Disclosure of Invention
The application aims to provide a video data-based safety protection method and device, which can realize safety protection of a monitoring scene based on an analysis result of a monitoring video, ensure the safety of the monitoring scene and improve the applicability of the analysis result of the monitoring video.
The embodiment of the application provides a video data-based safety protection method, which comprises the steps of obtaining a monitoring video and associated data of the monitoring video, wherein the associated data comprise evaluation data of various types of users aiming at the monitoring video, determining a plurality of analysis keywords corresponding to the monitoring video based on the associated data, determining an analysis strategy of the monitoring video based on the analysis keywords, wherein the analysis strategy comprises at least one of a target analysis object, a target analysis behavior and additional analysis information, wherein the plurality of analysis keywords, the target analysis object, the target analysis behavior and the additional analysis information are associated, determining analysis results corresponding to the analysis keywords respectively based on the analysis strategy, the associated relation and the monitoring video, determining a safety protection strategy of a monitoring scene corresponding to the monitoring video based on the analysis results, and feeding back the safety protection strategy.
In a possible implementation manner, the determining of the plurality of analysis keywords corresponding to the monitoring video based on the association data comprises determining a first analysis keyword corresponding to the monitoring video based on a preset keyword library and the association data, obtaining an association monitoring video corresponding to the monitoring video, wherein the association monitoring video corresponds to the analysis data, determining a second analysis keyword corresponding to the monitoring video according to the analysis data, and determining the plurality of analysis keywords corresponding to the monitoring video based on the first analysis keyword and the second analysis keyword.
In one possible implementation manner, the determining of the plurality of analysis keywords corresponding to the monitoring video based on the first analysis keywords and the second analysis keywords includes determining identical analysis keywords and non-identical analysis keywords in the first analysis keywords and the second analysis keywords, judging whether the occurrence frequency of the identical analysis keywords in the associated data is greater than a preset occurrence frequency for the identical analysis keywords, and determining associated analysis keywords corresponding to the identical analysis keywords if the occurrence frequency of the identical analysis keywords in the associated data is greater than the preset occurrence frequency, wherein the associated analysis keywords do not belong to the non-identical analysis keywords, and determining final analysis keywords based on the identical analysis keywords, the associated analysis keywords and the non-identical analysis keywords.
In one possible implementation manner, the method for determining the analysis strategy of the monitoring video based on the plurality of analysis keywords comprises the steps of determining keyword types corresponding to the plurality of analysis keywords respectively, determining the object, object behaviors, object evaluation and object behavior evaluation as target analysis objects according to the object types, determining the object behaviors and the object behaviors with association relations with the object behaviors as target analysis behaviors according to the object behavior types, and determining the object evaluation and the object behavior evaluation as additional analysis information.
In a possible implementation manner, the video data-based security protection method further comprises determining a first association according to the keyword types respectively corresponding to the plurality of analysis keywords, wherein the first association is used for representing the association between the analysis keywords and the target analysis objects and representing the association between the analysis keywords and the target analysis behaviors and representing the association between the analysis keywords and the additional analysis information, determining a second association according to the first association, wherein the second association is used for representing the association between the target analysis objects and the target analysis behaviors and representing the association between the target analysis objects and the additional analysis information, wherein the target analysis objects and the target analysis behaviors have the association if the analysis keywords associated with the target analysis objects are the same, and the target analysis objects and the additional analysis behaviors have the association if the types of the analysis keywords associated with the target analysis objects and the additional analysis information are the same.
In one possible implementation manner, the determining the analysis results respectively corresponding to the plurality of analysis keywords based on the analysis strategy and the association relation comprises determining an object analysis result based on an object analysis model and the monitoring video if the analysis strategy comprises an object analysis object, an object analysis behavior and additional analysis information, determining an action analysis result based on an action analysis model and the monitoring video, determining an additional information analysis result based on an event analysis model and the monitoring video, determining a final analysis result based on the object analysis result, the action analysis result and the additional information analysis result, and determining the analysis results respectively corresponding to the plurality of keywords based on the association relation and the final analysis result.
