Intelligent campus audio and video help seeking processing method

Abstract

The invention discloses an intelligent campus audio and video help seeking processing method, which belongs to the technical field of video processing, and comprises the following steps: collecting audio and video streams of a help seeker in a help seeking scene in real time, performing noise cancellation on the audio and video stream based on reverse sound waves, calculating a picture jitter trajectory based on a motion vector and in combination with inter-frame difference to perform reverse compensation, identifying equipment abnormal characteristics in a single-side / double-side detection picture of a segmentation point and keywords in audio, and associating the equipment abnormal characteristics and the keywords to metadata, and meanwhile, performing data processing on the equipment abnormal characteristics and the keywords in the audio; generating a structured help-seeking request containing the identity, the position and the help-seeking type of a help seeker, encrypting a data packet header, a data packet body and a data packet tail of the processed audio and video stream based on an encrypted transmission channel, transmitting the encrypted data packet header, the encrypted data packet body and the encrypted data packet tail, and carrying out whole-course encrypted storage on the call content by adopting a segmented storage strategy. And performing sentiment analysis on the evaluation data and associating the evaluation data to a service file of a corresponding administrator. The overall efficiency of campus help seeking processing is improved, and a safe and convenient campus environment is created for teachers and students.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of video processing, and particularly relates to an intelligent campus audio and video help processing method. BACKGROUND

[0002] In today's campus environment, ensuring the safety of teachers and students, providing convenient and efficient help services, and optimizing campus management have become the focus of attention of educational institutions. With the continuous expansion of the campus scale, the campus scene is becoming increasingly complex, and problems such as equipment failure, safety hazards, and sudden emergencies frequently occur in the campus. The traditional help processing method has exposed many drawbacks.

[0003] On the one hand, when teachers and students encounter problems in the campus and need help, it is often difficult to quickly and accurately convey their own location, help problem type, and actual situation on the spot to the relevant processing personnel. For example, in some large campuses, students may be in remote corners, and even if they call for help, it is difficult to describe the specific location clearly, making it difficult for rescue personnel to quickly locate and delaying the best opportunity to solve the problem.

[0004] On the other hand, the types of equipment failure and safety problems in different areas of the campus are diverse, such as IT equipment failure, logistics facility damage, and campus safety incidents, and each type of problem requires professional personnel for processing. However, the traditional method cannot accurately and efficiently match the help with the corresponding professional processing personnel, resulting in slow response speed and low processing efficiency.

[0005] In addition, the campus network environment is complex and variable, and during the transmission of the help audio and video, network instability often occurs, causing audio and video to lag and interrupt, seriously affecting the quality of help communication and processing effect. Moreover, the safe storage and management of help call content also face challenges. How to ensure that the call content is not leaked and is convenient for subsequent inquiry and tracing is a problem that needs to be solved urgently.

[0006] Furthermore, for the service quality evaluation after help processing, the traditional method lacks effective data collection and analysis means, and cannot deeply understand the needs of teachers and students, making it difficult to improve service levels in a targeted manner. For example, after processing the equipment maintenance help, it is difficult to know the satisfaction of teachers and students with the maintenance service and what problems exist in the service process that need to be improved.

[0007] Although there are some campus help related technologies and products on the market at present, there are generally problems such as single function and low intelligent degree. For example, some simple help button devices can only realize one-key calling function, cannot automatically obtain position, analyze problem type, and cannot sort help priority. In terms of audio and video processing, the existing technologies cannot effectively eliminate environmental noise and stabilize picture jitter, affecting accurate transmission of help information. In the face of these problems, an intelligent, efficient and comprehensive campus audio and video help processing method is urgently needed to improve the overall efficiency of campus help processing and create a safe and convenient campus environment for teachers and students. SUMMARY

[0008] In order to solve at least one of the above technical problems, the present application provides an intelligent campus audio and video help processing method.

[0009] The present application provides an intelligent campus audio and video help processing method, which comprises:

[0010] Starting the audio and video acquisition module based on the mobile terminal scanning the two-dimensional code of the preset site in the campus, and collecting the audio and video stream of the helper in the help site in real time;

[0011] Determine the noise feature vector of each frame in the audio and video stream, and input it into the pre-trained neural network model to obtain the corresponding frame factor. The loss function of the maximum aggregated factor after global source analysis and aggregated analysis of all factors is used to generate a reverse sound wave to cancel the noise of the audio and video stream;

[0012] According to the correction pair of each deflection angle in the deflection angle set of each frame of the acquisition device collecting the audio and video stream, generate a motion vector and combine the inter-frame difference to calculate the picture jitter track for reverse compensation. The correction pair includes jump correction and jitter correction.

[0013] According to the connection dependence factor and information dependence factor of each frame in the audio and video stream, determine the key frame and find the segmentation point, identify the device abnormal feature in the detection picture and the key word in the audio on the single side / double side of the segmentation point and associate them to the metadata, and at the same time, generate a structured help request containing the helper's identity, position and help type;

[0014] After receiving the structured help request, the data packet header, data packet body and data packet tail of the processed audio and video stream are encrypted and transmitted based on the encrypted transmission channel, and the network transmission state is monitored in real time. If the network transmission state does not meet the normal standard, the processed audio stream is transmitted preferentially at this time, and the processed video stream is retransmitted after the normal standard is met again. At the same time, the session recording module is started, the conversation content is stored by using the segmented storage strategy, the storage path is associated with the unique identifier of the structured help request, and the help answer is realized.

[0015] After the end of the help solution, the evaluation data is analyzed, the keywords are extracted and associated with the corresponding administrator's service profile.

[0016] Preferably, the key frame is determined according to the continuity dependence factor and the information dependence factor of each frame in the audio and video stream, comprising:

[0017] The spatio-temporal continuity parameter of the current frame is determined by extracting the RGB features of the current frame and the single-sided / bilateral frame, and the continuity dependence factor is obtained according to the spatio-temporal continuity parameter and the corresponding position of the campus space environment;

[0018] The preset key target of the current frame is identified based on the pre-trained target detection model, and the product of the target confidence and the area ratio is output as the visual weight. The audio key word of the current frame is identified by combining the pre-trained audio recognition model, and the audio weight of the audio key word and the audio energy value is output. The sum of the visual weight and the audio weight is taken as the information dependence factor;

[0019] The frame satisfying the double dynamic standard of the double dependence factor is regarded as the key frame.

