Intelligent Building Security Alarm System and Method Based on AI Visual Analysis

Abstract

The present invention discloses an intelligent building security alarm system and method based on AI visual analysis, which relates to the field of security technology. It acquires security building information, performs three-dimensional construction on the security building information to obtain a building model, acquires security building environment information and constructs the security building environment information into the building model to obtain a security model; security information is set corresponding to the security model, wherein the security information includes a plurality of security spaces and security items corresponding to the security spaces, and a cloud server chain corresponding to the security space is set based on the security information. The present invention can quickly detect the anomalies of objects to be security managed and perform subsequent security anomaly alarm processing in a timely manner, greatly improving the sensitivity and accuracy of building security alarms.

Description

Technical Field

[0001] The present invention relates to the field of security technologies, and particularly to an intelligent building security alarm system and method based on AI visual analysis. Background Art

[0002] The security industry is an industry that emerged in response to the security needs of modern society and can be divided into three categories: human security, physical security, and technical security. Among them, human security and physical security are traditional security means that have existed since ancient times and are the basis of security prevention. Among them, the prevention of fires is an important situation in buildings. Currently, in the early stage of a fire, due to the lack of obvious smoke and flames, it is difficult to quickly respond to fire hazards, which will lead to the continuous development of the fire and cause greater harm. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent building security alarm system and method based on AI visual analysis to solve the deficiencies in the background art.

[0004] To achieve the above purpose, the present invention provides the following technical solution: An intelligent building security alarm method based on AI visual analysis, comprising the following steps:

[0005] Obtain security building information, perform three-dimensional construction on the security building information to obtain a building model, obtain security building environment information, and construct the security building environment information into the building model to obtain a security model;

[0006] Set security information corresponding to the security model. Among them, the security information includes multiple security spaces and security items corresponding to the security spaces, and set a cloud server chain corresponding to the security space based on the security information;

[0007] Lay a security network corresponding to the security model. Among them, the security network includes point clusters, unit spaces, and unit space code positions, and construct the security network in the security model;

[0008] Obtain monitoring information in the building corresponding to the security item based on the cloud server chain. Among them, the monitoring information includes monitoring images and the positions where the monitoring images are located, and construct the monitoring information in the security model;

[0009] Track the monitoring information based on the security network to obtain a status index, and match the status index that does not meet the preset conditions with the corresponding alarm level for alarm.

[0010] In a preferred embodiment, the step of obtaining security building environment information and constructing the security building environment information into the building model to obtain a security model includes:

[0011] Obtain the building structure information to be security managed and the space information corresponding to the building structure as security building information;

[0012] Perform three-dimensional construction on the security building information to obtain a building model;

[0013] Obtain the item information and personnel information in the building to be security managed as the security building environment information, perform three-dimensional processing on the security building environment information and build it into the building model to obtain a security model.

[0014] In a preferred embodiment, the step of setting up a cloud server chain corresponding to the security space based on the security information includes:

[0015] Obtain the security requirement information of the building space to be security managed based on the security model, and divide the building space to be security managed based on the security requirement information to obtain multiple security spaces;

[0016] Configure corresponding security items for the corresponding security spaces, where the security items include the objects to be security managed and the corresponding security management rules;

[0017] Take the multiple security spaces and the security items corresponding to the multiple security spaces respectively as security information, set up sub-cloud servers corresponding to the security spaces through the security information, and communicate and connect the sub-cloud servers corresponding to the multiple security spaces to obtain a cloud server chain.

[0018] In a preferred embodiment, the step of laying out a security network corresponding to the security model includes:

[0019] Divide the multiple security spaces respectively into grid spaces based on the security model to obtain multiple unit spaces, and perform identity coding on the multiple unit spaces to obtain multiple unit space code positions;

[0020] Extract the object features of the objects to be security managed in the security space, where the object features include the object shape features and the object positions;

[0021] Configure multiple tracking prime points for the corresponding security spaces, lay out the tracking prime points based on the object features of the objects to be security managed, and associate the tracking prime points on a single object to be security managed to obtain a prime point group;

[0022] Obtain the unit space code positions where the tracking prime points in the prime point group are located, and bind the information of the tracking prime points and the unit space code positions;

[0023] Take the prime point group, the unit space and the unit space code positions as the security network, and build the security network in the security model.