In one possible implementation manner, the analysis result is used for representing whether abnormal objects exist in the monitoring scene and representing whether the number of objects in the monitoring scene is abnormal or not, and the determining the safety protection strategy of the monitoring scene corresponding to the monitoring video based on the analysis result comprises determining that the safety protection strategy is to evacuate objects in the monitoring scene and dispatch safety protection personnel to the monitoring scene if the analysis result represents that the abnormal objects exist in the monitoring scene and the number of objects in the monitoring scene is abnormal, and determining that the safety protection strategy is to limit the number of objects entering the monitoring scene and play first preset prompt information, wherein the first preset prompt information is used for indicating to leave the monitoring scene.
In one possible implementation manner, the video data-based safety protection method further comprises determining the safety protection policy to evacuate objects in the monitoring scene and send safety protection personnel to the monitoring scene and stop new objects from entering the monitoring scene if the analysis result indicates that abnormal objects exist in the monitoring scene and the number of objects in the monitoring scene are abnormal, and playing second preset prompt information for indicating that the monitoring scene is left under the guidance of the safety protection personnel, and determining the safety protection policy to play third preset prompt information for indicating that the monitoring scene is not abnormal currently if the analysis result indicates that no abnormal objects exist in the monitoring scene and the number of objects in the monitoring scene are not abnormal.
In one possible implementation manner, the video data-based safety protection method further comprises the steps of judging, for first evaluation data of the monitoring video, of a first type of user, whether a support rate corresponding to second evaluation data is higher than a preset support rate or not, determining, based on browsing amount and/or praise amount, if the support rate corresponding to the second evaluation data is higher than the preset support rate, taking the analysis result corresponding to the second evaluation data as reply data of the second evaluation data, displaying the first evaluation data to a watching user of the monitoring video in a target display form, which is different from display forms of other evaluation data, of the second evaluation data, judging, for second evaluation data of a second type of user, whether the support rate corresponding to the second evaluation data is higher than the preset support rate or not, determining, if the support rate corresponding to the second evaluation data is higher than the preset support rate, taking the analysis result corresponding to the second evaluation data as reply data of the second evaluation data, and if the support rate corresponding to the second evaluation data is lower than the preset support rate, hiding the support rate corresponding to the second evaluation data when the second evaluation data is displayed, and hiding the support rate corresponding to the second evaluation data when the display is required to be displayed, and the information is required to be hidden.
The embodiment of the application also provides a safety protection system based on video data, which comprises a video analysis module, a safety protection module and a safety protection module, wherein the video analysis module is configured to acquire a monitoring video and associated data of the monitoring video, the associated data comprise evaluation data of various types of users for the monitoring video, determine a plurality of analysis keywords corresponding to the monitoring video based on the associated data, determine an analysis strategy of the monitoring video based on the analysis keywords, wherein the analysis strategy comprises at least one of a target analysis object, a target analysis behavior and additional analysis information, an association relationship is arranged among the analysis keywords, the target analysis object, the target analysis behavior and the additional analysis information, and determine analysis results corresponding to the analysis keywords based on the analysis strategy, the association relationship and the monitoring video respectively, and the safety protection module is configured to determine a safety protection strategy of a monitoring scene corresponding to the monitoring video based on the analysis results and feed back the safety protection strategy.
Compared with the prior art, the video data-based safety protection method and system provided by the embodiment of the application acquire the associated data of the monitoring video, wherein the associated data comprises evaluation data of multiple types of users aiming at the monitoring video, multiple analysis keywords can be determined based on the associated data, and then the analysis strategy is determined by utilizing the multiple analysis keywords, so that the monitoring video is analyzed by utilizing the association relation between the analysis strategy and various information in the analysis strategy, and the analysis results respectively corresponding to the multiple analysis keywords are determined. By the analysis mode, more targeted analysis can be realized, and the safety of the analysis result of the monitoring video is improved. And the obtained analysis result is associated with the analysis keyword, so that the analysis result can be used for determining the safety protection strategy of the monitoring scene corresponding to the monitoring video, and the safety protection of the monitoring scene is realized. Therefore, the technical scheme can realize the safety protection of the monitoring scene based on the analysis result of the monitoring video, and improve the applicability of the analysis result of the monitoring video while ensuring the safety of the monitoring scene.
Detailed Description
The following detailed description of embodiments of the application is, therefore, to be taken in conjunction with the accompanying drawings, and it is to be understood that the scope of the application is not limited to the specific embodiments.