[0020] Preferably, the data packet header, data packet body and data packet tail of the processed audio and video stream are encrypted based on an encrypted transmission channel, comprising:

[0021] The data packet header is encrypted by using an elliptic curve encryption algorithm, and a 12-byte random number is attached as a salt value after the encrypted packet header, and a first hash value is generated together with the encrypted packet header data;

[0022] In the data packet body encryption stage, a random key is selected from the symmetric key pool according to the key index in the data packet header, and the data packet body is encrypted by combining the sequence number of each frame in the data packet body and the disturbed pseudo number, to generate a second hash value;

[0023] The first-level check code is obtained by cutting the joint hash value of the encrypted packet header and the encrypted packet body, and the second-level check code is obtained based on the authentication key and combined with the sequence number, and the first-level check code and the second-level check code are spliced to obtain a third hash value according to the global byte;

[0024] The encryption processing is realized according to the first hash value, the second hash value and the third hash value.

[0025] Preferably, after scanning the two-dimensional code of the preset location in the campus based on the mobile terminal, the position information of the requester and the basic attribute data of the location are automatically obtained, the basic attribute data including the type of the location, the affiliated campus, and the list of associated equipment; the requester selects the type of help on the mobile terminal interface, the type of help including IT operation and maintenance, logistics operation and maintenance, and campus security, each type of help corresponding to a preset service range and online personnel state; after the selection is completed, the audio and video collection module is started.

[0026] Preferably, before the help is answered, the following steps are included:

[0027] A multi-dimensional priority sorting engine is started, the input parameters of the engine including the emergency degree weight of the type of help, the frequency of emergency events in the historical help records of the requester, the help congestion coefficient of the current period in the area, and the preset priority level of the location; the parameters are weighted and calculated by the analytic hierarchy process to obtain the priority score of the help request, and based on the priority score, the online administrator list corresponding to the type of help is called from the permission management module, the administrators in the online administrator list are filtered according to the preset location distribution range and the priority of answering, and a candidate handler pool is formed;

[0028] The load value of each administrator in the candidate handler pool is calculated according to the current number of calls, the historical processing time, and the number of backlog of work orders, and the structured help request is preferentially distributed to the administrator with the lowest load value and the matching priority; if no response is obtained within a preset second, the distribution range is automatically expanded to other administrators with the same type and priority, and a multi-level reminding mechanism is triggered, including interface pop-up window, sound alarm, and SMS notification.

[0029] Preferably, the distribution range is automatically expanded to other administrators with the same type and priority, including:

[0030] The analytic hierarchy process is used to construct a judgment matrix, four parameters of teaching area, training area, dormitory area, and public area are taken as the criterion layer, the priority score is taken as the target layer, the relative importance of each parameter is determined, and a weight vector is generated;

[0031] The standardized values of the parameters involved in the audio and video stream are multiplied by the corresponding weights and summed to obtain the initial priority score;

[0032] The initial priority score is corrected according to the emergency keywords contained in the structured help request, the correction value is divided into three levels of low, medium, and high, and the matching online administrator is determined according to the level result to realize automatic expansion.

[0033] The application provides an electronic device, comprising a processor and a memory, the memory is used for storing computer program code, the computer program code comprises computer instructions, when the processor executes the computer instructions, the electronic device executes the intelligent campus audio and video help processing method.

[0034] The application provides a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program comprises program instructions, when the program instructions are executed by a processor of an electronic device, the processor executes the intelligent campus audio and video help processing method.

[0035] Compared with the prior art, the application has the following beneficial effects:

[0036] By optimizing the audio and video acquisition quality (noise cancellation, picture anti-shake), accurately extracting key information (key frame identification, structured request generation), and ensuring transmission and storage safety (multi-layer encryption, priority transmission), in combination with the role permission management and work order process in the operation manual, the whole process of campus help from initiation, processing to recording is realized, and the efficiency and safety of the whole process are improved, the help response speed, problem identification accuracy and data security are significantly improved, and reliable technical support is provided for campus operation and maintenance.

[0037] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, but not limiting the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the application or the background art, the drawings needed to be used in the embodiments of the application or the background art will be described below.

[0039] The drawings herein are incorporated into the specification and form part of the specification, which show embodiments consistent with the present disclosure, and together with the specification, serve to explain the technical solutions of the present disclosure.

[0040] Figure 1 A flowchart of an intelligent campus audio and video help processing method provided by the embodiments of the application;

[0041] Figure 2 A hardware structure schematic diagram of an electronic device in the embodiments of the application;

[0042] Figure 3 An interface diagram of a help type in the embodiments of the application. DETAILED DESCRIPTION

[0043] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of the present application.

[0044] The terms "first", "second", and the like in the description and claims of the present application and the above drawings are used to distinguish different objects, rather than to describe a particular order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device including a series of steps or units is not limited to the listed steps or units, but can optionally include other steps or units not listed or can optionally include other steps or units inherent to the process, method, product, or device.

[0045] The term "and / or" herein is only a description of the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent three cases of A alone, A and B together, and B alone. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0046] In this document, the term "embodiment" means that the specific features, structures, or characteristics described in connection with the embodiment can be included in at least one embodiment of the present application. The phrase appears at various places in the specification does not necessarily all refer to the same embodiment, nor is it necessarily independent or alternative embodiments to each other. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0047] In addition, in order to better illustrate the present application, a large number of specific details are given in the specific embodiments below. Those skilled in the art should understand that the present application can also be implemented without certain specific details. In some examples, methods, means, elements and circuits well known to those skilled in the art are not described in detail, in order to highlight the main idea of the present application.

