JP2017168885A - Imaging control device and camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow objects to be tracked to be naturally tracked successively in an image to be taken when a plurality of cameras track an object sequentially.SOLUTION: The imaging control device determines one camera of a plurality of cameras as a main camera; determines some or all of the cameras other than the main camera as sub cameras; transmits a tracking instruction to the main camera and transmits monitoring instructions to the sub cameras; receives from the main camera a report that an object disappears from a visual field of the main camera without acquiring predetermined feature amounts and a report about an imaging state of the main camera at the timing when the object disappears from the visual field; redetermines a camera other than the determined main camera of the plurality of cameras as the main camera and redetermines some or all of cameras other than the redetermined main camera as the sub cameras; transmits to the redetermined main camera a tracking instruction using the received imaging state and transmits to the redetermined sub cameras monitoring instructions.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an imaging control device and a camera.

In general, one or a plurality of monitoring cameras monitor a space such as a store. When one camera monitors the space, the resolution may be changed. When a plurality of monitoring cameras monitor a space in parallel, role sharing may be performed among them.
In Patent Document 1, one camera reads an entire image from an image sensor at a first resolution, and reads a partial area of the entire image at a second resolution. Further, the camera reads out a feature in a partial area at the third resolution. At this time, since the first resolution <the second resolution <the third resolution, the camera can detect the feature object with high accuracy without increasing the processing load.

  In Patent Document 2, when a detection camera including a reflecting mirror capable of imaging a wide range monitors a space and detects a moving body, the position information of the moving body is transmitted to the determination camera. On the other hand, the determination camera captures the moving body while performing pan, tilt, and zoom as necessary, and determines whether the moving body is an object. While the predetermined determination process is being performed, the detection camera continues to be monitored. Therefore, it is not overlooked that a mobile body appears.

  In Patent Document 3, a plurality of monitoring cameras monitor the same space in parallel. When a moving body is detected while a certain camera is performing wide-angle imaging, the camera determines whether another camera is performing wide-angle imaging. When another camera is performing wide-angle imaging, the camera tracks the moving body while performing pan, tilt, and zoom. If other cameras are not performing wide-angle imaging, wide-angle imaging is continued. Therefore, one of the cameras always performs wide-angle imaging, and there is no detection omission of the moving body.

  In Patent Document 4, a plurality of cameras capable of panning, tilting, and zooming monitor a certain space in parallel. When a camera detects a moving object, the moving object is tracked as it is. When the moving body moves to such an extent that the camera does not fit within the angle of view even when the imaging direction is changed, the camera takes over tracking of the moving body to another camera, and then enters a normal monitoring state (wide angle). Return. Accordingly, it is possible to prevent omission of detection of a new moving body without interrupting tracking of the already-discovered moving body.

  In patent document 5, a guard goes around in a certain space. Four cameras capable of panning, tilting and zooming are installed in the space. During normal times, one of them tracks the security guard, and the other three cameras share and monitor the space so as not to overlap. When one of the three units detects an abnormal state (for example, keeps watching the direction of a security guard), the other one of the three switches the wide-angle imaging to the zoom imaging, and the security guard The suspicious object ahead of the line of sight is imaged. At this time, the camera that is imaging the suspicious object cannot monitor its own shared space. Therefore, the remaining two units are shared and monitored so as not to overlap in the space. Therefore, even if a suspicious object is found, the entire space is continuously monitored.

JP 2010-130087 A JP-A-7-37100 JP 2007-104223 A JP 2005-33827 A JP 2012-156752 A

  For example, when searching for a human being arranged, (1) it is required in the field to accurately track a moving human and acquire features such as a face and a hairstyle with high accuracy. On the other hand, (2) it is important to prevent a wide range of monitoring from becoming obscured because the focus is on local tracking.

