CN111988563B

CN111988563B - Multi-scene video monitoring method and device

Info

Publication number: CN111988563B
Application number: CN202010682724.4A
Authority: CN
Inventors: 张珈毓; 陆振善; 李浙伟
Original assignee: Zhejiang Dahua Technology Co Ltd
Current assignee: Zhejiang Dahua Technology Co Ltd
Priority date: 2020-07-15
Filing date: 2020-07-15
Publication date: 2021-08-31
Anticipated expiration: 2040-07-15
Also published as: WO2022012002A1; CN111988563A

Abstract

The application provides a multi-scene video monitoring method and a device, wherein the method is applied to a monitoring device for monitoring N scenes through an analysis monitoring module for monitoring the number of paths by M paths, wherein N is greater than M; the method comprises the following steps: acquiring video stream data of N scenes; receiving a maximum value Y of the number of scenes in which a predetermined target may exist simultaneously in the N scenes; extracting a frame of picture from video stream data of each scene at intervals of a first preset time length to form a first picture stream; judging whether a preset target exists in the pictures in the first picture stream through an X-path analysis monitoring module; wherein, the number of the X paths is not more than the difference value of M and Y; when the pictures in the first picture stream have the preset target, analyzing and monitoring video stream data of a scene corresponding to the pictures with the preset target. The method does not need to keep the analysis starting state of each video channel all the time, can save resources and greatly improve the intelligent analysis performance.

Description

Multi-scene video monitoring method and device

Technical Field

The invention relates to the technical field of intelligent video monitoring, in particular to a multi-scene video monitoring method and device.

Background

With the wide application of video monitoring services, abnormal events in some scenes are often monitored by using the video monitoring services.

At present, in order to monitor abnormal events in some scenes, a plurality of cameras are generally arranged in the scenes to acquire real-time videos in corresponding scenes, and then a plurality of independent real-time video channels are respectively monitored and analyzed; specifically, the tracking relationship of the targets in the front frame and the rear frame in the video channel is utilized to identify whether the image picture of each real-time video channel has an abnormal event violating the established rule or not.

However, the existing method needs to keep analyzing the on state of each video channel all the time, so that for some video channels with low target occurrence frequency or video channels without targets, resource waste is caused.

Disclosure of Invention

The multi-scene video monitoring method and the multi-scene video monitoring device can solve the problem that in the prior art, each video channel needs to be kept in an analysis starting state all the time, and then resource waste is caused for some video channels with low target occurrence frequency or video channels without targets.

In order to solve the above technical problem, the first technical solution adopted by the present application is: the method is applied to a monitoring device for monitoring N scenes by an analysis monitoring module for monitoring the number of paths by M paths, wherein N is greater than M; the method comprises the following steps: acquiring video stream data of N scenes; receiving a maximum value Y of the number of scenes in which a predetermined target may exist simultaneously in the N scenes; extracting a frame of picture from video stream data of each scene at intervals of a first preset time length to form a first picture stream; judging whether a preset target exists in the pictures in the first picture stream through an X-path analysis monitoring module; wherein, the number of the X paths is not more than the difference value of M and Y; when the pictures in the first picture stream have the preset target, analyzing and monitoring video stream data of a scene corresponding to the pictures with the preset target.

In order to solve the above technical problem, the second technical solution adopted by the present application is: providing a multi-scene video monitoring device, wherein the device is provided with M monitoring paths; the device comprises a video acquisition module, a receiving module, a first picture stream generation module, an M-path analysis monitoring module and a control module; the video acquisition module is used for acquiring video stream data of N scenes; wherein N > M; the receiving module is used for receiving the maximum value Y of the scene quantity of the preset targets which possibly exist in the N scenes at the same time; the first picture stream generation module is used for extracting a frame of picture from video stream data of each scene at intervals of a first preset time length to form a first picture stream; the M paths of analysis monitoring modules are used for analyzing and monitoring the picture stream or the video stream; the control module is used for controlling the video acquisition module to acquire video stream data of the N scenes; the control receiving module receives the scene quantity Y with the preset target in the N scenes; controlling a first picture stream generation module to extract one frame of picture from video stream data of each scene at intervals of a first preset time length to form a first picture stream; controlling an X-path analysis monitoring module in the M-path analysis monitoring modules to judge whether a preset target exists in the pictures in the first picture stream; wherein, the number of the X paths is not more than the difference value of M and Y; when the pictures in the first picture stream have the preset target, analyzing and monitoring video stream data of a scene corresponding to the pictures with the preset target.

The method is applied to a monitoring device for monitoring N scenes through an analysis monitoring module for monitoring the number of paths by M paths, wherein N is greater than M; the method comprises the steps of obtaining video stream data of N scenes and receiving the maximum value Y of the number of the scenes in the N scenes, wherein the scenes may have a preset target at the same time; then extracting a frame of picture from the video stream data of each scene at intervals of a first preset time length to form a first picture stream; then, judging whether a preset target exists in the pictures in the first picture stream by using an X-path analysis monitoring module; when a preset target exists in a picture in the first picture stream, analyzing and monitoring video stream data of a scene corresponding to the picture with the preset target; therefore, the method not only can monitor abnormal conditions in a plurality of scenes, but also can monitor N paths of video stream data in real time under the condition of only M paths of monitoring capacity because the method extracts one frame of picture from each scene to form a picture stream and only analyzes and monitors the video stream corresponding to the picture in the picture stream with a preset target; compared with the prior art, the method needs to use N monitoring paths to respectively monitor N video stream data in real time, does not need to keep the analysis starting state of each video channel all the time, can save resources and greatly improve the intelligent analysis performance.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a multi-scene video monitoring method according to a first embodiment of the present application;

FIG. 2 is a sub-flowchart of step S13 in FIG. 1;

FIG. 3 is a sub-flowchart of step S134 in FIG. 2;

FIG. 4 is a sub-flowchart of step S15 in FIG. 1;

fig. 5 is a flowchart of a multi-scene video monitoring method according to a second embodiment of the present application;

fig. 6 is a flowchart of a multi-scene video monitoring method according to a third embodiment of the present application;

FIG. 7 is a sub-flowchart of step S36 in FIG. 6;

fig. 8 is a flowchart of a multi-scene video monitoring method according to a fourth embodiment of the present application;

fig. 9 is a flowchart of a multi-scene video monitoring method according to a fifth embodiment of the present application;

fig. 10 is a flowchart of a multi-scene video monitoring method according to a sixth embodiment of the present application;

fig. 11 is a flowchart of a multi-scene video monitoring method according to a seventh embodiment of the present application;

fig. 12 is a flowchart of a multi-scene video monitoring method according to an eighth embodiment of the present application;

fig. 13 is a schematic structural diagram of a multi-scene video monitoring apparatus according to an embodiment of the present application;

fig. 14 is a schematic structural diagram of a multi-scene video monitoring apparatus according to another embodiment of the present application.