[0048] The present application provides an intelligent campus audio and video help processing method, as shown in Figure 1 The method comprises the steps of:

[0049] Step 1: Start the audio and video collection module based on the mobile terminal scanning the two-dimensional code of the preset site in the campus, and collect the audio and video stream of the helper in the help site in real time;

[0050] Step 2: Determine the noise feature vector of each frame in the audio and video stream, and input it into the pre-trained neural network model to obtain the corresponding frame factor. The loss function of the maximum aggregated factor after global source analysis and aggregated analysis of all factors is used to generate a reverse sound wave to cancel the noise of the audio and video stream;

[0051] Step 3: According to the correction pair of each deflection angle in the deflection angle set of each frame of the collection device collecting the audio and video stream, generate a motion vector and combine the inter-frame difference to calculate the picture jitter trajectory for reverse compensation, wherein the correction pair includes jump correction and jitter correction;

[0052] Step 4: Determine the key frame according to the connection dependence factor and information dependence factor of each frame in the audio and video stream, find the segmentation point, identify the device abnormal features in the detection picture and the keywords in the audio on the single side / double side of the segmentation point and associate them to the metadata, and at the same time, generate a structured help request containing the helper's identity, location, and help type;

[0053] Step 5: After receiving the structured help request, encrypt the data packet header, data packet body, and data packet tail of the processed audio and video stream based on the encrypted transmission channel, and then transmit and monitor the network transmission state in real time. If the network transmission state does not meet the normal standard, the processed audio stream is transmitted first, and the processed video stream is retransmitted when the normal standard is met again. At the same time, start the session recording module, use the segmented storage strategy to store the call content, and store the path associated with the unique identifier of the structured help request to realize help answering;

[0054] Step 6: After the help answer is finished, perform sentiment analysis on the evaluation data, extract keywords and associate them to the corresponding administrator's service archive.

[0055] Preferably, the key frame is determined according to the connection dependence factor and information dependence factor of each frame in the audio and video stream, comprising:

[0056] Extract the RGB features of the current frame and the single / double side frame to determine the spatiotemporal continuity parameter of the current frame, and obtain the connection dependence factor according to the spatiotemporal continuity parameter and the corresponding position of the campus space environment;

[0057] The preset key target of the current frame is identified based on a pre-trained target detection model, a product of a target confidence and an area ratio is output as a visual weight, an audio keyword of the current frame is identified based on a pre-trained audio recognition model, an audio weight of the audio keyword and an audio energy value is output, and a sum of the visual weight and the audio weight is taken as an information dependency factor.

[0058] The frame satisfying the double dynamic criterion is regarded as a key frame.

[0059] Preferably, the data packet header, the data packet body and the data packet tail of the processed audio and video stream are encrypted based on an encrypted transmission channel, including:

[0060] The data packet header is encrypted by using an elliptic curve encryption algorithm, and 12 bytes of random numbers are attached as salt values after the encrypted packet header, and a first hash value is generated together with the encrypted packet header data.

[0061] In the data packet body encryption stage, a random key is selected from a symmetric key pool according to the key index in the data packet header, and the data packet body is encrypted in combination with the sequence number of each frame and the disturbed pseudo number, to generate a second hash value.

[0062] The first-level check code is obtained by cutting the joint hash value of the encrypted packet header and the encrypted packet body, the second-level check code is obtained based on the authentication key and in combination with the sequence number, and the third hash value is obtained by splicing the first-level check code and the second-level check code according to the global byte.

[0063] The encryption processing is realized according to the first hash value, the second hash value and the third hash value.

[0064] The mobile terminal refers to a smart phone, a tablet and the like used by a seeker, supports code scanning and audio and video acquisition, for example, a smart phone used by a student when scanning a two-dimensional code by using WeChat, specifically: the device camera and microphone permissions are called by using a WeChat applet.

[0065] The two-dimensional code of the preset site is a unique identification two-dimensional code of each site (such as a classroom and a training building) in a campus, which is associated with site basic information, specifically: the site management module is used by an administrator to generate and post in batches.

[0066] The audio and video acquisition module is a functional module in the mobile terminal for acquiring real-time sound and images, which is composed of hardware (a camera and a microphone) and software (acquisition algorithm), and can automatically call the device hardware based on the applet to capture audio and video data in real time.

[0067] The audio and video stream is a continuous audio and video data sequence, which is transmitted in time sequence to form dynamic pictures and sound, for example, is transmitted to the background in real time through a network, and supports real-time receiving by an administrator end.

[0068] Frame is a single static image in a video stream, which is the basic unit of video. The background system parses the video stream frame by frame.

[0069] Noise feature vector is a digital sequence that describes the characteristics of noise (such as frequency, amplitude), which is used to distinguish noise from effective sound. For example, the 50Hz low-frequency noise generated by the running air conditioner in the classroom has a feature vector of {50, 0.3, 0.1} (frequency, amplitude, duration). Specifically, the audio features are extracted by Fourier transform and converted into vector form.

[0070] The neural network model is an AI model trained in advance with a large amount of campus environmental noise data, which is used to analyze noise features and output processing parameters. It is deployed on the server side and receives noise feature vectors in real time and calculates processing factors.

[0071] The factor is a parameter output by the model to measure the degree of noise influence. The higher the value, the greater the interference of noise on effective information. For example, the factor of construction noise in a certain frame is 0.8 (full score 1.0), indicating strong interference. Global source analysis is a cross-frame analysis of all frame factors to locate the common source of noise. For example, after analyzing the factors of 200 frames, it is found that 90% of the high-factor frames contain 50Hz low-frequency characteristics, and it is determined that the noise comes from the air conditioner. Aggregated analysis is to aggregate all factors and calculate the overall noise distribution (such as mean, peak). For example, after aggregating the factors of 100 frames, the maximum factor is 0.9 and the average factor is 0.4, indicating the existence of intermittent strong noise. Specifically, the factor sequence is fitted with a mathematical model (such as Gaussian distribution).

[0072] The maximum aggregated factor is the largest value in the aggregated analysis, representing the most serious noise interference. For example, after aggregated analysis, the maximum factor is 0.95, corresponding to a sudden outdoor construction sound in a certain frame. Specifically, the sorted aggregated factor sequence is sorted and the maximum value is taken.