Throughout the specification and claims, unless explicitly stated otherwise, the term "comprise" or variations thereof such as "comprises" or "comprising", etc. will be understood to include the stated element or component without excluding other elements or components.
The technical scheme provided by the embodiment of the application can be applied to various analysis scenes of the monitoring video, and in the analysis scenes, the safety protection is carried out on the monitoring scenes by utilizing the analysis results of the monitoring video.
It is understood that the monitoring scenario is a variety of monitorable environments, such as squares, malls, etc.
In the embodiment of the application, the monitoring video may be monitoring data of different scenes and may relate to a plurality of objects.
In some application scenarios, the surveillance video may have different types of viewers, some of which may be professional, and some of which are not professional and only subjective decisions. Different comment data may be available for the surveillance video for different viewing users.
At present, in the analysis mode of the monitoring video, an off-line analysis mode is usually adopted, and the analysis mode is not related to a real monitoring scene, so that the application of the analysis result of the monitoring video is poor.
Based on the above, the embodiment of the application provides a safety protection scheme based on monitoring video analysis, and based on the analysis result of the monitoring video, the safety protection of the monitoring scene is realized, the safety of the monitoring scene is ensured, and the applicability of the analysis result of the monitoring video is improved.
Further, the hardware running environment of the technical scheme of the embodiment of the application can be a safety protection platform for monitoring a scene.
In some embodiments, as shown in fig. 1, the security protection platform is communicatively coupled to the video surveillance platform, such that the security protection platform may obtain data from the video surveillance platform to enable processing of surveillance video. And the safety protection platform can also feed back the processing result to the video monitoring platform so as to apply the processing result.
Referring next to fig. 2, a flowchart of a video data based security protection method according to an embodiment of the present application includes:
step 201, acquiring a monitoring video and associated data of the monitoring video. The monitoring video is a monitoring video, and the associated data comprise evaluation data of various types of users aiming at the monitoring video.
In some embodiments, multiple types of users, such as viewing users of surveillance videos, may differ in their professionals, e.g., for surveillance scenes of a game, the viewing users may be game commentators, game coaches, fans of the game, etc.
In some embodiments, the evaluation data may be real-time evaluation data, such as a bullet screen, a fly screen, etc., or non-real-time evaluation data, such as a video comment, a video message, etc. Or other forms of available evaluation data.
In some embodiments, the evaluation data may be from a surveillance video playback platform, so that the data may be easily obtained. The evaluation data corresponds to various information such as an evaluation user, an evaluation user type, an evaluation method, and an evaluation content.
Step 202, determining a plurality of analysis keywords corresponding to the monitoring video based on the associated data.
It will be appreciated that the plurality of analysis keywords are information extracted from these associated data. Thus, the plurality of analysis keywords may represent analysis information of the monitoring video that different types of users want to see.
As an optional implementation manner, step 202 includes determining a first analysis keyword corresponding to a monitoring video based on a preset keyword library and associated data, obtaining an associated monitoring video corresponding to the monitoring video, wherein the associated monitoring video corresponds to the analysis data, determining a second analysis keyword corresponding to the monitoring video according to the analysis data, and determining a plurality of analysis keywords corresponding to the monitoring video based on the first analysis keyword and the second analysis keyword.
In some embodiments, based on a preset keyword library, extracting keywords from each piece of evaluation data, so as to determine analysis keywords corresponding to each piece of evaluation data, and performing deduplication processing on the analysis keywords, where the obtained analysis keywords are first analysis keywords.
In some embodiments, the preset keyword library is a preset keyword library, and a plurality of keywords associated with the surveillance video are stored in the keyword library. Such as keywords corresponding to information such as object names, object behaviors, object events, etc.
In some embodiments, the associated surveillance video corresponding to the surveillance video may be obtained from a large data platform (a platform storing a large amount of surveillance video).
In some embodiments, the surveillance video and the associated surveillance video have some features in common, for example, for a game surveillance scene, the features may be the same party, the same form of the game, the same referee, etc. For example, if the surveillance video is a bottom half surveillance video, the associated surveillance video may be a top half surveillance video, etc.
In some embodiments, the associated monitoring video corresponds to analysis data, where the analysis data may be analysis data provided by a professional analyst, or may be analysis data provided by a large data platform, and an analysis manner of the analysis data is not limited.
