Data transfer system

Abstract

To provide a data transfer system capable of properly checking a notable action which is detected in a monitoring video, while reducing a data transfer cost or the like.SOLUTION: A data transfer system 1 detects a feature point of an object action body Z shown in an original video Y, and determines whether or not a first notable action or a second notable action has been performed, on the basis of the detected feature point. When it is determined that the first notable action has been performed, a portion equivalent to the first notable action in the original video Y is regarded as a short clip video and is transmitted to a user terminal B over a cloud A. When it is determined that the second notable action has been performed, a push notification is transmitted to the user terminal B over the cloud A without transmitting the short clip video.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a data transfer system capable of appropriately confirming an attention behavior detected in a surveillance video while reducing data transfer costs and the like.

Background Art

[0002] Conventionally, when the behavior of an actor shown in a time-series image is different from a predetermined "normal behavior", a technique for extracting the actor who has performed the behavior different from the "normal behavior" from the time-series image is known. (For example, see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By using the above technique, it is conceivable to configure to transmit only the video related to the abnormal behavior to a user such as an administrator and prompt confirmation only when an abnormal behavior occurs in the video captured by a surveillance camera or the like.

[0005] By the way, these days, data management is often performed on the cloud for its convenience. When the above configuration is performed using the cloud, it is conceivable to transfer all video shootings in which an abnormal behavior is detected to the cloud and make the user view the video on the cloud. However, since it costs according to the transfer amount to transfer the video on the cloud, it will cost a lot if all the videos in which an abnormal behavior is detected are transferred.

[0006] Therefore, an object of the present invention is to provide a data transfer system capable of appropriately confirming a target action detected in surveillance video while reducing data transfer costs and the like.

Means for Solving the Problems

[0007] The present invention includes an acquisition unit that acquires an original video captured by a photographing unit, a video storage unit that stores the original video, a first target action with a high degree of confirmation required, and a second target action with a low degree of confirmation required. A target action storage unit that stores the above, a detection unit that detects feature points of a target moving body shown in the original video, and based on the detected feature points, whether the first target action or the second target action has been performed. A determination unit that determines whether or not, and when it is determined that the first target action has been performed, a portion of the original video corresponding to the first target action is used as a short clip video and transmitted to a user terminal via the cloud. When it is determined that the second target action has been performed, a transmission unit that transmits a push notification to the user terminal via the cloud without transmitting the short clip video is provided. A data transfer system is provided.

[0008] According to such a configuration, when it is determined that the first target action has been performed, a portion of the original video corresponding to the first target action is used as a short clip video and transmitted to the user terminal via the cloud. When it is determined that the second target action has been performed, a push notification is transmitted to the user terminal via the cloud without transmitting the short clip video, and the user terminal can access the original video stored in the video storage unit using the P2P communication method.

[0009] Also, from another aspect of the present invention, there is provided an acquisition unit that acquires an original video captured by a photographing means, a video storage unit that stores the original video, a first attention behavior with a high degree of confirmation required, and a second attention behavior with a low degree of confirmation required. A data transfer system is provided, comprising: an attention behavior storage unit that stores the above, a detection unit that detects feature points of a target behavior body shown in the original video, a pseudo-video generation unit that can generate a pseudo-video in which displacement of the detected feature points is caused in a pseudo behavior body, and a determination unit that determines whether the first attention behavior or the second attention behavior has been performed based on the detected feature points; and a transmission unit that, when it is determined that the first attention behavior has been performed, transmits a portion of the pseudo-video corresponding to the first attention behavior as a short-clip video to a user terminal via the cloud, and when it is determined that the second attention behavior has been performed, transmits a push notification to the user terminal via the cloud without transmitting the short-clip video.

[0010] Also, from another aspect of the present invention, there is provided a data transfer program and a data transfer method corresponding to the above data transfer system.

Advantages of the Invention

[0011] According to the data transfer system of the present invention, it is possible to appropriately confirm the attention behavior detected in the surveillance video while reducing data transfer costs and the like.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Best Mode for Carrying Out the Invention

[0013] Hereinafter, the data transfer system 1 according to an embodiment of the present invention will be described with reference to FIGS. 1-4.