  Although the camera of Patent Document 1 certainly reduces the processing load, the whole image read at the first insufficient resolution (low resolution) is high enough to identify the feature. Often not. In addition, when the feature is not detected immediately but is reproduced after long-time imaging / recording, it is costly to capture the entire image with a sufficiently high resolution.

  At first glance, the techniques of Patent Documents 2 to 5 seem to satisfy the above-mentioned (1) and (2). However, when a plurality of cameras share roles and track, a plurality of images acquired by the plurality of cameras are not smoothly continuous. It is not guaranteed that the person to be tracked appears at the same position on the screen before and after the camera switching, and that the size on the screen is the same. Therefore, even if the apparatus extracts human features immediately or afterwards, extra time is required regardless of whether the user visually confirms the human features by viewing the screen.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to make a tracking target naturally continuous in an acquired image when a plurality of cameras sequentially track the target.

An imaging control apparatus of the present invention determines a camera among a plurality of cameras as a main camera, and determines a part or all of the plurality of cameras other than the main camera as sub cameras, and a target A tracking instruction unit that tracks an object and transmits an instruction to acquire a predetermined number of feature quantities of the object to the main camera, and transmits an instruction to continuously image a predetermined area to the sub camera; Report that the target object has come out of the main camera field of view without being able to acquire a predetermined feature quantity from the number of feature quantities, and the main point when the target object has come out of the main camera field of view Re-determine the imaging state of the camera as the main camera from the imaging state reception unit that receives from the main camera and a camera other than the determined main camera among the multiple cameras. A role re-determination unit that re-determines some or all of the cameras other than the re-determined main camera as sub-cameras, and tracks the object using the received imaging state, and a predetermined number of objects A continuous tracking instruction unit that transmits to the re-determined main camera an instruction to acquire the feature amount of the image and transmits an instruction to continuously image a predetermined area to the sub-camera that has been re-determined. .
Other means will be described in the embodiment for carrying out the invention.

  According to the present invention, when a plurality of cameras track an object in order, the tracking target can be naturally continued in the acquired image.

(A) is a figure explaining imaging object space. (B) is a figure explaining the structure of an imaging control apparatus. It is a figure explaining the object tracking (existing technique) using three cameras. It is a figure explaining object tracking (this embodiment) using three cameras. (A) is a figure showing an example of feature quantity information. (B) is a figure showing an example of imaging condition information. It is a figure which shows an example of imaging log | history information. It is a figure explaining an envelope rectangular parallelepiped. It is a sequence diagram of a processing procedure.

  Hereinafter, a mode for carrying out the present invention (referred to as “the present embodiment”) will be described in detail with reference to the drawings.

  The imaging target space will be described along FIG. A typical imaging target space S of the present embodiment is a store or the like where an unspecified number of people come and go. Cameras CM1, CM2, and CM3 are arranged in the imaging target space S. These three cameras CM1 etc. are attached to the wall surface of a store, for example. Each of the three cameras can pan and tilt on the free pan head, and can also change (zoom) the angle of view.

  Each of the three cameras has a unique imaging range. For example, the camera CM1 captures an image around a store entrance, the camera CM2 captures an image near a store shelf, and the camera CM3 captures an image near a store cash register. Therefore, each of the three cameras has a unique blind spot. However, all points in the imaging target space S can be imaged by at least one of the three cameras. That is, the blind spot of the imaging target space S is eliminated by the cooperation of the three cameras.

  The cameras CM1, CM2, and CM3 are connected to the imaging control device 1 by wire or wirelessly. The imaging control apparatus 1 transmits a control signal to each camera, and receives an image captured from each camera and other necessary information. The imaging control device 1 is normally arranged near the outside of the imaging target space S.

  A configuration of the imaging control device 1 will be described with reference to FIG. The imaging control device 1 is a general computer, a central control device 11, an input device 12 such as a keyboard, an output device 13 such as a display, a main storage device 14, an auxiliary storage device 15, and a communication device 16 (communication with a camera). ). These devices are connected to each other by a bus. The auxiliary storage device 15 stores a map 31, an image 32, feature amount information 33, imaging condition information 34, and imaging history information 35.

