JP6701402B2 - Automatic tracking device, automatic tracking method, imaging device, and program - Google Patents

Description

  The present invention relates to control of an image pickup apparatus having a camera turning function for changing an image pickup direction, and particularly to pan/tilt control of an image pickup direction for tracking a detected moving body.

  The surveillance camera, which has a moving object detection function and is capable of pan/tilt control in the imaging direction, automatically performs pan/tilt control in the imaging direction so as to keep capturing the moving object detected in the imaging area near the center of the display screen. Some have a tracking function. For example, an imaging apparatus has been proposed which has a camera turning mechanism, detects a moving object from a captured image, and automatically controls the camera rotating mechanism so that the detected moving object is automatically tracked and continuously captured in the captured image. Reference 1). In addition, by setting a dead area around the position where the moving object should be displayed on the display screen, and when the center position of the moving object moves within the dead area, the motion of the moving object is not followed so that unnecessary image fluctuations occur. A tracking method that does not cause the noise has been proposed (see Patent Document 2).

JP 2004-40514 A JP 7-23270 A

  By the way, a surveillance camera with a camera turning function having a high-power telephoto function is generally required to have high stop position accuracy after the turning operation. Therefore, in order to improve the stop accuracy near the stop position of the camera when the pan/tilt operation is performed, the stop position is finely adjusted immediately before the pan/tilt operation is stopped. However, the time required for the fine adjustment of the stop position depends on the mechanical (mechanical) deviation of the pan/tilt mechanism in the vicinity of the stop position. Will fall.

  Further, since the stop position adjustment of the camera needs to be performed for each of the pan operation and the tilt operation, when one of the position adjustments is performed, the other needs to be stopped. Therefore, the stop alignment is not performed simultaneously in the pan direction (horizontal direction) and the tilt direction (vertical direction), but must be performed in order. In this case, the time required to adjust the stop position when the pan operation and the tilt operation are simultaneously driven is longer than that when only one of the pan operation and the tilt operation is performed, and the followability deteriorates. Will end up.

  It is an object of the present invention to provide a technique for improving followability when automatically tracking a detected moving body.

An automatic tracking device according to the present invention is configured to detect a moving body from an image picked up by an image pickup means capable of changing an image pickup direction by a panning operation and a tilting operation, and to track the moving body detected by the detecting means. and a control means for performing at least one of the pan control and tilt control of the imaging means, wherein, when the center of the moving object is in the first area of the image, the panning operation or the the tilt operation is stopped, the shape of the first region is characterized Rukoto is changed according to a tilt position of the imaging means.

  According to the automatic tracking device of the present invention, it is possible to improve the tracking performance when automatically tracking the detected moving body.

It is a block diagram showing a schematic structure of a surveillance camera concerning an embodiment of the present invention. FIG. 3 is a diagram schematically illustrating a dead region and a zero correction region when the tilt position is near the horizontal direction in the surveillance camera in FIG. 1. It is a figure which shows typically the pan/tilt operation when the surveillance camera of FIG. 1 is installed in the ceiling. FIG. 6 is a diagram schematically illustrating a dead region and a zero correction region set when the tilt position is near 90 degrees in the first control method of automatically tracking a moving object by the surveillance camera in FIG. 1. 6 is a flowchart of a first control method for automatically tracking a moving body by the surveillance camera of FIG. 1. 6 is a flowchart of a pan/tilt movement amount correction method in step S506 of FIG. 5. FIG. 9 is a diagram schematically illustrating a dead region and a zero correction region that are set when the tilt position is near 90 degrees in the second control method for automatically tracking a moving object by the surveillance camera in FIG. 1. 6 is a flowchart of a second control method for automatically tracking a moving body by the surveillance camera of FIG. 1. It is a figure which shows the relationship between a dead region and the imaging field angle which considered the size of a moving body. 9 is a flowchart of a third control method for automatically tracking a moving body by the surveillance camera of FIG. 1.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a block diagram showing a schematic configuration of a surveillance camera 100 according to the embodiment of the present invention.

  The surveillance camera 100 is an example of an imaging device having a camera turning function for changing the imaging direction. The surveillance camera 100 includes an imaging unit 101, an image processing unit 102, a system control unit 103, a lens drive unit 104, a lens control unit 105, a pan operation drive unit 106, a pan operation control unit 107, a tilt operation drive unit 108, and a tilt operation control. The unit 109 is provided. The surveillance camera 100 also includes an image analysis unit 110 and a communication interface unit 111.

