JP6340113B2 - Measuring device and program thereof - Google Patents

Description

  The present invention relates to a measuring apparatus that measures the distance between a measurement point and a measurement target point using a differential dynamic coordinate system and a program thereof.

  Conventionally, a three-point method using a distance from three points is known as a measurement technique. This measurement technique is realized using measurement means such as wireless, infrared, laser, and camera. In recent years, along with the development of the IT industry, such measurement techniques are also used in the fields of, for example, human body structure generation technology for video game production and moving image processing technology (Non-Patent Documents 1 and 2).

  Conventionally, measurement techniques other than the three-point method are also known (Non-Patent Document 3). The technique described in Non-Patent Document 3 obtains the position and dimensions of a measurement object on the basis of this scale by photographing a scale having a known size in the vicinity of the measurement object.

LB Co., Ltd., motion capture & animation integration software “PV STUDIO 3D”, http://www.privatestudio.co.jp/pvs3d.html Library Co., Ltd., “Nanohana Series” dedicated to video processing, http://www.library-inc.co.jp/products/nanohana/03.htm IT, Inc., Real-time 3D measurement with a single-camera video camera "Single View 3D", http://www.ittc.co.jp/sproduct/photocalc/SingleView3D/SingleView3D.htm

  However, since the measurement technique described above uses a stationary coordinate system, problems relating to errors such as bias errors and random errors, and problems relating to the number of measurement points and indefiniteness arise. When one of the measurement point (reference point) or the measurement target point moves, it is affected by these problems. In particular, when both the measurement point and the measurement target point move, these problems become significant.

First, problems related to errors will be described.
The true value of the position of the arbitrary point P is (x ^, y ^, z ^), and the measurement error is (ε _x , ε _y , ε _z ). In this case, the measurement value at the position of the arbitrary point P is (x ^ + ε _x , y ^ + ε _y , z ^ + ε _z ), and an error is added to the true value of each measurement value in the stationary coordinate system. .

Next, problems related to the number of measurement points and indefiniteness will be described.
In the techniques described in Non-Patent Documents 1 and 2, three or more points are required due to the principle of the three-point method, and many measurement points (for example, the technique described in Non-Patent Document 2 are used to reduce errors). Then 8 points) are required. In the technique described in Non-Patent Document 3, since the number of measurement points is less than 3, a plurality of scales with known dimensions (for example, 5 points) are required. And in the technique of a nonpatent literature 1-3, when each measurement condition is not prepared, there exists a problem that a measured value becomes indefinite.

  Then, this invention makes it a subject to solve the above-mentioned problem and to provide the measuring device which enabled the exact measurement, and its program.

In view of the above-described problems, the measurement apparatus according to the first invention of the present application performs measurement when at least one of a measurement reference object that is a distance measurement reference and a measurement object that is a distance measurement target is moving. A measuring apparatus for measuring a distance between a measurement point where a reference object is located and a measurement object point where a measurement object is located, and is characterized by comprising a dynamic coordinate basis vector decomposition means and a distance calculation means.

According to such a configuration, the measuring apparatus uses the dynamic coordinate basis vector decomposing means to measure the scale information indicating the dimension of the scale for measuring the difference in position or distance of the measurement reference object, and the direction of the measurement reference object with respect to the measurement object. Alternatively, the difference between the directions is obtained by using a first dynamic coordinate basis vector in a three-dimensional dynamic coordinate system, a second dynamic coordinate basis vector orthogonal to the first dynamic coordinate basis vector, and a first dynamic coordinate basis vector. And a third dynamic coordinate basis vector orthogonal to the second dynamic coordinate basis vector .

In addition, the measurement device may calculate the relative distance between the measurement point and the measurement target point by using a distance calculation unit including a distance calculation formula including a difference in direction or direction from the dynamic coordinate base vector decomposed by the dynamic coordinate base vector decomposition unit. Is calculated.

Here, the true value of the position of the arbitrary point P is (x ^, y ^, z ^), and the true value of the position (P + ΔP) is (x ^ + Δx ^, y ^ + Δy ^, z ^ + Δz ^). And (ε _x , ε _y , ε _z ) is a measurement error of the position between the point P and the point ΔP. In this case, the difference in position between the (P + ΔP) point and the P point becomes (Δx ^, Δy ^, Δz ^), and the bias error is canceled out. Thus, the measuring device can cancel the error by calculating the relative angle.

Further, the measuring apparatus according to the present Application the second invention, the dynamic coordinate basis vector decomposition unit, scale information _{(Δx t, Δy t, Δz} t) and the direction (phi _t, theta _t) and the formula (5) As shown in FIG. 4, by decomposing the first dynamic coordinate basis vector E1 _t , the second dynamic coordinate basis vector E2 _t , and the third dynamic coordinate basis vector E3 _t , The angles Δφ _t and Δθ _t indicating the change in direction are calculated, and the distance calculation means calculates the second dynamic coordinate base vector E2 _t and the third dynamic coordinate base vector according to the distance calculation formula represented by Expression (6). Based on E3 _t and the angles Δφ _t and Δθ _t , the distance r _t between the measurement point and the measurement target point is calculated using a differential dynamic coordinate system .

