A risk feature identification and analysis method for campus security monitoring systems

Abstract

The present invention discloses a risk feature identification and analysis method for a campus security monitoring system, which relates to the field of campus risk identification technology, including: using the campus security monitoring system to collect campus risk data to be identified to obtain campus risk features to be identified; performing risk feature identification processing based on the campus risk features to be identified to obtain campus risk features; and obtaining risk feature identification and analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features. The present invention realizes the intelligence of campus security monitoring by conducting in-depth analysis of data in the campus security monitoring system, and can automatically and accurately identify and analyze risk features, reducing the labor intensity and omission risk of manual monitoring, reducing the false alarm rate, improving monitoring efficiency, accuracy, and the intelligent level of campus safety management, and enhancing the ability to identify campus safety risks, thereby ensuring campus safety.

Description

Technical Field

[0001] The present invention relates to the technical field of campus risk identification, and in particular to a risk feature identification and analysis method for a campus security monitoring system. Background Art

[0002] With the increasing attention paid to campus safety, the real-time identification of various campus risks (such as fires, suspicious individuals, and emergencies) through security monitoring systems has become a pressing issue. However, existing campus security monitoring systems primarily rely on video surveillance, access control, and alarm systems to manage campus safety. However, these systems suffer from several issues, such as low monitoring efficiency, inaccurate risk identification, and high false alarm rates, making them ineffective in timely and effective detection and resolution of campus safety hazards. Therefore, a risk signature identification and analysis method for campus security monitoring systems is urgently needed to address these shortcomings. Summary of the Invention

[0003] The purpose of this invention is to propose a risk feature identification and analysis method for campus security monitoring systems. By automatically and accurately identifying and analyzing the risk features in the campus security monitoring system, the monitoring efficiency and accuracy are improved, the false alarm rate is reduced, and the ability to identify campus safety risks is enhanced to ensure campus safety.

[0004] To achieve the above objectives, the present invention provides a risk feature identification and analysis method for a campus security monitoring system, comprising the following steps:

[0005] S1. Use the campus security monitoring system to collect campus risk data to obtain campus risk characteristics;

[0006] S2. Perform risk feature identification processing based on the campus risk features to be identified to obtain campus risk features;

[0007] S3. Obtain risk feature identification and analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features.

[0008] Optionally, using the campus security monitoring system to collect campus risk data to obtain campus risk characteristics to be identified includes:

[0009] S1-1. Utilizing the campus security monitoring system to collect campus map information, campus video surveillance information, and campus facility monitoring information;

[0010] S1-2. Acquire the campus video surveillance information and the campus facility monitoring information as the campus risk data to be identified;

[0011] S1-3. Obtaining campus risk characteristics based on the campus risk data and the campus map information;

[0012] Among them, the campus facility monitoring information includes campus facility operating status data and campus facility operating environment data. The campus facility operating status data includes experimental equipment data, fire-fighting equipment data and sports facilities data. The campus facility operating environment data includes temperature and humidity data and air quality data around the campus facilities. The experimental equipment data is the operating parameters and usage parameters of the experimental equipment. The fire-fighting equipment data is the inspection records and operating status data of the fire-fighting equipment. The sports facility data is the operating status and usage parameters of the sports facilities.

[0013] Optionally, obtaining campus risk features to be identified based on the campus risk data to be identified and according to the campus map information includes:

[0014] S1-3-1. Separate and process the campus video surveillance information to obtain image data and audio data of the campus video surveillance information as campus video risk data to be identified;

[0015] S1-3-2. Obtain risk features of the campus videos to be identified based on the campus video risk data and the campus map information;

[0016] S1-3-3. Obtain risk characteristics of the campus facilities to be identified based on the campus facility monitoring information and the campus map information;

[0017] S1-3-3-1. Match the campus facility operation status data with the campus facility operation environment data using the campus map information according to the campus facility monitoring information to obtain a data pair of campus facility monitoring information;

[0018] S1-3-3-2. Perform time stamp sorting on the data pairs of the campus facility monitoring information to obtain continuous data pairs of the campus facility monitoring information;

[0019] S1-3-3-3. Perform linear regression analysis on the continuous data of the campus facility monitoring information to obtain a change trend of the campus facility monitoring information as the risk feature of the campus facility to be identified;

[0020] S1-3-4. Obtain the risk characteristics of the campus video to be identified and the risk characteristics of the campus facilities to be identified as the risk characteristics of the campus to be identified;

[0021] The data pair of the campus facility monitoring information is campus facility operation status data and campus facility operation environment data corresponding to the same location information.

