smoothly tracking a moving body.

In view of the above-mentioned problems, the moving object tracking device according to the present invention is a moving object tracking device that causes a photographing camera to track a moving object by using position information indicating the position of the moving object with respect to the photographing camera, and has a state value input unit and a state value input unit. It is configured to include a prediction information request unit, a prediction information generation unit, a position data acquisition unit, a difference calculation unit, a control information generation unit, and a shooting camera control unit.

According to this moving object tracking device, the state value input unit inputs the state value of the shooting camera at a predetermined input interval while the state of the shooting camera is changing (feedback signal).
The prediction information request unit requests the prediction of the position information when the state value is input to the state value input unit or when the preset waiting time elapses without inputting the state value.
Based on the input position information, the prediction information generation unit obtains the predicted position from the past position and the current position of the moving object with respect to the photographing camera, and generates the prediction information including the past position, the current position, and the predicted position as position data. do.
The position data acquisition unit acquires the position data of the delay time set in advance with respect to the current position.

The difference calculation unit calculates the difference between the predicted position at the control time point when controlling the photographing camera and the predicted position at the time when the update interval set in advance is added to the control time point.
The control information generation unit generates control information indicating the operating speed of the photographing camera based on the difference calculated by the difference calculation unit, and the state value input to the state value input unit and the position acquired by the position data acquisition unit. Correct the control information based on the data.
The shooting camera control unit controls the shooting camera based on the control information.
In this way, the moving object tracking device considers the delay when the state value (feedback signal) is input from the photographing camera and the delay when obtaining the predicted position as the delay time, so that the direction of the photographing camera can be accurately determined. I can grasp it.

According to the present invention, the following excellent effects are obtained.
Since the moving object tracking device according to the present invention generates control information in consideration of the delay time and the update interval, the moving object can be tracked smoothly.

In the embodiment of the present invention, it is an overall block diagram of the aircraft tracking photography system. It is explanatory drawing explaining the installation method of the sensor camera of FIG. It is a block diagram which shows the structure of the moving body tracking apparatus of FIG. It is a flowchart which shows the operation of the moving body tracking apparatus of FIG. It is explanatory drawing explaining the tracking of a moving body in embodiment of this invention. It is explanatory drawing explaining the tracking of a moving body in the modification of this invention.

[Outline of aircraft tracking shooting system]
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.
An outline of the aircraft tracking imaging system 1 will be described with reference to FIG.
The aircraft tracking photography system 1 tracks an aircraft taking off and landing at an airport with a robot camera (shooting camera) 40 and photographs the aircraft. As shown in FIG. 1, the sensor cameras 10 (10 _R , 10 _C , 10 _L ) A moving object detecting device 20 (20 _R , 20 _C , 20 _L ), a moving object tracking device 30, and a robot camera 40 are provided.

The moving object is a moving object or a person, and is a tracking target of the moving object tracking device 30. In the present embodiment, an aircraft taking off and landing at the airport is illustrated as a moving body, but it goes without saying that the moving body is not particularly limited.

The sensor camera 10 captures an aircraft and generates a sensor image 90. For example, the sensor camera 10 is a fixed camera having a constant shooting direction and shooting angle of view. Then, the sensor camera 10 outputs the generated sensor image 90 to the moving object detection device 20.
Here, one or more sensor cameras 10 can be installed as needed. In this embodiment, as shown in FIG. 2, three sensor cameras 10 are installed on the roof of the building in the center of the airport so that the entire runway can be photographed. In FIG. 2, the sensor camera 10 _R faces the right side of the runway, the sensor camera 10 _C faces the center of the runway, and the sensor camera 10 _L faces the left side of the runway. Then, the sensor cameras 10 _R , 10 _C , and 10 _L generate sensor images 90 _R , 90 _C , and 90 _L in which the runway is photographed in each photographing direction.