  The image pickup unit 101 includes a lens, an image pickup device, and the like, picks up an image of a predetermined area, converts an image picked up by photoelectric conversion into an analog electric signal, and further A/D converts the analog electric signal to obtain digital data ( A LAW image) is generated. The digital data generated by the imaging unit is sent to the image processing unit 102. The image processing unit 102 performs predetermined image processing and compression encoding processing on the digital data acquired from the image pickup unit 101 to generate image data. In this way, the image data generated by the image processing unit 102 is distributed to the client device (described later) via the communication interface unit 111 and the network.

  The system control unit 103 includes a CPU that is an arithmetic unit, a memory that stores a program, a memory that has a program development area and a data storage area, and the like. The system control unit 103 executes a program to control the operation of each unit included in the surveillance camera 100, and thus controls the overall operation of the surveillance camera 100. For example, the system control unit 103 instructs the lens control unit 105 to perform zoom or focus control, instructs the pan operation control unit 107 and the tilt operation control unit 109 to automatically track a moving object, and instructs the image processing unit. The image quality adjustment instruction or the like is given to 102.

  The lens driving unit 104 includes a mechanical driving system for driving the focus lens and the zoom lens included in the image pickup unit 101, a motor that is a driving source thereof, and the like. The lens control unit 105 controls the operation of the lens driving unit 104 based on an instruction from the system control unit 103.

  The pan operation drive unit 106 is configured by a mechanical drive system that performs the pan operation of the image pickup unit 101, a motor that is a drive source thereof, and the like, and is configured to be capable of 360° infinite turning in the image pickup direction in the pan direction. .. The pan operation control unit 107 controls the operation of the pan operation drive unit 106 based on an instruction from the system control unit 103. In the present embodiment, the pan direction is set to the horizontal direction.

  The tilt operation drive unit 108 includes a mechanical drive system that performs the tilt operation of the imaging unit 101 and a motor that is a drive source thereof, and can rotate the imaging direction in the tilt direction from 0 degrees to +180 degrees in the horizontal direction. it can. The tilt operation control unit 109 controls the drive of the tilt operation drive unit 108 based on an instruction from the system control unit 103. Note that the tilt direction is set to the vertical direction (within the vertical plane) in the present embodiment.

  The image analysis unit 110 analyzes the image captured by the image capturing unit 101, that is, the captured image generated by the image processing unit 102, and detects, for example, a moving body (a subject to be tracked) as described later. Performs various analysis processes. The communication interface unit 111 enables communication connection to a network (not shown).

  In the present embodiment, the surveillance camera 100 is connected to a network such as a LAN or the Internet via the communication interface unit 111, and is configured to communicate with a plurality of client devices (not shown) via the connected network. .. Then, the client device can display a video image captured by the surveillance camera 100 and record and record it via the network. In addition, the client device transmits a camera control command to the monitoring camera 100, thereby performing setting of imaging conditions of the monitoring camera 100, pan operation, tilt operation, and the like.

  Next, an insensitive area of the surveillance camera 100 and an area for correcting the pan/tilt movement amount to zero (0) (hereinafter referred to as “zero correction area”) will be described. In the present embodiment, an area pattern including a dead area and a zero correction area is set according to the posture of the imaging unit 101, for example, the tilt position.

  First, the dead area when the tilt position is near the horizontal direction will be described. FIG. 2 is a diagram schematically illustrating the dead region and the zero correction region of the surveillance camera 100 when the tilt position is near the horizontal direction. The dead area 201 is an area in which the pan/tilt operation is not performed on the moving body existing near the center of the screen 200. The width and height of the dead area 201 are based on the number of pixels on the screen 200. Here, as an example, the width of the dead region 201 is 20% of the screen width and the height is 15% of the screen height, and the positions of the dead region 201 are equal in the vertical and horizontal directions (horizontal direction, vertical direction) from the screen center. Is set to be. However, the setting of the width, height and position of the dead area 201 is not limited to this.

  For example, when the center of the moving object detected by the image analysis unit 110 is located at the point P1 in the area 210 in the screen 200, the point P1 is outside the dead area 201, and therefore the system control unit 103 causes the pan/tilt. Execute control. On the other hand, when the moving object is located at the point P2, the system control unit 103 does not execute the pan/tilt control because the point P2 is in the dead region 201. When the moving object is at the point P3 on the boundary line of the dead area 201, the system control unit 103 does not execute the pan/tilt control because it is in the dead area 201.