  The role determination unit 21, the tracking instruction unit 22, the imaging state reception unit 23, the role redetermination unit 24, and the continuous tracking instruction unit 25 in the main storage device 14 are programs. Thereafter, when the subject is described as “XX section”, the central control device 11 reads out each program from the auxiliary storage device 15 and loads it into the main storage device 14, and then the function of each program (detailed later). Shall be realized.

The object tracking (existing technology) using three cameras will be described with reference to FIG. The vertical axis in FIG. The lower it is in FIG. The horizontal axis in FIG. 2 represents each camera. The image at the intersection of the vertical axis and the horizontal axis is an image captured by the camera at that time.
At time t1, the camera CM1 captures the person to be tracked (also referred to as “object”) in the field of view. Then, the camera CM 1 tracks the object while performing pan, tilt, and zoom. At the same time, it tries to acquire the feature amount (detailed later) of the object.

  At time t2, the camera CM1 has caught the object in the approximate center of the screen. At time t3, the camera CM1 captures the object at a field angle that is more zoomed than at time t2. However, the object moves from moment to moment and is unlikely to fall within the pan / tilt range of the camera CM1. From the time point t1 to the time point t3, the cameras CM2 and CM3 do not particularly track the object, and take an image of a predetermined space in the store with a wide angle of view. Although details will be described later, the imaging condition of the camera CM1 from time t1 to time t3 is referred to as “CN2” (tracking). The imaging condition of the camera CM2 and camera CM3 is referred to as “CN1” (monitoring).

  At time t4, the camera CM1 switches the imaging condition from “CN2” to “CN1”. Then, the camera CM1 stops tracking the object and starts imaging a predetermined space (near the entrance) in the store with a wide angle of view. At time t4, the camera CM2 switches the imaging condition from “CN1” to “CN2”. Then, the camera CM2 no longer captures a predetermined space in the store (near the product shelf) with a wide angle of view, and starts tracking the object.

  At time t4, the camera CM2 captures the object in the field of view. Then, the camera CM 2 tracks the object while performing pan / tilt / zoom. At the same time, the feature amount of the object is to be acquired. At time t5, the camera CM2 captures the object in the approximate center of the screen. At time t6, the camera CM1 captures the object at a field angle that is more zoomed than at time t5. However, the object moves from moment to moment and is unlikely to fall within the pan / tilt range of the camera CM2. From the time point t4 to the time point t6, the cameras CM1 and CM3 do not particularly track the object, and take an image of a predetermined space in the store with a wide angle of view.

  At time t7, the camera CM2 switches the imaging condition from “CN2” to “CN1”. Then, the camera CM2 stops tracking the object and starts imaging a predetermined space in the store (in the vicinity of the product shelf) with a wide angle of view. At time t7, the camera CM3 switches the imaging condition from “CN1” to “CN2”. Then, the camera CM3 no longer captures a predetermined space in the store (near the cash register) with a wide angle of view, and starts tracking the object.

  At time t7, the camera CM3 captures the object in the field of view. Then, the camera CM 3 tracks the object while performing pan, tilt, and zoom. At the same time, the feature amount of the object is to be acquired. At the time point t8, the camera CM3 captures the object at the approximate center of the screen. At time t9, the camera CM3 captures the object at a field angle that is more zoomed than at time t8. However, the object moves from moment to moment and is unlikely to be within the pan / tilt range of the camera CM3. From time t7 to time t9, the cameras CM1 and CM2 do not particularly track the object, and take an image of a predetermined space in the store with a wide angle of view.

  It can be seen from FIG. 2 that the three cameras CM1 and the like monitor the imaging target space S so as not to have a blind spot while tracking the object while sequentially changing the roles. One of the three cameras acquires the feature amount of the object while tracking the object. However, it is not easy to acquire feature amounts, and tracking of an object is often handed over to the next camera while one camera cannot acquire the feature amounts.

