JP2008153879A - Automatic tracking photographic apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an automatic tracking photographic apparatus capable of hand-held photographing by a camera. <P>SOLUTION: The automatic tracking photographic apparatus 2 comprises: a moving image input means 5 for inputting moving images outputted from the camera 1; an object detection means 6 for detecting the moving amount of an object within a frame on the basis of the image feature amount of the object within the moving image inputted by the moving image input means 5; a control means 7 for outputting a control signal indicating a torque amount predetermined for the moving amount detected by the object detection means 6; and a torque generation means 3 for generating torque on the basis of the control signal outputted by the control means 7. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an automatic tracking photographing apparatus that realizes automatic tracking photographing of a subject that is a photographing target in hand-held photographing of a camera.

2. Description of the Related Art Conventionally, in the case of automatic tracking imaging, a tracking imaging apparatus has been disclosed in which a camera is mounted on a tripod with a motor, a pan head, or a crane, and the attitude angle and position of the camera are automatically or electrically controlled ( Patent Document 1). In the case of automatic tracking photographing, this tracking photographing device changes the photographing direction by a tilt motor and a pan motor according to a movement amount of a characteristic image determined in advance according to a photographing magnification.
In hand-held cameras, an optical system that can be physically moved along the optical path, such as a lens, prism, or mirror, is set up and controlled electrically in order to guarantee minute shaking and rotation of the camera. There are devices that make it possible to do this.
Furthermore, a gyro stabilizer that stabilizes a posture by rotating a momentum wheel at high speed is disclosed in a vibration isolator for photographing including hand-held photographing (Patent Document 2).
Japanese Patent Laid-Open No. 11-122526 (paragraphs 0033 to 0043, FIG. 1) Japanese Patent Laying-Open No. 2006-17686 (paragraphs 0005 to 0010, FIG. 2)

  However, in the conventional tracking imaging device, for example, in the technique disclosed in Patent Document 1, since the camera is mounted on a tripod, a pan head, or a crane with a motor, the total weight including the camera is heavy and a person holds it. There is a problem that it is difficult to perform hand-held shooting.

  In addition, the method of moving along the optical path is small in the amount of movement of the subject image that can be tracked and can guarantee a small posture change such as camera shake, but for applications that require a large posture change to track a specific subject. There is a problem that it is not suitable.

  Furthermore, in the technique disclosed in Patent Document 2, some gyro stabilizers are compatible with hand-held shooting, but the effect is to generate a torque that cancels the constant acceleration from the outside and cuts off the vibration. It is intended and not intended to track the subject.

  As described above, the conventional technique has a problem that it is not possible to perform automatic tracking shooting of a subject that is a shooting target in handheld shooting of a camera.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide an automatic tracking photographing apparatus that realizes automatic tracking photographing of a subject that is a photographing target in hand-held photographing of a camera.

  The present invention has been developed to achieve the above object. First, the automatic tracking imaging apparatus according to claim 1 generates torque so that a hand-held video camera is directed in a shooting direction. Thus, an automatic tracking imaging device that tracks a subject that is a subject to be photographed from a video camera, and an automatic tracking photographing device that tracks a subject that is a subject to be photographed from the video camera, A moving image input unit, a subject detection unit, a control unit, and a torque generation unit are provided.

  According to this configuration, the automatic tracking imaging apparatus inputs the moving image output from the moving image camera by the moving image input unit, and outputs the moving image to the subject detection unit. The video camera here refers to a video camera (photographing camera). A moving image is a real image taken by a camera, and includes a fixed subject having a certain three-dimensional geometric shape.

  Next, the automatic tracking imaging device uses the subject detection unit to move the subject in the frame that is the imaging region based on the image feature amount of the subject that is the imaging target in the moving image input from the moving image input unit. Is detected.

  Next, in the automatic tracking imaging apparatus, the control unit outputs a control signal indicating a predetermined torque amount with respect to the movement amount input by the subject detection unit. The control signal indicating the torque amount represents a signal corresponding to the rotation axis of the torque to be generated and its magnitude.