The moving object detection device 20 detects an aircraft from the sensor image 90 input from the sensor camera 10, and outputs the detection result (moving object detection result) to the moving object tracking device 30. The motion detection device 20 can detect an aircraft by a general method. For example, as an aircraft detection method, a method using an ADS-B (Automatic Dependent Surveillance-Broadcast) signal of an aircraft and a method of predicting the course of an aircraft from a radar image can be mentioned.
In the present embodiment, three motion detection devices 20 are installed at different locations away from the airport so as to have a one-to-one correspondence with the sensor camera 10. Then, the moving object detecting device 20 _R analyzes the image of the sensor image 90 _R of the sensor camera 10 _R , the moving object detecting device 20 _C analyzes the image of the sensor image 90 _C of the sensor camera 10 _C , and the moving object detecting device 20 _L analyzes the image of the sensor image 90 C. Image analysis is performed on the sensor image 90 _L of 10 _L.

The moving object tracking device 30 uses the moving object detection result input from the moving object detecting device 20 to cause the robot camera 40 to track the aircraft. The moving object tracking device 30 outputs a command command to the robot camera 40 so that the robot camera 40 tracks the aircraft. In the present embodiment, one moving object tracking device 30 is installed at the same place as the moving object detecting device 20.

The robot camera 40 is driven according to a command command from the moving object tracking device 30, and generates a captured image 91 in which an aircraft is photographed. The robot camera 40 is a general robot camera capable of changing states such as pan and tilt. The robot camera 40 includes a camera body 41 that generates an image taken by an aircraft 91, and a pan head 43 that changes the posture of the mounted camera body 41. In the present embodiment, one robot camera 40 is installed on the roof of the building in the center of the airport, like the sensor camera 10.
In the present embodiment, the robot camera 40 and the moving object tracking device 30 are connected by a low-speed serial line. Therefore, it is difficult to input / output at a high bit rate between the robot camera 40 and the moving object tracking device 30.

Further, the robot camera 40 outputs a state value to the moving object tracking device 30 as a feedback signal at predetermined input intervals (for example, 0.5 seconds) during the state change. In the present embodiment, the robot camera 40 outputs the shooting direction (AZ angle, EL angle) of the robot camera 40 as a state value.
Here, when one or both of the AZ angle (azimuth angle) and the EL angle (elevation / depression angle) are changed, the robot camera 40 uses the changed AZ angle or EL angle value in the moving object tracking device 30. Output.
On the other hand, the robot camera 40 does not output the state value to the moving object tracking device 30 when the posture has not changed (that is, when both the AZ angle and the EL angle have not changed).

[Configuration of moving object tracking device]
The configuration of the moving object tracking device 30 will be described with reference to FIG.
As shown in FIG. 3, the moving object tracking device 30 controls the robot camera 40 by using the position specifying unit 31 that identifies the position of the aircraft and generates the prediction information thereof and the prediction information generated by the position specifying unit 31. The robot camera control unit 32 is provided.
The position specifying unit 31 includes a sensor information generation unit 310, a prediction information generation unit 311, a prediction information storage unit 312, a request input unit 313, and a prediction information output unit 314.

The sensor information generation unit 310 generates sensor information (position information) based on the motion detection result input from the motion detection device 20, and outputs the generated sensor information to the prediction information generation unit 311. In the present embodiment, the sensor information generation unit 310 inputs the moving object detection result at a predetermined generation interval (for example, 0.1 second interval), so that the sensor information is generated each time the moving object detection result is input. become.

The sensor information is information indicating the position (direction) of the aircraft with respect to the robot camera 40. For example, the sensor information is represented by an AZ angle and an EL angle when the aircraft is viewed from the robot camera 40, similar to the state value of the robot camera 40. This sensor information physically means a direction, but is logically expanded in a spherical polar coordinate system, so it may be called position information.

Here, the sensor information generation unit 310 can generate sensor information by any method. For example, the sensor information generation unit 310 sets in advance a conversion parameter for converting the image coordinates of the sensor image 90 into a polar coordinate system for each sensor camera 10. Then, the sensor information generation unit 310 refers to this conversion parameter and converts the image coordinates of the sensor image 90 indicated by the motion detection result into the sensor information.