Moreover, the measuring device according to the third invention of the present application extracts a measurement object included in a photographed image photographed by a photographing camera as a measurement reference object by template matching, and converts the extracted measurement object into an approximate figure. The measurement object recognition means for recognizing the measurement object included in the captured image, the angle calculated by the dynamic coordinate basis vector decomposition means , the distance calculated by the distance calculation means, and the measurement object recognition means Camera control means for controlling the photographing camera using the recognized area of the measurement object is further provided.
According to such a configuration, the measurement device can accurately control the photographing camera so as to face the measurement object.

Further, the measuring device according to the fourth invention of the present application provides the position and direction of the measurement point indicated by the position and direction information indicating the position of the measurement reference object and the direction of the measurement reference object with respect to the measurement object, and the distance calculated by the distance calculation means And a position calculation means for calculating the position of the measurement target point using a preset position calculation formula.

The measurement apparatus according to the first invention of the present application is a measurement program for causing hardware resources such as a CPU (Central Processing Unit), a memory, and an HDD (Hard Disk Drive) included in a computer to cooperate with each other as described above. It can also be realized. This measurement program may be distributed via a communication line, or may be distributed by writing in a recording medium such as a CD-ROM or a flash memory (the fifth invention of the present application).

According to the present invention, the following excellent effects can be obtained.
According to the first and fifth inventions of the present application, since the relative angle is calculated, the error is canceled and the measurement can be performed accurately.
According to the third aspect of the present invention, the photographing camera can be accurately controlled so as to face the measurement object.
According to the fourth invention of the present application, it is possible to present an accurate position of the measurement object.

It is a block diagram which shows the structure of the position measuring apparatus which concerns on 1st, 2nd embodiment of this invention. In 1st Embodiment of this invention, it is explanatory drawing explaining conversion of a target object, (a) shows the target object extracted from the camera image | video, (b) shows the figure of the shape approximated to a target object. In 1st Embodiment of this invention, it is explanatory drawing explaining camera position information and imaging | photography direction information. In 1st Embodiment of this invention, it is explanatory drawing explaining decomposition | disassembly to a dynamic coordinate basis vector. It is a flowchart which shows operation | movement of the position measuring apparatus of FIG. In 2nd Embodiment of this invention, it is explanatory drawing explaining a scale. In 2nd Embodiment of this invention, it is explanatory drawing explaining decomposition | disassembly to a dynamic coordinate basis vector. It is a block diagram which shows the structure of the camera control apparatus which concerns on 3rd Embodiment of this invention. In 3rd Embodiment of this invention, it is explanatory drawing explaining decomposition | disassembly to a dynamic coordinate basis vector. In 3rd Embodiment of this invention, it is explanatory drawing explaining calculation of an angle and distance, (a) shows two virtual imaging | photography cameras, (b) is an enlargement of the dashed-dotted line part of (a). It is a figure, (c) is an enlarged view of the intersection part of (b). In 3rd Embodiment of this invention, it is explanatory drawing explaining the moving average of distance. It is a block diagram which shows the structure of the measuring device which concerns on this invention. In the reference example, it is explanatory drawing explaining the operation principle of a measuring device.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate. In each embodiment, means having the same function are denoted by the same reference numerals and description thereof is omitted.

[Outline of position measuring device]
The outline of the position measuring apparatus 1 will be described with reference to FIG.
The position measuring device (measuring device) 1 measures the distance between the measurement point where the measurement reference object is located and the measurement target point where the object is located, and measures the position of the object based on this distance.

In the present embodiment, it is assumed that a measurement reference object serving as a distance measurement reference is one photographing camera 92 (FIG. 3). The measurement reference object is not limited to the photographing camera 92, and may be a wireless device, an infrared irradiation device, or a laser irradiation device.
Further, in the present embodiment, it is assumed that the object 90 (FIG. 3), which is a distance measurement target, is a subject photographed by the photographing camera 92.
The object 90 corresponds to the measurement object described in the claims.

  The position measuring apparatus 1 performs processing in either a normal mode in which both the photographing camera 92 and the object 90 move or a stationary mode in which the photographing camera 92 is stationary. Hereinafter, after describing the normal mode in the first embodiment, the stationary mode will be described in the second embodiment.

(First embodiment: normal mode)
[Configuration of position measuring device]
The configuration of the position measuring device 1 will be described.
As shown in FIG. 1, the position measurement apparatus 1 includes a mode selection unit 10, an object recognition unit (measurement target recognition unit) 11, an input unit 12, and a dynamic coordinate basis vector decomposition unit (angle calculation unit) 13. And an object position detection means 14 and a map display means 15.