[0022] Optionally, obtaining risk features of the campus video to be identified based on the campus video risk data and the campus map information includes:

[0023] S1-3-2-1. Use the campus map information to locate the campus video risk data to obtain the location information of the campus video risk data to be identified;

[0024] S1-3-2-2. Performing regional division processing on the campus video risk data to be identified according to the location information of the campus video risk data to be identified to obtain a regional division result of the campus video risk data to be identified;

[0025] S1-3-2-3. Perform time stamping processing based on the regional division result of the campus video risk data to be identified to obtain a timestamp of the regional division result;

[0026] S1-3-2-4. Perform time matching processing on the regional division results of the campus video risk data to be identified according to the timestamp of the regional division results to obtain a pair of campus video risk data to be identified;

[0027] S1-3-2-5. Obtain location information of campus image data and location information of campus audio data corresponding to the campus video risk data to be identified using the campus map information according to the campus video risk data to be identified;

[0028] S1-3-2-6. Determine whether the location information of the campus image data is consistent with the location information of the campus audio data. If so, directly execute S1-3-2-7. Otherwise, use the location information of the inconsistent campus image data and the location information of the campus audio data to update the location information of the campus video risk data to be identified, and return to S1-3-2-2.

[0029] S1-3-2-7. Obtaining the timestamp of the campus image data and the timestamp of the campus audio data corresponding to the campus video risk data pair to be identified according to the campus video risk data pair to be identified;

[0030] S1-3-2-8. Determine whether the timestamp of the campus image data is consistent with the timestamp of the campus audio data. If so, obtain the campus video risk data pair to be identified as the campus video risk feature to be identified. Otherwise, use the timestamp of the corresponding inconsistent campus image data and the timestamp of the campus audio data to update the timestamp of the region division result, and return to S1-3-2-4.

[0031] The pair of campus video risk data to be identified is campus video risk data to be identified corresponding to the same location information and the same time period.

[0032] Optionally, performing risk feature identification processing based on the campus risk features to be identified to obtain campus risk features includes:

[0033] S2-1. Obtaining campus video risk features according to a campus risk feature identification model based on the campus video risk features to be identified;

[0034] S2-2, performing risk feature identification based on the risk features of the campus facilities to be identified to obtain risk features of the campus facilities;

[0035] S2-2-1. Construct a historical campus facility risk feature database using the risk features of the campus facility to be identified and the corresponding historical campus facility risk features;

[0036] S2-2-2. Obtain risk characteristics of historical campus facilities using a campus facilities database based on the risk characteristics database of historical campus facilities to be identified;

[0037] S2-2-3. Obtaining initial campus facility risk characteristics based on the risk characteristics of the campus facilities to be identified and the historical campus facility risk characteristics;

[0038] S2-2-4. Determine whether the historical campus facility risk characteristics include the initial campus facility risk characteristics. If so, obtain the initial campus facility risk characteristics as the campus facility risk characteristics. Otherwise, execute S2-2-5.

[0039] S2-2-5. Determine whether the historical campus facility risk feature database contains the initial campus facility risk feature. If so, return to S2-2-3. Otherwise, update the campus facility risk feature based on the initial campus facility risk feature and return to S2-2-1.

[0040] S2-3. Obtain the campus risk characteristics based on the campus video risk characteristics and the campus facility risk characteristics;

[0041] Among them, the campus facility database includes the basic parameters and operating parameters of the campus facilities. The basic parameters of the campus facilities are the size specifications, rated voltage, rated power, carrying capacity and service life of the campus facilities. The operating parameters of the campus facilities are the operating time, operating environment requirements, maintenance cycle, failure frequency and safety indicators of the campus facilities.

[0042] Optionally, obtaining the campus video risk feature according to the campus risk feature identification model based on the campus video risk feature to be identified includes:

[0043] S2-1-1. Construct a historical campus video risk feature database based on the risk features of the campus video to be identified and the corresponding historical campus video risk features to be identified;

[0044] S2-1-2. Obtain risk features of a first campus video to be identified as a training set based on the historical campus video risk feature database;

[0045] S2-1-3. Obtain risk features of videos to be identified on a second campus based on the training set as a verification set;

[0046] S2-1-4. Using the training set as input and the campus video risk features corresponding to the training set as output, an initial campus risk feature recognition model is constructed based on a support vector machine;

[0047] S2-1-5. Input the verification set into the initial campus risk feature recognition model to obtain the campus video risk features corresponding to the verification set;

[0048] S2-1-6. Determine whether the campus video risk features corresponding to the verification set completely correspond to the campus video risk features corresponding to the training set. If so, obtain the initial campus risk feature recognition model as the campus risk feature recognition model and execute S2-1-7. Otherwise, update the training set according to the campus video risk features corresponding to the verification set and return to S2-1-2.

[0049] S2-1-7. Input the campus risk feature recognition model according to the risk feature of the campus video to be identified to obtain the campus video risk feature;

[0050] Among them, the campus risk feature identification model includes deep convolutional neural network, variational autoencoder, generative adversarial network, long short-term memory network and fusion network based on attention mechanism.