The prediction information generation unit 311 generates prediction information when sensor information is input from the sensor information generation unit 310. Specifically, the prediction information generation unit 311 predicts future sensor information from the current sensor information input from the sensor information generation unit 310 and the past sensor information stored in the prediction information storage unit 312. , Generate prediction information.
After that, the prediction information generation unit 311 writes the generated prediction information and the sensor information in the prediction information storage unit 312.

This prediction information includes position data of the past position, the current position, and the predicted position of the aircraft with respect to the robot camera 40. That is, the prediction information includes past sensor information (actual measurement value), current sensor information (actual measurement value), and future sensor information (prediction value) as position data.

For example, the prediction information generation unit 311 generates prediction information by linear prediction of sensor information. Specifically, the prediction information generation unit 311 linearly predicts the future aircraft position from the past and present sensor information. This prediction information includes, for example, 1.0 second (10 in 0.1 second) past position data and 2.0 seconds (20 in 0.1 second) from the present to the future. Includes location data. The number of past and future position data included in this prediction information is arbitrary, and is set in consideration of factors such as various delays in the aircraft tracking imaging system 1 and the moving speed of the aircraft.
The prediction information generation unit 311 may generate prediction information not only by linear prediction but also by a method using a quadratic approximation or a spline function.

The prediction information storage unit 312 is a storage device such as a memory for storing prediction information and sensor information, an HDD (Hard Disk Drive), and an SSD (Solid State Drive). In the prediction information storage unit 312, the latest prediction information and sensor information are sequentially written by the prediction information generation unit 311, and the sensor information before that is accumulated as the past sensor information.

The request input unit 313 inputs a prediction information request from the prediction information request unit 321 described later. Then, when the prediction information request is input from the prediction information request unit 321, the request input unit 313 notifies the prediction information output unit 314 of the input of the prediction information request (prediction information request notification).

The prediction information output unit 314 outputs the prediction information stored in the prediction information storage unit 312 to the prediction information input unit 322 in response to the request from the prediction information request unit 321. That is, when the prediction information request notification is input from the request input unit 313, the prediction information output unit 314 immediately outputs the prediction information to the prediction information input unit 322.

The robot camera control unit 32 includes a state value input unit 320, a prediction information request unit 321, a prediction information input unit 322, a prediction information storage unit 323, a position data acquisition unit 324, a difference calculation unit 325, and control information. It includes a generation unit 326 and a command conversion unit 327. The command conversion unit 327 corresponds to the shooting camera control unit according to the claim.

The state value input unit 320 inputs the state value of the robot camera 40 at a predetermined input interval (for example, 0.5 seconds) while the state of the robot camera 40 is changing. Then, the state value input unit 320 notifies the prediction information request unit 321 of the input of the state value (state value input notification). Further, the state value input unit 320 outputs the state value of the robot camera 40 to the control information generation unit 326.

The prediction information request unit 321 requests and inputs the latest prediction information when a state value input notification is input from the state value input unit 320, or when a predetermined waiting time has elapsed without inputting the state value input notification. Request to unit 313 (prediction information request).
In the present embodiment, the prediction information request unit 321 outputs a prediction information request to the request input unit 313 every 0.5 seconds every time there is an input while the state of the robot camera 40 is changing. Further, the prediction information request unit 321 determines that a time-out occurs when the waiting time elapses without changing the state of the robot camera 40, and outputs the prediction information request to the request input unit 313.
The waiting time is set in advance with an arbitrary value. For example, the standby time can be set at intervals at which the robot camera 40 outputs the state value, and the state value that should be input every 0.5 seconds, such as 0.55 seconds, is not input from the pan head 43. It is advisable to set an appropriate time to confirm that.

The prediction information input unit 322 inputs prediction information from the prediction information output unit 314 in response to the prediction information request of the prediction information request unit 321. Then, the prediction information input unit 322 writes the input prediction information in the prediction information storage unit 323.
The prediction information storage unit 323 is a storage device such as a memory, an HDD, or an SSD for storing prediction information. The prediction information stored in the prediction information storage unit 323 is referred to by the position data acquisition unit 324 and the difference calculation unit 325.