  The mode selection means 10 is for manually selecting either the normal mode or the stationary mode. In the present embodiment, it is assumed that the mode selection unit 10 has selected the normal mode in advance. That is, the position measuring apparatus 1 performs the process described in the first embodiment in the normal mode, and performs the process described in the second embodiment in the stationary mode.

  Further, the mode selection means 10 may manually select either the normal sampling mode in which the position measurement apparatus 1 performs normal sampling or the high-speed sampling mode in which high-speed sampling is performed. For example, the mode selection means 10 selects the high speed sampling mode when the object moves at high speed. Then, the mode selection unit 10 outputs selection mode information indicating the normal sampling mode or the high-speed sampling mode to the input unit 12.

  The object recognizing means 11 receives a camera image (captured image) obtained by photographing the object with the photographing camera 92 and recognizes the object included in the camera image. Then, the object recognition unit 11 outputs the recognition result of the object to the input unit 12.

  As shown in FIG. 2, it is preferable that the object recognition means 11 extracts the object 90 included in the camera image and converts the extracted object 90 into a figure 91 that approximates the object 90. For example, the object recognition unit 11 registers in advance pre-data (template images) obtained by photographing the object 90 from various angles. Moreover, the target object recognition means 11 performs template matching using prior data, and extracts the target object 90 from the camera video (FIG. 2A). Then, the object recognizing means 11 obtains a graphic 91 that is closest to the contour of the extracted object 90 (FIG. 2B). That is, the target object recognition means 11 converts the target object 90 into a figure 91 having a simple shape that approximates the target object 90, such as a circle, a rectangle, or a triangle. Thereby, the target object recognition means 11 can improve the recognition rate of the target object 90.

  The input unit 12 receives camera information, camera position information, and shooting direction information from the outside. The input unit 12 receives selection mode information from the mode selection unit 10 and receives a recognition result of the target 90 from the target recognition unit 11.

The camera information is information indicating the focus of the photographing camera 92.
The camera position information is information indicating the position of the photographing camera 92.
The shooting direction information is information indicating the direction of the shooting camera 92 with respect to the object 90.
The camera position information and the shooting direction information correspond to the position / direction information described in the claims.

  Note that the input unit 12 may instruct the object recognition unit 11 to recognize the object 90 at a high sampling frequency when the selection mode information input from the mode selection unit 10 indicates the high-speed sampling mode.

<Camera position information and shooting direction information>
With reference to FIG. 3, the camera position information and the shooting direction information will be specifically described (see FIG. 1 as appropriate).

In FIG. 3, the photographing camera 92 is illustrated by a black circle, and the object 90 is illustrated by a black triangle (the same applies to FIGS. 4, 6, and 7). Further, in FIG. 3, the position of the object at time t is illustrated as 90 _t, and the position of the object at time _{t + 1} is illustrated as 90 _{t + 1} (the same applies to FIG. 4).

As shown in FIG. 3, the input unit 12 uses the camera position information indicating the camera position P _t = (x _t , y _t , z _t ) and the shooting direction D _t = (φ _t, The shooting direction information indicating θ _t ) is input. This shooting direction D _t points from the camera position P _t to the position 90 _{t of the} object. Further, at time t, φ _t indicates the shooting direction on the X axis and the Y axis, and θ _t indicates the shooting direction on the X axis, the Y axis, and the Z axis.

In addition, the input unit 12 uses the camera position information indicating the camera position P _{t + 1} = (x _{t + 1} , y _{t + 1} , z _{t + 1} ) as the measurement value at time t + 1, and the shooting direction D _{t + 1.} = Shooting direction information indicating (φ _{t + 1,} θ _{t + 1} ). This shooting direction D _{t + 1} points from the camera position P _{t + 1} to the position 90 _{t + 1} of the object. In addition, at time t + 1, φ _{t + 1} indicates the shooting direction on the X axis and the Y axis, and θ _{t + 1} indicates the shooting direction on the X axis, the Y axis, and the Z axis.

Thus, the input means 12 receives the camera position information and the shooting direction information for two points of time t and t + 1. Then, the input means 12 converts the input camera positions P _t and P _{t + 1} for the two points, the shooting directions D _t and D _{t + 1} and the recognition result of the object 90 into a dynamic coordinate basis vector decomposition. Output to means 13.

Returning to FIG. 1, the description of the configuration of the position measuring device 1 will be continued.
The moving coordinate basis vector decomposition unit 13 decomposes the difference between the camera position information and the shooting direction information for two points input from the input unit 12 into a moving coordinate basis vector, so that the shooting direction of the shooting direction can be obtained in the differential type moving coordinate system. An angle indicating a change is calculated.
This angle is a difference (Δφ _{t + 1} , Δθ _{t + 1} ) described later.

<Decomposition into dynamic coordinate basis vectors>
With reference to FIG. 4, the decomposition into the dynamic coordinate basis vectors will be specifically described (see FIG. 1 as appropriate).