  Then, the automatic tracking imaging device generates torque based on the control signal output by the control means by the torque generation means. As a result, it is possible to turn a person's shoulder camera or a person's moving image camera toward a subject to be photographed by the generated torque, and automatic tracking photography can be performed.

  According to a second aspect of the present invention, in the automatic tracking photographing apparatus according to the first aspect, the torque generating means is detachably provided on the moving image camera.

  According to this configuration, the automatic tracking imaging device is formed separately from the moving image camera and mechanically mounted and fixed by the torque generating means. For example, it can be fixed via a tripod hole or a quick shoe that exists in the video camera.

  Furthermore, the automatic tracking imaging device according to claim 3 is the automatic tracking imaging device according to claim 1 or 2, wherein the torque generating means includes at least one reaction wheel.

  According to such a configuration, the automatic tracking photographing apparatus generates torque based on the control signal by the reaction wheel provided in the inside by one or more torque generating means, and changes the rotation angle of the attached moving image camera. Become.

  According to a fourth aspect of the present invention, there is provided the automatic tracking photographing apparatus according to the first or third aspect, further comprising unloading means for reducing the generated torque.

  According to this configuration, in the automatic tracking imaging device, when the unloading is necessary, the torque generating unit outputs a control signal to the unloading unit, and the unloading unit decreases the magnitude of the angular velocity. A control signal to be outputted to the torque generating means. Thereby, the torque generation means can reduce the magnitude of the angular velocity of the torque.

  The automatic tracking imaging apparatus according to claim 5 is the automatic tracking imaging apparatus according to any one of claims 1 to 4, wherein the control unit generates the detected movement amount. Storage means for storing the control value corresponding to the power torque in correspondence with each other, extracting the control value from the storage means using the detected movement amount, and based on the extracted control value A control signal was output.

  According to such a configuration, the automatic tracking imaging apparatus generates torque by outputting a control signal based on the control value (torque amount) corresponding to the detected movement amount from the storage unit by the control unit. Can do. The control value stored by the storage means is stored as a combination of the magnitude of torque to be generated and the rotating shaft.

  Furthermore, the automatic tracking imaging apparatus according to claim 6 includes a sensor camera different from a moving-image camera for hand-held shooting, and generates the torque so that the moving-image camera is directed in a shooting direction. Is an automatic tracking imaging device that tracks a subject that is an imaging target input from the above, and includes a moving image input unit, a subject detection unit, a control unit, and a torque generation unit.

According to this configuration, the automatic tracking imaging apparatus inputs a moving image output from a sensor camera different from the moving image camera by the moving image input unit, and outputs the moving image to the subject detection unit. The subject detection means, the control means, and the torque generation means are the same as in the first aspect.
As a result, the moving image camera can be directed to the subject to be imaged, and automatic tracking imaging can be performed.

  According to the first and sixth aspects of the invention, the moving image camera can be directed in the shooting direction (subject) by generating the torque determined using the movement amount of the subject that is the shooting target in the shooting region. In addition, the user's shooting posture can be changed, and automatic tracking shooting can be performed. Further, if the subject can be automatically tracked, the subject can be photographed even if the user does not completely look into the video camera. Furthermore, there is an advantage that the adverse effect of losing sight of the subject to be tracked is reduced.

  According to the second aspect of the present invention, it is possible to mount an automatic tracking imaging device even with a normal moving image camera, and it is possible to perform automatic tracking imaging with the camera.

  According to the third aspect of the present invention, it is possible to change the rotation angle of the attached moving image camera by generating torque by the reaction wheel. Further, torque can be generated without fixing the torque generating means to an object other than the moving image camera, such as a tripod, and the operability can be improved.

  According to the fourth aspect of the present invention, the reaction wheel angular velocity may increase and reach the limit as time elapses from the start of operation, but the magnitude of the angular velocity that may reach the limit can be suppressed.

  According to the fifth aspect of the present invention, it is possible to store the subject within the photographing region by storing the movement amount of the subject and the control value for converting into the control signal for generating the torque in association with each other. Since a linear or non-linear torque can be generated by a control signal based on the amount of movement, the storage means can be changed according to the application and user, and operability can be improved.