[0024] In a preferred embodiment, the step of laying out the tracking prime points based on the object features of the objects to be security managed and associating the tracking prime points on a single object to be security managed to obtain a prime point group includes:

[0025] Based on the external shape characteristics of the object, the contour is depicted to obtain a contour line, a three-dimensional model of the object to be security managed is acquired, and the contour line is marked on the three-dimensional model to obtain a monitoring body;

[0026] Based on the division of the monitoring body by the contour line, multiple monitoring surfaces are obtained. Multiple tracking prime points are arranged on the monitoring surfaces, and the tracking prime points are associated with each other in a single unit space. The distances and position angles between the associated tracking prime points are stored in the corresponding unit space;

[0027] The associated tracking prime points are used as prime point groups.

[0028] In a preferred embodiment, the step of obtaining the monitoring information in the building of the corresponding security project based on the cloud server chain includes:

[0029] Based on the cloud server chain, the monitoring images in the building of the corresponding security project and the positions where the monitoring images are located are obtained as monitoring information;

[0030] The monitoring information is updated in real time in the security model.

[0031] In a preferred embodiment, the step of obtaining the status index by tracking the monitoring information based on the security network and matching the status index that does not meet the preset conditions with the corresponding alarm level for alarm includes:

[0032] Based on the monitoring information, the distances between the mutually associated tracking prime points in the object to be security managed and the position angles between the mutually associated tracking prime points are obtained and normalized;

[0033] Based on the monitoring information, the status index of the object to be security managed is obtained, and the numerical values in the monitoring information are processed;

[0034] The status index that does not meet the preset conditions is matched with the corresponding alarm level and an alarm is issued.

[0035] The present invention also provides an intelligent building security alarm system based on AI visual analysis, including:

[0036] An acquisition module, configured to acquire security building information, perform three-dimensional construction on the security building information to obtain a building model, acquire security building environment information, and construct the security building environment information into the building model to obtain a security model;

[0037] A setting module, connected to the acquisition module, configured to set security information corresponding to the security model. Among them, the security information includes multiple security spaces and security projects corresponding to the security spaces, and a cloud server chain corresponding to the security space is set based on the security information;

[0038] The layout module is connected to the device module and is used to layout a security network corresponding to the security model. Among them, the security network includes prime point groups, unit spaces, and unit space code positions, and the security network is constructed in the security model;

[0039] The monitoring module is connected to the layout module and is used to obtain the monitoring information in the building of the corresponding security project based on the cloud server chain. Among them, the monitoring information includes monitoring images and the locations where the monitoring images are located, and the monitoring information is constructed in the security model;

[0040] The alarm module is connected to the monitoring module and is used to track the monitoring information based on the security network to obtain a status index, and match the status index that does not meet the preset conditions with the corresponding alarm level for alarm.

[0041] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0042] After the present invention binds the tracking prime points and the unit space code positions for information, it performs security monitoring on the objects to be security managed. Regarding the prime point groups, unit spaces, and unit space code positions as the security network and constructing the security network in the security model, it has a good effect on monitoring the objects to be security managed, can quickly detect the abnormalities of the objects to be security managed, and perform subsequent security anomaly alarm processing in a timely manner, greatly improving the sensitivity and accuracy of building security alarms. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0044] Figure 1 It is a flowchart of the method of the present invention.

[0045] Figure 2 It is a system block diagram of the present invention. Detailed Embodiments

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0047] Embodiment 1, please refer to Figure 1As shown in the figure, the intelligent building security alarm method based on AI visual analysis in this embodiment includes the following steps:

[0048] S1. Obtain the security building information, perform three-dimensional construction on the security building information to obtain a building model, obtain the security building environment information, and build the security building environment information into the building model to obtain a security model;

[0049] S2. Set security information corresponding to the security model. Among them, the security information includes multiple security spaces and security items corresponding to the security spaces, and set the cloud server chain corresponding to the security space based on the security information;

[0050] S3. Arrange a security network corresponding to the security model. Among them, the security network includes point clusters, unit spaces, and unit space code positions, and build the security network into the security model;

[0051] S4. Obtain the monitoring information in the building corresponding to the security item based on the cloud server chain. Among them, the monitoring information includes monitoring images and the positions where the monitoring images are located, and build the monitoring information into the security model;

[0052] S5. Track the monitoring information based on the security network to obtain a status index, and match the status index that does not meet the preset conditions with the corresponding alarm level for alarm;

