Path guidance method and system based on alarm point monitoring and tracing

Abstract

The present application relates to the technical field of path navigation based on video recognition, including a path guidance method and system based on alarm point monitoring and tracing, the method including obtaining the current description alarm point and alarm event type of the current alarm event, obtaining the alarm point monitoring screen, extracting the real-time monitoring image, and judging whether the real-time monitoring image contains the current event feature; if it is judged to be yes, extracting the actual alarm subject, extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current description alarm point, generating the alarm point guidance path and the alarm point containment path; guiding the building security personnel to the current description alarm point according to the alarm point guidance path and the alarm point containment path. The present invention analyzes the real-time monitoring data, traces the historical monitoring of the alarm point, and thus achieves the effect of more efficient and safer processing of the alarm event.

Description

Technical Field

[0001] The present application relates to the technical field of path navigation based on video recognition, and in particular to a path guidance method and system based on alarm point monitoring and tracing. Background Art

[0002] With the increasing application of intelligent technology in daily life, people have more rapid, accurate and efficient requirements for emergency handling of security alarms. Improving safety and handling efficiency has always been a top priority, especially in places with high population density, such as schools, hotels, hospitals, smart buildings, etc. How to assist security personnel in path navigation and rapid handling after an alarm event occurs has become an urgent problem to be solved.

[0003] At present, the security personnel of buildings have a relatively simple way of handling general alarms. When they receive alarm information in the building, they randomly arrange personnel to handle it, such as arranging idle staff in the monitoring room to handle it. This handling method has the problems of unclear understanding of the alarm situation, unclear route planning, and mismatched security personnel arrangements, which easily leads to low efficiency in handling alarm incidents. Summary of the invention

[0004] Based on this, it is necessary to provide a path-guiding method and system based on alarm point monitoring and tracing that can analyze alarm situations through real-time monitoring images and video data, and trace the historical monitoring of alarm points to address the above technical problems, thereby achieving high processing efficiency.

[0005] The technical solution of the present invention is as follows:

[0006] A path guidance method based on alarm point monitoring and tracing, the method comprising:

[0007] Acquire the current description alarm point and alarm event type of the current alarm event, and acquire the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event;

[0008] Extracting the real-time monitoring image in the alarm point monitoring screen, and determining whether the real-time monitoring image contains the current event feature;

[0009] If the judgment is yes, then extract the actual alarm subject according to the real-time monitoring image, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data that currently describes the alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data;

[0010] The building security personnel are guided to the currently described alarm point according to the alarm point guidance path and the alarm point containment path.

[0011] Optionally, obtain the current description alarm point and alarm event type of the current alarm event, including:

[0012] Acquire alarm description information of the current alarm event, and extract event description information and current location description information from the alarm description information;

[0013] Generate an alarm event type based on a preset event type mapping function according to the event description information;

[0014] Calculating the intersection-and-joint ratio of the current position description information and the standard position description information of each preset alarm point, wherein the number of the preset alarm points is multiple, and each preset alarm point corresponds to a standard position description information;

[0015] According to the intersection-and-combination ratio, a current description alarm point matching the current position description information is screened out from each of the preset alarm points.

[0016] Optionally, the alarm point monitoring screen includes a current moment screen and a previous monitoring screen;

[0017] Obtaining the alarm point monitoring screen of the current description alarm point, including:

[0018] Obtain the current moment picture of the current description alarm point at the current time point;

[0019] generating a previous time point according to the current time point and a first time interval, wherein the first time interval is preset;

[0020] Acquire the previous monitoring screen between the previous time point and the current time point.

[0021] Optionally, the real-time monitoring image includes a current moment image and a previous monitoring image;

[0022] Extracting a real-time monitoring image from the alarm point monitoring screen, and determining whether the real-time monitoring image contains the current event feature, including:

[0023] Generate a current moment image according to the current moment picture, and extract current moment features from the current moment image;

[0024] Extracting images from the previous monitoring screen according to a preset number of image extraction frames, and obtaining a previous monitoring image;

[0025] Extracting previously monitored features from each of the previously monitored images;

[0026] Determine whether the current moment feature and each of the previously monitored features include the current event feature.

[0027] Optionally, if the judgment is yes, then extracting the actual alarm subject according to the real-time monitoring image, extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data that currently describes the alarm point, and setting the alarm point guidance path and the alarm point containment path according to the historical behavior data, including:

[0028] If the judgment is yes, extracting an initial alarm subject from the real-time monitoring image based on the current event feature, wherein the number of the initial alarm subjects is multiple;

[0029] Acquire the identity feature information of the initial alarm subject, and screen out the actual alarm subject from the initial alarm subject according to the identity feature information;

[0030] Extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current alarm point description, and extracting the used historical path used by the actual alarm subject from the historical behavior data;

[0031] Generate a current candidate path according to the estimated starting point and each of the used historical paths;

[0032] Acquire a current monitoring picture of the front-end monitoring points included in the current candidate path, remove blocked paths from the current candidate path according to the current monitoring picture, and generate remaining candidate paths;

[0033] Calculating the travel time of the remaining paths to be selected, and setting the remaining path to be selected corresponding to the shortest travel time of the path to be selected as the alarm point guidance path;

[0034] The paths other than the alarm point guiding path in the remaining candidate paths are set as alarm point blocking paths.

[0035] Optionally, after determining whether the real-time monitoring image contains the current event feature, the method further includes:

[0036] If the judgment is no, then determining the actual location of the event based on the current event characteristics according to the real-time monitoring image and the current description alarm point;

[0037] Extracting data of people entering and leaving the actual location where the event occurred from the historical monitoring data of the alarm point currently describing the alarm point;

[0038] Eliminate the subjects who have left the actual place where the event occurred according to the data of the people who have entered and exited, and obtain the subjects who have not left the actual place where the event occurred;

[0039] Setting an estimated alarm subject according to the subject that has not left;

[0040] Extracting the estimated behavior data of the estimated alarm subject from the alarm point historical monitoring data of the current alarm point, and extracting the estimated historical path used by the estimated alarm subject from the estimated behavior data;

[0041] The alarm point guidance path and the alarm point containment path are set according to the estimated historical path.

[0042] Optionally, guiding the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path includes:

[0043] Generate an alarm emergency level of the current alarm event according to the alarm description information of the current alarm event, and generate a minimum number of security personnel according to the alarm emergency level;

[0044] Determining whether the alarm emergency level is greater than or equal to a preset alarm level threshold;

[0045] If the judgment is yes, a nearby deployment instruction is generated, and temporary security personnel matching the minimum number of security personnel are screened out from the alarm point diffusion area corresponding to the currently described alarm point according to the nearby deployment instruction;

[0046] Adjust the alarm point guidance path according to the real-time location of the temporary security personnel and generate a real-time adjustment path, wherein one real-time adjustment path corresponds to one temporary security personnel;

[0047] Instructing each of the temporary security personnel to go to the currently described alarm point according to each of the real-time adjusted paths;

[0048] The building security personnel are guided to the currently described alarm point according to the alarm point containment path.