Detailed Description

The present application will be described in further detail with reference to the following drawings and examples. It is to be noted that the following examples are only illustrative of the present application, and do not limit the scope of the present application. Likewise, the following examples are only some examples and not all examples of the present application, and all other examples obtained by a person of ordinary skill in the art without any inventive work are within the scope of the present application.

The terms "first", "second" and "third" in this application are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any indication of the number of technical features indicated. Thus, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of the feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless explicitly specifically limited otherwise. All directional indications (such as up, down, left, right, front, and rear … …) in the embodiments of the present application are only used to explain the relative positional relationship between the components, the movement, and the like in a specific posture (as shown in the drawings), and if the specific posture is changed, the directional indication is changed accordingly. The terms "comprising" and "having" and any variations thereof in the embodiments of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or may alternatively include other steps or elements inherent to such process, method, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments. The present application will be described in detail with reference to the accompanying drawings and examples.

Referring to fig. 1, fig. 1 is a flowchart of a multi-scene video monitoring method according to a first embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, where the method is applied to a monitoring device that monitors N scenes through an analysis monitoring module that monitors the number of paths by M paths, and a maximum value Y of the number of scenes in which a predetermined target exists in the N scenes is known, where N > M; in one embodiment, N is much greater than M; for example, N may be 300 and M may be 15. For example, a monitoring device of N scenes monitors a predetermined number Y of unmanned delivery trucks, where the number Y of scenes with a predetermined target in the N scenes is the maximum value of the number of delivery trucks that may be delivered simultaneously.

Specifically, the method comprises the following steps:

step S11: video stream data of N scenes is acquired.

Specifically, video stream data of N scenes are obtained through a video obtaining module; the video acquisition module specifically comprises a plurality of video acquisition units, the video acquisition units correspond to each scene one by one, namely, each scene is provided with a video acquisition unit so as to acquire video stream data of each scene through the corresponding video acquisition unit in each scene, and further acquire the video stream data of N scenes through N video acquisition units in the video acquisition module; the video acquisition unit may be a camera.

Step S12: a maximum value Y of the number of scenes in which a predetermined object may exist simultaneously among the N scenes is received.

It should be noted that, in this embodiment, Y is a preset value.

Step S13: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Specifically, referring to fig. 2, fig. 2 is a sub-flowchart of step S13 in fig. 1; step S13 specifically includes:

step S131: a first predetermined length of time is configured.

Specifically, in the process of configuring the first preset time length, the configured first preset time length is verified at the same time to ensure that the first preset time length is far longer than the frame interval time length of the video stream, so that the performance optimization is realized; the frame interval duration of the video stream refers to the interval duration of two adjacent frames in the video stream; in a specific implementation process, the first preset time length is an empirical value, and can be specifically configured according to an actual scene, such as 1-5 seconds; in one embodiment, the first predetermined length of time is 2 seconds.

S132: and judging whether the first preset time length is less than the interval time length of the I frame or not.

Specifically, if the first predetermined time length is less than the I-frame interval time length, step S133 is performed; if the first predetermined duration is greater than or equal to the I frame interval duration, the process proceeds to step S134.

S133: and decoding video stream data of each scene and extracting a frame of YUV pictures at intervals of a first preset time length to form a first picture stream.

S134: and extracting I frames in the video stream data of each scene at intervals of a first preset time length, decoding and converting YUV pictures to form a first picture stream.

Specifically, in step S134, decoding only the I frame saves more resources than directly decoding each video stream.

Specifically, referring to fig. 3, fig. 3 is a sub-flowchart of step S134 in fig. 2; step S134 further includes:

step S1341: and judging whether the video stream data in the first preset time comprises only one I frame.

Specifically, if the video stream data within the first predetermined time period only includes one I frame, step S1342 is performed; if the video stream data includes a plurality of I frames within the first predetermined time period, the process proceeds to step S1343.

Step S1342: and directly decoding the current I frame and converting YUV pictures to form a first picture stream.

Step S1343: one I frame is extracted from the plurality of I frames for decoding and converting YUV pictures to form a first picture stream.

It should be noted that the above mentioned I frame interval duration specifically refers to an interval duration between two adjacent I frames in the video stream data.

It can be understood that the first picture stream is assembled by extracting a frame of picture from the video stream data of each scene; and a new first picture stream is formed every interval of a first predetermined length.

Step S14: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, a target recognition unit in the X-path analysis monitoring module is used to perform target recognition on the pictures in the first picture stream, and determine whether the pictures in the first picture stream have a predetermined target. If the pictures in the first picture stream have the predetermined target, executing step S15; if the picture in the first picture stream does not have the predetermined target, the video stream data of the plurality of scenes is not analyzed and the process returns to step S11.

Specifically, in this embodiment, the first picture stream is set to X-way. The number of the X paths can be specifically selected according to the first picture stream, the first predetermined time, the specific hardware processing capability of each path, and the capability of the target identification unit; in the specific implementation process, the number of the X paths is not more than the difference between M and Y. It can be understood that the larger N, the larger the first picture stream, and the larger the required number of paths. In this embodiment, X is 3 or less.

Step S15: and analyzing and monitoring video stream data of a scene corresponding to the picture with the preset target.

Further, in a specific embodiment, when the pictures in the first picture stream have the predetermined target, the method further includes obtaining the number Y' of the first pictures in the first picture stream having the predetermined target; then judging whether the number Y' of the first pictures with the preset target in the pictures in the first picture stream is more than M-X; if the number Y' of the first pictures with the preset targets in the pictures in the first picture stream is greater than M-X, sending an abnormal prompt; if the number Y' of the first pictures with the predetermined target existing in the pictures in the first picture stream is not greater than M-X, the above step S15 is executed; specifically, the video stream data of the scene corresponding to the Y 'first picture with the predetermined target is analyzed and monitored by the Y' path analyzing and monitoring module.

Specifically, referring to fig. 4, fig. 4 is a sub-flowchart of step S15 in fig. 1; step S15 specifically includes:

step S151: and decoding video stream data of a scene corresponding to the picture with the preset target, and extracting one frame of picture at intervals of a second preset time length to form a second picture stream corresponding to the video stream data of the scene corresponding to the picture with the preset target.

Wherein the second predetermined duration is less than the first predetermined duration.

For example, when there are Y scenes corresponding to the pictures with the predetermined target, decoding the video stream data of each scene by a decoding frame extraction unit in the analysis monitoring module, and extracting one frame of picture from the video stream data of each scene at intervals of a second predetermined time length to form a second picture stream corresponding to the video stream data of the current scene; it can be understood that the second picture stream is obtained by extracting one frame of picture at a second predetermined time interval from the video stream data of one scene, and the scenes corresponding to the Y pictures with the predetermined target correspond to the Y second picture streams.

In a specific embodiment, step S151 further includes checking the number of paths of the second picture stream to ensure that the number of video paths that can be monitored in an abnormal state is not greater than M-X, and if the number of video paths monitored in an abnormal state is greater than M-X, indicating that the monitoring capability of the apparatus is exceeded, then throwing out the abnormal information.