[0053] As described in the above steps S1-S5, the security industry is an industry that emerged in response to the security needs of modern society and can be divided into three categories: human security, physical security, and technical security; among them, human security and physical security are traditional security means that have existed since ancient times, and they are the basis of security prevention. Among them, the prevention of fires is an important situation in buildings. At present, in the early stage of a fire, due to the lack of obvious smoke and flames, it is difficult to quickly respond to fire hazards, which will lead to the continuous development of the fire and cause greater harm. In this application, after binding the tracking points and unit space code positions for information, security monitoring is carried out on the objects to be security managed. The point clusters, unit spaces, and unit space code positions are used as the security network, and the security network is built into the security model, which has a good effect on monitoring the objects to be security managed, can quickly detect abnormalities of the objects to be security managed, and perform subsequent security anomaly alarm processing in a timely manner, greatly improving the sensitivity and accuracy of building security alarms;

[0054] In one embodiment, the step S1 of obtaining the security building environment information and building the security building environment information into the building model to obtain a security model includes:

[0055] S11. Obtain the building structure information to be security managed and the spatial information corresponding to the building structure as the security building information;

[0056] S12. Perform three-dimensional construction on the security building information to obtain a building model;

[0057] S13. Obtain the item information and personnel information in the building to be security managed as the security building environment information, perform three-dimensional processing on the security building environment information and build it into the building model to obtain a security model;

[0058] As described in the above steps S11 - S13, in order to better monitor the security in the building, it is necessary to understand the building structure information of the building and the spatial information corresponding to the building structure as the security building information. Among them, the building structure is the layout structure of the building, such as independent spaces and public spaces in the building, etc. The spatial information corresponding to the building structure is, for example, the size of the independent space in the building. After that, a building model is obtained through three-dimensional modeling of the security building information, which can better clearly display the buildings in the building, facilitating the manager to understand the security situation in the building. Then, obtain the item information in the building to be security managed. In the early stage of a fire, items will deform. For example, the fire hazard caused by the electrical contact problem of equipment is easily overlooked here, and when items that cannot be damaged are damaged by external forces, it is easy to cause safety hazards, which can also be monitored through this system. Cooperating with traditional smoke and flame monitoring can better improve the efficiency of fire detection. The item information here is the public facilities in the building or other items to be security controlled, as well as the personnel information in the building. Building the item information and personnel information into the building model to obtain a security model can better manage security;

[0059] In one embodiment, step S2 of setting a cloud server chain corresponding to the security space based on the security information includes:

[0060] S21. Obtain the security requirement information of the building space to be security managed based on the security model, and divide the building space to be security managed based on the security requirement information to obtain multiple security spaces;

[0061] S22. Configure corresponding security items for the corresponding security spaces, where the security items include the objects to be security managed and the corresponding security management rules;

[0062] S23. Take the multiple security spaces and the security items corresponding to the multiple security spaces respectively as security information, set up sub-cloud servers corresponding to the security spaces through the security information, and communicate and connect the sub-cloud servers corresponding to the multiple security spaces to obtain a cloud server chain;

[0063] As described in the above steps S21 - S23, security requirement information in the building space to be security - managed is obtained based on the security model. Here, the security requirement information can, for example, but is not limited to, monitoring the location of items, monitoring the status of items, monitoring the dangerous behaviors of personnel, and monitoring environmental risk factors (such as sudden situations like fires). These are the factors for demand monitoring in the building. Then, according to the security requirement information in the building, the building space to be security - managed is divided into multiple security spaces, and corresponding security - management objects to be security - managed and corresponding security - management rules are configured for each security space as security projects. For example, the security - management objects to be security - managed can be various items, and then corresponding security - management rules are set. For example, the security - management rule for an item can be one or more of rules such as not being able to be moved from its position or not being violently attacked by personnel. Then, the multiple security spaces and the security projects corresponding to the multiple security spaces are used as security information. Then, a sub - cloud server for the corresponding security space is set through the security information, and the sub - cloud servers corresponding to the multiple security spaces are communicatively connected to obtain a cloud - server chain. Here, a sub - cloud server is set for the corresponding security space through the security space. This sub - cloud server is dedicated to managing the corresponding security space and has better security - processing efficiency. Communicatively connecting the multiple sub - cloud servers to obtain a cloud - server chain can make the information coherent and better reflect the overall security situation of the building;

[0064] In one embodiment, step S3 of deploying a security network according to the corresponding security model includes:

