A monitoring transmission system based on edge computing gateway

Abstract

The present invention relates to the field of edge computing technology, and is used to solve the management and deployment problems and poor data consistency problems existing in the existing edge computing technology. In particular, a monitoring transmission system based on an edge computing gateway is disclosed, including an edge subsystem, a central database and a central monitoring module, wherein the edge subsystem includes a plurality of edge monitoring modules, and each edge monitoring module includes an edge monitoring unit, an edge computing unit and an edge gateway; the present invention obtains monitoring information through each independent edge computing module, and the edge computing module extracts abnormal changes in the monitoring information through data analysis, and only sends the monitoring information that produces abnormal changes to the central database, thereby improving the data security of the entire platform; the unnecessary parts of the monitoring information are identified and compressed by the edge gateway, thereby reducing unnecessary data transmission, improving data collection efficiency, and reducing the time cost wasted in manual monitoring.

Description

Technical Field

[0001] The present invention relates to the field of edge computing technology, and in particular to a monitoring transmission system based on an edge computing gateway. Background Art

[0002] Edge computing is a distributed computing paradigm that transfers data computing, storage, and application service tasks from the center of the network to the edge of the network, that is, near the device or data source. In this way, edge computing can provide real-time and fast data processing services nearby to meet the industry's needs in real-time business, application intelligence, security and privacy protection. Edge computing has a wide range of application scenarios, such as smart manufacturing, intelligent transportation systems, medical and health fields, and agriculture. In smart manufacturing, edge computing can improve the reliability and efficiency of industrial production; in intelligent transportation systems, by delegating processing tasks to roadside equipment, edge computing can monitor and manage vehicles and pedestrians in real time, improving traffic flow and safety; in the field of medical health, edge computing can provide faster and more accurate medical services, such as telemedicine and real-time image processing during surgery; in the field of agriculture, edge computing technology also shows great application potential.

[0003] However, edge computing also faces some challenges, such as security needs to be improved, high network latency, insufficient system management and deployment, and poor data consistency, etc. These need to be solved through technological innovation and improvement. Summary of the invention

[0004] The purpose of the present invention is to solve the problems existing in the background technology and to propose a monitoring transmission system based on edge computing gateway.

[0005] The purpose of the present invention can be achieved through the following technical solutions: a monitoring and transmission system based on edge computing gateway, comprising an edge subsystem, a central database and a central monitoring module, wherein the edge subsystem includes several edge monitoring modules, each edge monitoring module includes an edge monitoring unit, an edge computing unit and an edge gateway.

[0006] As a preferred embodiment of the present invention, the edge monitoring unit obtains video monitoring information of the monitoring area through a high-speed spherical camera, records sound when the ambient sound is detected to be greater than a preset decibel threshold through the sound sensor, obtains the temperature T of the monitoring area through the temperature sensor at every preset time t to generate a group of data points (t, T), and obtains the smoke concentration q of the monitoring area through the smoke sensor at every preset time t to generate a group of data points (t, q).

[0007] As a preferred embodiment of the present invention, the edge computing unit records the video surveillance video and audio recording information acquired by the edge monitoring unit, and analyzes the objects in the video surveillance video whose positions are different from the positions before time t, or which suddenly appear or disappear, based on the video motion analysis technology at preset time intervals t, and numbers the objects with a number i, i=1, 2,…, n; wherein n is the total number of objects with different positions, which suddenly appear or disappear. Obtain the moving distance di of each object and the projection area Si of the moving object in the video, and obtain the average speed vi of each object through the formula vi=di / t. When an object i with a projection area of ​​Si suddenly disappears or suddenly appears compared to the previous time t, let the average speed vi of the object be the preset value vimax. Obtain the hour moment h in the current time t, analyze the sound decibel B in the current recording information, and use the formula Calculate the impact factor k1, where h2 is the preset hourly threshold, λ1 and λ2 are preset weight factors, k0 is the preset impact factor reference value, and B is the sound decibel number.