The position data acquisition unit 324 acquires the position data of the delay time preset with respect to the current time point from the prediction information of the prediction information storage unit 323. Then, the position data acquisition unit 324 outputs the acquired position data to the control information generation unit 326.
This delay time is set based on the delay when the state value is input from the robot camera 40 (D ₁₁ in FIG. 5) and the delay when obtaining the predicted position (D ₁₂ in FIG. 5). For example, the delay time may be set to 0.4 seconds from D ₁₁ to D ₁₂ .

From the prediction information of the prediction information storage unit 323, the difference calculation unit 325 calculates the difference between the predicted position at the control time point in which the robot camera 40 is controlled and the predicted position at the time obtained by adding the update interval set in advance to the control time point. It is to be calculated. Then, the difference calculation unit 325 outputs the calculated difference to the control information generation unit 326.
As this update interval, a command update interval is set in consideration of the cycle when the robot camera 40 updates the command command from the command conversion unit 327 (D2 in _FIG . 5). For example, the update interval can be set by any value, and may be set to 0.1 seconds. It is preferable that this update interval is set so that the command command can be transmitted to the robot camera 40 without delay while the update interval is short enough to track a fast-moving aircraft.

The control information generation unit 326 generates control information representing the operating speed of the robot camera 40 based on the difference input from the difference calculation unit 325. This control information is the pan speed (change in pan angle per unit time) and tilt speed (change in tilt angle per unit time) of the robot camera 40. In the present embodiment, the control information generation unit 326 converts the difference (AZ angle, EL angle) calculated by the difference calculation unit 325 into the control information (pan speed, tilt speed) output to the robot camera 40.

Next, the control information generation unit 326 corrects the control information based on the state value Xu input from the state value input unit 320 and the position data Xd input from the position data acquisition unit 324. In the present embodiment, the control information generation unit 326 obtains the difference between the position data Xd and the state value Xu as in the following equation (1). Then, the control information generation unit 326 multiplies the obtained difference by the correction coefficient to obtain the correction value R. This correction coefficient is the product of the reciprocal of the correction time T and the hunting prevention coefficient K, which will be described later. Further, the control information generation unit 326 adds the obtained correction value R to the control information and corrects the control information. After that, the control information generation unit 326 outputs the corrected control information to the command conversion unit 327.
R = (Xd-Xu) / T × K ... Equation (1)

The correction time T represents the time (seconds) for correcting the error of automatic tracking. That is, the control information generation unit 326 corrects the automatic tracking error at a speed that corrects it over T seconds. The hunting prevention coefficient K is a coefficient for preventing hunting, and is usually 1 or less. Here, the correction speed is K times because of the hunting prevention coefficient. Therefore, it is actually corrected at a speed that corrects in T / K seconds. For example, when the correction time T is 1.0 second and the hunting prevention coefficient K is 0.6, it is equivalent to correcting the automatic tracking error over 1.67 seconds.
Further, the correction value R has an opportunity to be corrected every input interval (for example, 0.5 seconds) of the state value.

The command conversion unit 327 converts the control information input from the control information generation unit 326 into a command command to the robot camera 40. For example, the command command is represented by 127 numerical values from −63 to +63 for pan and tilt. Then, the command conversion unit 327 outputs a command command in which an identification command indicating that it is a pan or tilt instruction and these numerical values are set to the robot camera 40.
Here, the command conversion unit 327 converts the control information into a command command so that the robot camera 40 moves as intended. For example, the command conversion unit 327 can perform conversion by using a linear function, a LUT (look-up table), a power function (exponential function), or a function in which a power function is combined.

[Operation of moving object tracking device]
With reference to FIGS. 4 and 5, the tracking of the moving object by the moving object tracking device 30 will be specifically described (see FIG. 3 as appropriate).

Here, the robot camera 40 receives a command every 0.1 seconds and operates, the moving object tracking device 30 operates in units of 0.1 seconds, and the prediction information A is generated in units of 0.1 seconds. do.
As shown in FIG. 5, based on the delay D ₁₁ = 0.6 seconds when the state value is input from the robot camera 40 and the delay D ₁₂ = 0.2 seconds when the prediction information A is generated. From D ₁₁ to D ₁₂ , the delay time = 0.4 seconds is set.
Further, the command update interval D ₂ = 0.1 second is set.