[0051] Optionally, inputting the campus risk feature recognition model into the campus risk feature recognition model according to the campus video risk feature to be recognized includes:

[0052] S2-1-7-1. Input the risk features of the campus video to be identified into the deep convolutional neural network to obtain the image features and audio features of the risk features of the campus video to be identified;

[0053] S2-1-7-2. Perform anomaly extraction processing on the image features of the campus video risk features to be identified according to the variational autoencoder and the generative adversarial network to obtain campus abnormal image features;

[0054] S2-1-7-3. Perform abnormal extraction processing on the audio features of the campus video risk features to be identified according to the long short-term memory network to obtain abnormal campus audio features;

[0055] S2-1-7-4. Based on the abnormal campus image features and the abnormal campus audio features, the campus video risk features are obtained according to the fusion network based on the attention mechanism.

[0056] Optionally, obtaining the campus risk feature according to the campus video risk feature and the campus facility risk feature includes:

[0057] S2-3-1. Obtain the campus video risk features corresponding to the training set as the historical campus video risk features;

[0058] S2-3-2. Determine whether the campus video risk feature corresponds to the historical campus video risk feature. If so, execute S2-3-3. Otherwise, use the campus video risk feature to update the campus video risk feature to be identified, and return to S2-1-1.

[0059] S2-3-3. Determine whether the campus facility risk characteristics correspond to the historical campus facility risk characteristics. If so, obtain the campus video risk characteristics and campus facility risk characteristics as the campus risk characteristics. Otherwise, use the campus facility risk characteristics to update the campus facility risk characteristics to be identified, and return to S2-2-1.

[0060] Optionally, obtaining risk feature identification analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features includes:

[0061] S3-1. Constructing a list of campus risk features to be identified based on the campus risk data to be identified and the campus risk features to be identified;

[0062] S3-2. Determine whether the list of campus risk features to be identified contains the campus risk feature. If so, obtain the location information and timestamp of the campus risk feature using the list of campus risk features to be identified based on the campus risk feature, and execute S3-3. Otherwise, update the campus risk feature to be identified using the campus risk feature, and return to S2-1.

[0063] S3-3. Performing location correlation analysis on the campus risk characteristics based on the location information of the campus risk characteristics to obtain distribution of the campus risk characteristics;

[0064] S3-4. Perform risk trend analysis on the campus risk characteristics according to the timestamp of the campus risk characteristics to obtain a change trend of the campus risk characteristics;

[0065] S3-5. Obtain a risk report of the campus risk characteristics based on the distribution of the campus risk characteristics and the changing trend of the campus risk characteristics as the risk characteristic identification and analysis result.

[0066] Optionally, constructing a list of campus risk features to be identified based on the campus risk data to be identified and the campus risk features to be identified includes:

[0067] S3-1-1. Perform time stamping processing according to the campus risk feature to be identified to obtain a timestamp of the campus risk feature to be identified;

[0068] S3-1-2. Based on the risk features to be identified on campus, obtain location information of the risk features to be identified on campus according to the campus map information;

[0069] S3-1-3. Obtain the campus risk data to be identified, the timestamp of the campus risk feature to be identified, and the location information of the campus risk feature to be identified, corresponding to the campus risk feature to be identified;

[0070] S3-1-4. Construct a list of the campus risk features to be identified based on the campus risk features to be identified and the campus risk data to be identified, timestamp and location information corresponding to the campus risk features to be identified.

[0071] Compared with the closest prior art, the present invention has the following beneficial effects:

[0072] The present invention obtains campus video risk features through a campus risk feature recognition model and can meet the risk feature recognition needs in different scenarios. It has strong adaptability and robustness. Among them, the combination of variational autoencoder and generative adversarial network can effectively discover abnormal campus image features that are different from normal feature distributions. The long short-term memory network can well process the audio features of the campus video risk features to be identified. By introducing the attention mechanism, the network can pay more attention to feature information that is more important for risk judgment according to specific circumstances. The present invention realizes the intelligence of campus security monitoring by deeply analyzing the data in the campus security monitoring system. It can automatically and accurately identify and analyze risk features, reduce the labor intensity and omission risk of manual monitoring, reduce the false alarm rate, improve monitoring efficiency, accuracy and the intelligence level of campus safety management, and enhance the ability to identify campus safety risks, thereby ensuring the safety of the campus. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0074] Figure 1 The present invention provides a flow chart of a risk feature identification and analysis method for a campus security monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0075] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present invention and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0076] The terms used in the embodiments of the present invention are only used to explain the specific embodiments of the present invention and are not intended to limit the present invention.