  Attention is paid to the image 41a of the camera CM1 at the time point t3 and the image 41b of the camera CM2 at the time point t4. At time t3, the camera CM1 captures the appearance of the object in the screen. On the other hand, at the time point t4, the camera CM2 not only captures the appearance of the target object in the screen, but also the target object approaches the left side of the screen. Human feature quantities such as a face and a hairstyle are more easily acquired when the person is captured (zoomed) at a large size in the screen and is captured at the center of the screen. Then, the time until the camera CM2 captures a large object at the center of the screen is wasted. The same applies to the image 42a of the camera CM2 at the time point t6 and the image 42b of the camera CM3 at the time point t7.

  The object tracking (this embodiment) using three cameras is demonstrated along FIG. Attention is paid to the image 43 of the camera CM2 at the time point t4. The camera CM2 has a large object captured substantially in the center of the screen. Attention is paid to the image 44 of the camera CM3 at time t7. The camera CM3 has a large object captured substantially in the center of the screen. In this way, if the camera that has succeeded in tracking the object can catch the object largely in the center of the screen from the beginning, the speed of acquiring the feature amount of the object is improved.

(Feature information)
The feature amount information 33 will be described with reference to FIG. In the feature amount information 33, the feature amount is stored in the feature amount column 102 in association with the feature amount ID stored in the feature amount ID column 101.
The feature amount ID in the feature amount ID column 101 is an identifier that uniquely identifies the feature amount (immediately described later).
The feature amount in the feature amount column 102 is a partial image that contributes to specifying the target object among the partial images of the target object on the screen. For example, when the object is a human, the feature amount is a body part such as head hair, eyes, nose, mouth, outer garment, trousers, shoes, or a bag, a clothing part, or a portable object. For example, “hair” has a size, shape, and color sufficient to be recognized as “hair”. The imaging control apparatus 1 stores a plurality of samples (templates) of such hair in the auxiliary storage device 15 and can cut out the hair from the image. The same applies to eyes, nose, mouth, and so on.

(Imaging condition information)
The imaging condition information 34 will be described with reference to FIG. In the imaging condition information 34, in association with the imaging condition ID stored in the imaging condition ID field 111, the name field 112 has a name, the pan / tilt field 113 has a pan / tilt condition, and the angle of view field 114 has a name. The angle-of-view condition is stored in the feature amount detection column 115.

The imaging condition ID in the imaging condition ID column 111 is an identifier that uniquely identifies the imaging condition. The imaging condition is an instruction related to imaging that the imaging control device 1 transmits to the camera.
The name in the name column 112 is the name of the imaging condition, and here is either “monitoring” or “tracking” as described above.
The pan / tilt condition in the pan / tilt column 113 is either “optical axis fluctuation” in which the optical axis of the camera is changed in the vertical direction and horizontal direction, or “optical axis fixed” in which the optical axis is not changed.
The angle-of-view condition in the angle-of-view column 114 is “fixed wide” to fix the angle of view of the camera to a predetermined wide angle (for example, 30 degrees), or the angle of view of the camera so that the object is imaged on the full screen. One of the “zoom fluctuations” to zoom.
The feature amount detection condition in the feature amount detection column 115 is either “Yes” for cutting out the feature amount or “No” for not cutting out the feature amount.

(Imaging history information)
The imaging history information 35 will be described with reference to FIG. In the imaging history information 35, a CM1 column 122, an imaging state takeover column 123, a CM2 column 124, an imaging state takeover column 125, a CM3 column 126, and an acquired feature amount column 127 are stored in association with the imaging time point column 121. For example, the data in the CM1 column 122 is data about the camera CM1, and the other is the same.