First, the processing of the robot camera control unit 32 will be described.
As described above, the robot camera 40 outputs the state value Xu to the state value input unit 320 only every 0.5 seconds even during the state change. That is, the state value input unit 320 inputs the state value Xu from the robot camera 40 at the shortest interval of 0.5 seconds (step S1).

As shown in FIG. 5, the state value Xu input from the robot camera 40 includes, for example, a delay of 0.6 seconds (delay D ₁₁ ). Here, the time when the state value Xu is input to the state value input unit 320 is set as the state value input time point.

The state value input unit 320 outputs a state value input notification to the prediction information request unit 321. Then, the prediction information request unit 321 outputs the prediction information request to the request input unit 313 in response to the state value input notification input from the state value input unit 320 (step S2).

In parallel with the processing of steps S1 and S2, the position specifying unit 31 performs the following operations.
The sensor information generation unit 310 generates sensor information based on the motion detection result input from the motion detection device 20 (step S10).
The prediction information generation unit 311 predicts future sensor information from the current sensor information generated in step S10 and the past sensor information stored in the prediction information storage unit 312, and generates prediction information A (step). S11).

The prediction information generation unit 311 writes the prediction information A and the sensor information generated in step S11 into the prediction information storage unit 312 (step S12).
The request input unit 313 receives a prediction information request from the prediction information request unit 321 and outputs a prediction information request notification to the prediction information output unit 314 (step S13).
The prediction information output unit 314 outputs the prediction information A stored in the prediction information storage unit 312 to the prediction information input unit 322 in response to the prediction information request notification input from the request input unit 313 (step S14).

As described above, when the state value Xu is input from the robot camera 40, the robot camera control unit 32 requests the position specifying unit 31 for the latest prediction information A. The position specifying unit 31 constantly updates and holds the latest prediction information A, and if there is a request, the position specifying unit 31 immediately outputs the latest prediction information A to the robot camera control unit 32. As a result, the timing variation between the update of the prediction information A in the position specifying unit 31 and the command update in the robot camera control unit 32 causes changes in the image update cycle of the sensor camera 10, the position detection method, the prediction method, and the like. It falls within a certain range regardless.

As shown in FIG. 5, the prediction information A includes 10 past positions (position data a ₁ to a ₁₀ ), one current position (position data a ₁₁ ), and 19 predicted positions (position data a). ₁₂ to a ₃₀ ) and are included. That is, the prediction information A includes a total of 30 position data a ₁ to a ₃₀ generated in units of 0.1 seconds.

Hereinafter, the process returns to the processing of the robot camera control unit 32, and the description will be continued.
The prediction information input unit 322 receives prediction information from the prediction information output unit 314 (step S3), and writes the input prediction information A in the prediction information storage unit 323 (step S4).

Here, a delay of, for example, 0.2 seconds occurs from the time when the sensor camera 10 captures the sensor image 90 until the position specifying unit 31 generates the prediction information A (delay D ₁₂ in FIG. 5).
Therefore, when referring to the prediction information A, it is necessary to consider the delay D ₁₁ = 0.6 seconds when the state value Xu is input and the delay D ₁₂ = 0.2 seconds when the prediction information is generated. be. That is, based on the current position of the prediction information A (sensor information a ₁₁ ), the state value Xu input from the robot camera 40 is 0.4 seconds obtained by subtracting the delay D ₁₂ from the delay D ₁₁ and past information. It becomes.
Therefore, the position data acquisition unit 324 acquires the position data Xd (position data a ₇ ) at the time retroactive by the delay time (delay D ₁₁ -delay D ₁₂ ) from the current position from the prediction information A stored in step S4. (Step S5).

As shown in FIG. 5, after the robot camera control unit 32 outputs a command command, the robot camera 40 actually starts driving and reaches a specified speed, for example, 0.1 seconds or more, and in some cases, about 1 second. Delay occurs. Therefore, when controlling the robot camera 40, a deviation in the camera direction occurs due to this delay, but this can be continuously corrected by a method described later as a simple positional deviation.