[0077] like Figure 1 As shown, an embodiment of the present invention provides a risk feature identification and analysis method for a campus security monitoring system, comprising the following steps:

[0078] S1. Use the campus security monitoring system to collect campus risk data to obtain campus risk characteristics;

[0079] S2. Perform risk feature identification processing based on the campus risk features to be identified to obtain campus risk features;

[0080] S3. Obtain risk feature identification and analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features.

[0081] S1 specifically includes:

[0082] S1-1. Utilizing the campus security monitoring system to collect campus map information, campus video surveillance information, and campus facility monitoring information;

[0083] S1-2. Acquire the campus video surveillance information and the campus facility monitoring information as the campus risk data to be identified;

[0084] S1-3. Obtaining campus risk characteristics based on the campus risk data and the campus map information;

[0085] The campus facility monitoring information includes campus facility operation status data and campus facility operation environment data.

[0086] In this embodiment, the campus facility operation status data includes experimental equipment data, firefighting equipment data, and sports facility data. Experimental equipment data refers to the operating parameters and usage parameters of the experimental equipment, where the operating parameters include the operating time, voltage, current, power, temperature, etc. of the experimental equipment, and the usage parameters include the usage time and frequency of use. Firefighting equipment data refers to the inspection records and operation status data of firefighting equipment, where the inspection records include the inspection time, inspection personnel, equipment status, etc. of firefighting equipment such as fire extinguishers and fire hydrants, and the operation status data includes abnormal data and alarm data of firefighting equipment. Sports facility data refers to the operation status and usage parameters of sports facilities, where the operation status includes parameters such as whether the equipment is operating normally, the operation mode, and the operation time, and the usage parameters include the frequency of use and wear data of sports equipment.

[0087] Campus facility operating environment data includes temperature and humidity data and air quality data around campus facilities. Air quality data includes harmful gas concentrations, which are used to detect the concentrations of harmful gases such as formaldehyde, benzene, and carbon dioxide in the air around campus facilities, and particulate matter concentrations, which are used to detect the content of particulate matter such as PM2.5 and PM10 in the air. This helps understand the campus air quality status, avoid excessive particulate matter concentrations that affect the respiratory health of teachers and students, and reduce pollution and damage to some precision instruments and equipment.

[0088] S1-3 specifically includes:

[0089] S1-3-1. Separate and process the campus video surveillance information to obtain image data and audio data of the campus video surveillance information as campus video risk data to be identified. In this embodiment, the video processing tool FFmpeg is used to separate and process the campus video surveillance information;

[0090] S1-3-2. Obtain risk features of the campus videos to be identified based on the campus video risk data and the campus map information;

[0091] S1-3-3. Obtain risk characteristics of the campus facilities to be identified based on the campus facility monitoring information and the campus map information;

[0092] S1-3-3-1. Match the campus facility operation status data with the campus facility operation environment data using the campus map information according to the campus facility monitoring information to obtain a data pair of campus facility monitoring information;

[0093] S1-3-3-2. Perform time stamp sorting on the data pairs of the campus facility monitoring information to obtain continuous data pairs of the campus facility monitoring information;

[0094] S1-3-3-3. Perform linear regression analysis on the continuous data of the campus facility monitoring information to obtain a change trend of the campus facility monitoring information as the risk feature of the campus facility to be identified;

[0095] S1-3-4. Obtain the risk characteristics of the campus video to be identified and the risk characteristics of the campus facilities to be identified as the risk characteristics of the campus to be identified;

[0096] The data pair of the campus facility monitoring information is campus facility operation status data and campus facility operation environment data corresponding to the same location information.

[0097] S1-3-2 specifically includes:

[0098] S1-3-2-1. Use the campus map information to locate the campus video risk data to obtain the location information of the campus video risk data to be identified;

[0099] S1-3-2-2. Performing regional division processing on the campus video risk data to be identified according to the location information of the campus video risk data to be identified to obtain a regional division result of the campus video risk data to be identified;

[0100] S1-3-2-3. Perform time stamping processing based on the regional division result of the campus video risk data to be identified to obtain a timestamp of the regional division result;

[0101] S1-3-2-4. Perform time matching processing on the regional division results of the campus video risk data to be identified according to the timestamp of the regional division results to obtain a pair of campus video risk data to be identified;

[0102] S1-3-2-5. Obtain location information of campus image data and location information of campus audio data corresponding to the campus video risk data to be identified using the campus map information according to the campus video risk data to be identified;

[0103] S1-3-2-6. Determine whether the location information of the campus image data is consistent with the location information of the campus audio data. If so, directly execute S1-3-2-7. Otherwise, use the location information of the inconsistent campus image data and the location information of the campus audio data to update the location information of the campus video risk data to be identified, and return to S1-3-2-2.