[0073] The loss function is a function that measures the difference between the predicted noise and the actual noise, which is used to optimize the accuracy of the generated inverse sound waves. For example, the mean square error between the predicted noise and the actual noise is 0.02, indicating a small error. The inverse sound wave is a sound wave that is completely opposite to the noise waveform, and when superimposed, it can cancel out the noise (wave peak against wave trough). For example, an inverse sound wave is generated for a 50Hz air conditioner noise, and after superimposition, the noise amplitude is reduced by 80%. Noise cancellation is achieved by superimposing the inverse sound wave with the original noise to weaken or eliminate the noise and enhance the effective sound (such as the voice of a person seeking help). Specifically, the inverse sound wave is superimposed with the original audio stream in real time.

[0074] The collection device is a mobile terminal (such as a mobile phone) that captures video. Its posture (angle, position) will affect the stability of the picture. Specifically, the posture data is obtained through the device's gyroscope and acceleration sensor.

[0075] The deviation angle is the angle between the collection device and the horizontal / vertical direction, which reflects the degree of screen tilt. Based on real-time monitoring of the three-dimensional angle (pitch angle, yaw angle, roll angle) of the device, the parameter combination for correcting the deviation angle is corrected, including the threshold / value of jump correction and the coefficient of jitter correction. For example, the correction for (jump threshold of 5° and jitter coefficient of 0.3) indicates that when the angle changes suddenly by more than 5°, the jump correction is triggered, and the slight shaking is smoothed with a coefficient of 0.3. Jump correction is a correction for sudden changes in the angle of the device (such as sudden rotation of the mobile phone), which avoids dramatic jumps in the screen. Specifically, through inter-frame angle jump detection, an interpolation algorithm is used for smooth transition. Jitter correction is a correction for slight and continuous changes in the angle of the device (such as hand holding instability), which reduces screen blur. Specifically, a Kalman filter algorithm is used to smooth the angle data.

[0076] The motion vector is a vector that describes the direction and distance of pixel movement in adjacent frames, which reflects the movement of objects or devices in the screen. For example, the pixels on the edge of the blackboard in the first frame and the second frame move 3 pixels to the right, and the motion vector is (3, 0). Inter-frame difference is the calculation of the pixel difference between adjacent frames, which quantifies the degree of screen change. For example, the difference between the fifth frame and the sixth frame is 5%, indicating that the screen change is small. Specifically, the RGB values of adjacent frames are compared pixel by pixel, and the difference ratio is calculated. The screen jitter trajectory is the path formed by the motion vector and the change in the deviation angle of consecutive frames, which reflects the jitter pattern. Specifically, the motion vector and the deviation angle of consecutive frames are recorded, and the trajectory curve is fitted.

[0077] Reverse compensation is to generate a reverse adjustment amount based on the jitter trajectory to offset the screen jitter (such as reverse translation and rotation), and to adjust the screen in real time through image transformation algorithms (such as affine transformation).

[0078] The connection dependency factor is a parameter that measures the spatiotemporal continuity of the current frame and the previous and next frames (single-sided / double-sided). The higher the value, the more coherent the screen. For example, if the position change rate of the blackboard in the last 3 frames is less than 5%, the connection dependency factor is 0.9 (full score 1.0). The information dependency factor is a parameter that measures the importance of the effective information in the frame, which is obtained by summing the visual weight and the audio weight. For example, a frame contains a projector failure screen (visual weight 0.6) and a projector keyword (audio weight 0.3), and the information dependency factor is 0.9. A key frame is a frame that meets the preset threshold (double dynamic standard) in both dependency factors, which contains core information (such as a faulty device or a key voice). For example, a frame with a connection dependency factor greater than 0.7 and an information dependency factor greater than 0.6 is marked as a key frame. The split point is the mutation point of the key frame sequence in the audio and video stream, which marks the switching of information topics (such as from describing the problem to showing the fault). For example, the screen from the speaker (without a device) switches to the screen of the projector, which is the split point. Specifically, the visual / audio feature mutation of the key frame is detected to determine the location of the split point.

[0079] Device abnormality features are visual manifestations of device failures (such as no picture, flickering, or a disconnected cable). For example, a projector screen may be completely black (no picture) or the wireless microphone indicator light may be off (no power). Specifically, the preset fault features are identified through the target detection model.

[0080] The keywords in the audio are the core words related to the fault (such as device name, fault phenomenon) in the voice of the person seeking help. Specifically, they are extracted through the speech recognition model (ASR) and associated with the preset fault keyword library.

[0081] Metadata is additional descriptive information of the audio and video stream (such as acquisition time, device model, and venue ID), for example: {acquisition time: 2024-12-26 10:01:50, venue ID: BY-501, device model: Xiaomi 13}.

[0082] A structured help request is a standardized help information form that integrates key information such as identity, location, and help type. For example: {Assister: Mr. / Ms. XXX (employee number DG202413), Location: BY-501, Training Building, Baiyun Campus, Help Type: IT Operations and Maintenance, Fault: No picture on projector}. Specifically, it automatically extracts metadata, device anomaly features, and keywords to generate structured data.

[0083] An encrypted transmission channel is a network transmission path that uses an encryption protocol (such as SSL / TLS) to ensure data transmission security. The packet header is the head of the data packet, which contains transmission control information (such as sender ID, target ID, and key index), for example: {sender ID: requester terminal ID, target ID: administrator terminal ID, key index: 005}. The packet body is the core part of the data packet, which contains actual audio and video data (such as key frames and audio clips). The packet trailer is the tail of the data packet, which contains verification information (such as checksum and length) for verifying data integrity.

[0084] The network transmission status is an indicator that measures network stability and performance (such as bandwidth, latency, and packet loss rate). For example, bandwidth < 1 Mbps, latency > 200 ms, and packet loss rate > 5% are considered substandard. Prioritizing audio streaming transmission prioritizes audio data transmission when the network status is poor (audio has lower bandwidth requirements than video).