  Next, the zero correction area when the tilt position is near the horizontal direction will be described. A horizontally long zero correction area 202 that vertically divides the vicinity of the center of the screen 200 is an area in which the tilt movement amount is corrected to zero and only the pan operation is performed. Further, a vertically long zero correction region 203 that divides the vicinity of the center of the screen into left and right is a region in which the pan movement amount is corrected to zero and only the tilt operation is performed.

  Here, the height of the zero correction area 202 is set lower than the height of the dead area 201, and the width of the zero correction area 203 is set narrower than the width of the dead area 201. Each of the zero correction regions 202 and 203 has a region which intersects with the dead region 201 and overlaps in the vicinity of the center of the screen. In this overlapping region, the control defined for the dead region 201 has priority over the control defined for the zero correction regions 202 and 203. That is, with respect to the moving body located at the points P6, P7, and P8, neither pan operation nor tilt operation is performed unless the moving body moves outside the dead zone 201.

  Next, the case where the tilt position is in the zenith direction will be described. When the surveillance camera 100 is installed on the ceiling of a building or the like, when the tilt angle is around 90 degrees, the image pickup direction is directly downward. FIG. 3 is a diagram schematically showing the pan/tilt operation when the surveillance camera 100 is installed on the ceiling. As shown in FIG. 3A, the pan operation is a circular motion centered on the screen center on the display screen. On the other hand, the tilt operation is a vertical movement on the display screen.

  The movement of the moving body on the screen 200 by the pan/tilt control in this case will be described with reference to FIG. When the moving body detected by the image analysis unit 110 is located at the point P11, in order to capture the moving body at the center of the screen, the point P11 is moved to the position of the point P12 by the pan operation of 90 degrees, and then the moving body is moved. The tilting operation may be performed so that is at the center of the screen. Further, even when the detected moving body is at the point P13, the pan movement amount required to capture the moving body at the center of the screen is 90 degrees, which is the same as when the moving body was located at the point P11. Therefore, it is sufficient to move the moving body from the point P13 to the position of the point P14 by the pan operation, and then perform the tilt operation so that the moving body comes to the center of the screen. On the other hand, when the moving body is located at the point P15, the moving body can be moved to the point P12 with a slight pan movement amount. However, the tilt movement amount is the same as when the moving body is at the point P11, and is the movement amount from the point P12 to the center of the screen.

  Thus, the pan movement amount is determined by the angle between the vertical axis and the straight line connecting the position of the moving body and the screen center, and has a value in the range of +90 degrees to −90 degrees. The amount of tilt movement is determined by the distance from the position of the moving body to the center of the screen, and the maximum amount of movement is half the horizontal angle of view determined by the zoom position. The position of the moving body is the center position of the moving body, and is represented by, for example, the center position of a circumscribed rectangle surrounding the moving body, as described later.

  Specifically, assuming that the vertical angle of view at the wide end is 40 degrees and the vertical angle of view at the tele end is 2 degrees at a zoom of 20 times, the tilt movement amount at the wide end is +20 degrees. The range is from -20 to -20 degrees, and the range from +1 to -1 degrees at the telephoto end. In the above description, the pan operation is performed first, and then the tilt operation is performed, but in reality, the pan operation and the tilt operation are simultaneously started.

  Next, the dead area when the tilt position is near 90 degrees will be described. FIG. 4 is a diagram schematically illustrating the dead region and the zero correction region of the surveillance camera 100 when the tilt position is near 90 degrees. As described with reference to FIG. 3, since the pan operation is a rotational motion with the center of the screen as the center of rotation, a region with a small pan movement amount is a vertically symmetrical triangular region as shown in FIG. 4A. It becomes 401. On the other hand, since the tilt operation is a movement toward the center of the screen, the area where the amount of tilt movement is small becomes a circular area 402 having a small distance from the center of the screen as shown in FIG. 4B. The tilt movement amount for moving the detected moving body to the center of the screen decreases as the position of the moving body approaches the center of the screen, and increases as the position moves away from the center of the screen.

  A region in which the tilt movement amount is large and the pan movement amount is small is indicated by a region 403 shown in FIG. A region where the pan movement amount is large and the tilt movement amount is small is indicated by a region 404 shown in FIG. 4C. Therefore, as shown in FIG. 4D, a region in which the dead region 405 near the center of the screen is overlapped with the regions 403 and 404 is a dead region where the tilt position is around 90 degrees.