[0049] Optionally, generating the alarm urgency level of the current alarm event according to the alarm description information of the current alarm event includes:

[0050] Extracting current description words according to the alarm description information of the current alarm event, and extracting emergency sensitive words from the current description words, wherein the number of words in the current description words is N, and the number of the emergency sensitive words is n;

[0051] Extracting the alarm description time from the alarm description information, and obtaining a pre-stored normal speech speed, wherein the normal speech speed corresponds to a normal volume;

[0052] Obtaining the urgency score of each of the emergency sensitive words;

[0053] According to the preset volume collection points, the alarm description information is volume collected and multiple real-time collected volumes are obtained, and the real-time volume average of each of the real-time collected volumes is calculated;

[0054] Counting the real-time collected volumes that are greater than the real-time volume average among the real-time collected volumes, and marking them as current marked volumes, wherein the number of the current marked volumes is k;

[0055] Generate an alarm urgency value based on the urgency score, normal speech speed, real-time volume average, normal volume and current marked volume;

[0056] An alarm urgency level is generated according to the alarm urgency value.

[0057] The present invention also provides a path guidance system based on alarm point monitoring and tracing, the system comprising:

[0058] A monitoring screen acquisition module, used to acquire the current description alarm point and alarm event type of the current alarm event, and acquire the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event;

[0059] A monitoring screen confirmation module is used to extract the real-time monitoring image in the alarm point monitoring screen and determine whether the real-time monitoring image contains the current event feature;

[0060] A traceability path generation module is used to extract the actual alarm subject according to the real-time monitoring image if the judgment is yes, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data;

[0061] The path guidance module is used to guide the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path.

[0062] Optionally, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps described in the above-mentioned path guidance method based on alarm point monitoring and tracing when executing the computer program.

[0063] Optionally, a computer-readable storage medium is also provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps described in the above-mentioned path guidance method based on alarm point monitoring and tracing are implemented.

[0064] The technical effects achieved by the present invention are as follows:

[0065] The above-mentioned path guidance method and system based on alarm point monitoring and tracing includes obtaining the current description alarm point and alarm event type of the current alarm event, and obtaining the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event; extracting the real-time monitoring image in the alarm point monitoring screen, and judging whether the real-time monitoring image contains the current event characteristics; if it is judged to be yes, extracting the actual alarm subject according to the real-time monitoring image, extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current description alarm point, and setting the alarm point guidance path and the alarm point containment path according to the historical behavior data; guiding the building security personnel to the current description alarm point according to the alarm point guidance path and the alarm point containment path. In order to accurately locate and arrange security personnel to go to the current description alarm point for processing, the present invention needs to first obtain the real-time monitoring screen through the monitoring camera unit, specifically obtaining the alarm point monitoring screen of the current description alarm point. Different event types have different characteristics, and the corresponding characteristics can usually be analyzed through the monitoring screen. For example, in case of fire, it is possible to quickly confirm the presence of flames from the monitoring screen. In a fight, the people who are fighting can be extracted from the monitoring screen. Therefore, different alarm events correspond to different event features. Therefore, after obtaining the alarm event type, the current event feature of the current alarm event contained in the alarm event type can be obtained. Then, by extracting the real-time monitoring image in the alarm point monitoring screen and judging whether the real-time monitoring image contains the current event feature, it is judged whether the actual event is the same as the event described by the user. On the other hand, it is also based on this to confirm whether the current alarm situation is displayed in the real-time monitoring image. If it is judged to be yes, it is judged that the real-time monitoring image contains the current event feature. At this time, in order to understand the situation in advance before the building security personnel arrive at the alarm point, and then better handle the event, the actual alarm subject is extracted according to the real-time monitoring image. The actual alarm subject is the subject that causes the current alarm event. For example, the actual alarm subject is generally a party to the fight. Then, the historical behavior data of the actual alarm subject is extracted from the alarm point historical monitoring data of the current alarm point, and the alarm point guidance path and the alarm point containment path are set according to the historical behavior data. The historical behavior data is the data corresponding to the action trajectory of the actual alarm subject in the past time. The alarm point guidance path is used to indicate the path for some building security personnel to go to the current alarm point. When generating the alarm point guidance path, the main consideration is to enable the building security personnel to arrive as quickly as possible. The alarm point containment path is generated based on the historical behavior data, and is an estimated route for the actual alarm subject to escape after a fight.Next, by guiding the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path, some of the building security personnel can reach the currently described alarm point as soon as possible according to the alarm point guidance path to handle the alarm as soon as possible, and other building security personnel can pass through the possible escape route of the actual alarm-causing subject on the way to the currently described alarm point, so as to intercept the actual alarm-causing subject that may escape. On the one hand, the current event point is resolved, and on the other hand, possible events are prevented through the generation of an estimated path, that is, through the generation of the alarm point containment path. Therefore, when the present invention needs building security personnel to handle alarm events in the building, the present invention first confirms the event through real-time monitoring images, then traces the historical monitoring data of the alarm point, generates a path for the building security personnel to quickly go to the alarm point, and also generates a possible escape path for the alarm subject based on the traced historical data, and guides the building security personnel based on the path. Compared with the method of randomly arranging personnel to handle the alarm after receiving the alarm in the prior art, which is easy to bring about the problem of low efficiency in alarm event handling due to unclear understanding of the alarm situation, unclear path planning, and mismatched security personnel arrangements, the present invention analyzes the real-time monitoring data and traces the historical monitoring of the alarm point, thereby achieving the effect of more efficient and safer handling of alarm events. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flowchart of a path guidance method based on alarm point monitoring and tracing in one embodiment;

[0067] Figure 2 It is a structural block diagram of a path guidance system based on alarm point monitoring and tracing in one embodiment. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0069] In one embodiment, a terminal is provided, which is used to: obtain a currently described alarm point and an alarm event type of a current alarm event, and obtain an alarm point monitoring screen of the currently described alarm point, wherein the alarm event type includes current event features of the current alarm event; extract a real-time monitoring image in the alarm point monitoring screen, and determine whether the real-time monitoring image includes the current event features; if determined to be yes, extract the actual alarm subject according to the real-time monitoring image, extract historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the currently described alarm point, and set an alarm point guide path and an alarm point containment path according to the historical behavior data; and guide building security personnel to the currently described alarm point according to the alarm point guide path and the alarm point containment path.