[0065] S31: Based on the security model, multiple unit spaces are obtained by dividing each of the multiple security spaces into grid spaces, and multiple unit - space code positions are obtained by performing identity coding on the multiple unit spaces;

[0066] S32: Extract the object features of the security - management objects to be security - managed in the security space. Among them, the object features include the object's external shape features and the object's location;

[0067] S33: Configure multiple tracking prime points for the security space, deploy the tracking prime points based on the object features of the security - management objects to be security - managed, and associate the tracking prime points on a single security - management object to be security - managed to obtain a prime - point group;

[0068] S34: Obtain the unit - space code positions where the tracking prime points in the prime - point group are located, and bind the tracking prime points with the unit - space code positions;

[0069] S35: Use the prime - point group, the unit space, and the unit - space code positions as the security network, and construct the security network in the security model;

[0070] As described in the above steps S31 - S35, in order to better conduct security monitoring on the objects to be security - managed in the building space, it is necessary to divide the grid space of multiple security spaces respectively to obtain multiple unit spaces, perform identity coding on the corresponding multiple unit spaces to obtain multiple unit - space code positions. The unit - space code positions represent the spatial positions of the unit spaces. Then, when conducting security monitoring on the objects to be security - managed in the unit spaces, it is necessary to extract the object features of the objects to be security - managed in the security space. Among them, the object features include the object shape features and the object position features. Configure multiple tracking points for the security space, arrange the tracking points based on the object features of the objects to be security - managed, and associate the tracking points on a single object to be security - managed to obtain a point cluster. Here, configuring multiple tracking points for the security space means that a security space will be configured with multiple tracking points on standby. Mark the object shape features and the object position features of the objects to be security - managed, obtain the unit - space code positions where the tracking points in the point cluster are located, and after binding the information of the tracking points and the unit - space code positions, the security monitoring of the objects to be security - managed is completed. Take the point cluster, the unit space, and the unit - space code position as a security network, and build the security network in the security model, which has a good effect on monitoring the objects to be security - managed, can quickly detect the abnormalities of the objects to be security - managed and promptly conduct subsequent security - anomaly alarm processing, greatly improving the sensitivity and accuracy of building security alarms;

[0071] In one embodiment, the step S33 of arranging the tracking points based on the object features of the objects to be security - managed and associating the tracking points on a single object to be security - managed to obtain a point cluster includes:

[0072] S331. Conduct contour drawing based on the object shape features to obtain a contour line, obtain the three - dimensional model of the object to be security - managed, and mark the contour line in the three - dimensional model to obtain a monitoring body;

[0073] S332. Divide the monitoring body based on the contour line to obtain multiple monitoring surfaces, arrange multiple tracking points on the monitoring surfaces, associate the tracking points with each other in a single unit space respectively, and store the distances and position angles between the associated tracking points in the corresponding unit space;

[0074] S333. Take the associated tracking points as the point cluster;

[0075] As described in the above steps S331 - S333, based on the object's external shape features, a contour is depicted to obtain a contour line. A three - dimensional model of the object to be security - managed is acquired, and the contour line is marked on the three - dimensional model to obtain a monitoring body. It is necessary to divide the monitoring body based on the contour line to obtain multiple monitoring surfaces. Multiple tracking points are arranged on the monitoring surfaces, and the tracking points are associated with each other in a single unit space. The distances and position angles between the associated tracking points are stored in the corresponding unit space. Here, in the unit space, the object to be security - managed is security - monitored through the tracking points. For example, when there are changes in the distances and position angles between the mutually - associated tracking points in the unit space, it means that there are security problems with the security - management object. Conducting monitoring on the corresponding object to be security - managed within a small range of the unit space can enable precise security monitoring of the security - management object and has an efficient response ability. For example, when there are precursors to a fire in a device, first, the device will have flames and a small amount of smoke that is not detected. Here, through the tracking points, the slight state changes of the device can be obtained. The tracking points are mutually - associated and there are a large number of them, providing more references for monitoring. Therefore, the deformation problem of the device due to the precursors to the fire can be accurately monitored, enabling early detection of security problems, timely alarm, and timely response;

[0076] In one embodiment, step S4 of obtaining the monitoring information in the building corresponding to the security project based on the cloud server chain includes:

[0077] S41. Obtain the monitoring images in the building corresponding to the security project and the locations where the monitoring images are located as monitoring information based on the cloud server chain;

[0078] S42. Update the monitoring information in the security model in real - time;