[0008] As a preferred embodiment of the present invention, by formula Calculate the anomaly index et at the current moment, where k1 is the influencing factor. Record the anomaly index et at the current moment every preset time t, retrieve the anomaly index et-1 before time t and the anomaly index e2t-1 before time 2t, and form an anomaly index vector (et-2, et-1, et). By formula Calculate the overall anomaly factor and mutation factor p. When the overall abnormal factor Greater than the preset maximum threshold When the mutation factor p is greater than the preset maximum threshold pmax, a start recording signal is generated; when the overall abnormal factor Less than the preset minimum threshold When the mutation factor p is less than the preset threshold pmin, a termination recording signal is generated. The generation time t1 of the start recording signal and the generation time t2 of the termination recording signal are obtained, and the video surveillance video between the time t1 and t2 is automatically edited and recorded as an abnormal video, and the abnormal video is sent to the edge gateway.

[0009] As a preferred embodiment of the present invention, the edge computing module counts data points about temperature (t, T), counts data points about smoke concentration (t, q), generates a two-dimensional rectangular coordinate system with time t as the horizontal axis and temperature T as the vertical axis, fills all data points about temperature into the coordinate system and then connects all data points (t, T) with a smooth curve to generate a two-dimensional image of temperature T changing with time t, and generates a two-dimensional rectangular coordinate system with time t as the horizontal axis and smoke concentration q as the vertical axis, fills all data points about smoke concentration into the coordinate system and then connects all data points (t, q) with a smooth curve to generate a two-dimensional image of smoke concentration q changing with time t.

[0010] The analytical expression T=f(t1) is obtained in the two-dimensional image of temperature T changing with time t, where t1 is the time variable in the two-dimensional image. Every preset monitoring base t0, the formula Calculate the average temperature E(T). In the two-dimensional image of smoke concentration q changing with time t, the analytical expression of the image q=g(t2) is obtained, where t2 is the time variable in the two-dimensional image. Every preset monitoring base t0, the formula Calculate the average smoke concentration E(q). By the formula Construct a two-dimensional probability density function F(T, q; 0, t0) of temperature T and smoke concentration q. Calculate the correlation coefficient B (T, q). When the average temperature E (T) is greater than the preset temperature threshold E (T) max or the average smoke concentration E (q) is greater than the preset smoke concentration threshold E (q), further judge the value of the correlation coefficient B (T, q). If the correlation coefficient B (T, q) is greater than 0 at this time, generate a fire alarm signal and record the video surveillance information recorded in the previous preset monitoring base t0 as abnormal recording.

[0011] As a preferred embodiment of the present invention, the edge gateway obtains and stores all abnormal videos, pre-processes the abnormal videos, obtains the occupied space θ of each abnormal video, and compresses the abnormal videos whose occupied space is greater than a preset threshold value θ2.

[0012] The specific compression operation is as follows: obtain the mutation factor p at each moment in the abnormal video, mark the moment when the mutation factor is less than the preset threshold pmin as a non-feature moment, obtain the original resolution r1 (0), original bit rate r2 (0) and original frame rate r3 (0) of the non-feature moment, and use the formula Calculate compression parameters: compressed resolution r1, compressed bit rate r2 and compressed frame rate r3, and compress the video at non-feature moments in the abnormal video according to the calculated compression parameter values, where c1, c2 and c3 are preset parameters.

[0013] Abnormal videos that occupy less space than a preset threshold value θ1 are deleted, and the preset threshold value θ2 is greater than the preset threshold value θ1. Each edge gateway sends the pre-processed abnormal videos to the central database.

[0014] As a preferred embodiment of the present invention, the central database classifies the acquired abnormal videos according to the number of the edge computing module sent, and classifies the abnormal videos with the same edge computing module number into a video group. For the same video group, images of the front faces of people and vehicles are obtained based on live monitoring and image processing technology, and the central database is preset with a personnel data table and a vehicle data table. The personnel data table contains the name of the person and the face recognition information of the person, and the vehicle data table contains the vehicle number and license plate number. The front face image of the person is compared with the face recognition information stored in the personnel data table, and matched based on face recognition technology. When the face recognition result matches the personnel information in the personnel data table, the person is marked as green, when the personnel information in the personnel data table cannot be matched, the person is marked as yellow, and when the face recognition result matches the blacklist in the personnel data table, the person is marked as red. The personnel marked as green, yellow and red are counted to obtain the number of green personnel n1, the number of yellow personnel n2 and the number of red personnel n3. The personnel abnormality parameter A is obtained by formula A=n3 / (n1+n2+n3).