[0104] S1-3-2-7. Obtaining the timestamp of the campus image data and the timestamp of the campus audio data corresponding to the campus video risk data pair to be identified according to the campus video risk data pair to be identified;

[0105] S1-3-2-8. Determine whether the timestamp of the campus image data is consistent with the timestamp of the campus audio data. If so, obtain the campus video risk data pair to be identified as the campus video risk feature to be identified. Otherwise, use the timestamp of the corresponding inconsistent campus image data and the timestamp of the campus audio data to update the timestamp of the region division result, and return to S1-3-2-4.

[0106] The pair of campus video risk data to be identified is campus video risk data to be identified corresponding to the same location information and the same time period.

[0107] S2 specifically includes:

[0108] S2-1. Obtaining campus video risk features according to a campus risk feature identification model based on the campus video risk features to be identified;

[0109] S2-2, performing risk feature identification based on the risk features of the campus facilities to be identified to obtain risk features of the campus facilities;

[0110] S2-2-1. Construct a historical campus facility risk feature database using the risk features of the campus facility to be identified and the corresponding historical campus facility risk features;

[0111] S2-2-2. Obtain risk characteristics of historical campus facilities using a campus facilities database based on the risk characteristics database of historical campus facilities to be identified;

[0112] S2-2-3. Obtaining initial campus facility risk characteristics based on the risk characteristics of the campus facilities to be identified and the historical campus facility risk characteristics;

[0113] S2-2-4. Determine whether the historical campus facility risk characteristics include the initial campus facility risk characteristics. If so, obtain the initial campus facility risk characteristics as the campus facility risk characteristics. Otherwise, execute S2-2-5.

[0114] S2-2-5. Determine whether the historical campus facility risk feature database contains the initial campus facility risk feature. If so, return to S2-2-3. Otherwise, update the campus facility risk feature based on the initial campus facility risk feature and return to S2-2-1.

[0115] S2-3. Obtain the campus risk characteristics based on the campus video risk characteristics and the campus facility risk characteristics;

[0116] The campus facilities database includes basic parameters and operating parameters of campus facilities.

[0117] In this embodiment, the basic parameters of the campus facilities include the size specifications, rated voltage, rated power, carrying capacity, service life, etc. of the campus facilities, and the operating parameters of the campus facilities include the operating time, operating environment requirements, maintenance cycle, failure frequency, safety indicators, etc. of the campus facilities.

[0118] S2-1 specifically includes:

[0119] S2-1-1. Construct a historical campus video risk feature database based on the risk features of the campus video to be identified and the corresponding historical campus video risk features to be identified;

[0120] S2-1-2. Obtain risk features of a first campus video to be identified as a training set based on the historical campus video risk feature database;

[0121] S2-1-3. Obtain risk features of videos to be identified on a second campus based on the training set as a verification set;

[0122] S2-1-4. Using the training set as input and the campus video risk features corresponding to the training set as output, an initial campus risk feature recognition model is constructed based on a support vector machine;

[0123] S2-1-5. Input the verification set into the initial campus risk feature recognition model to obtain the campus video risk features corresponding to the verification set;

[0124] S2-1-6. Determine whether the campus video risk features corresponding to the verification set completely correspond to the campus video risk features corresponding to the training set. If so, obtain the initial campus risk feature recognition model as the campus risk feature recognition model and execute S2-1-7. Otherwise, update the training set according to the campus video risk features corresponding to the verification set and return to S2-1-2.

[0125] S2-1-7. Input the campus risk feature recognition model according to the risk feature of the campus video to be identified to obtain the campus video risk feature;

[0126] Among them, the campus risk feature identification model includes deep convolutional neural network, variational autoencoder, generative adversarial network, long short-term memory network and fusion network based on attention mechanism.

[0127] S2-1-7 specifically includes:

[0128] S2-1-7-1. Input the risk features of the campus video to be identified into the deep convolutional neural network to obtain the image features and audio features of the risk features of the campus video to be identified. Specifically:

[0129] Deep convolutional neural networks can learn the different characteristic patterns of audio and image in the risk features of campus videos to be identified. Using operations such as convolutional layers and pooling layers, they can separate the image features from the audio features of the risk features of campus videos to be identified.

[0130] S2-1-7-2. Perform abnormal extraction processing on the image features of the campus video risk features to be identified using the variational autoencoder and the generative adversarial network to obtain abnormal campus image features, specifically:

[0131] A variational autoencoder is used to learn the normal feature distribution corresponding to the image features of the risk features to be identified in campus videos. Through adversarial training with a generative adversarial network, it can effectively detect abnormal campus image features that differ from the normal feature distribution, such as abnormal crowds and vehicles on campus.