[0014] As shown in FIGS. 1 and 2, the data transfer system 1 estimates the behavior of the target moving body Z shown in the original video Y (frames constituting the video in FIG. 2) captured by the photographing means X, and when a notable behavior such as an abnormal behavior is performed, notifies the user terminal B to that effect via the cloud A.

[0015] In the present embodiment, it is assumed that the data transfer system 1 is provided integrally with the photographing means X.

[0016] Examples of the user terminal B include a smartphone and a PC.

[0017] Also, in the present embodiment, a human is adopted as the target moving body Z, and for easy understanding, the target moving body Z is simply displayed only by its skeleton.

[0018] As shown in FIG. 3, the data transfer system 1 includes an acquisition unit 2, a video storage unit 3, a notable behavior storage unit 4, a detection unit 5, a determination unit 6, a transmission unit 7, and a permission unit 8.

[0019] The acquisition unit 2 acquires the original video Y captured by the photographing means X.

[0020] The video storage unit 3 stores the original video Y.

[0021] The notable behavior storage unit 4 stores a first notable behavior with a high degree of confirmation required and a second notable behavior with a low degree of confirmation required.

[0022] In this embodiment, the attention behavior memory unit 4 stores the displacement of the feature points of the acting body when the acting body performs an attention behavior. For example, when detecting joints as feature points, it is conceivable to store the displacements of a plurality of joints.

[0023] As the first attention behavior, for example, those that require quickly confirming the "degree of attention behavior" and the "appearance of the target acting body Z" such as "violence", "theft", and "fall" are conceivable.

[0024] On the other hand, as the second attention behavior, for example, those that do not require quickly confirming the "actual degree of attention behavior" and the "appearance of the target acting body Z" such as "the movement of a person in a wheelchair" and "the movement of a person with a white cane" are conceivable.

[0025] Also, like "entry into a restricted area", although the action itself is a normal action (walking), it may become an abnormal action due to the combination with the restricted area (whether the feature point is located within the restricted area), and such a combination also corresponds to the "attention behavior" of the present invention.

[0026] Also, as the first attention behavior or the second attention behavior, "foreshadowing behaviors" such as shoplifting and suicide may be stored.

[0027] Furthermore, the attention behavior stored in the attention behavior memory unit 4 may be determined by learning. In this case, a sample video (a plurality of sample time-series images) captured by the imaging means X installed to capture a predetermined range is acquired, and based on a large number of actions detected in the sample video, one or more "normal actions" in the predetermined range are determined. When the displacement of the feature points detected from the original video Y does not correspond to the "normal action", a configuration for determining that an attention behavior ("abnormal action" or "uncomfortable action") has been performed is conceivable. In this case, it can be regarded that the attention behavior memory unit 4 stores "actions not included in normal actions".

[0028] "Actions not included in normal behavior" can be set as either the first notable action or the second notable action. However, if you want to "check what kind of action the 'action not included in normal behavior' was", it is conceivable to set it as the first notable action. However, for example, "actions not included in normal behavior" can also be classified into the first notable action and the second notable action according to the degree of separation from "normal behavior" (such as how much the displacement of the detected feature points is separated from the displacement of "normal behavior").

[0029] Also, even for notable actions of the same classification, whether confirmation is required or not varies depending on conditions such as the monitoring location, so it can be set appropriately according to the conditions. For example, even for "entry into a restricted area", in the case of a place where unauthorized entry often occurs unintentionally, it is conceivable to set it as the first notable action for an entry of more than a predetermined time and as the second notable action for an entry of less than the predetermined time.

[0030] The detection unit 5 detects the feature points of the target moving body Z shown in the original video.

[0031] Although various things can be considered as feature points, in this embodiment, an example of detecting joints as feature points will be used for explanation.

[0032] When detecting joints as feature points, for example, the following methods can be considered.

[0033] First, store the "joint identification criteria" and the "moving body identification criteria" in a storage unit (which can be the notable action storage unit 4 or another storage unit).