In some embodiments, based on these analysis data, keywords are extracted therefrom, and a second analysis keyword may be determined. Based on the analysis data, the determination of the second analysis keyword can also be realized by using a preset keyword library.
Further, based on the first analysis keyword and the second analysis keyword, a plurality of analysis keywords corresponding to the monitoring video are determined.
As an optional implementation manner, the method comprises the steps of determining a plurality of analysis keywords corresponding to a monitoring video based on a first analysis keyword and a second analysis keyword, determining the same analysis keyword and different analysis keywords in the first analysis keyword and the second analysis keyword, judging whether the occurrence frequency of the same analysis keyword in association data is larger than a preset occurrence frequency or not according to the same analysis keyword, determining the association analysis keyword corresponding to the same analysis keyword if the occurrence frequency of the same analysis keyword in association data is larger than the preset occurrence frequency, wherein the association analysis keyword does not belong to the different analysis keyword, and determining the final analysis keyword based on the same analysis keyword, the association analysis keyword and the different analysis keyword.
In some embodiments, the first analysis keyword and the second analysis keyword are compared to determine the same analysis keyword and the different analysis keyword. The similarity between the first analysis keyword and the second analysis keyword can be calculated, and if the similarity is greater than a preset similarity, the two keywords are regarded as the same. It is to be understood that the identity herein is not an absolute identity.
In some embodiments, for the same analysis keyword, determining the occurrence frequency of the same analysis keyword in the associated data, and if the occurrence frequency is greater than a preset occurrence frequency, determining the associated analysis keyword corresponding to the same analysis keyword. The preset occurrence frequency may be set according to different application scenarios, which is not limited herein.
In some embodiments, the associated analysis keywords corresponding to the same analysis keyword may be keywords that have a number of occurrences of the same analysis keyword together with the associated analysis keyword greater than a preset number of occurrences in the associated data. That is, each time the same analysis keyword appears in the associated data, the associated analysis keyword also appears. The preset occurrence times can be set according to different application scenarios, which is not limited herein.
It will be appreciated that the associated analysis keywords do not belong to different analysis keywords, and that the associated analysis keywords do not repeat with the different analysis keywords.
Further, a final analysis keyword is determined based on the same analysis keyword, the associated analysis keyword, and the different analysis keywords. The three keywords may be integrated, for example, a similarity between the keywords may be calculated, and if the similarity between the two keywords is greater than a threshold, only one of the keywords may be retained. For another example, it is determined whether the total number of keywords is greater than a threshold, and if so, some of the keywords are pruned. And, among the non-identical analysis keywords, preferentially deleting the analysis keyword determined based on the second analysis keyword.
Step 203, determining an analysis strategy of the monitoring video based on the plurality of analysis keywords.
The analysis strategy comprises at least one of a target analysis object, a target analysis behavior and additional analysis information, and a plurality of analysis keywords, the target analysis object, the target analysis behavior and the additional analysis information are associated.
In some embodiments, an analysis strategy of the monitoring video is determined based on a plurality of analysis keywords, wherein the analysis strategy comprises the steps of determining keyword types corresponding to the analysis keywords, wherein the keyword types comprise objects, object behaviors, object evaluations and object behavior evaluations, determining the objects and objects with association relations with the objects as target analysis objects according to the object types, determining the object behaviors and the object behaviors with association relations with the object behaviors as target analysis behaviors according to the object behavior types, and determining the object evaluations and the object behavior evaluations as additional analysis information.
In some embodiments, the keyword types corresponding to the plurality of analysis keywords may be determined based on a preset keyword classification model. The preset keyword classification model can be trained and determined based on a keyword set marked with a keyword type, and specific training modes are not described in detail herein, and can refer to mature technologies in the field.
In some embodiments, the keyword types relate to four types of objects, object behaviors, object ratings, and object behavior ratings. The system comprises a target evaluation, an object behavior evaluation and a behavior evaluation, wherein the target evaluation is used for representing the evaluation of the performance of a certain target, and the object behavior evaluation is used for representing the evaluation of a certain behavior.
In some embodiments, for an object type, the object and the object having an association relationship with the object are determined as target analysis objects. The object having the association relationship may be a teammate of the object or an opponent of the object.