  According to the sixth aspect of the present invention, automatic tracking shooting is performed even for a camera that does not have a moving image output by inputting a moving image output from a sensor camera fixed to a torque generating means different from the moving image camera. Can do.

[Embodiment 1]
Embodiment 1 of the present invention will be described below with reference to the drawings.
[Configuration of automatic tracking camera]
First, an automatic tracking imaging apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of an automatic tracking imaging apparatus.
As shown in FIG. 1, the automatic tracking imaging device 2 calculates the movement amount of the subject in the imaging region based on the image feature amount of the subject that is the imaging target existing in the moving image output from the camera 1. It detects, determines a torque using the detected moving amount, and generates the determined torque. Here, the automatic tracking imaging device 2 includes a torque generation unit 3, a torque determination unit 4, and an unloading unit 9. The torque determination unit 4 further includes a moving image input unit 5, a subject detection unit 6, and the like. , And control means 7.
Note that the camera 1 according to the first embodiment means a hand-held video camera. The moving image camera refers to, for example, a video camera (photographing camera).
Further, hereinafter, an imaging target tracked by the automatic tracking imaging device 2 is referred to as a tracking target.

  In addition, the automatic tracking imaging device 2 is mechanically attached and fixed to the camera 1. For example, the camera 1 can be fixed via a tripod hole or a quick shoe.

The torque generating means 3 is an actuator that receives a control signal indicating the torque amount output from the control means 7 and generates a magnitude and a torque in the direction of the rotation axis based on the input control signal.
The torque generating means 3 may be constituted by, for example, one or more reaction wheels 8 inside thereof, and control the respective angular velocities (hereinafter may be referred to as rotational speeds) and angular accelerations. In FIG. 1, the three reaction wheels 8 are arranged so that the rotation axes are orthogonal to each other. Further, a control system having reaction wheels 8 may be provided for each of the three axes, and the arrangement may be made by a three-axis attitude control system that controls the three axes independently. Furthermore, the rotation of only the rotation axis including the pan and tilt axis directions of the camera may be restricted.

Since the torque generating means 3 is an actuator that has already been disclosed, a description of the relationship between the angular velocity and angular acceleration of the reaction wheel 8 and the generated torque is omitted.
Further, hereinafter, when the torque generating means 3 inputs the rotating shaft and the generated torque by electric means (arbitrary values such as voltage value, current value, pulse shape, digital signal, etc.), the torque around the rotating shaft Can be generated.

In the case where the reaction wheel 8 is used, the torque generating means 3 may reach the electrical or mechanical limit because the magnitude of the angular velocity of the reaction wheel 8 increases with the passage of time from the start of the operation. When the limit is reached or when the limit is expected to be reached, unloading may be performed.
Here, unloading is an operation for reducing the magnitude of the angular velocity of the reaction wheel 8, and for example, by fixing the automatic tracking imaging device 2 in the present embodiment so that the user does not move, The reaction wheel 8 can be electrically or mechanically braked to reduce its angular velocity.
In situations where unloading is necessary, the torque generating means 3 may warn the user, for example, by displaying light, sound, vibration or symbols. Further, unloading may be executed automatically or may be executed manually by a user's switch operation. Furthermore, unloading may be automatically performed at the time of non-photographing.

In FIG. 1, the unloading means 9 is connected to the torque generating means 3. The torque generating means 3 outputs the angular speed of the reaction wheel 8 to the unloading means 9 as a control signal when unloading is necessary, and the output unloading means 9 determines the magnitude of the angular speed. A control signal (torque amount) to be reduced is output to the torque generating means 3.
As will be described later, unloading may be performed by the control means 7.

  The moving image input unit 5 inputs a moving image output from the camera 1. The camera 1 and the automatic tracking imaging device 2 are connected by a cable, and a moving image output from the camera 1 is transmitted from the camera 1 to the moving image input means 5 through the cable.

The subject detection unit 6 detects the amount of movement of the subject in the frame that is the shooting region based on the image feature amount of the subject that is the tracking target in the moving image input from the moving image input unit 5.
Here, the processing of the subject detection means 6 will be described with reference to FIG. FIG. 4 is a block diagram showing the configuration of the subject detection means.
As shown in FIG. 4, the subject detection unit 6 includes a feature extraction unit 61, a storage switch 62, a feature memory 63, and a feature tracking unit 64.