As shown in FIG. 4, the dynamic coordinate basis vector decomposing means 13 determines the difference ΔP _{t + 1 from} time t to time _{t + 1} = (x _{t + 1} −x _t , y) for the camera positions P _t and P _{t + 1. t +} 1 _{-y t,} calculates the z _{t +} 1 -z _t). Further, the dynamic coordinate basis vector decomposing means 13 has the difference (Δφ _{t + 1} , Δθ _{t + 1} ) = (φ _{t + 1} -φ _t ) between the time t and the time t + 1 with respect to the shooting directions D _t and D _{t + 1.} , Θ _{t + 1} −θ _t ).

Then, the dynamic coordinate basis vector decomposing means 13 calculates the camera position difference ΔP _{t + 1} and the imaging direction differences Δφ _{t + 1} , Δθ _{t + 1} as dynamic coordinates as shown in the following equation (1). It decomposes into basis vectors E1 _{t + 1} , E2 _{t + 1} , and E3 _{t + 1} .

A differential coordinate (differential or differential) dynamic coordinate system is a coordinate system using dynamic coordinate mathematics, and is described in, for example, the following references 1 and 2.
Reference 1: Akira Kurita, “Differential Form and Its Applications”, June 22, 2002, Hyundai Mathematics Reference 2: H. Flanders, “Theory of Differential Form”, June 30, 1967, Iwanami Shoten

The dynamic coordinate basis vector is a basis vector for representing a differential type coordinate system. Here, since the differential-form dynamic coordinate system is three-dimensional, three dynamic coordinate basis vectors E1 _{t + 1} , E2 _{t + 1} , and E3 _{t + 1} are used.

Thereafter, the dynamic coordinate basis vector decomposition unit 13 outputs the shooting direction _D t, the angle [Delta] [phi _{t + 1,} the object position detecting means 14 a [Delta] [theta] _{t + 1} showing the change in D _{t + 1.} Furthermore, the dynamic coordinate basis vector decomposition means 13 determines the dynamic coordinate basis vectors E1 _{t + 1} , E2 _{t + 1} , E3 _{t + 1} , the camera position P _{t + 1,} and the shooting direction D _{t + 1} as the object. Output to the position detection means 14.

Returning to FIG. 1, the description of the configuration of the position measuring device 1 will be continued.
The object position detection means 14 uses the dynamic coordinates based on the angles Δφ _{t + 1} and Δθ _{t + 1} input from the dynamic coordinate basis vector decomposition means 13 to calculate the measurement points and measurement target points in the differential dynamic coordinate system. The distance is calculated. Further, the object position detection unit 14 calculates the position of the measurement target point from the calculated distance, the camera position P _{t + 1,} and the shooting direction D _{t + 1} by a preset position calculation formula.
The object position detection means 14 corresponds to the distance calculation means and the position calculation means described in the claims.

Specifically, the object position detection unit 14 uses dynamic coordinate basis vectors E1 _{t + 1} , E2 _{t + 1} , E3 _{t + 1} and the distance calculation formulas expressed by the equations (2) and (3). Based on the angles Δφ _{t + 1} and Δθ _{t + 1} , the distance r _{t + 1} from the measurement point to the measurement target point is calculated.

Next, the object position detection means 14 captures the camera position P _{t + 1} = (x _{t + 1} , y _{t + 1} , z _{t + 1} ) according to the position calculation formula represented by Expression (4). Based on the direction D _{t + 1} = (φ _{t + 1,} θ _{t + 1} ) and the distance r _{t + 1} , the position Pm _{t + 1} of the object is calculated.

Thereafter, the object position detection unit 14 outputs the calculated position Pmt _{+ 1} of the object to the map display unit 15.

The map display means 15 displays the position Pmt _{+ 1} of the object input from the object position detection means 14 on the map. For example, the map display unit 15, synthesizes the position Pm t _{+ 1} of the object is converted into a map coordinate system, the marker indicating the position Pm t _{+ 1} of the transformed object, the map displayed on the display To do.

[Operation of position measuring device]
The operation of the position measuring apparatus 1 will be described with reference to FIG. 5 (see FIG. 1 as appropriate).
The position measuring device 1 recognizes the object 90 included in the camera image by the object recognition means 11 (step S1).

In the position measuring apparatus 1, various information such as camera information, camera position information, and shooting direction information is input by the input unit 12 (step S2).
The position measuring apparatus 1 performs decomposition into dynamic coordinate basis vectors by the dynamic coordinate basis vector decomposing means 13 (step S3).

The position measuring device 1 calculates the distance between the measurement point and the measurement target point in the differential dynamic coordinate system by the object position detection means 14 (step S4).
The position measurement apparatus 1 calculates the position of the measurement target point by the object position detection unit 14 (step S5).
The position measuring device 1 displays the position Pmt _{+ 1} of the object on the map by the map display means 15 (step S6).

[Action / Effect]
As described above, since the position measuring device 1 calculates the relative angles Δφ _{t + 1} and Δθ _{t + 1} , it can cancel the bias error and accurately calculate the distance r _{t + 1} . As a result, the position measuring apparatus 1 can display the accurate position Pmt _{+ 1} of the object on the map.