[0079] As described in the above steps S41 and S42, obtain the monitoring images in the building corresponding to the security project and the locations where the monitoring images are located as monitoring information based on the cloud server chain. Here, the monitoring images in the building corresponding to the security project and the locations where the monitoring images are located are collected by image - acquisition facilities and then transmitted to the corresponding sub - cloud servers. Then, the monitoring images in the building corresponding to the security project and the locations where the monitoring images are located are obtained from the sub - cloud servers, and the monitoring information is updated in the security model in real - time, enabling the administrator to clearly understand the situation;

[0080] In one embodiment, step S5 of obtaining a status index by tracking the monitoring information and triggering an alarm for the status index that does not meet the preset conditions by matching the corresponding alarm level includes:

[0081] S51. Obtain the distances between the mutually - associated tracking points in the object to be security - managed and the position angles between the mutually - associated tracking points based on the monitoring information and perform normalization processing;

[0082] S52. Obtain the status index for the object to be security - protected management corresponding to the monitoring information, and perform operations on the values in the monitoring information. Among them, the calculation formula of the status index is:

[0083] Among them, is the status index, ε is the security weight coefficient of the object to be security - protected management, L is the total distance difference between the mutually - related tracking prime points in the object to be security - protected management, V is the total difference value of the position angle values between the mutually - related tracking prime points, n is the number of tracking prime points with a distance difference between the mutually - related tracking prime points or a difference value of the position angle values between the mutually - related tracking prime points, μ is the number of unit spaces with a distance difference between the mutually - related tracking prime points or a difference value of the position angle values between the mutually - related tracking prime points, and γ and α are constants greater than zero;

[0084] S53. Match the status index that does not meet the preset conditions with the corresponding alarm level and give an alarm;

[0085] As described in the above steps S51 - S53, here is the monitoring of the tracking prime points based on the unit space, which can be more accurate. Based on the monitoring information, obtain the distance between the mutually - related tracking prime points and the position angle between the mutually - related tracking prime points in the object to be security - protected management and perform normalization processing. Here, obtain the status index for each object to be security - protected management one by one, match the corresponding alarm level according to the status index and give an alarm. The object to be security - protected management corresponding to the status index that does not meet the preset conditions represents an abnormal object, which can respond quickly and perform better security alarms;

[0086] Embodiment 2. Please refer to Figure 2 As shown, the intelligent building security alarm system based on AI visual analysis in this embodiment includes:

[0087] An acquisition module, used to acquire security building information, perform three - dimensional construction on the security building information to obtain a building model, acquire security building environment information and build the security building environment information into the building model to obtain a security model;

[0088] A setting module, connected to the acquisition module, used to set security information corresponding to the security model. Among them, the security information includes multiple security spaces and security items corresponding to the security spaces, and set the cloud server chain corresponding to the security space based on the security information;

[0089] A layout module, connected to the device module, used to layout a security network corresponding to the security model. Among them, the security network includes prime - point groups, unit spaces, and unit - space code positions, and build the security network into the security model;

[0090] The monitoring module, connected to the layout module, is used to obtain the monitoring information in the building of the corresponding security project based on the cloud server chain. The monitoring information includes the monitoring image and the location where the monitoring image is located, and constructs the monitoring information in the security model;

[0091] The alarm module, connected to the monitoring module, is used to obtain the status index by tracking the monitoring information based on the security network, and match the status index that does not meet the preset conditions with the corresponding alarm level for alarm.