The feature extraction unit 61 calculates the image feature amount of the subject that is the tracking target in the input moving image. The image feature amount can be in a predetermined area within the shooting area of the camera. FIG. 5 shows a central area among predetermined areas in the shooting area of the camera. Here, the imaging region 31 means the entire moving image input frame of the camera 1, and the central region 32 means a region near the preset center.
The image feature amount is represented by, for example, F (t, x, y), represents the image feature amount for a predetermined region from the coordinates (x, y) at the time point t, and the image at the time point t in the central region 32. The feature amount is indicated by F (t, 0, 0). The coordinates (x, y) can be (0, 0) at the upper left of the central region 32, and can be set to the x axis in the horizontal direction and the y axis in the vertical direction. Note that the upper left of the imaging region 31 may be (0, 0), the horizontal direction may be the x axis, and the vertical direction may be the y axis.

Then, the image feature amount F (t, x, y) can be a feature amount based on the color distribution of a predetermined region from the coordinates (x, y) at time t, and the color distribution normalization of the pixel value group A normalized histogram can be used.
Here, the pixel value group is, for example, a luminance value in the case of a monochrome moving image, a luminance value of each component of the three primary colors red, green, and blue in the case of a color moving image, or in the case of a multi-wavelength moving image. Can be the luminance value of each wavelength band. In the case of a color moving image, the pixel value may be expressed by an arbitrary color system such as a color system based on luminance and color difference and a color system based on hue, saturation and luminance.
In the following description, the image feature amount is represented by a pixel value group histogram in which the components of the three primary colors are represented by luminance values. That is, the image feature amount F (t, x, y) is obtained by converting the luminance values of red, green, and blue in a predetermined region from the coordinates (x, y) of the input moving image at time t, respectively R (t, x, y). , Y), G (t, x, y) and B (t, x, y).
Note that the image feature amount may be a feature amount based on the luminance gradient distribution.

As the normalized histogram, for example, as shown in FIG. 5, a histogram of frequency values (number of pixels) for (R, G, B) values is calculated for the central region 32 in the shooting region 31 of the camera. Normalization is performed by dividing the value by the total number of pixels in the central region 32.
Alternatively, the central area 32 may be cut out as a template, and the luminance pattern of the entire area may be used as the image feature amount.

  The image feature amount F (t, x, y) is stored in a storage device (not shown). The storage device is a storage device such as a RAM (Random Access Memory) or a flash memory.

The storage switch 62 is a switch for selecting a subject to be tracked, and stores the image feature amount of the selected subject in the feature memory 63 from the image feature amounts extracted by the feature extraction unit 61. It is. For example, the subject detection means 6 is triggered by the pressing of the storage switch 62, and the subject existing in the central region 32 of FIG. 63.
For example, a push switch, a rebound toggle switch, or a touch panel can be used as the storage switch 62, but the storage switch 62 is not limited to this.

As described above, the feature memory 63 stores the image feature amount F (t, 0, 0) of the central region 32 (see FIG. 5) using the time point when the storage switch 62 is pressed (time t) as a trigger.
The feature memory 63 is preferably a RAM (Random Access Memory) or a flash memory, but may be a magnetic disk device, an optical disk device, or a magneto-optical disk.

The feature tracking unit 64 obtains a similar region similar to the image feature amount F (t, 0, 0) in the central region 32 (see FIG. 5) stored in the feature memory 63 for the image feature amount from the current (from time t). A search is made from the input video of the future, and the movement amount (Vx, Vy) (translation amount) is output.
For example, when a similar region exists on the right side of the central region 32, Vx is set to a positive value. When a similar region exists on the left side of the central region 32, Vx is set to a negative value. If it exists, Vx = 0 can be set. Further, for example, when a similar region exists below the central region 32, Vy is a positive value, and when a similar region exists above the central region 32, Vy is a negative value, If a similar region exists, Vy = 0 can be set.
The movement amount (Vx, Vy) from the image feature amount of the central region 32 to a similar similar region is hereinafter referred to as a subject position (subject movement amount).