  Furthermore, the position measuring device 1 supplements (tracks) the moving object based on the recognition result of the object in the object recognizing means 11 by sequentially repeating the processing such as time t, time t + 1,. The position can be displayed on the map in real time.

  Note that the position measurement device 1 (object position detection means 14) may obtain the distance from the photographic camera to the object using camera information, for example, when the photographic camera 92 automatically focuses.

(Second embodiment: stationary mode)
[Configuration of position measuring device]
Returning to FIG. 1, the configuration of the position measurement device 1 </ b> B will be described while referring to differences from the first embodiment.
In the present embodiment, unlike the first embodiment, the photographing camera is stationary and the object is moving. Accordingly, in the mode selection means 10, the still mode is selected in advance.

In the present embodiment, the position measuring device 1 is denoted by reference numeral 1B, the input means 12 is denoted by reference numeral 12B, the dynamic coordinate basis vector decomposing means 13 is denoted by reference numeral 13B, and the object position detecting means 14 is denoted by reference numeral. 14B and distinguished from the first embodiment.
Note that the object recognition unit 11 and the map display unit 15 are the same as those in the first embodiment, and thus description thereof is omitted.

  The input means 12B receives scale information from the outside in addition to camera information, camera position information, and shooting direction information.

As shown in FIG. 6, the scale 93 is an object (for example, an ordinary automobile) that is near the object 90 and measures a distance.
The scale information is information indicating the dimensions of the scale 93. In the example of FIG. 6, the scale information is information that the height of a normal car is 1.5 meters and the width is 2 meters. In addition, the scale information may be information such that the height of the utility pole is 14 meters, the height of an adult is 1.6 meters, or the height of a two-story residence is 6 meters.

Then, the input unit 12B includes the camera position P _t = (x _t , y _t , z _t ), the shooting direction D _t = (φ _t, θ _t ), the recognition result of the object 90, and the scale ΔP _t = (Δx _t , Δy _t , Δz _t ) is output to the dynamic coordinate basis vector decomposing means 13B.

  The moving coordinate basis vector decomposing means 13B calculates the angle indicating the change in the shooting direction in the differential moving coordinate system by decomposing the shooting direction information and the scale information input from the input means 12B into a moving coordinate basis vector. To do.

<Decomposition into dynamic coordinate basis vectors>
With reference to FIG. 7, the decomposition into the dynamic coordinate basis vectors will be specifically described (see FIG. 1 as appropriate).

As shown in FIG. 7, the dynamic coordinate basis vector decomposing means 13B, as shown in the following equation (5), the shooting direction D _t = (φ _t, θ _t ), the recognition result of the object, and the scale information ΔP _t = (Δx _t , Δy _t , Δz _t ) is decomposed into dynamic coordinate basis vectors E1 _t , E2 _t , E3 _t .

Thereafter, the dynamic coordinate basis vector decomposition unit 13B, the angle [Delta] [phi _t showing changes in imaging direction D _t, and outputs the [Delta] [theta] _t the object position detecting means 14B. Further, the dynamic coordinate basis vector decomposition unit 13B outputs the dynamic coordinate basis vectors E1 _t , E2 _t , E3 _t , the camera position P _t, and the shooting direction D _t to the object position detection unit 14B.

Returning to FIG. 1, the description of the configuration of the position measurement apparatus 1B will be continued.
The object position detection means 14B calculates the distance between the measurement point and the measurement target point in the differential dynamic coordinate system by dynamic coordinate mathematics from the angles Δφ _t and Δθ _t input from the dynamic coordinate basis vector decomposition means 13B. To do. Furthermore, the object position detecting means 14B includes a calculated distance, a camera position P _t, and a photographing direction D _t, the predetermined position calculation formula to calculate the position of the measurement object point.

Specifically, the object position detection unit 14B calculates the dynamic coordinate base vectors E2 _t and E3 _t , the angles Δφ _t and Δθ _t, and the shooting direction φ _{t according} to the distance calculation formula represented by the formula (6). based on the calculated distance r _t from the measuring point to the measurement target point.

Next, the object position detection unit 14B calculates the camera position P _t = (x _t , y _t , z _t ) and the shooting direction D _t = (φ _t, The position Pm _{t of the} object is calculated based on θ _t ) and the distance r _t .

Then, the object position detecting means 14B includes a position Pm _t of the calculated object, and outputs the map display unit 15.

[Action / Effect]
As described above, the position measuring apparatus 1B has the same effect as that of the first embodiment even when the photographing camera is stationary.
In the first and second embodiments, the object has been described as moving. However, the photographing camera (measurement reference object) may be moved, and the object (measurement object) may be moved or stationary.

(Third embodiment)
[Configuration of camera control unit]
The configuration of the camera control device 2 will be described with reference to FIG.