[0085] The segmented storage strategy is to store the recorded audio and video data by time or size into multiple segments, which is convenient for management and traceability. For example, the content of each 5-minute call is stored as a file named "20241226100150_001" and "20241226100650_002". Encryption storage is to save the encrypted recorded audio and video segments to prevent unauthorized access. Unique identification is the unique code of the structured help request, which is used to associate the storage path and ticket information. For example, the ticket number GZ2412000058 and the storage path " / record / GZ2412000058 / ". RGB feature is the red (R), green (G), and blue (B) color values of each pixel in the image, which is used to describe the content of the picture. For example, the RGB value of the blackboard area is close to (0, 0, 0) (black), and the RGB value of the projector area is close to (255, 255, 255) (white). The spatiotemporal continuity parameter is a continuity index calculated by combining time (frame order) and space (picture content). For example, the spatial offset of the podium position in the last 5 frames is less than 2%, and the spatiotemporal continuity parameter is 0.9. The campus space environment is the physical property of the help site (such as classroom layout, equipment placement), such as the training building classroom usually contains projector, operation table, experimental equipment, and the layout is relatively fixed. The target detection model is an AI model used to identify pre-set targets (such as equipment) in the picture, which is based on the pre-trained YOLO model and can identify various equipment. Specifically, it is deployed on the server side, processes video frames in real time and outputs target position and type.

[0086] The pre-set key target is the equipment or object that needs to be focused on (related to common campus faults), such as projector, wireless microphone, central control equipment, air conditioner, etc. The target confidence is the confidence of the model identifying the target as the pre-set key target (0-1), such as the model identifying an object as a projector with a confidence of 0.95, indicating that it is very likely to be correct. The area ratio is the pixel area ratio of the target in the picture (target pixel number / total pixel number), such as the projector occupying 15% of the picture, the area ratio is 0.15. The visual weight is an index to measure the importance of visual information, which is the product of target confidence and area ratio. The audio recognition model is an AI model that converts speech into text and extracts keywords, which is based on the deep learning ASR model and can identify professional terms in the campus scene.

[0087] Audio energy value is an indicator of sound intensity (positively correlated with volume), for example, the audio energy value of the voice of the caller describing the fault is 0.7 (0-1), and the background noise is 0.2. Audio weight is an indicator of the importance of audio information, which is the keyword weight multiplied by the audio energy value. The double dynamic standard is a dynamic threshold that combines the connection dependency factor and the information dependency factor (adjusts with the scene), for example, the standard for the practice building scene is connection > 0.6, information > 0.5, and for the dormitory area, connection > 0.5, information > 0.4. Elliptic curve encryption algorithm (ECC) is a highly efficient asymmetric encryption algorithm that generates keys using points on an elliptic curve, providing high security and short keys, for example, a 256-bit key is generated using the secp256r1 curve to encrypt the data packet header. Salt value is a randomly generated string that, combined with encrypted data, generates a hash value to enhance collision resistance, for example, a 12-byte salt value 7f3a9d2b5c1e8 is combined with the encrypted header to calculate the hash. The first hash value is a hash value calculated based on the encrypted header and the salt value, used to verify the integrity of the header, calculated using the SHA256 algorithm (encrypted header + salt value). The symmetric key pool is a set of symmetric encryption keys (such as AES keys) maintained by the server to encrypt the data packet body, containing a pool of 1000 AES-256 keys, each key corresponding to a unique index. Random key is a randomly selected key from the symmetric key pool, used to encrypt the data packet body, for example, the key index 005 in the data packet header selects the 5th AES key from the pool.

[0088] Sequence number is the transmission sequence number of the frame in the data packet body, used to reassemble the data, pseudo number is the number after scrambling the sequence number, preventing attackers from guessing the transmission pattern, for example, sequence number 10 is scrambled to pseudo number 23 by algorithm, and sequence number 11 is scrambled to 27. The second hash value is a hash value calculated based on the encrypted packet body, used to verify the integrity of the packet body, calculated using the SHA256 algorithm on the encrypted packet body. Joint hash value is a hash value calculated based on the encrypted header and the encrypted packet body, used to verify the integrity of the entire data packet, the result of SHA256(encrypted header + encrypted packet body). The first check code is a part of the data extracted from the joint hash value, used for quick verification, for example, the first 16 bytes of the joint hash value are extracted as the first check code. The authentication key is a pre-shared key used to generate the second check code and verify the sender's identity, for example, the HMAC key pre-shared between the administrator and the server. The second check code is a check code calculated based on the authentication key and the sequence number, used for identity verification, calculated using the HMAC-SHA256 algorithm (authentication key + sequence number). The third hash value is a hash value calculated after concatenating the first check code and the second check code, used for final integrity and identity verification, for example, the result of SHA256(first check code + second check code).

[0089] In the help-seeking initiation stage, the help-seeker scans the two-dimensional code of the preset location in the campus through the mobile terminal, triggers the unified identity authentication process, and needs to input the account password for the first login to complete the authentication. Subsequent logins are automatically exempted from authentication. The system provides two interactive options: calling for help and filling out a work order. If the help-seeker chooses to call for help, the audio and video collection module is started. If the help-seeker chooses to fill out a work order, a list of fault types is displayed. The fault types in the list are dynamically matched with the list of associated equipment at the location. The help-seeker needs to select a specific fault type and upload a fault description and multimedia attachments. The system automatically generates a work order draft with a unique identifier.

[0090] After the administrator answers the call, the system provides an invitation assistance function, which supports real-time invitation of other online administrators to join the session. The new joiner automatically obtains the audio and video streams and historical interaction information of the current session, and the original session channel is seamlessly connected. For work order requests, the system automatically completes the work order information completion, including filling in the detailed address according to the location information, recommending knowledge base items associated with the fault type, and marking the work order status as unprocessed and pushing it to the work order list of the corresponding administrator.

[0091] In the work order closed-loop stage, the administrator receives the work order on the mobile terminal and can perform the confirm work order operation. The system updates the work order status to confirmed and records the claim time, and locks the editing permission of the work order. After the administrator finishes processing, the administrator needs to fill in the processing result, upload the multimedia attachments, and select the processing completion time. After submission, the system updates the work order status to completed. For situations where the fault has been reported through other channels and has been processed, the administrator can manually enter the fault information through the supplementary work order function. The supplementary work order does not need to go through the confirmation process, but needs to fill in the fault type, processing result, and timestamp completely. The system regularly analyzes the work order data, generates work order trend charts according to the time dimension, classifies and sorts the data according to the fault type, and forms the equipment fault ranking. All work order data supports export functions, including Excel and PDF formats, and the export content includes all fields of the work order and the processing track.