In some embodiments, for an object behavior type, an object behavior and an object behavior having an association with the object behavior are determined as target analysis behaviors. The object behaviors with the association relationship may be continuous object behaviors, for example, after the object behaviors are successfully intercepted, the object behaviors are usually passed, thrown, or the like, and the object behaviors can be regarded as having the association relationship.
In some embodiments, the object assessment and the object behavior assessment are determined directly as additional analysis information.
In some embodiments, since the analysis keywords, the target analysis objects, the target analysis behaviors, and the additional analysis information have an association relationship, the association relationship needs to be determined.
Therefore, as an optional implementation mode, the method further comprises the steps of determining a first incidence relation according to the types of keywords corresponding to the plurality of analysis keywords, wherein the first incidence relation is used for representing the incidence relation between the analysis keywords and the target analysis objects, representing the incidence relation between the analysis keywords and the target analysis behaviors, representing the incidence relation between the analysis keywords and the additional analysis information, determining a second incidence relation according to the first incidence relation, representing the incidence relation between the target analysis objects and the target analysis behaviors, representing the incidence relation between the target analysis objects and the additional analysis information, wherein the incidence relation between the target analysis objects and the target analysis behaviors is achieved if the analysis keywords associated with the target analysis objects are the same, and the incidence relation between the target analysis objects and the additional analysis information is achieved if the types of the analysis keywords associated with the target analysis objects and the additional analysis information are the same.
On the basis that the keyword types corresponding to the plurality of analysis keywords are known, the association relation among the analysis keywords, the target analysis objects, the target analysis behaviors and the additional analysis information can be directly determined. For example, if the keyword type corresponding to the analysis keyword is an object and the object is consistent with the target analysis object, the object and the target analysis object have an association relationship. If the keyword type corresponding to the analysis keyword is an object behavior and the object behavior is consistent with the target analysis behavior, the object behavior and the target analysis object have an association relation. The same is true for the additional analysis information.
Further, based on the first association, a second association may be determined. In some embodiments, if the analysis keywords associated with the target analysis object and the target analysis behavior are the same, the target analysis object and the target analysis behavior have an association relationship, and if the types of the analysis keywords associated with the target analysis object and the additional analysis information are the same, the target analysis object and the additional analysis information have an association relationship.
Step 204, determining analysis results corresponding to the analysis keywords respectively based on the analysis strategy, the association relation and the monitoring video.
In some embodiments, step 204 includes determining an object analysis result based on the object analysis model and the surveillance video, determining a behavior analysis result based on the behavior analysis model and the surveillance video, determining an additional information analysis result based on the event analysis model and the surveillance video, determining a final analysis result based on the object analysis result, the behavior analysis result and the additional information analysis result, and determining analysis results corresponding to the plurality of keywords based on the association relationship, if the analysis strategy includes the object analysis object, the object analysis behavior and the additional analysis information.
In some embodiments, an object analysis model, a behavior analysis model, and an event analysis model are preset, and each of the three analysis models may be a neural network model that is trained using a corresponding training data set. For specific training methods, reference may be made to the mature techniques in the art.
And extracting the analysis result of the target analysis object from the object analysis result, namely, the analysis result of the keyword related to the target analysis object.
And extracting the analysis result of the target analysis behavior from the behavior analysis result, namely, the analysis result of the keyword related to the target analysis behavior.
And extracting the additional information analysis result from the event analysis result, namely, the analysis result of the keywords related to the additional information analysis result.
Therefore, based on the object analysis result, the behavior analysis result and the additional information analysis result, a final analysis result can be determined, and based on the association relationship and the final analysis result, an analysis result corresponding to each of the plurality of keywords is determined.
In some embodiments, based on the association and the final analysis results, analysis results corresponding to each keyword may be determined.
In some embodiments, after determining the analysis results corresponding to the plurality of keywords, the analysis results may also be utilized.
Step 205, based on the analysis result, determining a security protection strategy of the monitoring scene corresponding to the monitoring video, and feeding back the security protection strategy.