  Here, the region 404 shown in FIG. 4C in which the tilt movement amount is corrected to zero is completely included in the dead region 405 in which the pan/tilt control is not performed shown in FIG. Region 404 no longer exists. The trapezoidal zero correction region 406 located above and below the dead region 405 is a region in which the pan movement amount is corrected to zero and only the tilt operation is performed.

  The dead zone 405 shown in FIG. 4D is a dead zone set when the tilt position is near 90 degrees in the first control method for automatically tracking a moving body, which will be described later. At this time, the dead area 201 in FIG. 2 and the dead area 405 in FIG. 4 are rectangular areas having the same size, so that the moving body is located inside or outside the dead area 201 (405) regardless of the tilt position. It can be determined.

  Next, a first control method for automatically tracking a moving body by the monitoring camera 100 will be described. FIG. 5 is a flowchart of a first control method for automatic tracking performed using the surveillance camera 100. Each process shown in the flowchart of FIG. 5 is realized by each unit configuring the monitoring camera 100 performing a predetermined operation under the control of the system control unit 103.

  In step S501, the image analysis unit 110 acquires an image from the image processing unit 102 and detects a moving object from the difference between the current frame image and the past frame image. In step S502, the system control unit 103 determines whether a moving body has been detected. When the moving body is detected (YES in S502), the system control unit 103 advances the process to step S503, and when the moving body is not detected (NO in S502), ends the process.

  In step S503, the system control unit 103 sets the moving body detected in step S501 as a tracking target object, obtains a circumscribed rectangle surrounding the detected moving body, and further obtains the center position of the moving body on the screen. Get the center position. It should be noted that the image analysis unit 110 can be configured to perform the actual calculation process of obtaining the circumscribed rectangle surrounding the detected moving body and further obtaining the center position thereof.

  In the following step S504, the system control unit 103 determines whether the center position of the moving object calculated in step S503 is inside the dead area. When the center position of the moving body is inside the dead area (including on the boundary line) (YES in S504), the system control unit 103 returns the process to step S501, and when the center position of the moving body is not inside the dead area ( (NO in S504), the process proceeds to step S505. Note that, in the present embodiment, as described above, the dead area 201 of FIG. 2 and the dead area 405 of FIG. 4 are set to the same size and at the same position, so that the dead area serving as the determination reference in step S504 is the dead area. It is not necessary to determine which of 201 and 405.

  In step S505, the system control unit 103 captures the moving object at the center of the screen based on the moving object position on the screen, the pan/tilt absolute position in which the imaging unit 101 is currently imaging, and the current zoom position. Calculate the required pan/tilt movement amount. The pan/tilt movement amount calculated in step S505 is a movement amount required to move the center position of the moving body to the screen center. Here, since the pan operation drive unit 106 can make an infinite turn of 360 degrees, the pan absolute position has a value in the range of 0 degrees to 360 degrees. Further, since the tilt operation drive unit 108 can be driven in the range of 0 degrees to 180 degrees, the absolute tilt position has a value in the range of 0 degrees to 180 degrees.

  In the following step S506, the system control unit 103 determines, based on the pan/tilt movement amount calculated in step S505 and the current zoom position, the pan/tilt movement amount and the zoom movement amount required to capture the moving object in the center of the screen. to correct. The correction processing of the pan/tilt movement amount in step S506 will be described with reference to the flowchart of FIG. FIG. 6 is a flowchart of a method for correcting the pan/tilt movement amount.

  In step S601, the system control unit 103 obtains the horizontal angle of view of the imaging range based on the current zoom position. In subsequent step S602, the system control unit 103 compares the pan movement amount with the horizontal angle of view, and determines whether the pan movement amount is equal to or more than a predetermined ratio of the horizontal angle of view. Here, the predetermined ratio serving as the determination standard is set to 10%, but the present invention is not limited to this. When the pan movement amount is less than 10% of the horizontal angle of view (NO in S602), the system control unit 103 advances the process to step S603, and when the pan movement amount is 10% or more of the horizontal angle of view (YES in S602). , The process proceeds to step S604.