Here, the search for similar regions will be described. As a premise for searching for similar regions, the feature extraction means 61 executes the following processing.
In the feature extraction means 61, first, a plurality of search areas moved by a predetermined vector from the central area 32 are set in advance in the imaging area 31 (see FIG. 5). For example, in FIG. 6, the search area 33 is an area moved from the central area 32 by (Δx, Δy).
Next, the image feature amount F (t ′, Δx, Δy) in the preset search area 33 is calculated for the currently input moving image (time t ′).

  Then, the feature tracking unit 64 calculates the image feature amount F (t ′, Δx, Δy) in the calculated current search area 33 (see FIG. 6) (time t ′) and the central area 32 in the feature memory 63. The distance from the image feature amount F (t, 0, 0) (movement amount of the subject) is calculated as shown in the following equation (1), and the movement amount (Δx, Δy) with the smallest distance is calculated as follows: Calculation is performed as shown in equation (2), and the result is output as the subject position (Vx, Vy).

  Next, when a normalized histogram is used as the image feature amount, for example, the Bhattacharya distance can be used as the distance d (Δx, Δy). In this case, the distance d between the normalized histogram h (R, G, B) of the central region 32 (see FIG. 6) and the normalized histogram h ′ (R, G, B) of the search region 33 is (3 ) Is calculated as shown in the equation.

In addition, when using the luminance pattern of a template obtained by cutting out a predetermined area as an image feature amount, an absolute value error is calculated for each pixel at the same position between templates, and the sum of absolute value errors over the entire template size is calculated. Can be a distance d.
Furthermore, the distance d may be calculated with d = −ln (r) for the cross-correlation r between templates.
When the template luminance pattern is used as the image feature amount, the above-described method for calculating the distance d is called block matching.

As described above, the case where the feature tracking unit 64 in the present embodiment outputs the subject position (Vx, Vy) (translation amount) has been described, but in addition to this, the image rotation amount may also be output. good. In this case, for example, block matching that searches both the subject position (translation amount) and the rotation amount can be used. Further, the feature tracking unit 64 stores the image feature amount of the selected subject to be tracked in the feature memory 63, and the feature extraction unit 61 calculates the image feature amount at each of the plurality of positions. (Translation amount) and rotation amount can be calculated.
Returning to FIG. 1, the description will be continued.

The control means 7 is a means for outputting a control signal based on the torque amount using the subject position input by the subject detection means 6.
Here, the processing of the control means 7 will be described with reference to FIG. FIG. 7 is a block diagram showing the configuration of the control means.
As shown in FIG. 7, the control unit 7 includes a control value calculation unit 71.

First, the control value calculation means 71 controls the subject position (Vx, Vy) input by the subject detection means 6 or the control value to be generated based on the subject position (Vx, Vy) and the rotation amount φ, that is, the magnitude of torque. And calculate the rotation axis.
For example, the control value calculation means 71 calculates the rotation vector ω in the direction in which the subject is tracked based on the subject position (Vx, Vy) and the rotation amount φ (φ = 0 in the case where the rotation amount is not calculated). The calculation is performed as shown in the following equation (4).

  Next, the control value calculation unit 71 generates and outputs a control signal so as to generate a torque having a magnitude N with the calculated rotation vector ω as the rotation axis, for example. The magnitude N of the torque is calculated as shown in the following equation (5).

  The torque magnitude N calculated based on the above equation (5) can be controlled more quickly as the constant k is larger. However, since it may cause destabilization such as overshoot, A positive integer.

  Further, instead of the above equation (5), as shown in the following equation (6), the magnitude N of the torque may be calculated by On-Off control.

  Nmax in the equation (6) is a magnitude of torque generated when On-Off control is turned on, and is a positive integer. For example, Nmax = 5 [N · m] can be set.