[0070] The terminal may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices.

[0071] In one embodiment, Figure 1 As shown, a path guidance method based on alarm point monitoring and tracing is provided, and the method includes:

[0072] Step S100: obtaining the current description alarm point and the alarm event type of the current alarm event, and obtaining the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event feature of the current alarm event;

[0073] Step S200: extracting the real-time monitoring image in the alarm point monitoring screen, and determining whether the real-time monitoring image contains the current event feature;

[0074] Step S300: If the judgment is yes, extract the actual alarm subject according to the real-time monitoring image, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data that currently describes the alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data;

[0075] Step S400: guiding building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path.

[0076] This application is mainly used to handle and respond to disputes, disasters and other events that need to be handled initiated by users in the monitored building. Such events are collectively referred to as alarm events. It should be understood that the alarm in this application refers to the user reporting an alarm to the security personnel.

[0077] In this embodiment, a building security monitoring system is pre-installed in the building to be monitored, and the building security monitoring system includes a plurality of monitoring camera units, each of which is respectively arranged at each front-end monitoring point of the building to be monitored. When a user makes an alarm through a telephone, an intercom, etc., the current description alarm point and the alarm event type of the current alarm event are first obtained, and then in order to accurately locate and arrange security personnel to go to the current description alarm point for processing, it is necessary to first obtain a real-time monitoring screen through the monitoring camera unit, specifically to obtain the alarm point monitoring screen of the current description alarm point. Different event types have different features, and the corresponding features can usually be analyzed through the monitoring screen. For example, in the event of a fire, it is possible to quickly confirm the presence of an image of a flame from the monitoring screen. In a fight, the people who are entangled can be extracted from the monitoring screen. Therefore, different alarm events correspond to different event features, so after obtaining the alarm event type, the current event features of the current alarm event contained in the alarm event type can be obtained. Next, by extracting the real-time monitoring image in the alarm point monitoring screen and judging whether the real-time monitoring image contains the current event feature, it is judged whether the actual event is the same as the event described by the user. On the other hand, it is also based on this to confirm whether the real-time monitoring image shows the current alarm situation. If it is judged to be yes, it is judged that the real-time monitoring image contains the current event feature. At this time, in order to understand the situation in advance before the building security personnel arrive at the alarm point, and then better handle the event, the actual alarm subject is extracted according to the real-time monitoring image. The actual alarm subject is the subject that causes the current alarm event, such as the actual alarm subject is generally a party in a fight. Then, the historical behavior data of the actual alarm subject is extracted from the alarm point historical monitoring data of the current description of the alarm point, and the alarm point guidance path and the alarm point containment path are set according to the historical behavior data. The historical behavior data is the data corresponding to the action trajectory of the actual alarm subject in the past time. The alarm point guidance path is used to indicate the path for some building security personnel to go to the current description of the alarm point. When generating the alarm point guidance path, the main consideration is to enable the building security personnel to arrive as quickly as possible at the moment. The alarm point containment path is generated based on the historical behavior data, and is an estimated route for the actual alarm subject to escape after a fight.Next, by guiding the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path, some of the building security personnel can reach the currently described alarm point as soon as possible according to the alarm point guidance path to handle the alarm as soon as possible, and other building security personnel can pass through the possible escape route of the actual alarm-causing subject on the way to the currently described alarm point, so as to intercept the actual alarm-causing subject that may escape. On the one hand, the current event point is resolved, and on the other hand, possible events are prevented through the generation of an estimated path, that is, through the generation of the alarm point containment path. Therefore, when the present invention needs building security personnel to handle alarm events in the building, the present invention first confirms the event through real-time monitoring images, then traces the historical monitoring data of the alarm point, generates a path for the building security personnel to quickly go to the alarm point, and also generates a possible escape path for the alarm subject based on the traced historical data, and guides the building security personnel based on the path. Compared with the method of randomly arranging personnel to handle the alarm after receiving the alarm in the prior art, which is easy to bring about the problem of low efficiency in alarm event handling due to unclear understanding of the alarm situation, unclear path planning, and mismatched security personnel arrangements, the present invention analyzes the real-time monitoring data and traces the historical monitoring of the alarm point, thereby achieving the effect of more efficient and safer handling of alarm events.

[0078] In one embodiment, step S100: obtaining the current description alarm point and alarm event type of the current alarm event includes:

[0079] Step S111: Acquire alarm description information of the current alarm event, and extract event description information and current location description information from the alarm description information;

[0080] Step S112: generating an alarm event type according to the event description information based on a preset event type mapping function;

[0081] Step S113: calculating the intersection-and-joint ratio of the current position description information and the standard position description information of each preset alarm point, wherein the number of the preset alarm points is multiple, and each preset alarm point corresponds to a standard position description information;

[0082] Step S114: selecting a current description alarm point that matches the current position description information from each of the preset alarm points according to the intersection-to-union ratio.

[0083] In this embodiment, the event type mapping function is pre-set, and the event type mapping function is established by pre-setting the correspondence between the description information and the event type. If the description information is "fire, smoke, temperature rise, fire extinguisher and smoke detector", the event type is set to the fire event type. A function is established based on the correspondence between the two to generate the event type mapping function. Therefore, the alarm description information of the current alarm event is first obtained, and the event description information and the current position description information are extracted from the alarm description information, and then the event description information is input into the event type mapping function to generate the alarm event type. Then, when calculating the intersection and union ratio of the current position description information and the standard position description information of each preset alarm point, the current position set is first generated according to the current position description information, and then the standard position set is generated according to each standard position description information. One preset alarm point corresponds to one standard position set. Then, the intersection and union ratio of the current position set and each standard position set is calculated respectively. The calculation method of the intersection and union ratio is: divide the intersection of the current position set and the standard position set by the union of the current position set and the standard position set. For example, the current location set is {conference room on the third floor}, the number of standard location sets is two, the first standard location set is {first floor lobby, first floor entrance, front desk, elevator side}, the second standard location set is {second floor conference room, second floor tea room, utility room, third floor conference room, third floor tool room}; then, the intersection and union ratio of the current location set and the first standard location set is 0. The intersection of the current location set and the second standard location set is {third floor conference room}, and the union of the current location set and the second standard location set is {second floor conference room, second floor tea room, utility room, third floor conference room, third floor tool room}, then the intersection and union ratio of the two is 1 / 5. Obviously, 1 / 5 is greater than 0, then the preset alarm point corresponding to the second standard location set is set to the current description alarm point that matches the current location description information, that is, according to the intersection and union ratio, the current description alarm point that matches the current location description information is screened out from each preset alarm point.