[0132] S2-1-7-3. Perform abnormal extraction processing on the audio features of the campus video risk features to be identified based on the long short-term memory network to obtain abnormal campus audio features, specifically:

[0133] The LSTM network can effectively process the time series information corresponding to the audio features of the risk features of the campus video to be identified, remember the long-term dependencies in the audio features, and has a good extraction effect on abnormal campus audio features with time-varying characteristics, such as continuous abnormal noise.

[0134] S2-1-7-4. Based on the abnormal campus image features and the abnormal campus audio features, the campus video risk features are obtained according to the fusion network based on the attention mechanism, specifically:

[0135] By introducing an attention mechanism, we dynamically assign weights to abnormal campus image features and abnormal campus audio features, enabling the network to focus more on feature information that is more important for risk judgment based on specific circumstances. For example, when judging fire risks, we may pay more attention to the alarm sound in the audio and the delay features in the image. The attention mechanism can highlight the role of these key features in fusion.

[0136] S2-3 specifically includes:

[0137] S2-3-1. Obtain the campus video risk features corresponding to the training set as the historical campus video risk features;

[0138] S2-3-2. Determine whether the campus video risk feature corresponds to the historical campus video risk feature. If so, execute S2-3-3. Otherwise, use the campus video risk feature to update the campus video risk feature to be identified, and return to S2-1-1.

[0139] S2-3-3. Determine whether the campus facility risk characteristics correspond to the historical campus facility risk characteristics. If so, obtain the campus video risk characteristics and campus facility risk characteristics as the campus risk characteristics. Otherwise, use the campus facility risk characteristics to update the campus facility risk characteristics to be identified, and return to S2-2-1.

[0140] S3 specifically includes:

[0141] S3-1. Constructing a list of campus risk features to be identified based on the campus risk data to be identified and the campus risk features to be identified;

[0142] S3-2. Determine whether the list of campus risk features to be identified contains the campus risk feature. If so, obtain the location information and timestamp of the campus risk feature using the list of campus risk features to be identified based on the campus risk feature, and execute S3-3. Otherwise, update the campus risk feature to be identified using the campus risk feature, and return to S2-1.

[0143] S3-3. Performing location correlation analysis on the campus risk characteristics based on the location information of the campus risk characteristics to obtain distribution of the campus risk characteristics;

[0144] S3-4. Perform risk trend analysis on the campus risk characteristics according to the timestamp of the campus risk characteristics to obtain a change trend of the campus risk characteristics;

[0145] S3-5. Obtain a risk report of the campus risk characteristics based on the distribution of the campus risk characteristics and the changing trend of the campus risk characteristics as the risk characteristic identification and analysis result.

[0146] S3-1 specifically includes:

[0147] S3-1-1. Perform time stamping processing according to the campus risk feature to be identified to obtain a timestamp of the campus risk feature to be identified;

[0148] S3-1-2. Based on the risk features to be identified on campus, obtain location information of the risk features to be identified on campus according to the campus map information;

[0149] S3-1-3. Obtain the campus risk data to be identified, the timestamp of the campus risk feature to be identified, and the location information of the campus risk feature to be identified, corresponding to the campus risk feature to be identified;

[0150] S3-1-4. Construct a list of the campus risk features to be identified based on the campus risk features to be identified and the campus risk data to be identified, timestamp and location information corresponding to the campus risk features to be identified.

[0151] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0152] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0153] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0154] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A risk feature identification and analysis method for a campus security monitoring system, characterized in that: The specific steps include: S1. Use the campus security monitoring system to collect campus risk data to obtain campus risk characteristics; Using the campus security monitoring system to collect campus risk data to obtain campus risk characteristics to be identified includes: S1-1. Utilizing the campus security monitoring system to collect campus map information, campus video surveillance information, and campus facility monitoring information; S1-2. Acquire the campus video surveillance information and the campus facility monitoring information as the campus risk data to be identified; S1-3. Obtaining campus risk characteristics based on the campus risk data and the campus map information; The campus facility monitoring information includes campus facility operating status data and campus facility operating environment data. The campus facility operating status data includes experimental equipment data, firefighting equipment data, and sports facility data. The campus facility operating environment data includes temperature, humidity, and air quality data around the campus facilities. The experimental equipment data includes operating parameters and usage parameters of the experimental equipment. The firefighting equipment data includes inspection records and operating status data of the firefighting equipment. The sports facility data includes operating status and usage parameters of the sports facilities. Acquiring campus risk features based on the campus risk data to be identified and the campus map information includes: S1-3-1. Separate and process the campus video surveillance information to obtain image data and audio data of the campus video surveillance information as campus video risk data to be identified; S1-3-2. Obtain risk features of the campus videos to be identified based on the campus video risk data and the campus map information; S1-3-3. Obtain risk characteristics of the campus facilities to be identified based on the campus facility monitoring information and the campus map information; S1-3-3-1. Match the campus facility operation status data with the campus facility operation environment data using the campus map information according to the campus facility monitoring information to obtain a data pair of campus facility monitoring information; S1-3-3-2. Perform time stamp sorting on the data pairs of the campus facility monitoring information to obtain continuous data pairs of the campus facility monitoring information; S1-3-3-3. Perform linear regression analysis on the continuous data of the campus facility monitoring information to obtain a change trend of the campus facility monitoring information as the risk feature of the campus facility to be identified; S1-3-4. Obtain the risk characteristics of the campus video to be identified and the risk characteristics of the campus facilities to be identified as the risk characteristics of the campus to be identified; The data pair of the campus facility monitoring information is campus facility operation status data and campus facility operation environment data corresponding to the same location information; S2. Perform risk feature identification processing based on the campus risk features to be identified to obtain campus risk features; S3. Obtain risk feature identification and analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features.