The control means 7 can be configured as shown in FIG. 8 instead of FIG.
Here, the processing of the control means 7 will be described with reference to FIG. FIG. 8 is a block diagram showing the configuration of the control means different from FIG.
As shown in FIG. 8, the control unit 7 includes a quantization unit 72, a table reference unit 73, and a lookup table 74. Further, a lookup table 74 is stored in the removable memory 75.

The quantization unit 72 detects the subject position (Vx, Vy) detected by the feature tracking unit 64 (see FIG. 4) of the subject detection unit 6 or the combination of the subject position (Vx, Vy) and the rotation amount φ. Convert to discrete numbers (index numbers).
For example, the quantizing unit 72 converts Vx of the subject position (Vx, Vy) to an integer value mx of Mx gradation of 0 to Mx-1, and Vy to an integer value my of My gradation of 0 to My-1. To do. When the range of the X coordinate Vx and the Y coordinate Vy of the subject position is Vxmin to Vxmax and Vymin to Vymax, respectively, the integer values mx and my are obtained by linear quantization as shown in the following equation (7). Can be converted.

Thereby, a set (mx, my) of integer values mx and my can be used as an index number.
Note that if the subject position (Vx, Vy) is already a discrete value, the quantization unit 72 may not be provided.

The table reference unit 73 extracts a control value (torque amount) from the lookup table 74 using the index number converted from the subject position, and outputs the control value as a control signal.
The lookup table 74 is composed of an index number and a control value, and a control value is stored for each index number. The control value is stored as a combination of the magnitude of the torque to be generated by the torque generating means 3 and the rotating shaft.

  The lookup table 74 is stored in a ROM (Read Only Memory), a flash memory, a hard disk storage element, or a flexible disk storage medium. These storage elements or storage media may be read-only or rewritable. In particular, it may be stored in a removable memory 75 including a flash memory and a micro drive so that the look-up table 74 can be switched according to the application and user.

  Note that the control means 7 uses the control value calculation means 71 according to the absolute value of the rotation vector ω in the direction in which the subject is tracked and its time history in calculating the torque magnitude N of the control value calculation means 71 in FIG. You may build. For example, PID (proportional operation / integration operation / differentiation operation) control can be applied based on the time history of the absolute value of ω. When PID control is applied, when the look-up table 74 of FIG. 8 is used, proportional gain, integration time, and differentiation time, which are parameters of PID control, may be stored for each index number.

Further, the control means 7 may control the unloading operation when the angular velocity of the reaction wheel 8 is expected to reach the limit.
Here, FIG. 2 is a block diagram showing a configuration when the unloading operation is executed by the control means in the automatic tracking imaging apparatus according to Embodiment 1 of the present invention.
In FIG. 2, the difference from the configuration of FIG. 1 is that the automatic tracking imaging apparatus 2 does not perform the unloading operation by the unloading means (see FIG. 1) connected to the torque generating means 3, but by the control means 7. Is to do.

The torque generating means 3 outputs the angular speed of the reaction wheel 8 as a control signal to the control means 7 when unloading is necessary, and the control means 7 determines the angular speed based on the output control signal. A control signal (torque amount) for reducing the magnitude can be output to the torque generating means 3. Further, the torque generating means 3 outputs the current angular speed of the reaction wheel 8 as a control signal to the control means 7 at regular time intervals, and the control means 7 determines in advance the angular speed corresponding to the output control signal. When the limit value of the angular velocity is exceeded, a control signal for reducing the magnitude of the angular velocity can be output to the torque generating means 3.
The other components in FIG. 2 are the same as those in FIG.

  As described above, by configuring the automatic tracking imaging device 2, the camera 1 mounted on the automatic tracking imaging device 2 can be directed in the imaging direction (subject), and the user's imaging posture can be changed. Automatic tracking shooting is possible.