Basically, in the prediction information A stored in step S4, the difference calculation unit 325 determines the predicted position (position data a ₁₃ ) at the control time point and the predicted position at the time when the update interval D2 is added to the control _time point (position data a 13). The difference from the position data _a14 ) is calculated (step S6).
In the example of FIG. 5, it is assumed that 0.2 seconds have elapsed from the current position of the forecast information A (the time when the forecast information A is input) to the control time point.

The control information generation unit 326 generates control information for the robot camera 40 based on the difference calculated in step S6. Then, as shown in FIG. 5, the control information generation unit 326 obtains a correction value by the above equation (1) in order to correct the deviation between the state value of the robot camera 40 and the sensor information, and the obtained correction value R Correct the control information with (step S7).
Here, since the control information generation unit 326 expresses both the state value Xu and the position data Xd by the AZ angle and the EL angle, the equation (1) is applied to each of the AZ angle and the EL angle, and the pan speed and the pan speed and the pan speed are applied. The correction value R for each tilt speed is obtained. Then, the control information generation unit 326 adds the correction values R of the pan speed and the tilt speed to the pan speed and the tilt speed of the control information, respectively.
The command conversion unit 327 converts the control information generated in step S7 into a command command to the robot camera 40 (step S8).

[Action / Effect]
As described above, since the moving object tracking device 30 considers the delay time and the update interval, it is possible to accurately grasp the direction of the robot camera 40 and smoothly track a moving object moving at high speed such as an aircraft.
Further, the moving object tracking device 30 can suppress the amount of data input / output even when the robot camera 40 and the moving object tracking device 30 are connected by a low-speed serial line, and the input / output of state values and command commands is not delayed. ..
Further, since the moving object tracking device 30 corrects the control information in a fixed time by the above-mentioned equation (1), the posture of the robot camera 40 does not change suddenly, and the tracking can be performed more smoothly.

(Modification 1)
Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the above-described embodiments, and includes design changes and the like within a range that does not deviate from the gist of the present invention.
As shown in FIG. 6, in the modification 1, the delay time is defined as follows in consideration of the adjustment time D ₁₃ . The adjustment time D ₁₃ is a time for adjusting the position of the aircraft in the captured image captured by the robot camera 40, and can be arbitrarily set (for example, the adjustment time D ₁₃ = 0.9 seconds).
Delay time = (delay D ₁₁ -delay D ₁₂ -adjustment time D ₁₃ )

Here, when the delay time becomes large, the robot camera 40 faces the rear side in the moving direction of the aircraft (that is, it is delayed). On the other hand, when the delay time becomes small, the robot camera 40 faces the front side in the moving direction of the aircraft. Further, if the adjustment time D ₁₃ is set so that the delay time becomes a negative value, the robot camera 40 faces the future position of the aircraft. In a normal shot image, a composition in which the front side in the moving direction of the aircraft is wide is preferable. Therefore, the adjustment time D ₁₃ can be set so that the composition can be adjusted to a desired composition.

In FIG. 6, the position data acquisition unit 324 acquires the predicted position Xd of the time further advanced by the adjustment time D13 _in the prediction information A. That is, the position data acquisition unit 324 acquires the position data a ₁₆ at the time obtained by adding the delay D ₁₂ and the adjustment time D ₁₃ and subtracting the delay D ₁₁ .
Since the processing after acquiring the position data a ₁₆ is the same as that of the above-described embodiment, the description thereof will be omitted.

(Other variants)
In the above-described embodiment, it has been described that the moving body is an aircraft, but in the present invention, the moving body is not particularly limited, and even a moving body moving at high speed can be smoothly tracked.
In the above-described embodiment, it has been described that a moving object is detected by a sensor camera, but the present invention is not particularly limited. For example, a moving object may be detected by a GPS (Global Positioning System), an infrared sensor, or the like.

In the above-described embodiment, it has been described that three sensor cameras and three motion detection devices are provided, but in the present invention, the number of these sensors is not particularly limited and may be one or more.
In the above-described embodiment, the delay time, the update interval, the input interval, and the waiting time are exemplified, but in the present invention, these values are not particularly limited. For example, these values can be set as arbitrary values according to the performance and specifications of the robot camera and the network environment.