The CM1 column 122 includes an imaging condition ID column 122a, an optical axis position column 122b, an object occupancy rate column 122c, and an angle of view column 122d.
The imaging condition ID in the imaging condition ID column 122a is the same as the imaging condition ID in FIG.
The optical axis position in the optical axis position column 122b is the coordinate value of the center of gravity of the object on the image in the row where the imaging condition ID is “CN2”. The optical axis position is the coordinate value of a fixed object that is imaged at the intersection of the diagonal lines of the screen in the row where the imaging condition ID is “CN1”. The origin of the three-dimensional coordinate axis can be set at an arbitrary position inside or outside the imaging target space S. In the example of FIG. 1, one vertex of the imaging target space S (cuboid) is the origin. When changing the imaging condition from “monitoring” to “tracking”, each camera takes over the optical axis position from the camera that was “tracking” at the previous time point so that its own optical axis passes through the optical axis position. The direction of the optical axis can be determined.

The object occupancy in the object occupancy column 122c is the ratio (percentage) of the area of the object to the area (number of pixels) of the entire screen.
The angle of view in the angle of view column 122d is the angle of view of the camera.
The same applies to the CM2 column 124 and the CM3 column 126.

In the imaging state takeover column 123, “◎” is stored across the rows at time t3 and time t4. This indicates the following.
The imaging control device 1 has received “optical axis position = X ₃ , Y ₃ , Z ₃ ” and “object occupancy = 20%” from the camera CM 1 at time t ₃ .
The imaging control device 1 has transmitted “optical axis position = X ₃ , Y ₃ , Z ₃ ” and “object occupancy = 20%” to the camera CM 2 at time t ₃ .

As a result, at time t4, the camera CM2 images the object under the conditions of “optical axis position = X ₃ , Y ₃ , Z ₃ ” and “object occupation ratio = 20%”.
The angle of view of camera CM1 at time t3 was “7 degrees”, but the angle of view of camera CM2 at time t4 was “8 degrees”. As a matter of course, since the distance between the object and the camera is different for each camera, the angle of view will change if an attempt is made to maintain the same object occupation rate.
The same applies to “◎” in the imaging state takeover column 125.

The acquired feature value column 127 stores the feature value ID when the feature value of the object is acquired, and stores “none” when the feature value is not acquired.
When the imaging history information 35 is viewed again as a whole, the following can be understood.
The camera CM1, the camera CM2, and the camera CM3 track the object in this order.
The camera CM1 tracks the object from the time point t1 to the time point t3 and widely monitors the imaging target space S at other time points.
The camera CM2 tracks the object from time t4 to time t6 and widely monitors the imaging target space S at other time points.

The camera CM3 tracks the object from time t7 to time t9 and widely monitors the imaging target space S at other time points.
From time t3 to time t4 and from time t6 to time t7
The optical axis position and object occupancy must be passed on to the next camera.
The camera CM1 acquires a feature value F1 (hair) and a feature value F2 (eyes), the camera CM2 acquires a feature value F3 (nose) and a feature value F6 (trousers), and the camera CM3 acquires a feature value F4 (mouth). And the feature amount F5 (outerwear) has been acquired.
・ Other features (shoes and bags) were not acquired.

(Distance from camera to object)
Each camera can acquire “optical axis position = X ₃ , Y ₃ , Z ₃ ” etc. because each camera has its own coordinate value in the imaging target space S, pan / tilt angle, and target from the camera. This is because the distance to the object is recognized. Currently, commercially available cameras include a camera that can measure the distance to an object (a three-dimensional camera) and a camera that does not (a two-dimensional camera). Even if a two-dimensional camera is used, the imaging control device 1 stores the three-dimensional map (reference numeral 31 in FIG. 1B) of the imaging target space S in the auxiliary storage device 15. Each camera can measure the distance to the target portion.