[0034] The "joint identification criteria" are for identifying a plurality of joints of a human, and for each joint, they show the shape, direction, size, etc. for identifying each of them.

[0035] The "action identification criteria" indicate the "basic postures" of various variations of a human (such as "walking", "standing upright", etc.), the "range of motion of each joint", the "distance between each joint" in one human, and so on.

[0036] After detecting a plurality of joints corresponding to the above "joint identification criteria", by referring to the "action identification criteria" and identifying the plurality of joints included in one target action body Z, it becomes possible to identify the joints included in each of the target action bodies Z.

[0037] Note that in this embodiment, the feature points are detected collectively from a plurality of frames constituting the original video Y (after associating the time-series order), but they may also be detected individually for each frame.

[0038] The determination unit 6 determines whether the first attention action or the second attention action has been performed based on the detected feature points.

[0039] For example, when the displacement of the detected feature points coincides with the displacement of the feature points of the first attention action or the second attention action stored in the attention action memory unit 4 by a predetermined amount or more, it can be considered that "the said attention action has been performed" is determined.

[0040] When it is determined that the first attention action has been performed, the transmission unit 7 transmits, as a short-clip video, the portion of the original video Y corresponding to the first attention action to the user terminal B via the cloud A. When it is determined that the second attention action has been performed, a push notification is transmitted to the user terminal B via the cloud A without transmitting the short-clip video.

[0041] As a result, the original video Y is transmitted to the user terminal B only in the case of the first notable action with a high degree of confirmation required, and when the second notable action with a low degree of confirmation required is performed, only a push notification is sent. Therefore, it is possible to significantly reduce the costs related to data transfer from the data transfer system 1 to the cloud A, data transfer from the cloud A to the user terminal B, and data storage in the cloud A. In addition, since the original video Y is not transmitted except for the first notable action with a high degree of confirmation required, the risk of personal information leakage is also reduced.

[0042] By the way, for the second notable action for which a push notification has been sent and other actions as well, cases where the user wants to view the actual video are assumed.

[0043] Therefore, in the present embodiment, the user terminal B can confirm the original video Y stored in the video storage unit 3 using the P2P communication method.

[0044] In order for the user terminal B to be able to confirm the original video Y stored in the video storage unit 3 using the P2P communication method, the user terminal B only needs to acquire the IP address on the data transfer system 1 side in advance.

[0045] As a result, while eliminating the risk of personal information leakage through the cloud A, it becomes possible for the user to confirm the original video Y.

[0046] However, if it is simply made possible for the user terminal B to confirm the original video Y stored in the video storage unit 3 using the P2P communication method, it may become a state that allows intrusion into the data transfer system 1 (especially the video storage unit 3) from various networks, and a security risk may occur.

[0047] Therefore, in this embodiment, when it is determined that the second attention action has been performed, the permission unit 8 permits the user terminal B to check, using the P2P communication method, the portion of the original video Y corresponding to the second attention action. As methods for this permission, for example, the following two patterns can be considered. (1) Pattern 1

[0048] First, when it is determined that the second attention action has been performed, a random key generation unit (not shown) generates a random key and transmits it to both the data transfer system 1 side and the user terminal B. In this embodiment, it is assumed that the random key generation unit is arranged on the cloud A and has previously acquired the IP address of the data transfer system 1 side and the IP address of the user terminal B.

[0049] Then, when the user terminal B requests access to the portion of the original video Y corresponding to the second attention action to the data transfer system 1 (the IP address of the data transfer system 1) together with the random key, if the random keys match, the permission unit 8 permits the user terminal B to check, using the P2P communication method, the portion of the original video Y corresponding to the second attention action.

[0050] Specifically, a configuration can be considered in which the video storage unit 3 functions as the permission unit 8 and permits access from the user terminal B to the video storage unit 3 when the random keys match.

[0051] Also, a configuration can be considered in which the transmission unit 7 functions as the permission unit 8 and transmits, to the user terminal B, the portion of the original video Y stored in the video storage unit 3 corresponding to the second attention action when the random keys match.