[0092] In the feedback optimization stage, after the audio and video call ends or the work order is completed, the system automatically pushes the evaluation request to the help-seeker, including technical ability, working attitude, and comment input. After the help-seeker submits the evaluation, the system performs sentiment analysis on the evaluation data, extracts keywords, and associates them to the service archives of the corresponding administrator. The administrator can create, edit, and query operation and maintenance knowledge in the knowledge base module. The knowledge types include documents and videos, and support classification and archiving according to fault types and device models. The knowledge content is accessed differently through permission control. The system dynamically adjusts the priority weight of the help-seeking type, the site allocation range of the administrator, and the parameters of the knowledge base recommendation algorithm based on historical help-seeking data and evaluation results, forming a self-optimizing closed loop.

[0093] By optimizing the audio and video acquisition quality (noise cancellation, picture anti-shake), accurately extracting key information (key frame identification, structured request generation), ensuring transmission and storage safety (multi-layer encryption, priority transmission), combined with the role permission management, work order process and other functions in the operation manual, the whole process of campus help from initiation, processing to recording is realized. Efficient and safe, significantly improving the help response speed, problem identification accuracy and data security, providing reliable technical support for campus operation and maintenance.

[0094] The application provides a kind of intelligent campus audio and video help processing method, after scanning the two-dimensional code of preset site in campus based on mobile terminal, the position information of helper and the basic attribute data of site are automatically obtained, the basic attribute data includes site type, belonging campus, associated equipment list;Helper selects help type in mobile terminal interface, the help type includes IT operation and maintenance, logistics operation and maintenance and campus security, each help type corresponds to preset service range and online personnel state;After selection is completed, start audio and video acquisition module.

[0095] Preferably, before realizing help answer, it includes:

[0096] Start multi-dimensional priority sorting engine, the input parameters of the engine include the emergency degree weight of help type, the emergency event frequency in the historical help record of helper, the help congestion coefficient of the current period in the area and the preset site priority level;The priority score of the help request is obtained by weighting calculation of parameters by analytic hierarchy process, and the online administrator list of the corresponding help type is called from the permission management module based on the priority score, the administrators in the online administrator list are filtered according to the preset site allocation range and the listening priority, forming a candidate handler pool;

[0097] According to the load value of the current call number, the historical processing time and the work order backlog of each administrator in the candidate handler pool, the structured help request is preferentially distributed to the administrator with the lowest load value and the matching priority;If no response is obtained within a preset second, the distribution range is automatically expanded to other priority administrators, and a multi-level reminding mechanism is triggered, including interface pop-up window, sound alarm and SMS notification.

[0098] Preferably, the distribution range is automatically expanded to other priority administrators of the same type, including:

[0099] The analytic hierarchy process is used to construct a judgment matrix, four parameters of teaching area, training area, dormitory area and public area are taken as criterion layer, priority score is taken as target layer, relative importance of each parameter is determined, and weight vector is generated;

[0100] The sum of the standardized values of the parameters involved in the audio and video stream multiplied by the corresponding weight is obtained, and the initial priority score is obtained.

[0101] The initial priority score is modified according to the emergency keywords contained in the structured help request, the modified value is divided into low, medium and high three levels, and the matching online administrator is determined according to the level result to realize automatic expansion.

[0102] The basic attribute data is the core information describing the characteristics of the site, including site type, belonging to the school district, and associated equipment list, wherein the site type is the functional classification of the site (such as classroom, training building, comprehensive building).

[0103] The school district is the campus partition where the site is located, for example, Baiyun campus, Qingyuan campus, etc.

[0104] The associated equipment list is a list of repairable equipment in the site, such as projectors, wireless microphones, computers, etc. in the training building classroom, which is specifically: the administrator binds the device information in the site management.

[0105] The help type is the service category selected by the helper according to the nature of the problem, including IT operation and maintenance, logistics operation and maintenance, and campus security, wherein IT operation and maintenance is to handle technical faults of computers, networks, projectors and other devices. Logistics operation and maintenance is to handle maintenance requirements of water, electricity, doors and windows, air conditioning and other facilities. Campus security is to handle security incidents, emergencies, etc. Specifically: the helper selects on the mobile terminal interface, corresponding to the administrator's authority, as shown in Figure 3 .

[0106] The service range is the responsibility boundary and processing scenario corresponding to each help type, for example, the service range of IT operation and maintenance is the fault handling of all electronic devices in the campus, and the logistics operation and maintenance is the infrastructure maintenance. The online personnel state is the number of administrators / maintenance personnel online under the current help type, specifically: the administrator login state is monitored in real time and synchronized to the mobile terminal.

[0107] The multi-dimensional priority sorting engine is a system module that calculates the priority of the help request by integrating multiple parameters, which is used to determine the request processing order, for example, when receiving a training building device failure and a dormitory area light failure at the same time, the engine calculates and preferentially allocates the former, specifically: integrated in the background system, real-time processing and sorting of requests.

[0108] The input parameter is the core index for the engine to calculate the priority, including the following four categories:

[0109] The emergency degree weight of the help type is the preset weight of the urgency of different help types (such as campus security > IT operation and maintenance > logistics operation and maintenance), which is preset by the administrator in the system configuration.

[0110] The frequency of emergency events in the helper's historical help records is the number of times the helper has initiated emergency help in the past, and the higher the frequency, the higher the current request priority, which is retrieved from the statistical analysis module.

[0111] The current period of the area of the help congestion coefficient is the density of the number of help in a certain area per unit of time (the more the number, the higher the coefficient, and the priority may be reduced to balance the load), and the system calculates the number of requests in the area in real time.

[0112] The preset site priority level is the importance level of different sites (such as training building > teaching building > public area), which is set by the administrator in site management.

[0113] The analytic hierarchy process is a mathematical method of determining parameter weights by pairwise comparison, which decomposes complex decision-making problems into multiple levels (target layer, criterion layer, and scheme layer), and the system built-in algorithm realizes parameter weighting calculation.

[0114] The priority score is the comprehensive score calculated by the analytic hierarchy process, which is used to measure the emergency level of the request (the higher the score, the higher the priority), and the system automatically calculates and sorts.