[0015] As a preferred embodiment of the present invention, for the same video recording group, the vehicle license plate number in the vehicle image in the same video recording group is obtained based on image processing technology, and matched with the license plate number stored in the vehicle data table. When the vehicle license plate number result matches the vehicle information in the vehicle data table, the vehicle is marked as green, when the vehicle data in the vehicle data table cannot be matched, the vehicle is marked as yellow, and when the vehicle license plate number recognition result matches the blacklist in the vehicle data table, the vehicle is marked as red. The vehicles marked as green, yellow and red are counted to obtain the number of green vehicles m1, the number of yellow vehicles m2 and the number of red vehicles m3. The vehicle abnormality parameter B is obtained by the formula B=m3 / (m1+m2+m3).

[0016] As a preferred embodiment of the present invention, when the personnel abnormality parameter A is greater than the preset threshold value A', a personnel alarm signal is generated; when the vehicle abnormality parameter B is greater than the preset threshold value B', a vehicle alarm signal is generated. For the video recording group that generates a fire alarm signal, obtain the number k1 of its edge monitoring module and mark it as abnormal module one. For the video recording group that generates a personnel alarm signal, obtain the number k2 of its edge monitoring module and mark it as abnormal module two. For the video recording group that generates a vehicle alarm signal, obtain the number k3 of its edge monitoring module and mark it as abnormal module three.

[0017] The central monitoring module obtains the numbers of abnormal module one, abnormal module two and abnormal module three and the abnormal videos recorded by them and plays them through the display screen.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] (1) The present invention obtains monitoring information through each independent edge computing module. The edge computing module extracts abnormal changes in the monitoring information through data analysis and only sends the monitoring information with abnormal changes to the central database, thereby improving the data security of the entire platform;

[0020] (2) The present invention identifies unnecessary parts of monitoring information through an edge gateway and extracts their original resolution, original bit rate and original frame rate, and compresses the unnecessary parts based on data analysis, thereby reducing unnecessary data transmission, improving data collection efficiency, and reducing the time cost wasted in finding and locating special moments in monitoring information during manual monitoring;

[0021] (3) The present invention matches the face recognition results and vehicle license plate number recognition results in the monitoring information with the data stored in the personnel data table and the vehicle data table, and obtains personnel abnormality parameters and vehicle abnormality parameters by performing data operations on the matching results. It determines whether to generate a corresponding alarm signal based on the values ​​of the personnel abnormality parameters and the vehicle abnormality parameters, and performs data analysis from both personnel and vehicle aspects, thereby improving the interpretability of the data of the monitoring transmission system and the efficiency of data calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0023] Figure 1 This is the overall system block diagram of the present invention. DETAILED DESCRIPTION

[0024] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. 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 creative work are within the scope of protection of the present invention.

[0025] See also Figure 1 As shown, a monitoring and transmission system based on an edge computing gateway includes an edge subsystem, a central database and a central monitoring module, wherein the edge subsystem includes several edge monitoring modules, and each edge monitoring module includes an edge monitoring unit, an edge computing unit and an edge gateway.

[0026] The edge monitoring unit obtains video monitoring information of the monitoring area through a high-speed spherical camera, and records sound when the ambient sound is detected to be greater than a preset decibel threshold through an acoustic sensor. The temperature T of the monitoring area is obtained through a temperature sensor at preset time intervals t to generate a set of data points (t, T). The smoke concentration q of the monitoring area is obtained through a smoke sensor at preset time intervals t to generate a set of data points (t, q).

[0027] The edge computing unit records the video surveillance footage and audio recording information acquired by the edge monitoring unit, and analyzes the objects in the video surveillance footage whose positions are different from the positions before time t, or which suddenly appear or disappear, based on the video motion analysis technology at preset time intervals t, and numbers the objects with a number i, i=1, 2,…, n; wherein n is the total number of objects with different positions, which suddenly appear or disappear. Obtain the moving distance di of each object and the projection area Si of the moving object in the video, and obtain the average speed vi of each object through the formula vi=di / t. When an object i with a projection area of ​​Si suddenly disappears or appears compared to the previous time t, let the average speed vi of the object be the preset value vimax. Obtain the hour moment h in the current time t, analyze the sound decibel B in the current recording information, and use the formula Calculate the impact factor k1, where h2 is the preset hourly threshold, λ1 and λ2 are the preset weight factors, k0 is the preset impact factor reference value, and B h It is the value of the sound decibel corresponding to the hour.