[0084] Therefore, when a user issues an alarm, the alarm description information is obtained and analyzed in real time, so as to locate the accurate alarm location in time, and provide accurate endpoint data for subsequent effective processing and precise path navigation.

[0085] In one embodiment, the alarm point monitoring screen includes a current moment screen and a previous monitoring screen;

[0086] In step S100, obtaining the alarm point monitoring picture of the current alarm point description includes:

[0087] Step S121: obtaining the current moment picture of the current description alarm point at the current time point;

[0088] Step S122: generating a previous time point according to the current time point and a first time interval, wherein the first time interval is preset;

[0089] Step S123: Acquire previous monitoring images between the previous time point and the current time point.

[0090] In this embodiment, in order to obtain as much information as possible about the current event, information is not only obtained for the current time point, but also by presetting the first time interval, the time point corresponding to the first time interval before the current time point is the previous time point. For example, if the current time point is 6 o'clock and the first time interval is 5 minutes, then the previous time point is the time point corresponding to 5 minutes before 6 o'clock, that is, 5:55. Then the previous monitoring screen is the previous monitoring screen between 5:55 and 6 o'clock. Therefore, the alarm point monitoring screen not only includes the current time screen corresponding to the current time point, but also includes the previous monitoring screen before that, so as to obtain as much event information as possible by extending the monitoring time setting.

[0091] It should be noted that the first time interval is generally set to about 5 minutes, and usually less than 5 minutes. This will prevent the forward tracing time from being too long, and will prevent the event features from being extracted in time due to excessive data processing, or the subsequent path guidance from being unable to be carried out in time due to long data processing time.

[0092] In addition, the time length of the current time point compared to the time length of the first time interval can be understood as a time point. However, the current time point is not specifically limited to a certain second, such as not limited to only 3:42:12. It should be understood that the current time point can also be a shorter time period, such as between 3:42:12 and 3:42:16. Therefore, the current moment picture is not just a single picture accurate to the second, but a dynamic picture within a shorter time period.

[0093] In one embodiment, the real-time monitoring image includes a current moment image and a previous monitoring image;

[0094] In step S200, extracting the real-time monitoring image in the alarm point monitoring screen and determining whether the real-time monitoring image contains the current event feature includes:

[0095] Step S210: generating a current moment image according to the current moment picture, and extracting current moment features from the current moment image;

[0096] Step S220: extracting images from the previous monitoring screen according to a preset number of image extraction frames, and obtaining a previous monitoring image;

[0097] Step S230: extracting previously monitored features from each of the previously monitored images;

[0098] Step S240: Determine whether the current moment feature and each of the previously monitored features include the current event feature.

[0099] In this embodiment, in order to perform a refined comparison, image extraction and feature extraction are performed on the current moment screen and the previous monitoring screen respectively. First, the current moment image can be generated according to the current moment screen according to the image extraction frame, and then the current moment feature is extracted from the current moment image. Next, the image extraction is performed on the previous monitoring screen according to the preset number of image extraction frames, and the previous monitoring image is obtained. Then, the previous monitoring feature is extracted from each of the previous monitoring images, and finally, it is determined whether the current moment feature and each of the previous monitoring features contain the current event feature.

[0100] The current event feature matches the alarm event type. For example, when the alarm event type is a fight event, the current event feature is a punch feature, a scuffle feature, a kick feature, and a grappling feature. The punch feature, the scuffle feature, the kick feature, and the grappling feature are generated by collecting images of punching, scuffling, kicking, and grappling when the tester simulates a fight, and performing feature extraction.

[0101] For another example, when the alarm event type is fire, the current event feature is flame color feature, smoke feature, etc. Of course, fire is only an example, and this application mainly provides path guidance for fighting events.

[0102] Therefore, after extracting the current moment feature and the previously monitored feature, it is determined whether the current moment feature and each of the previously monitored features contain the current event feature.

[0103] In one embodiment, in step S300, if the judgment is yes, the actual alarm subject is extracted according to the real-time monitoring image, the historical behavior data of the actual alarm subject is extracted from the alarm point historical monitoring data that currently describes the alarm point, and the alarm point guidance path and the alarm point containment path are set according to the historical behavior data, including:

[0104] Step S310: if the judgment is yes, extracting the initial alarm subject from the real-time monitoring image based on the current event feature, wherein the number of the initial alarm subjects is multiple;

[0105] In this step, if the judgment is yes, it means that the subject causing the current alarm event can be seen through the monitored screen, so the subsequent judgment can be directly carried out, and the subject causing the alarm is confirmed first.

[0106] Step S320: obtaining identity feature information of the initial alarm subject, and screening out the actual alarm subject from the initial alarm subjects according to the identity feature information;

[0107] In this step, the identity feature information is obtained by analyzing the monitoring screen and image, including identification of clothing, body shape, posture, face recognition, height, etc. There are two situations for the identity feature information of the initial alarm subject. The first situation is that both or more parties in the fight belong to the area where the current description alarm point is located or the user or owner of the house near the current description alarm point. At this time, both or more parties in the fight are set as the actual alarm subject.

[0108] The second situation is that one of the two or more parties in the fight belongs to the user or owner of the house in the area where the current description alarm point is located or near the current description alarm point, and the other party belongs to a person outside the area where the current description alarm point is located and its vicinity, which can be understood as an outsider. In this case, the outsider enters the building to be monitored and has a dispute with the user in the building to be monitored. At this time, it is necessary to analyze the danger of the outsider, so at this time, the outsider is set as the actual alarm subject.

[0109] It should be noted that the fire incident is not the focus of this application, so it will not be described again, and only the fight incident will be explained.

[0110] Step S330: extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data describing the current alarm point, and extracting the used historical path used by the actual alarm subject from the historical behavior data;

[0111] In this step, the used historical path is the path that the actual alarm subject has walked in the past. Generally speaking, there are multiple used historical paths.

[0112] Step S340: generating a current candidate path according to the estimated starting point and each of the used historical paths;

[0113] In this embodiment, the estimated starting point is the location of the building security personnel waiting for instructions, which is generally the monitoring room. That is, the building security personnel waiting for instructions generally start from the estimated starting point and go to the currently described alarm point. The end point of each of the used historical paths is the currently described alarm point, but the starting point may be different because the actual alarm subject usually enters the monitored building from different doors. Therefore, in this case, the starting point of the used historical path is replaced, specifically replaced with the estimated starting point, and the current path to be selected is generated.