2. A risk feature identification and analysis method for a campus security monitoring system according to claim 1, characterized in that: Acquiring risk features of campus videos to be identified based on the campus video risk data and the campus map information includes: S1-3-2-1. Use the campus map information to locate the campus video risk data to obtain the location information of the campus video risk data to be identified; S1-3-2-2. Performing regional division processing on the campus video risk data to be identified according to the location information of the campus video risk data to be identified to obtain a regional division result of the campus video risk data to be identified; S1-3-2-3. Perform time stamping processing based on the regional division result of the campus video risk data to be identified to obtain a timestamp of the regional division result; S1-3-2-4. Perform time matching processing on the regional division results of the campus video risk data to be identified according to the timestamp of the regional division results to obtain a pair of campus video risk data to be identified; S1-3-2-5. Obtain location information of campus image data and location information of campus audio data corresponding to the campus video risk data to be identified using the campus map information according to the campus video risk data to be identified; S1-3-2-6. Determine whether the location information of the campus image data is consistent with the location information of the campus audio data. If so, directly execute S1-3-2-7. Otherwise, use the location information of the inconsistent campus image data and the location information of the campus audio data to update the location information of the campus video risk data to be identified, and return to S1-3-2-2. S1-3-2-7. Obtaining the timestamp of the campus image data and the timestamp of the campus audio data corresponding to the campus video risk data pair to be identified according to the campus video risk data pair to be identified; S1-3-2-8. Determine whether the timestamp of the campus image data is consistent with the timestamp of the campus audio data. If so, obtain the campus video risk data pair to be identified as the campus video risk feature to be identified. Otherwise, use the timestamp of the corresponding inconsistent campus image data and the timestamp of the campus audio data to update the timestamp of the region division result, and return to S1-3-2-4. The pair of campus video risk data to be identified is campus video risk data to be identified corresponding to the same location information and the same time period.

3. The risk feature identification and analysis method for a campus security monitoring system according to claim 2, characterized in that: Performing risk feature identification processing based on the campus risk features to be identified to obtain campus risk features includes: S2-1. Obtaining campus video risk features according to a campus risk feature identification model based on the campus video risk features to be identified; S2-2, performing risk feature identification based on the risk features of the campus facilities to be identified to obtain risk features of the campus facilities; S2-2-1. Construct a historical campus facility risk feature database using the risk features of the campus facility to be identified and the corresponding historical campus facility risk features; S2-2-2. Obtain risk characteristics of historical campus facilities using a campus facilities database based on the risk characteristics database of historical campus facilities to be identified; S2-2-3. Obtaining initial campus facility risk characteristics based on the risk characteristics of the campus facilities to be identified and the historical campus facility risk characteristics; S2-2-4. Determine whether the historical campus facility risk characteristics include the initial campus facility risk characteristics. If so, obtain the initial campus facility risk characteristics as the campus facility risk characteristics. Otherwise, execute S2-2-5. S2-2-5. Determine whether the historical campus facility risk feature database contains the initial campus facility risk feature. If so, return to S2-2-3. Otherwise, update the campus facility risk feature based on the initial campus facility risk feature and return to S2-2-1. S2-3. Obtain the campus risk characteristics based on the campus video risk characteristics and the campus facility risk characteristics; Among them, the campus facility database includes the basic parameters and operating parameters of the campus facilities. The basic parameters of the campus facilities are the size specifications, rated voltage, rated power, carrying capacity and service life of the campus facilities. The operating parameters of the campus facilities are the operating time, operating environment requirements, maintenance cycle, failure frequency and safety indicators of the campus facilities.