[0052] Note that when the P2P communication is once released, the above permission is also released. When performing P2P communication again, it is preferable to generate a random key again and require permission again. Also, it is preferable to perform the above-described communication in an encrypted manner as a whole. (2) Pattern 2

[0053] First, when it is determined that the second attention action has been performed, a transmission unit (not shown) transmits the IP address of the user terminal B to the data transfer system 1 side. In the present embodiment, it is assumed that the transmission unit is arranged on Cloud A and has acquired in advance the IP address on the data transfer system 1 side and the IP address of the user terminal B.

[0054] Then, when the user terminal B requests access to the portion of the original video Y corresponding to the second attention action to the data transfer system 1 (the IP address of the data transfer system 1), if the IP address from which the access request has been made matches the previously transmitted IP address, the permission unit 8 permits the user terminal B to confirm the portion of the original video Y corresponding to the second attention action using the P2P communication method.

[0055] Specifically, it is conceivable that the video storage unit 3 functions as the permission unit 8 and permits access from the user terminal B to the video storage unit 3 when the IP addresses match.

[0056] Also, it is conceivable that the transmission unit 7 functions as the permission unit 8 and transmits to the user terminal B the portion of the original video Y stored in the video storage unit 3 corresponding to the second attention action when the IP addresses match.

[0057] Note that when the P2P communication is once released, the above permission is also released. When performing the P2P communication again, it is preferable to transmit the IP address again and require re - permission. Also, it is preferable to perform the above - described communication in an encrypted manner as a whole.

[0058] Thereby, access to videos other than the second attention action cannot be performed by the P2P communication method, and the access by the P2P communication method is limited. Thus, risks such as malware infection and network bandwidth congestion caused by using the P2P communication method are reduced.

[0059] Next, with reference to the flowchart of FIG. 4, the data transfer process by the data transfer system 1 of the present embodiment will be described.

[0060] First, when the original video Y is acquired (S1), the feature points of the target actor Z shown in the original video Y are detected (S2).

[0061] Subsequently, based on the feature points detected in S2, it is determined whether the first attention action or the second attention action has been performed (S3).

[0062] When it is determined that the first attention action has been performed (S3: first attention action), the short - clip video is transmitted to the user terminal B via the cloud A (S4).

[0063] On the other hand, when it is determined that the second attention action has been performed (S3: second attention action), a push notification is transmitted to the user terminal B via the cloud A (S5).

[0064] As described above, in the data transfer system 1 according to the present embodiment, when it is determined that the first attention action has been performed, the part of the original video Y corresponding to the first attention action is used as a short - clip video and transmitted to the user terminal B via the cloud A. When it is determined that the second attention action has been performed, a push notification is transmitted to the user terminal B via the cloud A without transmitting the short - clip video, and the user terminal B can access the original video Y stored in the video storage unit 3 using the P2P communication method.

[0065] According to such a configuration, for the user terminal B, the original video Y is transmitted only in the case of the first attention action with a high degree of confirmation required. When the second attention action with a low degree of confirmation required is performed, only a push notification is transmitted. Therefore, it is possible to significantly reduce the costs related to data transfer from the data transfer system 1 to the cloud A, data transfer from the cloud A to the user terminal B, and data storage in the cloud A. In addition, since the original video Y is not transmitted except for the first attention action with a high degree of confirmation required, the risk of personal information leakage is reduced.

[0066] Also, in the data transfer system 1 according to the present embodiment, the user terminal B can confirm the original video Y stored in the video storage unit 3 using the P2P communication method.

[0067] According to such a configuration, since the user terminal B can confirm the original video Y stored in the video storage unit 3 using the P2P communication method, the user can confirm the original video Y while eliminating the risk of personal information leakage through the cloud A.

[0068] Also, in the data transfer system 1 according to the present embodiment, when it is determined that the second attention action has been performed, the permission unit 8 permits the user terminal B to confirm, using the P2P communication method, the portion of the original video Y corresponding to the second attention action.