  For example, a plurality of product shelves are neatly arranged vertically and horizontally in a plan view in a store. All the product shelves have a predetermined height, a predetermined number of steps, and a predetermined product arrangement width. Each product placement space on the product shelf is hung with a label indicating the product name, price, and the like. Then, such a label is arranged in an orderly manner at a lattice point in the three-dimensional space. A three-dimensional coordinate value is assigned to each of these labels. Then, the camera can estimate the distance between itself and the target object based on a state in which the target object (human being) is hiding a certain label in an image captured by a certain camera.

  The envelope cuboid will be described with reference to FIG. The envelope cuboid 45 is a cuboid having the smallest volume that includes all the points constituting the object among the cuboids having sides in the coordinate axis direction of the imaging target space S. A technique for the camera to specify such an envelope cuboid 45 is widely known. The imaging control apparatus 1 can take over the pan / tilt angle and the angle of view to the next camera by specifying the envelope cuboid 45 instead of the method described in FIG.

(Processing procedure)
The sequence of the processing procedure will be described with reference to FIG. In FIG. 7, solid arrows indicate the flow of processing, and broken arrows indicate the flow of information. The “first camera” in FIG. 7 does not necessarily mean the camera CM1. The same applies to the “second camera” and the “third camera”. That is, FIG. 7 shows that there are three cameras that transmit and receive information when viewed from the imaging control apparatus 1. Of course, the number of cameras may be “2” or “4” or more.

  In step S201, the role determination unit 21 determines a main camera and a sub camera. Here, it is assumed that the role determination unit 21 determines the first camera as the main camera, and determines the second camera and the third camera other than the first camera as sub cameras.

  In step S202, the tracking instruction unit 22 transmits a tracking instruction to the first camera. The tracking instruction is an instruction for tracking an object and acquiring a feature amount of the object. The feature values here are, for example, “F1” to “F8” described above. At this time, the tracking instruction unit 22 transmits the imaging condition CN2 to the first camera as a specific condition for realizing the tracking instruction.

  In step S203, the tracking instruction unit 22 transmits a monitoring instruction to the second camera and the third camera. The monitoring instruction is an instruction for continuously imaging a predetermined area in the imaging target space S without tracking the object. At this time, the tracking instruction unit 22 transmits the imaging condition CN1 to the second camera and the third camera as specific conditions for realizing the monitoring instruction.

  In step S204, the tracking instruction unit 22 receives an image of an object from the first camera. The image of the object includes, for example, a feature amount F1. The tracking instruction unit 22 stores the feature value F1 as an image 32 (FIG. 1B) in time series in the auxiliary storage device 15 in association with the received time point.

  In step S205, the imaging state reception unit 23 receives a report and an imaging state from the first camera. The report is a report that the object has come out of its own field of view (view angle) without obtaining all the feature quantities “F1” to “F8” of the object. The imaging state is the optical axis position and the object occupancy when the object comes out of its own field of view. Note that the imaging state may be information that can identify the envelope cuboid described above.

  In step S206, the role redetermining unit 24 redetermines the next main camera and sub camera. Here, it is assumed that the role re-determination unit 24 re-determines the second camera as the main camera and re-determines the first camera and the third camera other than the second camera as sub-cameras.

  In step S207, the continuous tracking instruction unit 25 transmits a tracking instruction to the second camera. The tracking instruction is an instruction for tracking an object using an imaging state and acquiring a feature amount of the object. The feature amounts here are “F2” to “F8” (excluding F1 that has already been acquired). At this time, the continuous tracking instruction unit 25 transmits the imaging condition CN2 and the imaging state received in step S205 as specific conditions for realizing the tracking instruction to the second camera.

  In step S208, the continuous tracking instruction unit 25 transmits a monitoring instruction to the first camera and the third camera. The monitoring instruction is an instruction for continuously imaging a predetermined area in the imaging target space S without tracking the object. At this time, the continuous tracking instruction unit 25 transmits the imaging condition CN1 to the first camera and the third camera as a specific condition for realizing the monitoring instruction.