4. The risk feature identification and analysis method for a campus security monitoring system according to claim 3, characterized in that: Acquiring campus video risk features based on the campus risk feature identification model based on the campus video risk features to be identified includes: S2-1-1. Construct a historical campus video risk feature database based on the risk features of the campus video to be identified and the corresponding historical campus video risk features to be identified; S2-1-2. Obtain risk features of a first campus video to be identified as a training set based on the historical campus video risk feature database; S2-1-3. Obtain risk features of videos to be identified on a second campus based on the training set as a verification set; S2-1-4. Using the training set as input and the campus video risk features corresponding to the training set as output, an initial campus risk feature recognition model is constructed based on a support vector machine; S2-1-5. Input the verification set into the initial campus risk feature recognition model to obtain the campus video risk features corresponding to the verification set; S2-1-6. Determine whether the campus video risk features corresponding to the verification set completely correspond to the campus video risk features corresponding to the training set. If so, obtain the initial campus risk feature recognition model as the campus risk feature recognition model and execute S2-1-7. Otherwise, update the training set according to the campus video risk features corresponding to the verification set and return to S2-1-2. S2-1-7. Input the campus risk feature recognition model according to the risk feature of the campus video to be identified to obtain the campus video risk feature; Among them, the campus risk feature identification model includes deep convolutional neural network, variational autoencoder, generative adversarial network, long short-term memory network and fusion network based on attention mechanism.

5. The risk feature identification and analysis method for a campus security monitoring system according to claim 4 is characterized in that: Inputting the campus risk feature recognition model into the campus risk feature recognition model to obtain the campus video risk feature includes: S2-1-7-1. Input the risk features of the campus video to be identified into the deep convolutional neural network to obtain the image features and audio features of the risk features of the campus video to be identified; S2-1-7-2. Perform anomaly extraction processing on the image features of the campus video risk features to be identified according to the variational autoencoder and the generative adversarial network to obtain campus abnormal image features; S2-1-7-3. Perform abnormal extraction processing on the audio features of the campus video risk features to be identified according to the long short-term memory network to obtain abnormal campus audio features; S2-1-7-4. Based on the abnormal campus image features and the abnormal campus audio features, the campus video risk features are obtained according to the fusion network based on the attention mechanism.

6. The risk feature identification and analysis method for a campus security monitoring system according to claim 4, characterized in that: Acquiring the campus risk characteristics according to the campus video risk characteristics and the campus facility risk characteristics includes: S2-3-1. Obtain the campus video risk features corresponding to the training set as the historical campus video risk features; S2-3-2. Determine whether the campus video risk feature corresponds to the historical campus video risk feature. If so, execute S2-3-3. Otherwise, use the campus video risk feature to update the campus video risk feature to be identified, and return to S2-1-1. S2-3-3. Determine whether the campus facility risk characteristics correspond to the historical campus facility risk characteristics. If so, obtain the campus video risk characteristics and campus facility risk characteristics as the campus risk characteristics. Otherwise, use the campus facility risk characteristics to update the campus facility risk characteristics to be identified, and return to S2-2-1.

7. The risk feature identification and analysis method for a campus security monitoring system according to claim 1, characterized in that: Obtaining risk feature identification analysis results based on the campus risk data to be identified, the campus risk features to be identified, and the campus risk features includes: S3-1. Constructing a list of campus risk features to be identified based on the campus risk data to be identified and the campus risk features to be identified; S3-2. Determine whether the list of campus risk features to be identified contains the campus risk feature. If so, obtain the location information and timestamp of the campus risk feature using the list of campus risk features to be identified based on the campus risk feature, and execute S3-3. Otherwise, update the campus risk feature to be identified using the campus risk feature, and return to S2-1. S3-3. Performing location correlation analysis on the campus risk characteristics based on the location information of the campus risk characteristics to obtain distribution of the campus risk characteristics; S3-4. Perform risk trend analysis on the campus risk characteristics according to the timestamp of the campus risk characteristics to obtain a change trend of the campus risk characteristics; S3-5. Obtain a risk report of the campus risk characteristics based on the distribution of the campus risk characteristics and the changing trend of the campus risk characteristics as the risk characteristic identification and analysis result.

8. The risk feature identification and analysis method for a campus security monitoring system according to claim 7, characterized in that: Constructing a list of campus risk features to be identified based on the campus risk data to be identified and the campus risk features to be identified includes: S3-1-1. Perform time stamping processing according to the campus risk feature to be identified to obtain a timestamp of the campus risk feature to be identified; S3-1-2. Based on the risk features to be identified on campus, obtain location information of the risk features to be identified on campus according to the campus map information; S3-1-3. Obtain the campus risk data to be identified, the timestamp of the campus risk feature to be identified, and the location information of the campus risk feature to be identified, corresponding to the campus risk feature to be identified; S3-1-4. Construct a list of the campus risk features to be identified based on the campus risk features to be identified and the campus risk data to be identified, timestamp and location information corresponding to the campus risk features to be identified.

Images