[0114] Step S350: obtaining a current monitoring picture of the front-end monitoring point included in the current candidate path, removing a blocked path from the current candidate path according to the current monitoring picture, and generating a remaining candidate path;

[0115] In this step, the front-end monitoring points included in the current candidate path are counted, and the current monitoring images collected by the monitoring camera units installed at the included front-end monitoring points are analyzed, and according to whether there are objects that hinder the passage of people in the current monitoring images, if there are, the current candidate path is marked as a blocked path, and the blocked path is eliminated. After the elimination process is completed, the remaining candidate paths are generated, and the remaining candidate paths are the paths that can be passed by people at present.

[0116] Step S360: calculating the travel time of the remaining paths to be selected, and setting the remaining path to be selected corresponding to the shortest travel time of the path to be selected as the alarm point guidance path;

[0117] Step S370: setting the paths other than the alarm point guiding path in the remaining candidate paths as alarm point blocking paths.

[0118] In this embodiment, when setting the remaining paths to be selected, by selecting from the paths that the actual alarm subject has walked, the actual alarm subject's habit of using the path is fully considered, and the probability of a part of the building security personnel meeting the escaping actual alarm subject in the process of going to the current described alarm point is increased, and the probability of the actual alarm subject successfully escaping after creating a fight is reduced, and the efficiency of processing the current alarm time is improved. At the same time, by setting the alarm point blocking path, another part of the building security personnel will then go to the current described alarm point, which can also prevent the probability of the actual alarm subject escaping. If the actual alarm subject escapes, it will generally escape according to its habit of using the path. By setting the alarm point guide path and the alarm point blocking path, and by guiding the two parts of the building security personnel to go to the current described alarm point in batches and by different paths, it is achieved that on the premise of rushing to the alarm point in the fastest time, the actual alarm subject is prevented from escaping with a high probability, thereby improving safety and event processing efficiency.

[0119] In addition, the method further includes a step of expanding the blocking path of the alarm point, which is specifically as follows:

[0120] The images collected by the monitoring camera units installed at the preset three-dimensional building model and all the front-end monitoring points are used for calculation, and multiple recommended guide paths are generated. The recommended guide paths that are different from the remaining paths to be selected are also set as alarm point containment paths. Of course, this case is used when there are enough building security personnel. That is, it is necessary to first calculate whether the number of building security personnel can meet the alarm point containment path. If so, they can be supplemented. If not, there is no need to supplement. It is only necessary to leave enough building security personnel stationed in the monitoring room or at the door.

[0121] In another embodiment, in step S360, after calculating the travel time of the remaining paths to be selected, the remaining paths to be selected corresponding to the travel time of the path to be selected with the shortest time are not directly set as the alarm point guidance path, but another method is used to set the alarm point guidance path, which is as follows:

[0122] Step S361: Calculate the path using the preset three-dimensional building model and the images collected by the monitoring camera units installed at all the front monitoring points, and generate multiple recommended guide paths;

[0123] Step S362: Calculating the recommended path travel time of each of the recommended guidance paths;

[0124] Step S363: Filter out the shortest travel time from the travel times of the recommended paths and the travel times of the candidate paths;

[0125] Step S364: setting the path corresponding to the shortest travel time as the alarm point guidance path;

[0126] Step S365: setting the paths among the plurality of recommended guiding paths and the remaining candidate paths except the alarm point guiding path as alarm point blocking paths.

[0127] In this embodiment, the three-dimensional building model is pre-modeled and generated based on the monitoring images collected by the monitoring camera units corresponding to the front-end monitoring points, and based on BIM technology. Specifically, software including but not limited to Bentley AECOsim Building Designer can be used for modeling. By setting the three-dimensional building model, after obtaining the current description alarm point, the images collected by the monitoring camera units installed at all the front-end monitoring points can be obtained and analyzed, and the path analysis combining three-dimensional modeling with actual monitoring can be realized, so as to screen out the recommended path that meets the actual use of the current building, and solve the problem in the prior art that the path cannot be used due to the current occupation of the path due to the setting of a fixed shortest path. In addition, after obtaining multiple recommended guide paths, the recommended path travel time of the recommended guide path is calculated, and then the shortest travel time is selected from the recommended path travel time and the selected path travel time. At this time, the path corresponding to the shortest travel time is set as the alarm point guide path. This setting is based on the first principle of reaching the current description alarm point to set the alarm point guide path. At this time, whether the first batch of building security personnel can encounter the actual alarm subject when going to the current description alarm point is no longer considered as a factor, but only how to quickly reach the current description alarm point is considered. However, because the alarm point containment path is also set, there is still a second batch of building security personnel who can carry out containment, which still does not affect the prevention of the actual alarm subject from escaping.

[0128] Furthermore, in actual use, the method of setting the alarm point guide path and the alarm point blocking path is set by technical personnel in this field according to actual production conditions and different considerations. The above two setting methods are only examples and are not specifically limited.

[0129] In one embodiment, in step S200, it is determined whether the real-time monitoring image contains the current event feature, and then the following steps are further included:

[0130] Step S510: if the judgment is no, then determining the actual location of the event based on the current event characteristics according to the real-time monitoring image and the current description alarm point;

[0131] Step S520: extracting data of people entering and leaving the actual location where the event occurred from the alarm point historical monitoring data describing the current alarm point;

[0132] Step S530: Eliminate the subjects who have left the actual place where the event occurred according to the data of the people who have entered and exited, and obtain the subjects who have not left the actual place where the event occurred;

[0133] Step S540: setting an estimated alarm subject according to the subject that has not left;

[0134] In step S510-step S540, if the judgment is no, that is, it is judged that the real-time monitoring image does not contain the current event feature, then the person who caused the current alarm event cannot be obtained from the current monitoring screen. For example, the fight occurred indoors, that is, the actual location of the event is indoors near the current alarm point, and general indoor monitoring cannot be linked to the monitoring room where the building security personnel are located. Therefore, it is necessary to analyze the unobtainable data and estimate the person who caused the current alarm event.

[0135] Furthermore, the actual location of the event is located first, and then the data of people entering and leaving the actual location of the event within a second time interval before the current time point is extracted from the historical monitoring data of the alarm point that currently describes the alarm point. The data of people entering and leaving the actual location of the event includes multiple people entering and multiple people leaving. The second time interval is preset, generally set to 10 minutes.

[0136] Next, the body features and faces of each of the entering and exiting persons are identified through the monitoring screen, thereby obtaining multiple actual entry and exit entities, each of which includes entry data and exit data.

[0137] Then, according to the entry data and exit data of each of the actual entry and exit subjects, the subjects who have left are eliminated, and the subjects who have not left are obtained. The subjects who have left are people who have left the place where the event occurred. The subjects who have not left are people who are still at the place where the event occurred, such as still staying indoors.