  In step S603, the system control unit 103 corrects the pan movement amount to zero, and this processing ends. The processing of step S603 is executed when the moving body is within the zero correction area 203 in FIG. 2 when the tilt position is near the horizontal position, for example, when the center position of the moving body is the point P5. .. Further, when the tilt position is near 90 degrees, it means that the moving body is in the zero correction area 406 of FIG. 4D.

  In step S604, the system control unit 103 obtains the vertical angle of view of the imaging range based on the current zoom position. In subsequent step S605, the system control unit 103 compares the tilt movement amount with the vertical angle of view, and determines whether the tilt movement amount is equal to or more than a predetermined ratio of the vertical angle of view. Here, the predetermined ratio serving as the determination standard is set to 10%, but the present invention is not limited to this. When the tilt movement amount is 10% or more of the vertical angle of view (YES in S605), the system control unit 103 ends the present process. This processing ends here when the tilt position is near the horizontal when the moving body is in any one of the areas 210, 220, 230, and 340 in FIG. 2, and for example, the center position of the moving body. Is at the point P1. Further, when the tilt position is near 90 degrees, it means that the moving body is neither in the dead region 405 nor the zero correction region 406 in FIG. 4D.

If it is determined in step S605 that the tilt movement amount is less than 10% of the vertical angle of view (NO in S605), the system control unit 103 advances the process to step S606. Then, in step S606, the system control unit 103 corrects the tilt movement amount to zero, and this processing ends. The process of step S606 is executed, if the tilt position is near horizontal, moving object is when you are in the zero correction region 20 2 in FIG. 2, for example, in the case the center position of the moving object at the point P4 is there. Further, when the tilt position is near 90 degrees, it means that the moving body is in the area 404 of FIG. 4D, but since the area 404 is included in the dead area 405 of FIG. It is already determined by the inside/outside determination of the dead region that the pan/tilt operation is not performed.

  Returning to the explanation of FIG. In step S507, the system control unit 103 branches the process according to the corrected pan movement amount and tilt movement amount. When the corrected pan movement amount is not zero and the tilt movement amount is zero, the system control unit 103 advances the process to step S508, and the corrected pan movement amount is zero and the tilt movement amount is zero. If the amount of movement is not zero, the process proceeds to step S510. If the corrected pan movement amount and tilt movement amount are not zero, the system control unit 103 advances the process to step S512.

  In step S508, the system control unit 103 only controls the pan operation. Then, in step S509, the system control unit 103 determines whether or not the pan operation has ended. If the pan operation has not ended (NO in S509), the system control unit 103 repeats the determination in step S509, and if the pan operation has ended (YES in S509), returns the process to step S501.

  In step S510, the system control unit 103 only controls the tilt operation. Then, in step S511, the system control unit 103 determines whether or not the tilt operation has ended. If the tilt operation has not ended (NO in S511), the system control unit 103 repeats the determination in step S511, and if the tilt operation has ended (YES in S511), returns the process to step S501.

  In step S512, the system control unit 103 controls both the pan operation and the tilt operation. Then, in step S513, the system control unit 103 determines whether the pan/tilt operation is completed. If the pan/tilt operation has not ended (NO in S513), the system control unit 103 repeats the determination of step S513, and if the pan/tilt operation has ended (YES in S513), returns the process to step S501.

  As described above, according to the first control method of automatically tracking a moving object, the pan operation and the tilt operation are simultaneously executed by providing the areas for forcibly setting the pan movement amount and the tilt movement amount to zero. To reduce the frequency of its execution. As a result, the time required for the pan/tilt operation for automatic tracking can be shortened and the ability to follow a moving object can be improved.

Next, a second control method for automatically tracking a moving body by the surveillance camera 100 will be described. In the second control method for automatically tracking a moving object, the tilt position in the case of around 90 degrees, zero correction region described in FIG. 4 (d), the further adding a zero correction region for correcting the tilt movement amount to zero .

  FIG. 7 is a diagram schematically illustrating a dead area and a zero correction area that are set when the tilt position is near 90 degrees in the second control method of automatically tracking a moving object by the monitoring camera 100. In the triangular area 701 shown in FIG. 7A, since the movement angle of the pan operation is close to 90, the operation time becomes long. Therefore, even if there is a little distance from the vicinity of the center of the screen, it is not desirable to perform the tilt operation as much as possible. Therefore, the fan-shaped zero correction area 702 shown in FIG. 7B is added to the pattern (dead area 405, zero correction area 406) shown in FIG. 4D as an area in which the tilt operation is not performed. FIG. 7C is a diagram showing a dead region and a zero correction region in which the pattern of FIG. 7B and the pattern of FIG. 4D are superimposed. An area in the zero correction area 702 that does not overlap the rectangular dead area 405 is a zero correction area in which the tilt movement amount is corrected to zero and only the pan operation is performed.