As an alternative implementation manner, the analysis result is used for representing whether an abnormal object exists in the monitoring scene and representing whether the number of objects in the monitoring scene is abnormal, and step 205 may include determining that the safety protection policy is to evacuate the object in the monitoring scene and dispatch safety protection personnel to the monitoring scene if the analysis result represents that the abnormal object exists in the monitoring scene and the number of objects in the monitoring scene is not abnormal, and determining that the safety protection policy is to limit the number of objects entering the monitoring scene and play first preset prompt information for indicating to leave the monitoring scene if the analysis result represents that the abnormal object does not exist in the monitoring scene and the number of objects in the monitoring scene is abnormal.
In other embodiments, if the analysis result indicates that an abnormal object exists in the monitoring scene and the number of objects in the monitoring scene is abnormal, determining a safety protection strategy to evacuate the object in the monitoring scene, sending a safety protection personnel to the monitoring scene, stopping the new object from entering the monitoring scene, and playing a second preset prompt message, wherein the second preset prompt message is used for indicating to leave the monitoring scene under the guidance of the safety protection personnel, and if the analysis result indicates that no abnormal object exists in the monitoring scene and the number of objects in the monitoring scene is not abnormal, determining the safety protection strategy to play a third preset prompt message, wherein the third preset prompt message is used for indicating that no abnormality exists in the monitoring scene currently.
In some embodiments, the security protection policy may be fed back to the playing end of the surveillance video, or the security protection end of the surveillance scene, etc.
In some embodiments, the analysis related condition of the monitoring video can be fed back at the front end, so that the watching user has corresponding feedback information. The method further comprises the steps of judging the matching degree between the first evaluation data and analysis results corresponding to the first evaluation data aiming at the first evaluation data of the monitoring video aiming at the first type of user, displaying the first evaluation data to a watching user of the monitoring video in a target display form, wherein the target display form is different from the display form of other evaluation data, judging whether the support rate corresponding to the second evaluation data is higher than the preset support rate aiming at the second evaluation data of the monitoring video, determining the support rate based on the browsing amount and/or the praise amount, displaying the analysis results corresponding to the second evaluation data as reply data of the second evaluation data if the support rate corresponding to the second evaluation data is higher than the preset support rate, and hiding the analysis results corresponding to the second evaluation data as hidden information of the monitoring video if the support rate corresponding to the second evaluation data is lower than the preset support rate, wherein the support rate corresponding to the second evaluation data is used for displaying response to a hidden request.
In some embodiments, the first type of user may be the basketball professional user described previously, such as a commentator, coach, etc.
In some embodiments, the second type of user may be the aforementioned general user, such as basketball lovers, etc.
In some embodiments, the preset matching degree may be set according to different application scenarios, which is not limited herein.
In some embodiments, if the matching degree between the first evaluation data and the analysis result corresponding to the first evaluation data is greater than the preset matching degree, it indicates that the reference value of the first evaluation data is relatively high, and the evaluation data may be displayed in a target display form. The target presentation may be specific, unlike presentation of other evaluation data. Such as a fly screen display, an enlarged display, etc.
In some embodiments, if the matching degree between the first evaluation data and the analysis result corresponding to the first evaluation data is smaller than the preset matching degree, it is indicated that the reference value of the first evaluation data is not so large, and at this time, the first evaluation data and the analysis result corresponding to the first evaluation data may be displayed to the watching user of the monitoring video, so that the watching user can see the analysis data from different layers. The display priority of the first evaluation data is lower than that of the analysis result corresponding to the first evaluation data. That is, the analysis result corresponding to the first evaluation data is preferentially displayed.
In some embodiments, for second evaluation data of a monitoring video of a second type, judging whether a support rate corresponding to the second evaluation data is higher than a preset support rate or not, wherein the support rate is determined based on browsing amount and/or praise amount, if the support rate corresponding to the second evaluation data is higher than the preset support rate, displaying an analysis result corresponding to the second evaluation data as reply data of the second evaluation data, if the support rate corresponding to the second evaluation data is lower than the preset support rate, displaying the second evaluation data and the analysis result corresponding to the second evaluation data as hidden information of the monitoring video, wherein the hidden information is used for displaying when responding to a display request.
In some embodiments, the support rate is determined based on the browsing volume and/or the praise volume, e.g., weighted averaging, arithmetic averaging, etc. of the browsing volume and the praise volume.
In some embodiments, the preset support rate may be set according to different application scenarios, which is not limited herein.