The camera control device (measuring device) 2 calculates the angle at which the direction between the measurement point and the measurement target point has changed, and the distance between the measurement point and the measurement target point, and based on the angle and the distance, the photographing camera 92 is calculated. Is to control. As shown in FIG. 8, the camera control device 2 includes an object recognition unit 20, angle calculation units 21 ₁ and 21 ₂ , a distance calculation unit 22, an area calculation unit 23, a zoom control unit 24, And tilt control means 25.

In the present embodiment, it is assumed that the measurement reference object is a stationary subject 90 (FIG. 3) and the object is a photographing camera 92.
Further, in this embodiment, and the imaging camera 92 of one is moved, the imaging camera 92 ₁ before the movement, and to handle the imaging camera 92 ₂ after movement virtually as photographing camera 92 two To do.
Hereinafter, reference numeral 92 ₁ on one eye imaging camera will be described by reference numeral 92 ₂ onto a second imaging camera.

The photographing camera 92 includes a camera body 92a and a pan head 92b.
The camera body 92a drives the zoom of the lens based on the zoom control signal.
The camera platform 92b mounts the camera body 92a and drives the camera body 92a in the pan direction and the tilt direction based on the pan / tilt control signal.

The object recognition unit 20 includes image recognition / tracking units 20 ₁ and 20 ₂ .
Image recognition and tracking means 20 ₁ is input photographing camera 92 ₁ of the camera image (captured image) is one that recognizes an object 90 included in the camera image. Then, the object recognition unit 20 ₁ outputs the recognition result of the object 90 (coordinates on the screen) to the angle calculating means 21 ₁ and the area calculation section 23.

Image recognition and tracking means 20 ₂ is omitted except that target camera image of the imaging camera 92 _2, the same as in image recognition and tracking means 20 _1, the description.

The angle calculation means 21 ₁ calculates an angle indicating a change in direction in the differential dynamic coordinate system by decomposing the difference between the camera position information and the shooting direction information into a dynamic coordinate basis vector for the shooting camera 92 _1. It is.
In this embodiment, since the position of the subject 90 is known (fixed) and the shooting direction is determined as a result of the attitude control with respect to the shooting camera 92, it is not necessary to input camera position information and shooting direction information from the outside. .

Specifically, the angle calculating means 21 _1, as shown in FIG. 9, using the following equation (8) to Formula (10), the degradation of the motion coordinate basis vector _e x1 _{to e z3,} angle Calculate φ _and θ.

In FIG. 9, the photographing camera 92 ₁ is located at the origin O = (0,0,0), photographing camera 92 ₂ (omitted in FIG. 9) is to be in any position. Further, the position P of the subject 90 is represented by (x, y, z) in Cartesian coordinates. Further, the distance between the points OP is r. That is, FIG. 9 shows the correspondence between Cartesian coordinates (x, y, z) and the spherical coordinate system (r, θ, φ). Here, the distance r and the distance difference dr are unknown.

Then, the angle calculating means 21 ₁ includes a moving coordinate base vector _e x1 _{to e z3} calculated, the angle _phi, and theta, and outputs the distance calculating unit 22.
Angle calculating means 21 _2, except that target imaging camera 92 _2, the same as in the angle calculating unit 21 _1, the description thereof is omitted.

The distance calculation means 22 calculates the distance between the measurement point and the measurement target point in the differential dynamic coordinate system by the dynamic coordinate mathematics from the angles input from the angle calculation means 21 ₁ and 21 ₂ .

  Here, in order to simplify the description for obtaining the distance r, consider a case where dx and dy in Equation (9) change and dz is constant. That is, dz = 0, dθ = 0, and sin θ = 1. In this case, Expression (12) can be derived from Expression (11) below.

  Then, the equation (12) can be substituted into the following equation (13) to derive the equation (14). Furthermore, equation (15) can be derived by solving equation (14) with distance r. Therefore, the distance calculation means 22 calculates the distance r using the distance calculation formula represented by Formula (15).

  Here, when dφ is minute and dφ approximates to sindφ, the following equation (16) is established. Therefore, the distance calculation means 22 calculates the distance difference dr using the following equations (17) and (18). Here, ds represents the distance between any two points.

As shown in FIG. 10 (a), the imaging camera 92 ₁ at time t shown by reference numeral 92 _t, the imaging camera 92 ₂ at time t + 1 shown by reference numeral 92 _{t + 1.} In this case, as shown in FIG. 10 (b), the distance calculating means 22 extracts the motion by the base at the intersection of the optical axes of the photographing cameras 92t _and 92t _{+ 1} , and the base in the distance direction at time t. The vector e _rt and the angular direction base vectors e _θt and e _φt are extracted (the same applies to the time t + 1). As shown in FIG. 10C, the distance calculating unit 22 performs motion connection by dynamic coordinate mathematics, and calculates a difference between the base vector e _rt and the base vector e _{rt + 1} as a distance difference dr. .

Thereafter, the distance calculation unit 22 outputs the distance r and the distance difference dr to the zoom control unit 24 and outputs the angles φ 1 _and θ to the pan / tilt control unit 25.
Note that the orthogonal coordinates input to the distance calculation means 22 are used to represent the positions of the photographing camera 92 and the subject 90 in the Cartesian coordinate system.