Step S152: and analyzing and monitoring the second picture stream and judging whether the pictures in the video stream data of each scene are abnormal or not.

Specifically, the second picture stream is analyzed and monitored through an analysis and monitoring unit, and whether the picture in the video stream data of each scene is abnormal or not is judged; when there is no abnormality in the pictures in the video stream data of each scene, step S153 is executed; when there is an abnormality in the picture in the video stream data of each scene, step S154 is performed.

Specifically, whether a preset target in the picture in the second picture stream is abnormal or not can be judged through target identification, and whether the picture in the video stream data of each scene is abnormal or not is further judged; the specific determination of whether the predetermined target is abnormal or not refers to whether the behavior of the predetermined target is abnormal or not, for example, whether the predetermined target has abnormal behavior such as fighting or not.

Step S153: and returning to execute the acquisition of the video stream data of the N scenes.

Step S154: and sending an exception report.

In particular, the exception report may be an alarm signal.

Further, in a specific embodiment, step S15 further includes determining whether the corresponding predetermined target in the scene disappears; and when the corresponding preset target in the scene disappears, closing the analysis monitoring module corresponding to the video stream data of the scene with the disappeared preset target.

The multi-scene video monitoring method provided by this embodiment is applied to a monitoring device that monitors N scenes through an analysis monitoring module that monitors the number of paths by M paths, and the number Y of scenes in which a predetermined target exists in the N scenes is known, where N > M. Specifically, the method comprises the steps of obtaining video stream data of N scenes and receiving a maximum value Y of the number of scenes in the N scenes, wherein the scenes may have a preset target at the same time; then extracting a frame of picture from the video stream data of each scene at intervals of a first preset time length to form a first picture stream; then, judging whether a preset target exists in the pictures in the first picture stream by using an X-path analysis monitoring module; when a preset target exists in a picture in the first picture stream, analyzing and monitoring video stream data of a scene corresponding to the picture with the preset target; therefore, the method not only can monitor abnormal conditions in a plurality of scenes, but also can monitor N paths of video stream data in real time under the condition of only M paths of monitoring capacity because the method extracts one frame of picture from each scene to form a picture stream and only analyzes and monitors the video stream corresponding to the picture in the picture stream with a preset target; compared with the prior art, the method needs to use N monitoring paths to respectively monitor N video stream data in real time, does not need to keep the analysis starting state of each video channel all the time, can save resources and greatly improve the intelligent analysis performance.

Referring to fig. 5, fig. 5 is a flowchart of a multi-scene video monitoring method according to a second embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, where N scenes are monitored by an analysis monitoring module with Y + X being M paths of monitoring paths, also based on the fact that Y is known in advance; wherein N > M; specifically, the method comprises the following steps:

step S21: video stream data of N scenes is acquired.

Step S22: a maximum value Y of the number of scenes in which a predetermined object may exist simultaneously among the N scenes is received.

Step S23: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Step S24: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, if the picture in the first picture stream has the predetermined target, the target identification unit in the X-path analysis monitoring module notifies the switching module of the channel ID where the picture with the predetermined target appears, and executes step S25; if the picture in the first picture stream does not have the predetermined target, the video stream data of the plurality of scenes is not analyzed and the process returns to step S21.

Specifically, the specific implementation process of the step S21 to the step S24 is the same as or similar to the specific implementation process of the step S11 to the step S14 in the multi-scene video monitoring method provided in the first embodiment, and the same or similar technical effects can be achieved.

Step S25: and judging whether the video stream data of the scene corresponding to the picture with the preset target is analyzed and monitored.

Specifically, step S25 is executed by the switching module; specifically, after receiving the channel ID notification sent by the target identification unit and including the picture of the predetermined target, the switching module executes step S25; specifically, if the video stream data of the scene corresponding to the picture with the predetermined target is already being analyzed and monitored, step S26 is executed; if there is no video stream data of the scene corresponding to the picture of the predetermined target yet to be analyzed and monitored, step S27 is executed.

Step S26: and continuously analyzing and monitoring the video stream data of the scene corresponding to the picture with the preset target.

Step S27: and switching the video stream data of the scene corresponding to the picture with the preset target and starting analyzing and monitoring the video stream data of the scene corresponding to the picture with the preset target.

Specifically, the switching module is used to switch the video stream data of the scene corresponding to the picture with the predetermined target to the number of idle analysis channels (i.e., the number of idle analysis channels), and start analyzing and monitoring the video stream data of the scene corresponding to the picture with the predetermined target.

Specifically, for the specific implementation process of analyzing and monitoring the video stream data of the scene corresponding to the picture with the predetermined target, reference may be made to the specific implementation process of step S15 in the multi-scene video monitoring method provided in the first embodiment, which is the same or similar, and the same or similar technical effect may be achieved, and reference may be specifically made to the description of the related text, which is not repeated herein.

Compared with the multi-scene video monitoring method provided by the first embodiment, the multi-scene video monitoring method provided by the embodiment further switches the video stream according to the notification, so that the video stream data of the scene corresponding to the picture with the predetermined target can be directly analyzed and monitored, further resources can be further saved, the intelligent analysis performance is improved, and the normal abnormal event detection rate is still maintained on the premise of optimizing the performance.

Referring to fig. 6, fig. 6 is a flowchart of a multi-scene video monitoring method according to a third embodiment of the present application; it should be noted that, in the process of performing object identification on a picture, although the time consumption for identifying an object by using the picture is short and the speed is high, the time consumption still exists and there is an interval between pictures, therefore, in the multi-scene video monitoring method provided in this embodiment, in order to ensure that before a picture of a predetermined object is first monitored, if the predetermined object has appeared, the predetermined object can be recorded for behavior monitoring, and meanwhile, a cache module is used to perform cache processing on video stream data.

Specifically, in this embodiment, the multi-scene video monitoring method specifically includes:

step S31: video stream data of N scenes is acquired.

Step S32: a maximum value Y of the number of scenes in which a predetermined object may exist simultaneously among the N scenes is received.

Step S33: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Step S34: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, if the picture in the first picture stream has the predetermined target, step S36 is executed; if the picture in the first picture stream does not have the predetermined target, the video stream data of the plurality of scenes is not analyzed and the process returns to step S31.

Specifically, the specific implementation process of the step S31 to the step S34 is the same as or similar to the specific implementation process of the step S11 to the step S14 in the multi-scene video monitoring method provided in the first embodiment, and the same or similar technical effects can be achieved.

Specifically, the above is performed at the same time of performing steps S32 to S34, and also performing step S35, step S35: and carrying out caching processing on the video stream data of each scene.

Specifically, an N-path video frame buffer queue is established for storage; the buffer size of each piece of video stream data is the product of the first preset time and the frame rate of the video stream data; for example, the first predetermined time is 2 seconds, and the buffer queue size is 2 times the current video frame rate.

Step S36: and analyzing and monitoring the video stream data of the scene corresponding to the picture with the predetermined target.