  FIG. 8 is a flowchart of a second control method for automatic tracking executed using the surveillance camera 100. Each process shown in the flowchart of FIG. 8 is realized by each unit configuring the surveillance camera 100 performing a predetermined operation under the control of the system control unit 103. Note that the description will be made here on the assumption that the moving object to be tracked is detected by the image analysis unit 110.

  In step S801, the system control unit 103 acquires the center position of the moving object on the screen. The processing content of step S801 is the same as the processing content of step S503 of FIG. Subsequently, the system control unit 103 acquires the current tilt position and determines whether or not the tilt position is right below. Here, when the tilt position is in the range of 85 degrees to 95 degrees, it is determined to be near the bottom, and when it is in a range other than this, it is determined to be not near the bottom. If the tilt position is not directly below (NO in S802), the system control unit 103 proceeds with the process to step S803, and if the tilt position is near directly below (YES in S802), the process proceeds to step S805.

  In step S803, the system control unit 103 sets the area pattern (dead area 201, zero correction areas 202 and 203) near the horizontal direction described with reference to FIG. In subsequent step S804, the system control unit 103 determines the height of the zero correction area 202 and the width of the zero correction area 203 based on the current zoom position. For example, the vertical angle of view and the horizontal angle of view of the screen are obtained from the current zoom position, 10% of the vertical angle of view is the height of the zero correction area 202, and 10% of the horizontal angle of view is the width of the zero correction area 203.

  In step S805, the system control unit 103 sets the area pattern (dead area 405, zero correction areas 406 and 702) near the area directly underneath shown in FIG. 7C. Then, in step S806, the system control unit 103 determines the radius of the zero correction region 702 in which the tilt movement amount is zero based on the current zoom position. Here, the horizontal angle of view of the screen is obtained from the current zoom position, and 15% of the horizontal angle of view is set as the radius of the zero correction region 702, but the present invention is not limited to this. The center angles of the zero correction regions 702 and 406 are fixed regardless of the zoom position.

  The system control unit 103 advances the processing to step S807 after the processing of steps S804 and S806, and in step S807, determines whether or not the center position of the moving body is within the dead zone 405. If the center position of the moving body is within the dead region 405 (YES in S807), the system control unit 103 advances the process to step S808. If the center position of the moving body is not within the dead region 405 (NO in S807), the process is performed. To Step S809.

  In step S808, the system control unit 103 ends this processing without performing the pan/tilt operation. On the other hand, in step S809, the system control unit 103 determines the amount of pan/tilt movement required to capture the moving object at the center of the screen by the horizontal angle of view/angle calculated from the position of the moving object on the screen, the current pan/tilt position, and the zoom position. Calculated from the vertical angle of view. Also in this case, the calculated pan/tilt movement amount is a movement amount required to move the center position of the moving body to the center of the screen.

  In subsequent step S810, the system control unit 103 corrects the tilt movement amount to zero and determines whether or not the center position of the moving body is in the zero correction region 702 in which only the pan operation is performed. When the center position of the moving object is within the zero correction area 702 (YES in S810), the system control unit 103 advances the process to step S811, and when the center position of the moving object is not within the zero correction area 702 (NO in S810). , The process proceeds to step S812.

  In step S811, the system control unit 103 corrects the tilt movement amount to zero, and ends this processing. If the center position of the moving body is within the zero correction area 702 of FIG. 7 and also within the area overlapping the dead area 405, step S807 has already become YES and the process proceeds to step S808. It is progressing. That is, the case where the determination in step S810 is YES is the case where the center position of the moving body is within the zero correction region 702 of FIG. 7 and within the region that does not overlap with the dead region 405.

  In step S812, the system control unit 103 corrects the pan movement amount to zero and determines whether or not the center position of the moving object is in the zero correction region 406 where only the tilt operation is performed. When the center position of the moving object is within the zero correction area 406 (YES in S812), the system control unit 103 advances the process to step S813, and when the center position of the moving object is not within the zero correction area 406 (NO in S812). , The process proceeds to step S814.