Returning to FIG. 8, the description of the configuration of the camera control device 2 will be continued.
The area calculation unit 23 calculates the area of the subject 90 input from the angle calculation units 21 ₁ and 21 ₂ . For example, the area calculating unit 23 extracts the edge of the subject 90 and obtains the area of the edge region. Then, the area calculation unit 23 outputs the calculated area of the subject 90 to the zoom control unit 24.

  The zoom control unit 24 zooms in the photographing camera 92 (camera body 92 a) based on the distance r and the distance difference dr input from the distance calculation unit 22 and the area of the subject 90 input from the area calculation unit 23. Is to control.

  For example, the zoom control unit 24 generates a zoom control signal such that the photographing camera 92 zooms in as the distance r increases, and the photographing camera 92 zooms out as the distance r decreases, and outputs the zoom control signal to the camera body 92a. To do.

  At this time, the zoom control unit 24 preferably performs zoom control using a moving average of the distance r. As shown in FIG. 11, the zoom control means 24 calculates a moving average 94 of the distance r from the calculated value of the distance r (illustrated by a black circle). Then, the zoom control unit 24 generates a zoom control signal based on the distance r represented by the moving average 94. Thereby, the camera control device 2 can reduce an error in zoom control.

  In addition, for example, the zoom control unit 24 generates a zoom control signal so that the photographing camera 92 zooms in as the area of the subject 90 becomes smaller, and the photographing camera 92 zooms out as the area of the subject 90 becomes larger. And output to the camera body 92a.

The pan / tilt control means 25 controls the pan and tilt of the photographing camera 92 (the pan head 92 b) based on the angles φ 1 _and θ input from the distance calculation means 22. That is, the pan / tilt control means 25 generates a pan / tilt control signal such that the photographing camera 92 faces the subject 90 based on the angles φ 1 _and θ, and outputs the pan / tilt control signal to the pan head 92b.
The zoom control means 24 and the pan / tilt control means 25 correspond to the camera control means described in the claims.

[Action / Effect]
As described above, the camera control unit 2, in order to calculate the relative angle phi, theta _t, bias error are canceled, it is possible to accurately calculate the distance r. As a result, the camera control device 2 can accurately control the photographing camera 92 so as to face the subject 90.

In the above-described embodiment, the position measurement devices 1 and 1B and the camera control device 2 have been described as examples of the measurement device according to the present invention. However, the present invention is not limited to these. That is, as shown in FIG. 12, the measuring device 3 according to the present invention can be realized by including angle calculation means 21 ₁ , 21 ₂ and a distance calculation means 22.

  In the third embodiment, it has been described that the position of the subject 90 is not measured. However, the position of the subject 90 may be measured by the same method as the object position detection unit 14 (14B) in FIG. In this case, the camera control device 2 can control the zoom ratio of the photographing camera 92 based on the measured position of the subject 90.

(Reference example 1: Operating principle of measuring device)
With reference to FIG. 13, the operation principle of the measuring apparatus according to the present invention will be described as Reference Example 1.

As shown in FIG. 13, imaging camera 92 ₁ of first unit is arranged at the origin (0,0,0), placed onto a second imaging camera 92 ₂ coordinates _{(x 0, y 0, z} 0) Suppose that it is done.

Further, it is assumed that the position of the object moves from P ₁ to P ₂ and the movement amount ΔP = (Δx, Δy, Δz) is unknown.
The distance from the photographing camera 92 ₁ to the position P ₁ is r ₁ , and the distance from the photographing camera 92 ₁ to the position P ₂ is r ₁ ′.

The distance from the imaging camera 92 ₂ to the position _{P 1} is _{r 2,} the distance from the photographic camera 92 ₂ to the position _{P 2} is _{r 2} '.
Further, Δr ₁ = r ₁ −r ₂ and Δr ₂ = r ₁ ′ −r ₂ ′. Further, the angle of the photographing camera 92 ₁ is φ ₁ , θ ₁ , and the angle of the photographing camera 92 ₂ is φ ₂ , θ ₂ .

When viewed from the photographic camera 92 _1, [Delta] P can be expressed by the following equation (19). Also, as viewed from the imaging camera 92 _2, [Delta] P can be expressed by the following equation (20).

From the equations (19) and (20), the distances r ₁ and r ₂ can be expressed by the following equation (21).

Further, decomposing the photographing camera 92 _1, as shown in equation (22) to (24), the angle phi _1, the theta ₁ to the moving coordinate basis vectors _{e 11, e 21, e 31} .

Further, decomposing the photographing camera 92 _2, as shown in equation (25) to (27), the angle phi _2, the theta ₂ to the moving coordinate basis vectors _{e 12, e 22, e 32} .

The distances r ₁ and r ₂ and the movement amount ΔP can be obtained by solving the above-described equations (21) to (27).