In some embodiments, different users may configure different hidden information display interfaces in a customized manner, when the user opens the hidden information display interface, the user may be regarded as initiating a display request, and at this time, an analysis result corresponding to the second evaluation data may be displayed as a hidden monitor video.
According to the embodiment of the application, the associated data of the monitoring video is obtained, the associated data comprises evaluation data of various types of users aiming at the monitoring video, a plurality of analysis keywords can be determined based on the associated data, and then the analysis strategy is determined by utilizing the analysis keywords, so that the monitoring video is analyzed by utilizing the association relation between the analysis strategy and various information in the analysis strategy, and the analysis results respectively corresponding to the analysis keywords are determined. By the analysis mode, more targeted analysis can be realized, and the safety of the analysis result of the monitoring video is improved. And the obtained analysis result is associated with the analysis keyword, so that the analysis result can be used for determining the safety protection strategy of the monitoring scene corresponding to the monitoring video, and the safety protection of the monitoring scene is realized. Therefore, the technical scheme can realize the safety protection of the monitoring scene based on the analysis result of the monitoring video, and improve the applicability of the analysis result of the monitoring video while ensuring the safety of the monitoring scene.
Referring next to fig. 3, an embodiment of the present application provides a video data-based security protection system, which includes a video analysis module 301 configured to obtain a surveillance video and associated data of the surveillance video, where the surveillance video is a surveillance video, the associated data includes evaluation data of multiple types of users for the surveillance video, determine multiple analysis keywords corresponding to the surveillance video based on the associated data, determine an analysis policy of the surveillance video based on the multiple analysis keywords, where the analysis policy includes at least one of a target analysis object, a target analysis behavior, and additional analysis information, and have an association relationship among the multiple analysis keywords, the target analysis object, the target analysis behavior, and the additional analysis information, determine an analysis result corresponding to the multiple analysis keywords based on the analysis policy, the association relationship, and the surveillance video, and a security protection module 302 configured to determine a security policy of a surveillance scene corresponding to the surveillance video based on the analysis result, and feed back the security policy.
The implementation of the video data based security system may refer to the foregoing examples and will not be repeated here.
As shown in fig. 4, the embodiment of the present application further provides a terminal device, which includes a processor 401 and a memory 402, where the processor 401 and the memory 402 are communicatively connected, and the terminal device may be used as an execution body of the foregoing video data security protection method.
The processor 401 and the memory 402 are directly or indirectly electrically connected to each other to realize data transmission or interaction. For example, electrical connections may be made between these elements through one or more communication buses or signal buses. The aforementioned video data based security methods each include at least one software functional module that may be stored in the memory 402 in the form of software or firmware (firmware).
The processor 401 may be an integrated circuit chip having signal processing capabilities. The processor 401 may be a general purpose processor including a CPU (Central Processing Unit ), NP (Network Processor, network processor), etc., and may also be a digital signal processor, an application specific integrated circuit, an off-the-shelf programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. Which may implement or perform the disclosed methods, steps, and logic blocks in embodiments of the invention. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 402 may store various software programs and modules, such as program instructions/modules corresponding to the image processing methods and apparatuses provided in the embodiments of the present application. The processor 401 executes various functional applications and data processing, i.e., implements the methods of embodiments of the present application, by running software programs and modules stored in the memory 402.
Memory 402 may include, but is not limited to, RAM (RandomAccess Memory ), ROM (Read Only Memory), PROM (Programmable Read-Only Memory, programmable Read Only Memory), EPROM (Erasable Programmable Read-Only Memory, erasable Read Only Memory), EEPROM (Electric Erasable Programmable Read-Only Memory), etc.
It will be appreciated that the configuration shown in fig. 4 is merely illustrative, and that the terminal device may also include more or fewer components than shown in fig. 4, or have a different configuration than shown in fig. 4.
It will be appreciated by those skilled in the art that 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, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
It will be appreciated by those skilled in the art that 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, CD-ROM, 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 flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations 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 foregoing descriptions of specific exemplary embodiments of the present application are presented for purposes of illustration and description. It is not intended to limit the application to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain the specific principles of the application and its practical application to thereby enable one skilled in the art to make and utilize the application in various exemplary embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the application be defined by the claims and their equivalents.