  In step S813, the system control unit 103 corrects the pan movement amount to zero, and ends this processing. Further, in step S814, the system control unit 103 ends this processing without correcting the pan/tilt movement amount.

  According to such a second control method, as compared with the first control method, the region in which the tilt operation is not performed is increased, so that it is possible to further improve the followability to the moving body.

  Next, a third control method for automatically tracking a moving body by the monitoring camera 100 will be described. In the above-described first and second control methods, the drive control of the pan/tilt operation is performed based on the position of the moving body. On the other hand, in the third control method, not only the position of the moving body but also the size of the moving body is used to perform the drive control of the pan/tilt operation, so that the display screen (image pickup angle of view) of the moving body is changed. Prevent the protrusion.

  FIG. 9 is a diagram showing the relationship between the dead region and the imaging angle of view in consideration of the size of the moving body. The dead area 901 shown in FIG. 9A can be the same as the dead area 201 described with reference to FIG. 2 or the dead area 405 described with reference to FIG. When the image analysis unit 110 detects a moving object that is a tracking target within the imaging angle of view, the image analysis unit 110 (or the system control unit 103) sets the center position 903 of the moving object to the circumscribed rectangle 902 with respect to the moving object. Find the center position.

  When the center position 903 of the moving body is within the dead zone 901, the drive control of the pan/tilt operation was not performed by the first and second control methods. However, when the size of the moving body is large and the center position 903 of the moving body is in the dead area 901 and the drive control of the pan/tilt operation is not performed, when the moving body further moves in the dead area 901, As shown in FIG. 9(a), the moving object may run off the display screen. Therefore, in the third control method of automatically tracking a moving body, such a protruding of the moving body from the display screen is prevented. 9B and 9C will be described together in the description of the flowchart of FIG. 10 described below.

  FIG. 10 is a flowchart of a third control method for automatic tracking executed by the surveillance camera 100. Each process shown in the flowchart of FIG. 10 is realized by each unit configuring the monitoring camera 100 performing a predetermined operation under the control of the system control unit 103. It should be noted that in the description here, the reference numerals used in FIG. 9 will be used as appropriate, and the description will be given on the assumption that the moving object to be tracked is detected by the image analysis unit 110.

  In step S1001, the system control unit 103 acquires the detected position of the moving body on the screen. Subsequently, in step S1002, the system control unit 103 obtains the center position 903 of the moving body and determines whether the center position 903 is within the dead zone 901. When the center position 903 of the moving body is in the dead region 901 (YES in S1002), the system control unit 103 advances the process to step S1003, and when the center position 903 of the moving body is not in the dead region 901 (YES in S1002). , The process proceeds to step S1004.

  In order to obtain the center position 903 of the moving body in step S1002, the circumscribed rectangle 902 of the moving body has already been obtained. Therefore, in step S1003, the system control unit 103 determines whether the distance from the circumscribed rectangle 902 to the screen edge is smaller than a predetermined threshold value.

  The determination in step S1003 will be specifically described with reference to FIGS. 9B and 9C. In the case of FIG. 9B, the distances hu1 and hd1 between the circumscribing rectangle 902 and the upper and lower ends of the screen are compared with the height H of the screen, and the distances wl1 and wr1 between the circumscribing rectangle 902 and the left and right ends of the screen are calculated respectively. Compare with the width W of. For example, when at least one of hu1 and hd1 is smaller than the predetermined threshold value Th_h, it is determined that the distance on the screen between the circumscribed rectangle 902 and the angle of view is smaller than the predetermined threshold value. Also, when at least one of wl1 and wr1 is smaller than the predetermined threshold value Th_w, it is determined that the distance on the screen between the circumscribed rectangle 902 and the angle of view is smaller than the predetermined threshold value.

  Similarly, in the case of FIG. 9C, the distances hu2 and hd2 between the circumscribed rectangle 902 and the upper and lower ends of the screen are compared with the height H of the screen, respectively, and the distances wl2 and wr2 between the circumscribed rectangle 902 and the left and right ends of the screen are compared. Respectively with the width W of the screen. If at least one of hu2 and hd2 is smaller than the predetermined threshold value Th_h, it is determined that the distance on the screen between the circumscribed rectangle 902 and the angle of view is smaller than the predetermined threshold value. Further, even when at least one of the values wl2 and wr2 is smaller than the predetermined threshold value Th_w, it is determined that the distance on the screen between the circumscribed rectangle 902 and the angle of view is smaller than the predetermined threshold value.