(Reference Example 2: Another solution of Reference Example 1)
As Reference Example 2, another solution of Reference Example 1 will be described.

Here, the dynamic coordinate basis vectors e _1, e _{2, and} e ₃ are expressed by the following equation (28). Further, the movement amount dP = P ₂ −P ₁ is expressed by the following formula (29).

When the equations (28) and (29) are solved by simultaneous equations, the following equation (30) is established, and the movement amount dP is equal when viewed from the _two photographing cameras 92 _{1 and} 92 ₂ . This equation (30) is solved as a differential equation of equation (31).

Φ _A, θ _A, e _{1A to} e _3A are the angles φ _and θ and the dynamic coordinate basis vectors e _{1 to} e ₃ with respect to the photographing camera 92 ₁ . _{Also, φ B, θ B, e} 1B ~e 3B , the angle _phi, and theta for photographing camera 92 _2, a motion coordinate basis vectors _e 1 to e _3.
Here, since Expression (32) is satisfied and r = r ₁ when t = 0, Expression (33) is satisfied.

  By substituting ‘C’ in Expression (33) into Expression (32), Expression (34) is obtained, and this Expression is converted into Expression (35) and Expression (36) corresponding to the two photographing cameras.

  Here, when the position P is defined as in Expression (37) and considered in the X-axis direction, Expression (38) and Expression (39) are established.

  Further, when considered in the Y-axis direction as in the X-axis direction, Expressions (40) and (41) are established.

  Similarly to the X-axis direction, Expression (42) and Expression (43) are established when considered in the Z-axis direction.

The distances r _Ax , r _Bx , r _Ay , r _By , r _Az , and r _Bz can be obtained from the above equations (39), (41), and (43). As described above, in both Reference Examples 1 and 2, the calculation results match.

1, 1B Position measurement device 10 Mode selection means 11, 20 Object recognition means (measurement object recognition means)
12, 12B input means 13, 13B dynamic coordinate basis vector decomposition means (angle calculation means)
14, 14B Object position detection means (distance calculation means and position calculation means)
15 Map display means 2 Camera control device 20 ₁ , 20 ₂ Image recognition / tracking means 21 ₁ , 21 ₂ Angle calculation means 22 Distance calculation means 23 Area calculation means 24 Zoom control means (camera control means)
25 Pan / tilt control means (camera control means)

Claims (5)

When at least one of a measurement reference object that is a distance measurement reference and a measurement target that is the distance measurement object is moving , the measurement point where the measurement reference object is located and the measurement object are located A measuring device for measuring a distance from a measurement target point,
Scale information indicating a difference in position of the measurement reference object or a scale for measuring the distance, and a direction of the measurement reference object with respect to the measurement object or a difference in the direction in a three-dimensional dynamic coordinate system A first moving coordinate basis vector; a second moving coordinate basis vector orthogonal to the first moving coordinate basis vector; a first moving coordinate basis vector; and a second moving coordinate basis vector orthogonal to the first moving coordinate basis vector. Dynamic coordinate basis vector decomposition means for decomposing into three dynamic coordinate basis vectors;
A distance calculation means for calculating a distance between the measurement point and the measurement target point by a distance calculation formula including the dynamic coordinate base vector decomposed by the dynamic coordinate base vector decomposition means and the direction or a difference between the directions ;
A measuring device comprising: The dynamic coordinate basis vector decomposing means is configured to generate the scale information (Δx _t , Δy _t , Δz _t ) And the direction (φ _t, θ _t ) As shown in Expression (5), the first moving coordinate basis vector E1 _t And the second dynamic coordinate basis vector E2 _t And the third dynamic coordinate basis vector E3 _t And the angle Δφ indicating the change in the direction in the differential dynamic coordinate system. _t , Δθ _t To calculate

The distance calculation means calculates the second dynamic coordinate basis vector E2 by the distance calculation formula represented by the formula (6). _t And the third dynamic coordinate basis vector E3 _t And the angle Δφ _t , Δθ _t Based on the above, the distance r between the measurement point and the measurement target point in the differential dynamic coordinate system _t Calculate

The measuring apparatus according to claim 1. A measurement object included in a photographed image photographed by a photographing camera as the measurement reference object is extracted by template matching, and the extracted measurement object is converted into a figure that approximates, and the measurement object contained in the photographed image Measurement object recognition means for recognizing
The photographing camera is controlled using the angle calculated by the moving coordinate basis vector decomposition means , the distance calculated by the distance calculation means, and the area of the measurement object recognized by the measurement object recognition means. Camera control means,
The measuring device according to claim 2 , further comprising: A position calculation set in advance from the position and direction of the measurement point indicated by the position and direction information indicating the position of the measurement reference object and the direction of the measurement reference object with respect to the measurement object, and the distance calculated by the distance calculation means A position calculating means for calculating the position of the measurement target point by an equation;
Measurement apparatus according to any one of claims 1 to 3, characterized by further comprising a.   A measurement program for causing a computer to function as the measurement apparatus according to claim 1.

Images