[0069] According to such a configuration, it is not possible to access videos other than the second attention action by the P2P communication method, and the access by the P2P communication method is limited. Therefore, risks such as malware infection and network bandwidth compression caused by using the P2P communication method are reduced.

[0070] Note that the data transfer system of the present invention is not limited to the above-described embodiment, and various modifications and improvements are possible within the scope described in the claims.

[0071] For example, in the above embodiment, when it is determined that the first attention behavior has occurred, the original video Y (short clip video) is transmitted to the user terminal B. However, a configuration may be adopted in which a pseudo-video is transmitted instead of the original video Y.

[0072] In this case, as shown in FIG. 5, the data transfer system 1 further includes a pseudo-video generation unit 9 capable of generating a pseudo-action video in which the detected displacement of the feature points is caused to be performed by a pseudo-action body. When it is determined that the first attention behavior has occurred, the transmission unit 7 transmits a pseudo-video of the portion corresponding to the first attention behavior as a short clip video to the user terminal B via the cloud A. When it is determined that the second attention behavior has occurred, a configuration may be considered in which a push notification is transmitted to the user terminal B via the cloud A without transmitting the short clip video.

[0073] As the pseudo-action body, those that cannot identify personal information (those that are difficult to identify), for example, those that show only the bones of the action body Z or those replaced with a doll or the like, can be considered.

[0074] According to such a configuration, since a pseudo-video instead of the original video Y is transmitted to the user terminal B via the cloud A, while reducing the risk of leakage of personal information (videos including faces, etc.), it is possible for the user to quickly grasp "what kind of (to what extent) problem has occurred" from the pseudo-video.

[0075] Also, in the above embodiment, the photographing means X is provided integrally with the data transfer system 1, but it does not exclude the case where it is a separate body.

[0076] Also, in the above embodiment, each member is provided integrally, but it does not exclude the case where, for example, some members are provided at different locations (such as the cloud A).

[0077] For example, it is conceivable to provide the attention behavior memory unit 4 and the determination unit 6 on the cloud A, provide the other members at the edge, have the transmission unit 7 transmit the feature points detected by the detection unit 5 onto the cloud A, and have the determination unit 6 determine on the cloud A whether or not an attention behavior has occurred.

[0078] And when it is determined that the first attention behavior has been performed, a notice to that effect is transmitted from the cloud A to the edge, and thereafter, the short clip video is transmitted from the transmission unit 7 to the user terminal B via the cloud A.

[0079] In addition, in the above embodiment, an example of data transfer to one user terminal B for the video captured by one imaging means X has been described, but it goes without saying that data may be transferred to a plurality of user terminals B for the videos captured by a plurality of imaging means X.

[0080] However, when using the data transfer system 1 for a plurality of users, it is necessary to associate the imaging means X with the user terminal B so that transmission of short clip videos and permission to access the original video using the P2P communication method are not performed for another user (a user belonging to another organization).

[0081] Further, the present invention is also applicable to a program and method corresponding to the processing performed by each member as a controller, and a recording medium storing the program. In the case of a recording medium, the program is installed in a computer or the like. Here, the recording medium storing the program may be a non-transitory recording medium. Examples of non-transitory recording media include, but are not limited to, CD-ROMs.

Explanation of Reference Numerals

[0082] 1 Data transfer system 2 Acquisition unit 3 Video memory unit 4 Attention behavior memory unit 5 Detection unit 6 Determination unit 7 Transmission unit 8 Permit Department 9. Pseudo image generation section A Cloud B User terminal X Shooting method Y original video Z Target Action

Claims

1. An acquisition unit that acquires an original video captured by a photographing means; A video storage unit that stores the original video; A attention behavior storage unit that stores a first attention behavior with a high degree of confirmation required and a second attention behavior with a low degree of confirmation required; A detection unit that detects feature points of a target behavior body shown in the original video; A determination unit that determines whether the first attention behavior or the second attention behavior has been performed based on the detected feature points; When it is determined that the first attention behavior has been performed, a part of the original video corresponding to the first attention behavior is used as a short clip video and transmitted to a user terminal via the cloud. When it is determined that the second attention behavior has been performed, a transmission unit that transmits a push notification to the user terminal via the cloud without transmitting the short clip video; A data transfer system characterized by comprising the above.