[0115] The permission management module is a system module that manages administrator roles, permissions, and online status, and administrators configure through permission management-role management.

[0116] The online administrator list is a list of administrators currently logged in to the system and in an online state, and the system monitors the login status in real time.

[0117] The site allocation range is the responsible site pre-assigned to the administrator, which only handles help requests within that range, and the administrator configures in site allocation.

[0118] The call priority is the level of administrators handling requests (high > medium > low), and high-priority administrators receive requests first.

[0119] The candidate handler pool is a set of administrators filtered from the online administrator list that meet the site allocation range and call priority, for example, a request corresponding to the candidate pool of the site is 2 high-priority, 1 medium-priority IT operation and maintenance administrator, and the system automatically filters according to the rules.

[0120] The load value is an indicator that measures the current work saturation of the administrator, which is calculated by the current call number, historical processing time, and backlog number.

[0121] The current call number is the number of ongoing audio and video calls by the administrator. The historical processing time is the average time the administrator spent on handling similar requests in the past (the longer the time, the higher the load value). The backlog number is the number of unprocessed work orders by the administrator. The preset second is the maximum time waiting for the administrator's response (if not responded within the timeout, the allocation range will be expanded), for example: 5 seconds. The multi-level reminder mechanism is a multi-level notification method triggered when the timeout is not responded, ensuring that the request is received.

[0122] The judgment matrix is a matrix used in the analytic hierarchy process to compare the relative importance of parameters in the criterion layer, for example, comparing the importance of the teaching area, the training area, the dormitory area, and the public area, and the training area > the teaching area is recorded as 3 (1-9 scale). The criterion layer is an intermediate factor that affects the target layer in the analytic hierarchy process, which is the four types of sites of the teaching area, the training area, the dormitory area, and the public area. The target layer is the decision target of the analytic hierarchy process, which is the priority score, and the weight vector is a numerical vector reflecting the relative importance of parameters in the criterion layer, and the weight vector is {0.2, 0.4, 0.3, 0.1}, indicating that the training area (0.4) has the highest weight, and the public area (0.1) has the lowest weight. The standardized value is to convert the parameters involved in the audio and video stream (such as the type of site and the degree of equipment failure) into a unified numerical value between 0 and 1 (eliminate the dimension effect), and the system normalizes the parameters according to the preset rules.

[0123] The initial priority score is the basic score obtained by multiplying and summing the standardized value and the corresponding weight. The emergency keyword is a word (such as fire, injury, and large-scale failure) in the structured help request indicating an emergency situation. The correction value is the final score after adjusting the initial priority score with the emergency keyword, which is divided into three levels of low, medium, and high, for example, the initial score is 0.64, and the correction value is 0.9 (high level) due to the fire keyword.

[0124] By automatically associating the site information through scanning the code, accurately matching the service according to the help type, multi-dimensionally calculating the priority and dynamically allocating the administrator, and combining the multi-level reminding mechanism to ensure efficient response to the request, the accuracy and timeliness of the campus help processing are significantly improved, and the whole process of intelligentization from the help initiation, the priority sorting to the administrator allocation is realized, thereby providing scientific and efficient technical support for the campus operation and maintenance.

[0125] The application further provides an electronic device, comprising a processor, a sending device, an input device, an output device and a memory, the memory is used for storing computer program code, the computer program code comprises computer instructions, when the processor executes the computer instructions, the electronic device executes the method of any one of the possible implementation manners described above.

[0126] The application further provides a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program comprises program instructions, when the program instructions are executed by the processor of an electronic device, the processor executes the method of any one of the possible implementation manners described above.

[0127] Please refer to Figure 2 , Figure 2 A hardware structure schematic diagram of an electronic device provided by the embodiment of the application.

[0128] The electronic device 2 comprises a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, the memory 22, the input device 23, and the output device 24 are coupled through a connector, which comprises various interfaces, transmission lines, buses, or the like, and embodiments of the present application are not limited in this regard. It should be understood that in various embodiments of the present application, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, various interfaces, transmission lines, buses, or the like.

[0129] The processor 21 can be one or more graphics processing units (GPUs). In the case where the processor 21 is a GPU, the GPU can be a single-core GPU or a multi-core GPU. Alternatively, the processor 21 can be a processor group comprising a plurality of GPUs, and the plurality of processors are coupled to each other through one or more buses. Alternatively, the processor can be another type of processor, and embodiments of the present application are not limited in this regard.

[0130] The memory 22 can be used to store computer program instructions and various types of computer program codes for executing the schemes of the present application. Alternatively, the memory comprises, but is not limited to, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read only memory (EPROM), or a compact disc read-only memory (CD-ROM), which is used for storing relevant instructions and data.

[0131] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The input device 23 and the output device 24 can be independent devices or an integral device.

[0132] It can be understood that in embodiments of the present application, the memory 22 can be used to store not only relevant instructions, but also data stored in the memory, which is not limited in embodiments of the present application.

[0133] It can be understood that, Figure 2 Only a simplified design of an electronic device is shown. In actual applications, the electronic device can also include other necessary elements, including but not limited to any number of input / output devices, processors, memories, and the like, and all video analysis devices that can implement embodiments of the present application are within the scope of protection of the present application.

[0134] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0135] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here. Those skilled in the art can also clearly understand that each embodiment of the present application describes each with emphasis, and for the convenience and brevity of description, the same or similar parts in different embodiments can not be described in detail, and therefore, the parts not described or not described in detail in a certain embodiment can be referred to the description of other embodiments.

[0136] In several embodiments provided by the present application, it should be understood that the disclosed system, device and method can be implemented by other ways. For example, the above-described device embodiments are only schematic, and the division of the units is only a logical function division, and there can be another division way in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0137] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0138] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0139] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general purpose computer, a special purpose computer, a computer network or other programmable apparatus. The computer instructions can be stored in or transmitted by a computer readable storage medium. The computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through a wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a digital versatile disc (DVD)), or a semiconductor medium (for example, a solid state disk (SSD)) and the like.