Specifically, in a specific embodiment, when the pictures in the first picture stream have the predetermined target, the method further includes obtaining the number Y' of the first pictures in the first picture stream having the predetermined target; then judging whether the number Y' of the first pictures with the preset target in the pictures in the first picture stream is more than M-X; if the number Y' of the first pictures with the preset targets in the pictures in the first picture stream is greater than M-X, sending an abnormal prompt; if the number Y' of the first pictures in the first picture stream having the predetermined target is not greater than M-X, the above step S36 is performed.

Specifically, referring to fig. 7, fig. 7 is a sub-flowchart of step S36 in fig. 6; step S36 specifically includes:

step S361: and decoding the video stream data of the scene corresponding to the picture which is cached and has the preset target, extracting one frame of picture at intervals of a second preset time length, and forming a second picture stream corresponding to the video stream data of the scene corresponding to the picture which is cached and has the preset target.

And the second preset time length is also smaller than the first preset time length.

Step S362: and analyzing and monitoring the second picture stream and judging whether the pictures in the video stream data of each scene are abnormal or not.

Specifically, if there is no abnormality in the picture in the video stream data of each scene, step S363 is executed; when there is an abnormality in the picture in the video stream data of each scene, step S364 is performed.

Step S363: and returning to execute the acquisition of the video stream data of the N scenes.

Step S364: and sending an exception report.

Specifically, other specific implementation processes of step S36 and its sub-step may refer to the specific implementation processes of step S15 and its sub-step in the multi-scene video monitoring method provided in the first embodiment, and may achieve the same or similar technical effects, and refer to the related text descriptions specifically, and are not described herein again.

Compared with the multi-scene video monitoring method provided by the first embodiment, the multi-scene video monitoring method provided by the embodiment further analyzes and monitors the video stream data of the scene corresponding to the picture which has the predetermined target and is cached by using a video caching method so that the normal abnormal event detection rate can be ensured on the premise of optimizing the performance; meanwhile, before the first picture of the preset target is monitored, if the preset target appears, the preset target can be recorded for behavior monitoring.

Referring to fig. 8, fig. 8 is a flowchart of a multi-scene video monitoring method according to a fourth embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, which specifically includes:

step S41: video stream data of N scenes is acquired.

Step S42: a maximum value Y of the number of scenes in which a predetermined object may exist simultaneously among the N scenes is received.

Step S43: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Step S44: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, if the picture in the first picture stream has the predetermined target, the target identifying unit notifies the switching module of the channel ID where the picture with the predetermined target appears, and executes step S46; if the picture in the first picture stream does not have the predetermined target, the video stream data of the plurality of scenes is not analyzed and the process returns to step S41.

Specifically, after step S41, step S45 is performed while steps S42 to S44 are performed.

Step S45: and carrying out caching processing on the video stream data of each scene.

Specifically, other specific implementation processes of the steps S41 to S45 are the same as or similar to the specific implementation processes of the steps S31 to S35 in the multi-scene video monitoring method provided in the third embodiment, and the same or similar technical effects can be achieved.

Step S46: and judging whether the cached video stream data of the scene corresponding to the picture with the preset target is analyzed and monitored.

Specifically, if the cached video stream data of the scene corresponding to the picture with the predetermined target is already being analyzed and monitored, step S47 is executed; if the cached video stream data of the scene corresponding to the picture with the predetermined target has not been analyzed and monitored, step S48 is executed.

Step S47: and continuing to analyze and monitor the video stream data of the scene corresponding to the picture which is cached and has the preset target.

Step S48: and switching the video stream data of the scene corresponding to the picture with the preset target and starting to analyze and monitor the cached video stream data of the scene corresponding to the picture with the preset target.

Specifically, the specific implementation process of the step S47 to the step S48 is the same as or similar to the specific implementation process of the step S26 to the step S27 in the multi-scene video monitoring method provided by the second embodiment, and the same or similar technical effects can be achieved.

Compared with the multi-scene video monitoring method provided by the second embodiment, the multi-scene video detection method provided by the embodiment can still ensure the normal abnormal event detectable rate on the premise of optimizing the performance by using the video caching method; meanwhile, before the first picture of the preset target is monitored, if the preset target appears, the preset target can be recorded for behavior monitoring; compared with the multi-scene video monitoring method provided by the third embodiment, the video stream is further switched according to the notification, so that the video stream data of the scene corresponding to the picture with the predetermined target can be directly analyzed and monitored, further resources can be further saved, and the intelligent analysis performance can be improved.

It can be understood that the multi-scene video monitoring method according to the foregoing embodiment is all suitable for monitoring N scenes by monitoring the number of paths through Y + X — M paths when Y is known in advance. When the frequency of the targets is low or no targets appear, the method can achieve the number of analysis channels far exceeding the current hardware capacity, and greatly reduce the cost; meanwhile, the intelligent analysis performance is improved by using the technology of combining the picture stream and the video stream. Namely, the technology with small occupied performance is used for filtering a large amount of analysis requirements, and limited resources are given up to the technology with high occupied performance; in addition, the video cache and the switching method according to the notification are used, and the normal abnormal event detection rate can be still ensured on the premise of optimizing the performance.

Referring to fig. 9, fig. 9 is a flowchart of a multi-scene video monitoring method according to a fifth embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, and the method is specifically applied to a monitoring device having M monitoring paths and an analyzing and monitoring module for monitoring N scenes, where the number Y of scenes in the N scenes, in which a predetermined target may exist at the same time, is unknown, where N > M. For example, people of the ATM are monitored through monitoring devices of N scenes, and people do not need to be predicted to be present near the ATM in advance. The method specifically comprises the following steps:

step S51: video stream data of N scenes is acquired.

Step S52: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Specifically, the specific implementation processes of steps S51 to S52 may refer to the specific implementation processes of step S11 and step S13 in the multi-scene video monitoring method provided in the first embodiment, and the same or similar technical effects may be achieved, which are not described herein again.

Step S53: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, when the picture in the first picture stream has the predetermined target, step S54 is executed; when the picture in the first picture stream does not have the predetermined target, the process returns to perform step S51.

The number X of the paths of the analysis monitoring module is far smaller than the number of the scenes, so that the intelligent analysis path number is improved under the limited resource condition; and the path number X of the analysis monitoring module can be selected according to the first preset time length, the specific hardware capability and the capability of the target identification unit, and is not limited by Y and M.

Specifically, other specific implementation procedures of step S53 can be seen in the specific implementation procedure of step S14 in the multi-scene video monitoring method provided in the first embodiment.

Step S54: the number Y' of first pictures in which the pictures in the first picture stream have the predetermined target is acquired.

Specifically, in one embodiment, step S54 further includes determining whether Y' is greater than M-X; when Y' is larger than M-X, namely exceeding the analysis monitoring capability of the current device, sending an abnormal prompt; when Y' is less than or equal to M-X, step S55 is performed.

Step S55: and analyzing and monitoring the video stream data of Y ' scenes corresponding to Y ' first pictures with preset targets through a Y ' path analyzing and monitoring module.