[0138] Next, the subjects who stayed at the place where the incident occurred and did not leave were analyzed, and the estimated alarm subjects were screened out. Specifically, during the analysis, the identities of the subjects who did not leave were also divided into two situations. The first situation was that both parties or multiple parties in the fight were users or owners of the place where the incident occurred. At this time, both parties or multiple parties in the fight were set as actual alarm subjects. The second situation was that one of the two parties or multiple parties in the fight was a user or owner of the place where the incident occurred, and the other party was an outsider outside the place where the incident occurred. In this case, an outsider entered the building to be monitored and had a dispute with the users in the building to be monitored. At this time, it is necessary to analyze the danger of the outsider, so the outsider is set as the actual alarm subject.

[0139] Step S550: extracting the estimated behavior data of the estimated alarm subject from the alarm point historical monitoring data describing the current alarm point, and extracting the estimated historical path used by the estimated alarm subject from the estimated behavior data;

[0140] Step S560: setting an alarm point guidance path and an alarm point containment path according to the estimated historical path.

[0141] In this embodiment, the execution of steps S550-S560 is consistent with the method of executing steps S330-S370. It can be understood that the estimated behavior data in step S550 is equivalent to the historical behavior data in step S330, and the estimated historical path in step S550 is equivalent to the used historical path in step S330. For the convenience of distinction, they are named separately. However, since the data processing flow is consistent, those skilled in the art should be able to know how to execute steps S550-S560 through steps S330-S370. Therefore, this application will not repeat the steps S550-S56.

[0142] In one embodiment, step S400: guiding the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path includes:

[0143] Step S410: generating an alarm emergency level of the current alarm event according to the alarm description information of the current alarm event, and generating a minimum number of security personnel according to the alarm emergency level;

[0144] In this step, the corresponding relationship between the alarm emergency level and the minimum number of security personnel is preset. Therefore, after obtaining the current alarm emergency level, the current minimum number of security personnel can be generated.

[0145] Step S420: determining whether the alarm emergency level is greater than or equal to a preset alarm level threshold;

[0146] Step S430: if the judgment is yes, then generate a nearby deployment instruction, and screen out temporary security personnel matching the minimum number of security personnel from the alarm point diffusion area corresponding to the currently described alarm point according to the nearby deployment instruction;

[0147] Step S440: adjusting the alarm point guide path according to the real-time location of the temporary security personnel and generating a real-time adjustment path, wherein one real-time adjustment path corresponds to one temporary security personnel;

[0148] Step S450: guiding each of the temporary security personnel to the currently described alarm point according to each of the real-time adjusted paths;

[0149] Step S460: guiding the building security personnel to the currently described alarm point according to the alarm point containment path.

[0150] In this embodiment, in order to handle alarm events of different urgency separately, it is necessary to first generate the alarm emergency level of the current alarm event according to the alarm description information, and generate the minimum number of security personnel according to the alarm emergency level. The minimum number of security personnel is positively correlated with the alarm emergency level. The higher the alarm emergency level, the more urgent and difficult the event to be handled is, so the more the minimum number of security personnel is required to ensure that there are enough security personnel to handle it. Then, if it is judged to be yes, a nearby deployment instruction is generated, and temporary security personnel matching the minimum number of security personnel are screened out from the alarm point diffusion area corresponding to the current description alarm point according to the nearby deployment instruction. The alarm point diffusion area is an area formed by uniformly radiating outwards a preset diffusion distance with the current description alarm point as the center. The temporary security personnel are personnel who are in the alarm point diffusion area and the event being handled is not urgent, or the event being handled can be suspended. Then, the real-time location of the temporary security personnel is used as a new starting point to adjust the alarm point guide path and generate a real-time adjustment path. During the adjustment, the real-time location is taken as the starting point, and the fastest path leading to any path in the alarm point guidance path is designed, and merged with the remaining alarm point guidance paths to generate a real-time adjustment path. Then, the temporary security personnel are guided to the current description alarm point according to the real-time adjustment paths; and the building security personnel are guided to the current description alarm point according to the alarm point containment path, thereby realizing adaptive personnel deployment according to events of different urgency, thereby realizing a highly secure event handling method and highly efficient path guidance.

[0151] In one embodiment, in step S410, generating the alarm urgency level of the current alarm event according to the alarm description information of the current alarm event includes:

[0152] Step S411: extracting current description words according to the alarm description information of the current alarm event, and extracting emergency sensitive words from the current description words, wherein the number of words in the current description words is N, and the number of the emergency sensitive words is n;

[0153] Step S412: extracting the alarm description time from the alarm description information, and obtaining a pre-stored normal speech speed, wherein the normal speech speed corresponds to a normal volume;

[0154] In this step, the normal speaking speed and normal volume are preset. The speaking speed corresponding to the user's normal conversation is set to the normal speaking speed, and the volume of the normal conversation is set to the normal volume.

[0155] Step S413: Obtaining the urgency score of each of the emergency sensitive words;

[0156] In this step, different emergency sensitive words are pre-set with corresponding urgency scores to indicate different urgency levels. For example, if the emergency sensitive words are "need help", "come", "come quickly" and "come immediately", the corresponding urgency scores are 3 points, 4 points, 6 points and 8 points respectively. For other emergency sensitive words and scores, no examples will be given one by one.

[0157] Step S414: collecting the volume of the alarm description information according to the preset volume collection points and obtaining multiple real-time collected volumes, and calculating the real-time volume average of each of the real-time collected volumes;

[0158] In this step, the volume collection points are pre-set, such as setting a volume collection point every 2 seconds. The real-time volume average is the average value of each real-time collected volume.

[0159] Step S415: Counting the real-time collected volumes that are greater than the real-time volume average among the real-time collected volumes, and marking them as currently marked volumes, wherein the number of the currently marked volumes is k;

[0160] Step S416: Generate an alarm urgency value based on the following formula according to the urgency score, normal speech speed, real-time volume average, normal volume and current marked volume:

[0161] ;

[0162] Among them, Ae is the alarm urgency value, n is the number of emergency sensitive words, Pri is the urgency score of the i-th emergency sensitive word, is the preset weighting coefficient, N is the number of words in the current description word, Vsd is the normal speaking speed, T is the alarm description time, Vua is the real-time volume average, Voa is the normal volume, k is the number of currently marked volumes, and Vmj is the jth currently marked volume;

[0163] Step S417: Generate an alarm urgency level according to the alarm urgency value.

[0164] In this embodiment, the weighting coefficient is preset. The alarm urgency value is generated by combining the urgency score, the normal speaking speed, the real-time volume average, the normal volume and the current marked volume, and the alarm urgency value is compared with the pre-stored standard alarm urgency value, and the standard urgency level corresponding to the matching standard alarm urgency value is obtained, and the standard urgency level is set as the alarm urgency level.