Here, FIG. 3 is a diagram for explaining the operation content of the camera mounted on the automatic tracking imaging apparatus. When the user 30 of the handheld camera 1 selects a subject to be tracked, the automatic tracking imaging device 2 generates torque so that the camera 1 is directed in the direction in which the tracking target has moved, and the camera 1 is positioned in the shooting direction. It can be changed (rotated). The user 30 of the handheld camera 1 can change the shooting posture based on the rotating camera 1.
For example, in FIG. 3, when the subject running to the left in the left frame is to be tracked, the user 30 of the handheld camera 1 selects the subject in the central region 32 in the photographing region 31. Next, along with the movement of the subject to be tracked, the camera 1 automatically changes its orientation (rotates) in the direction of the subject based on the torque generated by the automatic tracking imaging device 2, and further changes the imaging orientation of the user 30. Let In FIG. 3, the automatic tracking imaging device 2 detects a subject traveling to the right in the right frame from the plurality of search areas 33, and the camera 1 can automatically perform tracking imaging.

[Operation of automatic tracking system]
Next, the operation of the automatic tracking imaging apparatus according to the present invention will be described with reference to FIG. FIG. 9 is a flowchart showing the overall operation of the automatic tracking imaging apparatus.

  First, the moving image input means 5 inputs a moving image taken by the camera 1 (step S1).

  Next, the subject detection means 6 calculates an image feature amount (step S2). That is, the feature extraction unit 61 inputs the moving image output from the moving image input unit 5 and calculates the image feature amount of the input moving image. Here, the image feature amount F (t, x, y) is an image feature amount for a predetermined region from each coordinate (x, y) at the time (time t) when the moving image is input. The feature extraction unit 61 stores the calculated image feature amount in a storage device (not shown).

  Next, the subject detection means 6 determines whether or not the user has pressed the storage switch 62 (step S3).

Here, when the storage switch 62 is pressed (Yes in step S3), the subject detection means 6 stores the image feature quantity to be tracked in the feature memory 62 (step S4). That is, the image feature amount F (t, 0, 0) of the current central region 32 (see FIG. 6) that is the tracking target is extracted from a storage device (not shown) and stored in the feature memory 63.
On the other hand, if the storage switch 62 is not pressed (No in step S3), the subject detection means 6 proceeds to step S5.

  Then, the subject detection means 6 searches for the amount of movement up to a similar region similar to the image feature amount to be tracked as the subject position (step S7). That is, the image feature amount F (t ′, x, y) of each predetermined region at the current time (time t ′) is collated with the image feature amount stored in the feature memory 63, and the subject position (Vx , Vy) (movement amount) is calculated.

  Next, the control means 7 converts the subject position into an index number (step S6). That is, the index number is converted based on the searched subject position (Vx, Vy) (movement amount).

  Then, the control means 7 extracts the torque amount from the lookup table 74 using the converted index number (step S7).

  Next, the torque generating means 3 generates torque based on the extracted torque amount (step S8).

Next, the automatic tracking imaging device 2 determines whether or not to end the operation (step S9).
Here, when the automatic tracking imaging device 2 determines that the automatic tracking imaging device 2 has ended by an external end signal or the like (Yes in step S9), the automatic tracking imaging device 2 ends the operation.
On the other hand, if it is not determined that the process has been completed (No in step S9), the process returns to step S1 to continue the operation.

  Even if the storage switch 62 has never been pressed, steps S5 to S8 are executed. Therefore, a tracking flag is set in a storage device (not shown) and tracking is started at the start of the automatic tracking imaging device 2. The flag is stored off (for example, “0”), and when the storage switch 62 is pressed (Yes in Step S3), the tracking flag is stored on (for example, “1”), and Steps S4 and S5 are performed. It is determined whether or not the tracking flag is on. If not, the process may proceed to step S9.

  With the above operation, the automatic tracking imaging device 2 generates torque for tracking the selected subject to be tracked, so that the mounted camera 1 can perform automatic tracking imaging.

[Embodiment 2]
Embodiment 2 of the present invention will be described below with reference to the drawings.
[Configuration of automatic tracking camera]
FIG. 10 is a block diagram showing the configuration of the automatic tracking imaging apparatus according to the second embodiment.
10, the difference from the first embodiment shown in FIG. 1 is that, in the second embodiment, the automatic tracking imaging device 2 includes a sensor camera 10 different from the camera 1 to be mounted.