  In step S209, the continuous tracking instruction unit 25 receives an image of the object from the second camera. The image of the target object includes, for example, a feature amount F2. The continuous tracking instruction unit 25 stores the feature amount F2 as an image 32 in the auxiliary storage device 15 in time series in association with the received time point.

  In step S210, the imaging state reception unit 23 receives a report and an imaging state from the second camera. The report is a report that the object has come out of its own field of view (view angle) without acquiring all of the feature quantities “F2” to “F8” of the object.

In step S211, the role redetermining unit 24 redetermines the next main camera and sub camera. Here, it is assumed that the role re-determination unit 24 re-determines the third camera as the main camera and re-determines the first camera and the second camera other than the third camera as sub-cameras.
Subsequent step information processing is the same as the above-described step, and is therefore omitted.

(Repeat process)
As is apparent from the above, the imaging control apparatus 1 repeats the processes of steps S201 to S205 until all of the predetermined feature amounts are acquired. Depending on the imaging environment, it may be difficult to acquire a specific feature amount. Also, not all feature quantities are absolutely necessary. Therefore, the imaging control apparatus 1 may determine in advance whether or not to repeat the processes in steps S201 to S205 with a predetermined number of threshold values (threshold values) to be acquired in advance and using the threshold value as a determination criterion. . Further, the processing in steps S201 to S205 may be repeated until a specific type of feature amount (for example, “eyes” and “mouth”) is acquired.

(Direction of movement)
The imaging state reception unit 23 may receive the direction in which the object moves from the main camera. In addition, the role redetermining unit 24 may select a camera to be redetermined as the main camera next in accordance with the received direction. For example, assume that the camera CM2 is the main camera in FIG. Then, it is assumed that the object is moving from the left to the right on the screen of the camera CM 2 (moving from the right to the left in FIG. 1). In this case, the role redetermining unit 24 determines the camera CM1 as the next main camera. This is because, when the camera CM1 and the camera CM3 are compared, the camera CM1 is closer to the object and can detect the feature quantity with a higher resolution.

(Alarm etc.)
The above example is based on the premise that acquiring the feature quantity of the object itself is meaningful. For example, the case of acquiring a feature quantity of a person who visits a store for the purpose of market research corresponds to the above example. In this case, the imaging control device 1 is not searching for a specific person. However, the imaging control device 1 may search for a specific person. This is the case, for example, when searching for a person who has left something in a store or when searching for a criminal such as shoplifting. In this case, the auxiliary storage device 15 stores the feature amount of the search target person in advance. In addition, the tracking instruction unit 22 (or the continuous tracking instruction unit 25) compares the feature amount received from the main camera with the feature amount stored in advance. Then, when the compared feature amounts match with each other so as to satisfy a predetermined standard, the tracking instruction unit 22 (or the continuous tracking instruction unit 25) may perform the following processing.

An alarm is issued to the output device 13. The alarm may be text or voice.
The acquired feature value (human hair, eyes, etc. currently in the store) is displayed on the output device 13 as an image.
-Sends an alarm and an image of the acquired feature value to a device installed in an external organization such as the police through a network (not shown).

(Camera with built-in device)
In the above description, the example in which the imaging control device 1 has a configuration different from that of the camera CM1 or the like has been described. However, for example, the camera CM1 may include the imaging control device 1 in its own casing. Then, the camera CM1 transmits and receives instructions and the like between other cameras and itself.

(Main camera and sub camera)
In the above description, an example in which all the cameras other than the camera determined (re-determined) as the main camera among the plurality of cameras is determined (re-determined) as the sub-camera has been described. However, some cameras other than the camera determined (re-determined) as the main camera may be determined (re-determined) as the sub-camera. That is, there may be a camera that is not determined (re-determined) by either the main camera or the sub camera.