[0165] In one embodiment, Figure 2 As shown, a path guidance system based on alarm point monitoring and tracing is also provided, and the system includes:

[0166] A monitoring screen acquisition module, used to acquire the current description alarm point and alarm event type of the current alarm event, and acquire the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event;

[0167] A monitoring screen confirmation module is used to extract the real-time monitoring image in the alarm point monitoring screen and determine whether the real-time monitoring image contains the current event feature;

[0168] A traceability path generation module is used to extract the actual alarm subject according to the real-time monitoring image if the judgment is yes, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data;

[0169] The path guidance module is used to guide the building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path.

[0170] In one embodiment, the monitoring screen acquisition module is also used to: obtain the alarm description information of the current alarm event, and extract the event description information and the current position description information from the alarm description information; generate an alarm event type based on the event description information and a preset event type mapping function; calculate the intersection-and-union ratio of the current position description information and the standard position description information of each preset alarm point, wherein the number of the preset alarm points is multiple and each of the preset alarm points corresponds to a standard position description information; and filter out the current description alarm point that matches the current position description information from each of the preset alarm points according to the intersection-and-union ratio.

[0171] In one embodiment, the alarm point monitoring screen includes a current moment screen and a previous monitoring screen; the monitoring screen acquisition module is also used to: obtain the current moment screen that currently describes the alarm point at the current time point; generate a previous time point based on the current time point and a first time interval, wherein the first time interval is pre-set; obtain the previous monitoring screen between the previous time point and the current time point.

[0172] In one embodiment, the real-time monitoring image includes a current moment image and a previous monitoring image; the monitoring screen confirmation module is also used to: extract the image of the previous monitoring screen according to a preset number of image extraction frames, and obtain the previous monitoring image; extract the previous monitoring features from each of the previous monitoring images; and determine whether the current moment features and each of the previous monitoring features contain the current event features.

[0173] The traceability path generation module is also used for: if the judgment is yes, extracting the initial alarm subject from the real-time monitoring image based on the current event characteristics, wherein the number of the initial alarm subjects is multiple; obtaining the identity feature information of the initial alarm subject, and screening the actual alarm subject from the initial alarm subject according to the identity feature information; extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the alarm point currently describing the alarm point, and extracting the used historical path used by the actual alarm subject from the historical behavior data; generating the current candidate path according to the estimated starting point and each of the used historical paths; obtaining the current monitoring screen of the front-end monitoring point included in the current candidate path, eliminating the blocked path from the current candidate path according to the current monitoring screen, and generating the remaining candidate paths; calculating the candidate path travel time of the remaining candidate paths, and setting the remaining candidate paths corresponding to the shortest candidate path travel time as the alarm point guide path; setting the paths other than the alarm point guide path in each of the remaining candidate paths as the alarm point blocking path.

[0174] The traceability path generation module is also used to: calculate through the preset three-dimensional building model and the images collected by the monitoring camera units installed at all the front-end monitoring points, and generate multiple recommended guidance paths, and also set the recommended guidance paths that are different from the remaining paths to be selected as alarm point containment paths.

[0175] The traceability path generation module is also used to: calculate the path through the preset three-dimensional building model and the images collected by the monitoring camera units installed on all the front-end monitoring points, and generate multiple recommended guidance paths; calculate the recommended path travel time of each recommended guidance path; select the shortest travel time from the travel time of each recommended path and the travel time of each candidate path; set the path corresponding to the shortest travel time as the alarm point guidance path; set the paths other than the alarm point guidance path in each of the multiple recommended guidance paths and the remaining candidate paths as alarm point containment paths.

[0176] The monitoring screen confirmation module is also used for: if the judgment is no, determining the actual location of the event based on the current event characteristics according to the real-time monitoring image and the current description alarm point; extracting the data of people entering and leaving the actual location of the event from the historical monitoring data of the alarm point of the current description alarm point; eliminating the subjects who have left the actual location of the event according to the data of people entering and leaving, and obtaining the subjects who have not left the actual location of the event; setting the estimated alarm subject according to the subject who has not left; extracting the estimated behavior data of the estimated alarm subject from the historical monitoring data of the alarm point of the current description alarm point, and extracting the estimated historical path used by the estimated alarm subject from the estimated behavior data; setting the alarm point guidance path and the alarm point containment path according to the estimated historical path.

[0177] The path guidance module is also used to: generate the alarm emergency level of the current alarm event according to the alarm description information of the current alarm event, and generate the minimum number of security personnel according to the alarm emergency level; determine whether the alarm emergency level is greater than or equal to the preset alarm level threshold; if it is determined to be yes, generate a nearby deployment instruction, and screen out temporary security personnel matching the minimum number of security personnel from the alarm point diffusion area corresponding to the currently described alarm point according to the nearby deployment instruction; adjust the alarm point guidance path according to the real-time location of the temporary security personnel and generate a real-time adjustment path, wherein one real-time adjustment path corresponds to one temporary security personnel; guide each temporary security personnel to the currently described alarm point according to each real-time adjustment path; guide the building security personnel to the currently described alarm point according to the alarm point blockade path.

[0178] The path guidance module is also used to: extract the current description word according to the alarm description information of the current alarm event, and extract the emergency sensitive word from the current description word, wherein the number of words in the current description word is N, and the number of the emergency sensitive words is n; extract the alarm description time from the alarm description information, and obtain the pre-stored normal speaking speed, wherein the normal speaking speed corresponds to the normal volume; obtain the urgency score of each of the emergency sensitive words; perform volume collection on the alarm description information according to the preset volume collection point and obtain multiple real-time collection volumes, and calculate the real-time volume average of each of the real-time collection volumes; count the real-time collection volumes in each of the real-time collection volumes that are greater than the real-time volume average, and mark them as the current marked volume, wherein the number of the current marked volumes is k; generate the alarm urgency value based on the following formula according to the urgency score, normal speaking speed, real-time volume average, normal volume and current marked volume:

[0179] ;

[0180] Among them, Ae is the alarm urgency value, n is the number of emergency sensitive words, Pri is the urgency score of the i-th emergency sensitive word, is a preset weighting coefficient, N is the number of words in the current description word, Vsd is the normal speaking speed, T is the alarm description time, Vua is the real-time volume average, Voa is the normal volume, k is the number of currently marked volumes, and Vmj is the jth currently marked volume; an alarm urgency level is generated according to the alarm urgency value.

[0181] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps described in the above-mentioned path guidance method based on alarm point monitoring and tracing are implemented.