For example, if the number of the first pictures with the predetermined target in the first picture stream is 10, 10 paths of analysis and monitoring modules respectively analyze and monitor video stream data of 10 scenes corresponding to 10 first pictures with the predetermined target.

Specifically, the specific implementation process of step S55 may refer to the specific implementation process of step S15 in the multi-scene video monitoring method provided in the first embodiment, and the same or similar technical effects may be achieved, and refer to the related text descriptions above specifically, which are not described herein again.

The multi-scene video monitoring method provided by the embodiment can monitor abnormal conditions in a plurality of scenes, and because the method extracts a frame of picture from video stream data of each scene to form a picture stream, and only analyzes and monitors the video stream corresponding to the picture in the picture stream with a preset target, each video channel does not need to be kept in an analysis starting state all the time, resources can be saved, and the intelligent analysis performance is greatly improved; meanwhile, the method does not need to preset a Y value, and compared with the multi-scene video monitoring method provided by any embodiment, the method is higher in automation degree.

Referring to fig. 10, fig. 10 is a flowchart of a multi-scene video monitoring method according to a sixth embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, and the method is specifically applied to a monitoring device having M monitoring paths and an analyzing and monitoring module for monitoring N scenes, where N > M. The method specifically comprises the following steps:

step S61: video stream data of N scenes is acquired.

Step S62: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Step S63: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, when the picture in the first picture stream has the predetermined target, step S64 is executed; when the picture in the first picture stream does not have the predetermined target, the process returns to perform step S61.

Specifically, the specific implementation processes of steps S61 to S63 may refer to the specific implementation processes of steps S51 to S53 in the multi-scene video monitoring method provided in the fifth embodiment, and the same or similar technical effects may be achieved.

Step S64: the number Y' of first pictures in which the pictures in the first picture stream have the predetermined target is acquired.

Specifically, the specific implementation process of step S64 may refer to the specific implementation process of step S54 in the multi-scene video monitoring method provided in the fifth embodiment, and the same or similar technical effects may be achieved, and refer to the description of the above related text specifically, which is not repeated herein.

Step S65: and judging whether the number Y' of the first pictures with the preset targets in the pictures in the first picture stream changes or not.

Specifically, the determining whether the number Y 'of the first pictures with the predetermined target in the pictures in the first picture stream changes or not specifically means whether the number Y' of the first pictures with the predetermined target in the first picture stream at this time is changed compared with the number Y 'of the first pictures with the predetermined target in the first picture stream formed at the last time, that is, whether the number Y' is increased, decreased or unchanged.

Specifically, when the number Y 'of first pictures in which the predetermined target exists among the pictures in the first picture stream is not changed, that is, the number Y' is not changed, the step S69 is directly performed; when the number Y' of first pictures in which the predetermined target exists among the pictures in the first picture stream is changed, step S66 is performed.

Step S66: it is determined whether the number Y' of first pictures in which a predetermined target exists among the pictures in the first picture stream increases.

Specifically, when the number Y' of first pictures in which the predetermined target exists among the pictures in the first picture stream increases, step S67 is executed; when the number Y' of first pictures in which the predetermined target exists among the pictures in the first picture stream decreases, step S68 is performed.

Step S67: and starting a new analysis detection module.

Specifically, step S67 specifically refers to opening a new monitoring path number.

Step S68: and closing the analysis monitoring module for analyzing the video stream data of the scene corresponding to the reduced first picture with the preset target.

Step S69: and analyzing and monitoring the video stream data of Y ' scenes corresponding to Y ' first pictures with preset targets through a Y ' path analyzing and monitoring module.

Specifically, the specific implementation process of step S69 can be referred to the specific implementation process of step S55 in the multi-scene video monitoring method provided in the fifth embodiment.

It should be noted that, when it is monitored that the predetermined target exists in the pictures in the first picture stream for the first time, there is no comparison between the number Y 'value corresponding this time and the Y' value corresponding to the last monitoring, so that the process directly proceeds to step S69 after step S64 without executing steps S65 to S68.

Compared with the method provided by the fifth embodiment, the multi-scene video monitoring method provided by the embodiment further adjusts the corresponding analysis monitoring module to be turned on or turned off at any time according to the change of the number Y' of the first pictures with the predetermined target in the pictures in the first picture stream, so that resources can be further saved, and the automation degree can be further improved.

In an embodiment, referring to fig. 11, fig. 11 is a flowchart of a multi-scene video monitoring method according to a seventh embodiment of the present application; in the present embodiment, unlike the sixth embodiment described above, step S69 is further followed by step S70 and step S71.

Step S70: it is determined whether a predetermined object in the video stream data of Y' scenes disappears.

Specifically, when the predetermined target disappears in the video stream data of Y' scenes, step S71 is executed; when the predetermined target in the video stream data of Y' scenes does not disappear, the execution returns to step S61.

Step S71: and closing the analysis monitoring module corresponding to the video stream data of the scene with the disappearance of the preset target.

According to the method, the corresponding analysis monitoring module can be closed in time after the preset target disappears, so that resources can be further saved.

In an embodiment, referring to fig. 12, fig. 12 is a flowchart of a multi-scene video monitoring method according to an eighth embodiment of the present application; in this embodiment, a multi-scene video monitoring method is provided, where the method includes:

step S81: video stream data of N scenes is acquired.

Step S82: one frame of picture is extracted from video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Step S83: and judging whether the pictures in the first picture stream have a preset target or not through an X-path analysis monitoring module.

Specifically, if the picture in the first picture stream has the predetermined target, step S85 is executed; if the pictures in the first picture stream do not have the predetermined target, the video stream data of the plurality of scenes are not analyzed and monitored, and the process returns to step S81.

Specifically, the above is performed at the same time as the steps S82 and S83, and the steps S84, S84: and carrying out caching processing on the video stream data of each scene.

Specifically, the specific implementation of step S84 can be referred to the specific implementation of step S35 in the third embodiment.

Step S85: the number Y' of first pictures in which the pictures in the first picture stream have the predetermined target is acquired.

Step S86: and analyzing and monitoring the video stream data of the scene corresponding to the picture with the preset target which is cached by the Y' path analyzing and monitoring module.

It is understood that, after the predetermined target is monitored for the first time, the steps S85 and S86 are directly performed; and after each monitoring of the predetermined object for the second time or thereafter, further performing step S87 after performing steps S81 to S83.

Specifically, the specific implementation of step S86 can be referred to the specific implementation of step S55 in the fifth embodiment.

Step S87: and judging whether the video stream data of the scene corresponding to the newly cached picture with the preset target is analyzed and monitored.

Specifically, step S87 is executed by the switching module; specifically, after monitoring the predetermined target, the target identification unit notifies the switching module of the channel ID where the picture with the predetermined target appears, and the switching module starts to execute step S87 after receiving the channel ID notification where the picture with the predetermined target appears, which is sent by the target identification unit; specifically, if the cached video stream data of the scene corresponding to the picture with the predetermined target is already being analyzed and monitored, the step S86 is continuously executed; if the cached video stream data of the scene corresponding to the picture with the predetermined target has not been analyzed and monitored, step S88 is executed.