[0140] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be instructed by a computer program to complete the relevant hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. The storage medium includes read-only memory (ROM) or random access memory (RAM), magnetic disk or optical disk and various media that can store program codes.

Claims

1. A method for processing an audio-video help-seeking in a smart campus, characterized in that, The method comprises the steps of: starting an audio and video collection module based on a mobile terminal scanning a two-dimensional code of a preset location in a campus to collect an audio and video stream of a helper at a help site in real time; determining a noise feature vector of each frame in the audio and video stream, inputting the noise feature vector into a pre-trained neural network model to obtain a factor of the corresponding frame, performing global source analysis on all factors, and combining a loss function of a maximum aggregated factor after aggregated analysis on all factors to generate a reverse sound wave to perform noise cancellation on the audio and video stream; generating a motion vector based on a correction pair of each deflection angle in a deflection angle set of a collection device collecting the audio and video stream, and performing reverse compensation on a picture jitter track in combination with inter-frame difference calculation, wherein the correction pair comprises jump correction and jitter correction; determining a key frame based on a connection-dependent factor and an information-dependent factor of each frame in the audio and video stream, finding a split point, identifying device abnormal features in a detection picture and keywords in audio on a single side / double side of the split point, and associating the device abnormal features and the keywords to metadata, and generating a structured help request including a helper identity, a location, and a help type; after receiving the structured help request, performing encryption processing on a data packet header, a data packet body, and a data packet tail of the processed audio and video stream based on an encrypted transmission channel, and transmitting the processed audio and video stream and monitoring a network transmission state in real time, if the network transmission state does not meet a normal standard, preferentially transmitting the processed audio stream and retransmitting the processed video stream when the normal standard is met again, starting a session recording module, and performing full-process encryption storage on a call content by using a segmented storage strategy, and storing a storage path in association with a unique identifier of the structured help request to realize help answering; after the help answering ends, performing sentiment analysis on evaluation data, extracting keywords, and associating the keywords to a service archive of a corresponding administrator.

2. The intelligent campus audio and video help processing method according to claim 1, characterized in that, The method for determining a key frame based on a connection-dependent factor and an information-dependent factor of each frame in the audio and video stream comprises the steps of: extracting RGB features of a current frame and a single side / double side frame to determine a space-time continuity parameter of the current frame, and obtaining a connection-dependent factor based on the space-time continuity parameter and a campus space environment of a corresponding position; identifying a preset key target of the current frame based on a pre-trained target detection model, outputting a product of a target confidence and an area ratio as a visual weight, identifying an audio keyword of the current frame based on a pre-trained audio recognition model, outputting an audio weight of the audio keyword and an audio energy value, and taking a sum of the visual weight and the audio weight as an information-dependent factor; regarding a frame satisfying a double dynamic standard based on the double-dependent factors as a key frame.

3. The intelligent campus audio and video help processing method according to claim 1, characterized in that, The method for performing encryption processing on a data packet header, a data packet body, and a data packet tail of the processed audio and video stream based on an encrypted transmission channel comprises the steps of: performing encryption on the data packet header by using an elliptic curve encryption algorithm, and appending a 12-byte random number as a salt value to the encrypted header to generate a first hash value together with the encrypted header data; in a data packet body encryption stage, selecting a random key from a symmetric key pool based on a key index in the data packet header, and performing encryption on the data packet body based on a sequence number of each frame in the data packet body and a disturbed pseudo number to generate a second hash value; The first check code is obtained by cutting according to the joint hash value of the encrypted header and the encrypted body, the second check code is obtained based on the authentication key and in combination with the sequence number, and the first check code and the second check code are spliced according to the global byte to obtain a third hash value; The first hash value, the second hash value and the third hash value are used to realize encryption processing.

4. The intelligent campus audio and video help processing method according to claim 1, characterized in that, After scanning the two-dimensional code of the preset location in the campus based on the mobile terminal, the position information of the person seeking help and the basic attribute data of the location are automatically obtained, and the basic attribute data includes the type of the location, the school district to which the location belongs, and the list of associated equipment; The person seeking help selects the help type on the mobile terminal interface, and the help type includes IT operation and maintenance, logistics operation and maintenance, and campus security, each help type corresponding to a preset service range and an online personnel state; After the selection is completed, the audio and video acquisition module is started.

5. The intelligent campus audio and video help processing method according to claim 1, characterized in that, Before the help answer is realized, the following steps are included: A multi-dimensional priority sorting engine is started, and the input parameters of the engine include the emergency degree weight of the help type, the emergency event frequency in the historical help records of the person seeking help, the help congestion coefficient of the current period in the region, and the preset location priority level; the parameters are weighted and calculated by the analytic hierarchy process to obtain a priority score of the help request, and based on the priority score, an online administrator list corresponding to the help type is called from the permission management module, the administrators in the online administrator list are filtered according to the preset location distribution range and the listening priority to form a candidate handler pool; According to the current call number, the historical processing time, and the work order backlog of each administrator in the candidate handler pool, a load value is calculated, and the structured help request is preferentially distributed to the administrator with the lowest load value and the matching priority; if no response is obtained within a preset second, the distribution range is automatically expanded to other priority administrators, and a multi-level reminding mechanism is triggered, including an interface pop-up window, a sound alarm, and a short message notification.

6. The intelligent campus audio and video help processing method according to claim 3, characterized in that, The distribution range is automatically expanded to other priority administrators of the same type, including: A judgment matrix is constructed by using the analytic hierarchy process, four parameters of teaching area, training area, dormitory area, and public area are taken as the criterion layer, the priority score is taken as the target layer, the relative importance of each parameter is determined, and a weight vector is generated; The standardization value of the parameter involved in the audio and video stream is multiplied by the corresponding weight and summed to obtain an initial priority score; The initial priority score is modified according to the emergency keywords contained in the structured help request, the modified value is divided into low, medium, and high levels, and the matching online administrator is determined according to the level result to realize automatic expansion.

7. An electronic device, comprising: It includes: A processor and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the intelligent campus audio and video help processing method as claimed in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, the computer program includes program instructions, the program instructions are executed by the processor of the electronic device, and the processor executes the intelligent campus audio and video help processing method as claimed in any one of claims 1 to 6.

Images