  Here, the example of FIG. 9B is an example in which all of the four distance values are equal to or greater than the corresponding threshold values, and thus NO is determined in step S1003. On the other hand, the example in FIG. 9C is an example in which hu2 of the four distance values is smaller than the threshold Th_w, and thus YES is determined in step S1003. For the determination in step S1003, the ratio value between the distances hu1, hd1, hu2, hd2 and the height H is compared with a predetermined threshold value, and the ratio value between the distances wl1, wr1, wl2, wr2 and the width W is calculated. May be compared with a predetermined threshold.

  When the distance from the circumscribed rectangle 902 to the screen edge is equal to or greater than the predetermined threshold (NO in S1003), the system control unit 103 advances the process to step S1004. In step S1004, the system control unit 103 performs a pan/tilt operation because the moving body does not protrude from the screen even if the moving body is moved within the screen, and this processing ends. On the other hand, when the distance from the circumscribed rectangle 902 to the screen edge is smaller than the predetermined threshold value (YES in S1003), the system control unit 103 advances the process to step S1005. In step S1005, the system control unit 103 does not perform the pan/tilt operation because the moving body may move out of the screen when the screen is moved, and this processing ends.

  As described above, the third control method for automatically tracking a moving object can also achieve the same effect as the first and second control methods. Further, in the third control method of automatically tracking a moving body, the size of the moving body is taken into consideration, the dead area is determined, and the drive control of the pan/tilt operation is not performed, so the moving body protrudes from the screen (image pickup angle of view). Can be prevented.

<Other embodiments>
Although the present invention has been described above in detail based on its preferred embodiments, the present invention is not limited to these specific embodiments, and various embodiments within the scope not departing from the gist of the present invention are also included in the present invention. included. Furthermore, each of the above-described embodiments is merely an example of the present invention, and the embodiments may be combined as appropriate.

  The present invention also supplies a program that implements one or more functions of the above-described embodiments to a system or apparatus via a network or a storage medium, and one or more processors in a computer of the system or apparatus read the program. It can also be realized by the processing to be executed. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

100 Monitoring Camera 101 Imaging Unit 102 Image Processing Unit 103 System Control Unit 106 Pan Operation Drive Unit 107 Pan Operation Control Unit 108 Tilt Operation Drive Unit 109 Tilt Operation Control Unit

Claims (7)

Detection means for detecting a moving body from a video imaged by an imaging means capable of changing the imaging direction by a pan operation and a tilt operation;
Control means for performing at least one of pan control and tilt control of the image pickup means so as to track the moving body detected by the detection means,
Wherein, when the center of the moving object is in the first area of the image, to stop the panning operation or the tilting operation,
Automatic tracking apparatus shape of the first region, wherein Rukoto is changed according to a tilt position of the imaging means. The control means, automatic tracking apparatus according to claim 1 in which the center of the moving object when in the second region of said image, characterized in that it stops the pan operation and the tilt operation. Before SL automatic tracking apparatus according the size of the first region to claim 1, characterized in Rukoto adjusted according to the zoom position of the imaging means. What is claimed is: 1. An automatic tracking method for automatically tracking a moving object imaged by an image pickup means capable of panning and tilting operations for changing an image pickup direction by controlling the image pickup means by a computer,
A detection step of detecting a moving object from the video imaged by the imaging means,
A control step of controlling at least one of the pan operation and the tilt operation of the imaging means so as to track the moving body detected in the detecting step,
In the control step, the stop panning operation or the tilting operation when the center of the moving object is within the first region of the image,
The automatic tracking method, wherein the shape of the first area is changed according to the tilt position of the image pickup means . A program that causes a computer to execute the automatic tracking method according to claim 4 . An image pickup means capable of changing the image pickup direction by a pan operation and a tilt operation;
A detection unit that detects a moving object from the video imaged by the imaging unit;
Control means for performing at least one of pan control and tilt control of the image pickup means so as to track the moving body detected by the detection means,
Wherein, when the center of the moving object detected by said detecting means is in a first region of the image, to stop the panning operation or the tilting operation,
Imaging apparatus shape of the first region, wherein Rukoto is changed according to a tilt position of the imaging means. 7. The image pickup apparatus according to claim 6 , wherein a control device having the detection unit and the control unit and the image pickup unit are communicably connected via a network.