Step S88: and switching the video stream data of the scene corresponding to the picture with the preset target which is cached, and analyzing and monitoring the video stream data of the scene corresponding to the picture with the preset target which is cached newly through a Y' path analysis and monitoring module.

Specifically, the video stream data of the scene corresponding to the picture with the predetermined target is switched to the number of idle channels by the switching module to perform the subsequent steps.

For analyzing and monitoring the video stream data of the scene corresponding to the picture with the predetermined target that has been cached by the Y' path analyzing and monitoring module, refer to the specific implementation process of step S86.

It should be noted that, in this embodiment, after the step S85 is executed, a step of determining whether Y' is greater than M-X may be further included in each monitoring process, and the steps S65 to S68 related to the sixth embodiment may be further included, and the relevant contents of the steps S70 to S71 related to the seventh embodiment may be further included after the step S88, which may be specifically referred to the above-mentioned relevant text, and will not be described herein again.

Compared with the multi-scene video monitoring method provided by the seventh embodiment, the multi-scene video monitoring method provided by this embodiment further analyzes and monitors the video stream data of the scene corresponding to the newly cached picture with the predetermined target by using a video caching method so that the predetermined target exists in the picture in the first picture stream, thereby still ensuring the normal abnormal event detection rate on the premise of optimizing the performance; meanwhile, before the first picture of the preset target is monitored, if the preset target appears, the preset target can be recorded for behavior monitoring; meanwhile, in the subsequent monitoring process, the video stream is further switched according to the notification, so that the video stream data of the scene corresponding to the picture with the preset target can be directly analyzed and monitored, further resources can be further saved, and the intelligent analysis performance is improved.

It can be understood that the multi-scene video monitoring methods provided in the fifth to eighth embodiments are applicable to the case where Y is not known in advance, and N scenes are monitored by selecting Y' paths from M paths of monitoring paths. Compared with the multi-scene video monitoring method provided by the first embodiment, the multi-scene video monitoring method provided by the embodiment can achieve the number of analysis channels far exceeding the current hardware capability when the occurrence frequency of the target is low, and the cost is greatly reduced; meanwhile, the intelligent analysis performance is improved by utilizing the technology of combining the picture stream and the video stream. Namely, the technology with small occupied performance is used for filtering a large amount of analysis requirements, and limited resources are given up to the technology with high occupied performance; in addition, the video cache and the switching method according to the notification are used, so that the normal abnormal event detection rate can be still ensured on the premise of optimizing the performance; in addition, when the number Y' of the first pictures with the preset targets in the pictures in the first picture stream is increased or decreased, a new analysis monitoring module can be automatically started or the corresponding analysis monitoring module can be automatically closed, so that the resources can be saved, the intelligent analysis performance is greatly improved, and the automation degree is high.

Referring to fig. 13, fig. 13 is a schematic structural diagram of a multi-scene video monitoring device according to an embodiment of the present application; in this embodiment, a multi-scene video monitoring apparatus 300 is provided, where the apparatus 300 has M monitoring paths for executing the multi-scene video monitoring methods provided in the first to fourth embodiments; specifically, the apparatus 300 includes a video acquisition module 301, a receiving module 302, a first picture stream generation module 303, an M-channel analysis monitoring module 304, and a control module 305.

The video acquiring module 301 is configured to acquire video stream data of N scenes; wherein N > M, and there may be a predetermined target in Y of the N scenes at the same time; specifically, the video acquisition module 301 includes a plurality of video acquisition units, and the video acquisition units correspond to each scene one by one, that is, a video acquisition unit is disposed in each scene, so as to acquire video stream data of each scene through the corresponding video acquisition unit in each scene, and further acquire video stream data of N scenes through N video acquisition units in the video acquisition module 301; the video acquisition unit may be a camera.

The receiving module 302 is configured to receive a maximum value Y of the number of scenes in which a predetermined target may exist simultaneously in the N scenes.

The first picture stream generating module 303 is configured to extract one frame of picture from the video stream data of each scene at intervals of a first predetermined length to form a first picture stream.

Specifically, the first picture stream generating module 303 is specifically configured to configure a first predetermined duration, and verify the configured first predetermined duration, so as to ensure that the first predetermined duration is much longer than a frame interval duration of the video stream, thereby playing a role in performance optimization; then judging whether the first preset time length is less than the interval time length of the I frame or not; if the first preset time length is less than the interval time length of the I frame, decoding video stream data of each scene and extracting a frame of YUV picture at intervals of the first preset time length to form a first picture stream; if the first preset time length is greater than or equal to the interval time length of the I frames, an I frame in the video stream data of each scene is directly extracted at intervals of the first preset time length to be decoded and converted into YUV pictures to form a first picture stream.

Further, the first picture stream generating module 303 is further configured to determine whether the video stream data in the first predetermined time period only includes one I frame when the first predetermined time period is greater than or equal to the I frame interval time period, and if the video stream data in the first predetermined time period only includes one I frame, directly use the current I frame for decoding and converting to YUV pictures to form a first picture stream; if the video stream data in the first preset time period comprises a plurality of I frames, one of the I frames is extracted from the I frames for decoding and converting into YUV pictures to form a first picture stream.

The M-channel analyzing and monitoring module 304 is configured to analyze and monitor a picture stream or a video stream; specifically, the X-path analyzing and monitoring module 304 of the M-path analyzing and monitoring modules 304 is configured to determine whether a predetermined target exists in a picture in the first picture stream; wherein, the number of the X paths is not more than the difference value of M and Y; and when the pictures in the first picture stream have the preset targets, analyzing and monitoring video stream data of scenes corresponding to the pictures with the preset targets.

In an embodiment, the X-path analysis monitoring module 304 includes a target identification unit, configured to perform target identification on the pictures in the first picture stream, and determine whether the pictures in the first picture stream have a predetermined target.

In an embodiment, the multi-scene video monitoring apparatus 300 further includes a switching module, where the switching module is configured to receive a channel ID notification that the picture with the predetermined target exists in the first picture stream, and receive the channel ID notification that the picture with the predetermined target exists and sent by the target identification unit, and determine whether video stream data of the scene corresponding to the picture with the predetermined target exists is already being analyzed and monitored, and at the same time, switch the video stream data of the scene corresponding to the picture with the predetermined target when the video stream data of the scene corresponding to the picture with the predetermined target does not already be analyzed and monitored, that is, switch the video stream data of the scene corresponding to the picture with the predetermined target to the number of idle channel paths for performing later-stage analysis and monitoring.