[0092] As mentioned above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. An intelligent building security alarm method based on AI visual analysis, characterized in that: The following steps are involved: Acquire security building information, construct the security building information in three dimensions to obtain a building model, acquire security building environment information and construct the security building environment information into the building model to obtain a security model; Setting security information corresponding to the security model, wherein the security information includes a plurality of security spaces and security items corresponding to the security spaces, and setting a cloud server chain corresponding to the security space based on the security information; A security network is arranged corresponding to the security model, wherein the security network includes a cluster of points, a unit space and a unit space code position, and the security network is constructed in the security model; Based on the cloud server chain, the monitoring information in the building corresponding to the security project is obtained, wherein the monitoring information includes the monitoring image and the location of the monitoring image, and the monitoring information is constructed in the security model; Based on the security network, the monitoring information is tracked to obtain the status index, and the status index that does not meet the preset conditions is matched with the corresponding alarm level to issue an alarm; The step of deploying a security network corresponding to the security model includes: Based on the security model, multiple security spaces are divided into grid spaces to obtain multiple unit spaces, and identity encoding is performed on the multiple unit spaces to obtain multiple unit space code positions; Extracting object features of the object to be managed in the security space, wherein the object features include object shape features and object position; A plurality of tracking points are configured in the corresponding security space, the tracking points are arranged based on the object features of the object to be managed for security, and the tracking points on a single object to be managed for security are associated to obtain a point group; Obtain the unit space code position where the tracking element point in the element point group is located, and bind the tracking element point to the unit space code position; The point cluster, unit space and unit space code are used as a security network, and the security network is built into the security model; The step of arranging the tracking points based on the object features of the object to be managed for security, and associating the tracking points on a single object to be managed for security to obtain a point cluster includes: Based on the appearance features of the object, the contour is drawn to obtain a contour line, a three-dimensional model of the object to be managed is obtained, and the contour line is marked in the three-dimensional model to obtain a monitoring body; Based on the contour line, the monitoring object is divided into multiple monitoring surfaces, multiple tracking points are arranged on the monitoring surfaces, and the tracking points are associated with each other on a single object to be managed for security, and the distance and position angle between the two associated tracking points are stored in the associated tracking points; The tracked points after association are regarded as point clusters; The step of obtaining a status index by tracking the monitoring information based on the security network and matching the status index that does not meet the preset conditions with the corresponding alarm level to issue an alarm includes: Based on the monitoring information, the distances between the tracking points that are associated with each other and the position angles between the tracking points that are associated with each other in the object to be managed for security are obtained and normalized; Obtain status index of the object to be managed based on the monitoring information, and perform operations on the values ​​in the monitoring information; The status index that does not meet the preset conditions is matched with the corresponding alarm level and an alarm is issued.

2. The intelligent building security alarm method based on AI visual analysis according to claim 1 is characterized in that: The step of obtaining the security building environment information and building the security building environment information into the building model to obtain the security model includes: Acquire the building structure information to be managed for security and the space information corresponding to the building structure as security building information; Construct the security building information in three dimensions to obtain a building model; The information of items and personnel in the building to be managed for security is obtained as the security building environment information, and the security building environment information is processed in three dimensions and constructed into the building model to obtain a security model.

3. The intelligent building security alarm method based on AI visual analysis according to claim 1 is characterized in that: The step of setting up a cloud server chain corresponding to the security space based on the security information includes: Based on the security model, security demand information of the building space to be managed is obtained, and based on the security demand information, the building space to be managed is divided into multiple security spaces; Configure corresponding security projects for corresponding security spaces, where the security projects include objects to be managed and corresponding security management rules; A plurality of security spaces and security items corresponding to the plurality of security spaces are used as security information, sub-cloud servers corresponding to the security spaces are set according to the security information, and the sub-cloud servers corresponding to the plurality of security spaces are communicatively connected to obtain a cloud server chain.

4. The intelligent building security alarm method based on AI visual analysis according to claim 1 is characterized in that: The step of obtaining monitoring information in a building corresponding to a security project based on a cloud server chain includes: Based on the cloud server chain, the monitoring images in the building of the corresponding security project and the locations of the monitoring images are obtained as monitoring information; Update the monitoring information in the security model in real time.

5. An intelligent building security alarm system based on AI visual analysis, used to implement the intelligent building security alarm method based on AI visual analysis according to any one of claims 1 to 4, characterized in that: include: An acquisition module is used to acquire security building information, construct the security building information in three dimensions to obtain a building model, acquire security building environment information and construct the security building environment information into the building model to obtain a security model; A setting module, connected to the acquisition module, for setting security information corresponding to the security model, wherein the security information includes a plurality of security spaces and security items corresponding to the security spaces, and setting a cloud server chain corresponding to the security space based on the security information; A deployment module is connected to the equipment module and is used to deploy a security network corresponding to the security model, wherein the security network includes a cluster of points, a unit space and a unit space code position, and the security network is constructed in the security model; A monitoring module, connected to the deployment module, is used to obtain monitoring information in the building of the corresponding security project based on the cloud server chain, wherein the monitoring information includes monitoring images and locations of the monitoring images, and the monitoring information is constructed in the security model; The alarm module is connected to the monitoring module and is used to track the monitoring information based on the security network to obtain the status index, and to match the status index that does not meet the preset conditions with the corresponding alarm level to issue an alarm.

Images