In the above-described embodiment, the position specifying unit and the robot camera control unit have been described as being mounted on the same device, but in the present invention, the position specifying unit and the robot camera control unit may be mounted on separate devices.
In the above-described embodiment, the state values of the robot camera are the AZ angle and the EL angle, and the pan and tilt of the robot camera are controlled, but the present invention is not limited thereto. That is, the moving object tracking device may control the zoom and / or the focus of the robot camera as well as the pan and tilt.

In the above-described embodiment, the moving object tracking device has been described as independent hardware, but the present invention is not limited thereto. For example, the present invention can also be realized by a program that cooperates with the hardware resources such as the CPU, memory, and hard disk of the computer as the above-mentioned moving object tracking device. This program may be distributed via a communication line, or may be written and distributed on a recording medium such as a CD-ROM or a flash memory.

1 Aircraft tracking shooting system 10 Sensor camera (sensor)
20 Motion detection device 30 Motion tracking device 31 Position identification unit 310 Sensor information generation unit 311 Prediction information generation unit 312 Prediction information storage unit 313 Request input unit 314 Prediction information output unit 32 Robot camera control unit 320 Status value input unit 321 Prediction information request Unit 322 Prediction information input unit 323 Prediction information storage unit 324 Position data acquisition unit 325 Difference calculation unit 326 Control information generation unit 327 Command conversion unit (shooting camera control unit)
40 Robot camera 41 Camera body 43 Pan head

Claims (7)

A moving object tracking device that causes the photographing camera to track a moving object by using position information indicating the position of the moving object with respect to the photographing camera.
A state value input unit in which the state value of the shooting camera is input at a predetermined input interval while the state of the shooting camera is changing.
When the state value is input to the state value input unit, or when a preset waiting time elapses without inputting the state value, a prediction information request unit that requests prediction information and a prediction information request unit.
Based on the input position information, the predicted position is obtained from the past position and the current position of the moving object with respect to the photographing camera, and the predicted information including the past position, the current position, and the predicted position as position data is obtained. The prediction information generation unit to be generated and
A position data acquisition unit that acquires position data with a delay time set in advance based on the current position, and a position data acquisition unit.
A difference calculation unit that calculates the difference between the predicted position at the control time point when controlling the photographing camera and the predicted position at the time when the update interval set in advance is added to the control time point.
Based on the difference calculated by the difference calculation unit, control information representing the operating speed of the photographing camera is generated, and based on the state value input to the state value input unit and the position data acquired by the position data acquisition unit. And the control information generation unit that corrects the control information,
A shooting camera control unit that controls the shooting camera based on the control information,
A moving object tracking device characterized by being equipped with. The moving object tracking device according to claim 1, wherein the prediction information generation unit obtains the predicted position by linear prediction of the position information, quadratic approximation, or a spline function. The control information generation unit obtains the difference between the position data acquired by the position data acquisition unit and the state value, and corrects the obtained difference by multiplying the obtained difference by a correction factor preset in consideration of correction time and hunting prevention. The moving object tracking device according to claim 1 or 2, wherein a value is obtained and the control information is corrected by the obtained correction value. The delay time is set in advance based on the delay when the state value is input from the photographing camera and the delay from the time when the sensor camera captures the sensor image to the time when the prediction information is generated .
The update interval is described in any one of claims 1 to 3, wherein the update interval is set in advance based on a cycle when the shooting camera updates a command from the shooting camera control unit. Motion tracking device.
The position data acquisition unit is characterized in that an adjustment time for adjusting the position of a moving object in a captured image captured by the photographing camera is further set, and position data is acquired at a time advanced from the delay time by the adjustment time. The moving object tracking device according to any one of claims 1 to 4. The moving object tracking device according to any one of claims 1 to 5, wherein the prediction information generation unit inputs the position information about the moving object detected by the sensor camera. A program for making a computer function as a moving object tracking device according to any one of claims 1 to 6.

Images