In another embodiment, the multi-scene video monitoring apparatus 300 further includes a buffering module, configured to perform buffering processing on the video stream data of each scene while the first picture stream generating module 303 extracts one frame of picture from the video stream data of each scene at intervals of a first predetermined time to form a first picture stream. Specifically, in this embodiment, the switching module is configured to receive, when a predetermined target exists in a picture in the first picture stream, a channel ID notification that the picture with the predetermined target is located and that is sent by the target identification unit, determine whether the cached video stream data of the scene corresponding to the picture with the predetermined target is already analyzed and monitored, and switch the cached video stream data of the scene corresponding to the picture with the predetermined target when the cached video stream data of the scene corresponding to the picture with the predetermined target is not already analyzed and monitored, that is, switch the cached video stream data of the scene corresponding to the picture with the predetermined target to the number of idle channel paths for performing the later-stage analysis and monitoring.

The analysis monitoring module 304 further includes a decoding frame-extracting unit and an analysis monitoring unit; the decoding frame extraction unit is used for decoding video stream data of a scene corresponding to the picture with the preset target, extracting one frame of picture at intervals of second preset time length, and forming a second picture stream corresponding to the video stream data of the scene corresponding to the picture with the preset target; the analysis monitoring unit is used for judging whether the pictures in the second picture stream are abnormal or not; if the pictures in the second picture stream are not abnormal, returning to execute the video stream data of the N scenes; and if the pictures in the second picture stream are abnormal, sending an abnormal report.

The control module 305 is configured to control the video obtaining module 301 to obtain video stream data of N scenes; the control receiving module 302 receives the number Y of scenes in which a predetermined target exists among the N scenes; controlling a first picture stream generation module 303 to extract one frame of picture from video stream data of each scene at intervals of a first predetermined time length to form a first picture stream; controlling an X-path analysis monitoring module 304 in the M-path analysis monitoring module 304 to determine whether a predetermined target exists in a picture in the first picture stream; wherein, the number of the X paths is not more than the difference value of M and Y; when the picture in the first picture stream has the predetermined target, the M-path analyzing and monitoring module 304 is controlled to analyze and monitor the video stream data of the scene corresponding to the picture with the predetermined target.

Referring to fig. 14, fig. 14 is a schematic structural diagram of a multi-scene video monitoring device according to another embodiment of the present application; in this embodiment, another multi-scene video monitoring apparatus 400 is provided, where the apparatus 400 has M monitoring paths, and the monitoring path M is much smaller than the number N of scenes, and is used to execute the multi-scene video monitoring methods provided in the fifth to eighth embodiments; specifically, the apparatus 400 includes a video acquisition module 401, a first picture stream generation module 402, an M-channel analysis monitoring module 403, and a control module 404.

The video acquiring module 401 is configured to acquire video stream data of N scenes; the first picture stream generating module 402 is configured to extract one frame of picture from video stream data of each scene at intervals of a first predetermined duration to form at least one first picture stream; the number of paths of the first picture stream is less than the number of scenes; specifically, the structures and functions of the video obtaining module 401 and the first picture stream generating module 402 are the same as or similar to those of the video obtaining module 301 and the first picture stream generating module 303 in the multi-scene video monitoring device 300 provided in the foregoing embodiment, and the same or similar technical effects can be achieved.

The M-channel analyzing and monitoring module 403 is configured to analyze and monitor a picture stream or a video stream.

In one embodiment, the X-ray analysis monitoring module 403 includes a target identification unit and an acquisition unit; the target identification unit is used for judging whether a preset target exists in the pictures in the first picture stream; the acquiring unit is used for acquiring the number Y' of the first pictures of the pictures in the first picture stream with the preset target when the pictures in the first picture stream have the preset target.

In an embodiment, the obtaining unit is further configured to determine whether the number Y' of first pictures with a predetermined target in the pictures in the first picture stream changes; when the number Y 'of first pictures with preset targets in the pictures in the first picture stream changes, judging whether the number Y' of the first pictures with the preset targets in the pictures in the first picture stream is increased or not; when the number Y' of first pictures with a predetermined target in the pictures in the first picture stream increases, a new path of analysis monitoring module 403 is started; when the number Y' of the first pictures with the predetermined target in the pictures in the first picture stream decreases, the analysis monitoring module 403 for analyzing the video stream data of the scene corresponding to the decreased first pictures with the predetermined target is turned off to further improve the automation degree.

In another embodiment, the multi-scene video monitoring apparatus 400 further includes a switching module, configured to determine whether video stream data of a scene corresponding to a picture with a predetermined target is already being analyzed and monitored when the picture in the first picture stream has the predetermined target, and switch the video stream data of the scene corresponding to the picture with the predetermined target when the video stream data of the scene corresponding to the picture with the predetermined target is not already analyzed and monitored.

In an embodiment, the multi-scene video monitoring apparatus 400 further includes a cache module; the buffering module is configured to perform buffering processing on the video stream data of each scene while the first picture stream generating module 402 extracts one frame of picture from the video stream data of each scene at intervals of a first predetermined time to form a first picture stream.

In this embodiment, the switching module is specifically configured to determine whether video stream data of Y 'scenes corresponding to Y' first pictures of a newly cached picture with a predetermined target is already being analyzed and monitored, and switch the video stream data of the scene corresponding to the newly cached picture with the predetermined target when the video stream data of the scene corresponding to the picture with the predetermined target is not already analyzed and monitored.

The Y ' path analyzing and monitoring module 403 is further configured to decode, by a decoding and frame-extracting unit therein, video stream data of Y ' scenes corresponding to Y ' first pictures in which the predetermined target exists, and extract one frame of picture every second predetermined time interval to form a second picture stream corresponding to video stream data of Y ' scenes corresponding to Y ' first pictures in which the predetermined target exists, and determine, by an analyzing and monitoring unit therein, whether a picture in the second picture stream is abnormal; if the pictures in the second picture stream are not abnormal, continuously analyzing and monitoring the video stream data of the cached scene; and if the pictures in the second picture stream are abnormal, sending an abnormal report.

The control module 404 is configured to control the video obtaining module 401 to obtain video stream data of N scenes; controlling a first picture stream generation module 402 to extract one frame of picture from video stream data of each scene at intervals of a first predetermined time length to form a first picture stream; controlling an X-path analyzing and monitoring module 403 in the M-path analyzing and monitoring module 403 to determine whether a predetermined target exists in a picture in the first picture stream; when the pictures in the first picture stream have the predetermined target, acquiring the number Y 'of the first pictures with the predetermined target in the pictures in the first picture stream, and controlling the Y' path analyzing and monitoring module 403 in the M path analyzing and monitoring module 403 to respectively analyze and monitor the video stream data of the Y 'scenes corresponding to the Y' first pictures with the predetermined target.

Specifically, it should be noted that the M-path analyzing and monitoring module mentioned above specifically refers to an analyzing and monitoring module for monitoring the number of M paths; the X-path analysis monitoring module is specifically an analysis monitoring module for monitoring the number of paths of the X-path; the Y 'path analysis monitoring modules are all analysis monitoring modules for monitoring the number of paths of the Y' path.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above embodiments are only examples of the present application, and not intended to limit the scope of the present application, and all equivalent devices or equivalent processes that are used in the present specification and drawings, or that are directly or indirectly applied to other related technical fields are also included in the scope of the present application.