[0182] In one embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned path guidance method based on alarm point monitoring and tracing are implemented.

[0183] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0184] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0185] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A path guidance method based on alarm point monitoring and tracing, characterized in that: The method comprises: Acquire the current description alarm point and alarm event type of the current alarm event, and acquire the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event; Extracting the real-time monitoring image in the alarm point monitoring screen, and determining whether the real-time monitoring image contains the current event feature; If the judgment is yes, then extract the actual alarm subject according to the real-time monitoring image, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data that currently describes the alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data; Instruct building security personnel to go to the currently described alarm point according to the alarm point guidance path and the alarm point containment path; Get the current description alarm point and alarm event type of the current alarm event, including: Acquire alarm description information of the current alarm event, and extract event description information and current location description information from the alarm description information; Generate an alarm event type based on a preset event type mapping function according to the event description information; Calculating the intersection-and-joint ratio of the current position description information and the standard position description information of each preset alarm point, wherein the number of the preset alarm points is multiple, and each preset alarm point corresponds to a standard position description information; According to the intersection-and-combination ratio, a current description alarm point matching the current position description information is screened out from each of the preset alarm points.

2. The path guidance method based on alarm point monitoring and tracing according to claim 1 is characterized in that: The alarm point monitoring screen includes a current moment screen and a previous monitoring screen; Obtaining the alarm point monitoring screen of the current description alarm point, including: Obtain the current moment picture of the current description alarm point at the current time point; generating a previous time point according to the current time point and a first time interval, wherein the first time interval is preset; Acquire the previous monitoring screen between the previous time point and the current time point.

3. The path guidance method based on alarm point monitoring and tracing according to claim 1 is characterized in that: If the judgment is yes, then extract the actual alarm subject according to the real-time monitoring image, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data that currently describes the alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data, including: If the judgment is yes, extracting an initial alarm subject from the real-time monitoring image based on the current event feature, wherein the number of the initial alarm subjects is multiple; Acquire the identity feature information of the initial alarm subject, and screen out the actual alarm subject from the initial alarm subject according to the identity feature information; Extracting the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current alarm point description, and extracting the used historical path used by the actual alarm subject from the historical behavior data; Generate a current candidate path according to the estimated starting point and each of the used historical paths; Acquire a current monitoring picture of the front-end monitoring point included in the current candidate path, remove the blocked path from the current candidate path according to the current monitoring picture, and generate the remaining candidate paths; Calculating the travel time of the remaining paths to be selected, and setting the remaining path to be selected corresponding to the shortest travel time of the path to be selected as the alarm point guidance path; The paths other than the alarm point guiding path in the remaining candidate paths are set as alarm point blocking paths.

4. The path guidance method based on alarm point monitoring and tracing according to claim 1 is characterized in that: Determining whether the real-time monitoring image contains the current event feature, and then further comprising: If the judgment is no, then determining the actual location of the event based on the current event characteristics according to the real-time monitoring image and the current description alarm point; Extracting data of people entering and leaving the actual location where the event occurred from the historical monitoring data of the alarm point currently describing the alarm point; Eliminate the subjects who have left the actual place where the event occurred according to the data of the people who have entered and exited, and obtain the subjects who have not left the actual place where the event occurred; Setting an estimated alarm subject according to the subject that has not left; Extracting the estimated behavior data of the estimated alarm subject from the alarm point historical monitoring data of the current alarm point, and extracting the estimated historical path used by the estimated alarm subject from the estimated behavior data; The alarm point guidance path and the alarm point containment path are set according to the estimated historical path.

5. The path guidance method based on alarm point monitoring and tracing according to claim 1 is characterized in that: Guiding building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path includes: Generate an alarm emergency level of the current alarm event according to the alarm description information of the current alarm event, and generate a minimum number of security personnel according to the alarm emergency level; Determining whether the alarm emergency level is greater than or equal to a preset alarm level threshold; If the judgment is yes, a nearby deployment instruction is generated, and temporary security personnel matching the minimum number of security personnel are screened out from the alarm point diffusion area corresponding to the currently described alarm point according to the nearby deployment instruction; Adjust the alarm point guidance path according to the real-time location of the temporary security personnel and generate a real-time adjustment path, wherein one real-time adjustment path corresponds to one temporary security personnel; Instructing each of the temporary security personnel to go to the currently described alarm point according to each of the real-time adjusted paths; The building security personnel are guided to the currently described alarm point according to the alarm point containment path.

6. The path guidance method based on alarm point monitoring and tracing according to claim 5 is characterized in that: Generating the alarm urgency level of the current alarm event according to the alarm description information of the current alarm event includes: Extracting current description words according to the alarm description information of the current alarm event, and extracting emergency sensitive words from the current description words; Extracting the alarm description time from the alarm description information, and obtaining a pre-stored normal speech speed, wherein the normal speech speed corresponds to a normal volume; Obtaining the urgency score of each of the emergency sensitive words; According to the preset volume collection points, the alarm description information is volume collected and multiple real-time collected volumes are obtained, and the real-time volume average of each of the real-time collected volumes is calculated; Counting the real-time collected volumes that are greater than the real-time volume average among the real-time collected volumes, and marking them as current marked volumes; Generate an alarm urgency value based on the urgency score, normal speech speed, real-time volume average, normal volume and current marked volume; An alarm urgency level is generated according to the alarm urgency value.

7. A path guidance system based on alarm point monitoring and tracing, characterized in that: The system comprises: A monitoring screen acquisition module, used to acquire the current description alarm point and alarm event type of the current alarm event, and acquire the alarm point monitoring screen of the current description alarm point, wherein the alarm event type includes the current event characteristics of the current alarm event; A monitoring screen confirmation module is used to extract the real-time monitoring image in the alarm point monitoring screen and determine whether the real-time monitoring image contains the current event feature; A traceability path generation module is used to extract the actual alarm subject according to the real-time monitoring image if the judgment is yes, extract the historical behavior data of the actual alarm subject from the alarm point historical monitoring data of the current alarm point, and set the alarm point guidance path and the alarm point containment path according to the historical behavior data; A path guidance module, used to guide building security personnel to the currently described alarm point according to the alarm point guidance path and the alarm point containment path; The monitoring screen acquisition module is also used to: obtain the alarm description information of the current alarm event, and extract the event description information and the current position description information from the alarm description information; generate an alarm event type based on the event description information and a preset event type mapping function; calculate the intersection-and-union ratio of the current position description information and the standard position description information of each preset alarm point, wherein the number of the preset alarm points is multiple and each of the preset alarm points corresponds to a standard position description information; and filter out the current description alarm point that matches the current position description information from each of the preset alarm points according to the intersection-and-union ratio.

Images