2. The data transfer system according to claim 1, wherein the user terminal can confirm the original video stored in the video storage unit using a P2P communication method.

3. The data transfer system according to claim 1, further comprising a permission unit that permits the user terminal to confirm, using a P2P communication method, a part of the original video corresponding to the second attention behavior when it is determined that the second attention behavior has been performed.

4. An acquisition unit that acquires an original video captured by a photographing means; A video storage unit that stores the original video; A attention behavior storage unit that stores a first attention behavior with a high degree of confirmation required and a second attention behavior with a low degree of confirmation required; A detection unit that detects feature points of a target behavior body shown in the original video; A pseudo-video generation unit capable of generating a pseudo-action video in which the displacement of the detected feature points is caused by a pseudo-action body; A determination unit that determines whether the first attention behavior or the second attention behavior has been performed based on the detected feature points; When it is determined that the first attention behavior has been performed, a part of the pseudo-action video corresponding to the first attention behavior is used as a short clip video and transmitted to a user terminal via the cloud. When it is determined that the second attention behavior has been performed, a transmission unit that transmits a push notification to the user terminal via the cloud without transmitting the short clip video; A data transfer system characterized by comprising

5. A program executed on a computer storing a first attention behavior with a high degree of confirmation required and a second attention behavior with a low degree of confirmation required, The step of acquiring the original video captured by the photographing means, The step of storing the original video, The step of detecting feature points of the target moving object shown in the original video, The step of determining whether the first attention behavior or the second attention behavior has been performed based on the detected feature points, When it is determined that the first attention behavior has been performed, a portion of the original video corresponding to the first attention behavior is used as a short clip video and transmitted to the user terminal via the cloud. When it is determined that the second attention behavior has been performed, a push notification is transmitted to the user terminal via the cloud without transmitting the short clip video. A data transfer program characterized by comprising

6. A method executed on a computer storing a first attention behavior with a high degree of confirmation required and a second attention behavior with a low degree of confirmation required, The step of acquiring the original video captured by the photographing means, The step of storing the original video, The step of detecting feature points of the target moving object shown in the original video, The step of determining whether the first attention behavior or the second attention behavior has been performed based on the detected feature points, When it is determined that the first attention behavior has been performed, a portion of the original video corresponding to the first attention behavior is used as a short clip video and transmitted to the user terminal via the cloud. When it is determined that the second attention behavior has been performed, a push notification is transmitted to the user terminal via the cloud without transmitting the short clip video. A data transfer method characterized by comprising

7. A program executed on a computer storing a first attention behavior with a high degree of confirmation required and a second attention behavior with a low degree of confirmation required, The step of acquiring the original video captured by the photographing means, The step of storing the original video, The step of detecting feature points of the target moving object shown in the original video, generating a pseudo-action video that causes the detected displacement of the feature points to be performed by a pseudo-agent; determining whether the first attention action or the second attention action has been performed based on the detected feature points; when it is determined that the first attention action has been performed, transmitting, as a short clip video, the part of the pseudo-action video corresponding to the first attention action to a user terminal via a cloud; when it is determined that the second attention action has been performed, transmitting a push notification to the user terminal via the cloud without transmitting the short clip video; A data transfer program characterized by comprising the above.

8. A method executed by a computer storing a first attention action with a high degree of confirmation required and a second attention action with a low degree of confirmation required, acquiring an original video captured by a photographing means; storing the original video; detecting feature points of a target agent shown in the original video; generating a pseudo-action video that causes the detected displacement of the feature points to be performed by a pseudo-agent; determining whether the first attention action or the second attention action has been performed based on the detected feature points; when it is determined that the first attention action has been performed, transmitting, as a short clip video, the part of the pseudo-action video corresponding to the first attention action to a user terminal via a cloud; when it is determined that the second attention action has been performed, transmitting a push notification to the user terminal via the cloud without transmitting the short clip video; A data transfer method characterized by comprising the above.

Images