(Effect of embodiment)
The imaging control device of this embodiment has the following effects.
(1) Since the main camera takes over the imaging state to the next main camera, the next main camera can track the object quickly and with high resolution.
(2) Since the main camera takes over the information for determining the direction of the optical axis and the ratio of the object in the screen to the next main camera, the next main camera can quickly cut out the feature amount of the object. Can do.
(3) Since the main camera takes over the direction in which the object moves, the imaging control device can select the next main camera most suitable for imaging the moving object.
(4) Since the object is a human, the imaging control device can be used for a wide range of applications such as product market research and criminal investigation.
(5) It becomes possible to search for an object to be searched.
(6) The feature amount can be visually recognized from the accumulated image even if it is not in real time.
(7) A camera that does not have a function of measuring distance can be used.
(8) The camera with a built-in imaging control device can save space for installing the imaging control device.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files that realize each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.
In addition, the control lines and information lines are those that are considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. In practice, it may be considered that almost all the components are connected to each other.

DESCRIPTION OF SYMBOLS 1 Imaging control device 11 Central control device 12 Input device 13 Output device 14 Main memory (storage part)
15 Auxiliary storage device (storage unit)
DESCRIPTION OF SYMBOLS 16 Communication apparatus 21 Role determination part 22 Tracking instruction | indication part 23 Imaging state reception part 24 Role redetermination part 25 Continuation tracking instruction | indication part 31 Map 32 Image 33 Feature-value information 34 Imaging condition information 35 Imaging history information CM1, CM2, CM3 camera

Claims (8)

A role determination unit that determines a camera among a plurality of cameras as a main camera, and determines a part or all of the plurality of cameras other than the main camera as sub cameras,
A tracking instruction unit that tracks an object and transmits an instruction to acquire a predetermined number of feature quantities of the object to the main camera and transmits an instruction to continuously image a predetermined area to the sub camera. When,
A report indicating that the object has come out of the field of view of the main camera without obtaining a predetermined amount of features among the predetermined number of features, and the field of view of the main camera An imaging state reception unit that receives the imaging state of the main camera at the time of exiting from the main camera;
A role of re-determining a camera other than the determined main camera among the plurality of cameras as a main camera, and re-determining some or all of the plurality of cameras other than the re-determined main camera as sub cameras. A redetermination part;
The received imaging state is used to track the object, and an instruction to acquire a predetermined number of feature quantities for the object is transmitted to the re-determined main camera, and a predetermined area is continuously A continuous tracking instruction section for transmitting an instruction to image to the re-determined sub camera;
An imaging control apparatus for controlling the plurality of cameras. The imaging state is
Including information for determining the direction of the optical axis and the proportion of the object in the screen;
The imaging control apparatus according to claim 1. The imaging state reception unit
Receiving the moving direction of the object from the main camera;
The role redetermining unit
Selecting the main camera to be redetermined based on the received direction;
The imaging control device according to claim 2. The object is
Being human,
The feature amount is
Identifying any one of the human body part, clothing part and portable items;
The imaging control apparatus according to claim 3. The imaging control device includes:
A storage unit in which the feature amount of the specific object is stored in advance;
The tracking instruction unit or the continuous tracking instruction unit,
Receiving the feature quantity of the object from the determined or re-determined main camera;
Comparing the received feature quantity with the pre-stored feature quantity;
If the compared feature quantities match with each other to meet a predetermined standard, sending an alarm to an external device;
The imaging control device according to claim 4. The storage unit
Storing the imaging state and the feature amount in time series;
The imaging control apparatus according to claim 5. When the plurality of cameras do not acquire the distance between the objects,
The storage unit
Storing a three-dimensional map of the space captured by the plurality of cameras;
The main camera is
Generating the imaging state by estimating a distance between the main camera and the object based on the three-dimensional map;
The imaging control device according to claim 6.   A camera provided with the imaging control device according to any one of claims 1 to 7.

Images