Claims

1. A multi-scene video monitoring method is characterized in that the method is applied to a monitoring device for monitoring N scenes through an analysis monitoring module for monitoring the number of paths by M paths, wherein N is greater than M; the method comprises the following steps:

acquiring video stream data of the N scenes;

receiving a maximum value Y of the number of scenes in which a predetermined target may exist simultaneously in the N scenes;

extracting a frame of picture from the video stream data of each scene at intervals of a first preset time length to form a first picture stream, and forming a new first picture stream at intervals of the first preset time length;

judging whether the pictures in the first picture stream have the preset target or not through an X-path analysis monitoring module in the M-path analysis monitoring modules; the number of the X paths is not more than the difference value of M and Y;

when the preset target exists in the pictures in the first picture stream, analyzing and monitoring video stream data of the scene corresponding to the pictures with the preset target.

2. The multi-scene video monitoring method according to claim 1, wherein said extracting a frame of picture from the video stream data of each of said scenes at intervals of a first predetermined duration, and the step of forming the first picture stream specifically includes:

configuring a first preset time length;

judging whether the first preset time length is less than the interval time length of the I frame or not;

when the first preset time length is less than the interval time length of the I frame, decoding video stream data of each scene and extracting a frame of YUV picture at the interval of the first preset time length to form a first picture stream;

and when the first preset time length is not less than the interval time length of the I frames, extracting the I frames in the video stream data of each scene at intervals of the first preset time length, decoding and converting YUV pictures to form a first picture stream.

3. The multi-scene video monitoring method according to claim 1, further comprising: when the predetermined target does not exist in the pictures in the first picture stream, the steps of analyzing and monitoring the video stream data of a plurality of scenes and returning to the step of acquiring the video stream data of the N scenes are not performed.

4. The multi-scene video monitoring method according to claim 1, further comprising:

extracting a frame of picture from the video stream data of each scene at intervals of a first preset time length, and performing cache processing on the video stream data of each scene while forming a first picture stream; the buffer size of each piece of video stream data is the product of the first preset time and the frame rate of the video stream data;

when the predetermined target exists in the picture in the first picture stream, the step of analyzing and monitoring the video stream data of the scene corresponding to the picture in which the predetermined target exists further includes:

when the predetermined target exists in the pictures in the first picture stream, analyzing and monitoring the video stream data of the scene corresponding to the pictures which have been cached and exist the predetermined target.

5. The multi-scene video monitoring method according to claim 4, wherein when the predetermined target exists in the pictures in the first picture stream, the step of analyzing and monitoring the video stream data of the scene corresponding to the pictures in which the predetermined target exists that have been cached further comprises:

acquiring the number Y' of first pictures with the preset target in the pictures in the first picture stream;

judging whether the number Y' of first pictures with the preset target in the pictures in the first picture stream is greater than M-X or not;

if the number Y' of the first pictures of the preset target in the pictures in the first picture stream is greater than M-X, sending an abnormal prompt; if the number Y' of the first pictures with the preset target in the pictures in the first picture stream is not more than M-X, analyzing and monitoring the video stream data of the scene corresponding to the pictures with the preset target which are cached.

6. The multi-scene video monitoring method according to claim 4, wherein the step of analyzing and monitoring the video stream data of the scene corresponding to the picture in which the predetermined target exists, when the predetermined target exists in the picture in the first picture stream, further comprises:

judging whether the cached video stream data of the scene corresponding to the picture with the preset target is analyzed and monitored;

if the video stream data of the scene corresponding to the picture with the preset target which is cached is analyzed and monitored, continuing to analyze and monitor the video stream data of the scene corresponding to the picture with the preset target which is cached;

if the cached video stream data of the scene corresponding to the picture with the preset target is not analyzed and monitored, switching the video stream data of the scene corresponding to the picture with the preset target and starting analyzing and monitoring the cached video stream data of the scene corresponding to the picture with the preset target.

7. The multi-scene video monitoring method according to claim 4, wherein the step of analyzing and monitoring the video stream data of the scene corresponding to the picture in which the predetermined target exists, which has been cached, further comprises:

decoding the video stream data of the scene corresponding to the picture which is cached and has the preset target, and extracting a frame of picture at intervals of second preset duration to form a second picture stream corresponding to the video stream data of the scene corresponding to the picture which is cached and has the preset target; wherein the second predetermined length of time is less than the first predetermined length of time;

analyzing and monitoring the second picture stream and judging whether the pictures in the video stream data of each scene are abnormal or not;

when the pictures in the video stream data of each scene are not abnormal, continuously analyzing and monitoring the cached video stream data of the scene;

and when the pictures in the video stream data of each scene have the abnormity, sending an abnormity report.

8. The multi-scene video monitoring method according to claim 4, wherein the step of analyzing and monitoring the video stream data of the scene corresponding to the picture in which the predetermined target exists, which has been cached, further comprises:

judging whether the corresponding preset target in the scene disappears or not;

and when the corresponding preset target in the scene disappears, closing the analysis monitoring module corresponding to the video stream data of the scene with the disappeared preset target.

9. A multi-scene video monitoring device is characterized in that the device is provided with M monitoring paths; the device comprises:

the video acquisition module is used for acquiring video stream data of N scenes; wherein N > M;

a receiving module, configured to receive a maximum value Y of the number of scenes in which a predetermined target may exist simultaneously in the N scenes;

a first picture stream generating module, configured to extract a frame of picture from video stream data of each scene at intervals of a first predetermined duration to form a first picture stream, and form a new first picture stream at intervals of the first predetermined duration;

the M paths of analysis monitoring modules are used for analyzing and monitoring the picture stream or the video stream;

the control module is used for controlling the video acquisition module to acquire video stream data of N scenes; controlling the receiving module to receive the scene number Y with a preset target in the N scenes; controlling the first picture stream generation module to extract one frame of picture from the video stream data of each scene at intervals of a first preset time length to form a first picture stream; controlling an X-path analysis monitoring module in the M-path analysis monitoring modules to judge whether the pictures in the first picture stream have the preset target or not; the number of the X paths is not more than the difference value of M and Y; when the preset target exists in the pictures in the first picture stream, analyzing and monitoring video stream data of the scene corresponding to the pictures with the preset target.

10. The multi-scene video monitoring device according to claim 9, wherein the X-path analysis monitoring module comprises a target recognition unit, and the target recognition unit is configured to determine whether the predetermined target exists in the pictures in the first picture stream.

11. The multi-scene video monitoring device according to claim 10, further comprising a switching module, wherein the switching module is configured to determine whether video stream data of the scene corresponding to the picture with the predetermined target is already being analyzed and monitored when the predetermined target exists in the picture in the first picture stream, and switch the video stream data of the scene corresponding to the picture with the predetermined target when the video stream data of the scene corresponding to the picture with the predetermined target does not already be analyzed and monitored.

12. The multi-scene video monitoring device according to claim 11, further comprising a buffering module for buffering the video stream data of each scene while the first picture stream generating module extracts one frame of picture from the video stream data of each scene at intervals of a first predetermined length to form a first picture stream.