The sensor camera 10 is fixed to the torque generating unit 3, inputs a moving image in a direction in which the lens of the camera 1 is directed, and transmits the output moving image to the moving image input unit 5.
The moving image input unit 5 inputs a moving image output from the sensor camera 10. The sensor camera 10 and the automatic tracking imaging device 2 are connected by a cable, and a moving image input from the sensor camera 10 is transmitted from the sensor camera 10 to the moving image input means 5 via the cable.
The other components in FIG. 10 are the same as those in FIG.

By configuring the automatic tracking imaging device 2 according to the second embodiment, the camera 1 attached to the automatic tracking imaging device 2 can be directed in the imaging direction (subject), and further, the user's imaging posture can be set. It can be changed, and automatic tracking shooting is possible.
Thereby, even the camera 1 which is not provided with a moving image output can perform automatic tracking photography. For example, automatic tracking shooting is possible even for a video camera, a digital still camera, or a silver salt type shooting device that cannot obtain a moving image output during shooting.

It is the block diagram which showed the structure of the automatic tracking imaging device which concerns on Embodiment 1 of this invention. It is the block diagram which showed the structure in the case of performing unloading operation in a control means in the automatic tracking imaging device which concerns on Embodiment 1 of this invention. It is a figure explaining the operation | movement content of the camera with which the automatic tracking imaging device which concerns on Embodiment 1 of this invention is mounted | worn. It is the block diagram which showed the structure of the to-be-photographed object detection means of the automatic tracking imaging device which concerns on Embodiment 1 of this invention. It is the example of the figure which showed the imaging | photography area | region of the camera, and the center area | region which sets the tracking target. It is the example of the figure which showed the imaging | photography area | region of the camera, the center area | region which sets a tracking object, and the search area | region which searches a tracking object. It is the block diagram which showed the structure of the control means of the automatic tracking imaging device concerning Embodiment 1 of this invention. FIG. 8 is a block diagram illustrating a configuration of a control unit of the automatic tracking imaging apparatus according to Embodiment 1 of the present invention instead of FIG. 7. It is a flowchart which shows the whole operation | movement of the automatic tracking imaging device which concerns on Embodiment 1 of this invention. It is the block diagram which showed the structure of the automatic tracking imaging device which concerns on Embodiment 2 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Camera 2 Automatic tracking imaging device 3 Torque generation means 4 Torque determination means 5 Moving image input means 6 Subject detection means 7 Control means 8 Reaction wheel 9 Unloading means 10 Sensor camera

Claims (6)

An automatic tracking imaging device that tracks a subject that is a subject to be photographed from a video camera by generating a torque so that the video camera for handheld shooting is directed in the shooting direction,
A moving image input means for inputting a moving image output from the camera;
Subject detection means for detecting the amount of movement of the subject in the frame based on the image feature amount of the subject in the moving image input by the moving image input means;
Control means for outputting a control signal indicating a predetermined torque amount with respect to the movement amount detected by the subject detection means;
Torque generating means for generating torque based on the control signal output by the control means;
An automatic tracking imaging device characterized by comprising: The torque generating means includes
The automatic tracking imaging apparatus according to claim 1, wherein the automatic tracking imaging apparatus is detachably provided on the camera. The torque generating means includes
The automatic tracking imaging device according to claim 1, further comprising at least one reaction wheel.   The automatic tracking imaging apparatus according to claim 1 or 3, further comprising unloading means for reducing the generated torque. The control means includes
Storage means for storing the detected movement amount in correspondence with the control value corresponding to the torque to be generated, and extracting the control value using the detected movement amount from the storage means; The automatic tracking imaging apparatus according to claim 1, wherein a control signal based on the extracted control value is output. Automatic tracking that tracks the subject that is the subject of photography input from the sensor camera by generating a torque so that the video camera is directed in the direction of photography, with a sensor camera that is different from the camera for hand-held photography A photographing device,
Moving image input means for inputting a moving image output from the sensor camera;
Subject detection means for detecting the amount of movement of the subject in the frame based on the image feature amount of the subject in the moving image input by the moving image input means;
Control means for outputting a control signal indicating a predetermined torque amount with respect to the movement amount detected by the subject detection means;
Torque generating means for generating torque based on the control signal output by the control means;
An automatic tracking imaging device characterized by comprising:

Images