[0028] By formula Calculate the anomaly index et at the current moment, where k1 is the influencing factor. Record the anomaly index et at the current moment every preset time t, retrieve the anomaly index et-1 before time t and the anomaly index e2t-1 before time 2t, and form an anomaly index vector (et-2, et-1, et). By formula Calculate the overall anomaly factor and mutation factor p. When the overall abnormal factor Greater than the preset maximum threshold When the mutation factor p is greater than the preset maximum threshold pmax, a start recording signal is generated; when the overall abnormal factor Less than the preset minimum threshold When the mutation factor p is less than the preset threshold pmin, a termination recording signal is generated. The generation time t1 of the start recording signal and the generation time t2 of the termination recording signal are obtained, and the video surveillance video between the time t1 and t2 is automatically edited and recorded as an abnormal video, and the abnormal video is sent to the edge gateway.

[0029] The edge computing module counts the data points about temperature (t, T), counts the data points about smoke concentration (t, q), generates a two-dimensional rectangular coordinate system with time t as the horizontal axis and temperature T as the vertical axis, fills all the data points about temperature into the coordinate system and then connects all the data points (t, T) with a smooth curve to generate a two-dimensional image of the temperature T changing with time t, and generates a two-dimensional rectangular coordinate system with time t as the horizontal axis and smoke concentration q as the vertical axis, fills all the data points about smoke concentration into the coordinate system and then connects all the data points (t, q) with a smooth curve to generate a two-dimensional image of the smoke concentration q changing with time t.

[0030] The analytical expression T=f(t1) is obtained in the two-dimensional image of temperature T changing with time t, where t1 is the time variable in the two-dimensional image. Every preset monitoring base t0, the formula Calculate the average temperature E(T). In the two-dimensional image of smoke concentration q changing with time t, the analytical expression of the image q=g(t2) is obtained, where t2 is the time variable in the two-dimensional image. Every preset monitoring base t0, the formula Calculate the average smoke concentration E(q). By the formula Construct a two-dimensional probability density function F(T, q; 0, t0) of temperature T and smoke concentration q. Calculate the correlation coefficient B (T, q). When the average temperature E (T) is greater than the preset temperature threshold E (T) max or the average smoke concentration E (q) is greater than the preset smoke concentration threshold E (q), further judge the value of the correlation coefficient B (T, q). If the correlation coefficient B (T, q) is greater than 0 at this time, generate a fire alarm signal and record the video surveillance information recorded in the previous preset monitoring base t0 as abnormal recording.

[0031] The edge gateway obtains and stores all abnormal videos, pre-processes the abnormal videos, obtains the occupied space θ of each abnormal video, and compresses the abnormal videos whose occupied space is greater than the preset threshold θ2. The specific compression operation is: obtain the mutation factor p at each moment in the abnormal video, mark the moment when the mutation factor is less than the preset threshold pmin as a non-feature moment, obtain the original resolution r1 (0), original bit rate r2 (0) and original frame rate r3 (0) of the non-feature moment, and use the formula Calculate compression parameters: compressed resolution r1, compressed bit rate r2 and compressed frame rate r3, and compress the video at non-feature moments in the abnormal video according to the calculated compression parameter values, where c1, c2 and c3 are preset parameters.

[0032] Abnormal videos that occupy less space than a preset threshold value θ1 are deleted, and the preset threshold value θ2 is greater than the preset threshold value θ1. Each edge gateway marks the pre-processed abnormal videos as videos to be analyzed.

[0033] The central database classifies the acquired abnormal videos according to the number of the edge computing module sent, and classifies the abnormal videos with the same edge computing module number into a video group. For the same video group, images of the front faces of people and vehicles are obtained based on live monitoring and image processing technology. The central database is preset with a personnel data table and a vehicle data table. The personnel data table contains the name of the person and the face recognition information of the person, and the vehicle data table contains the vehicle number and license plate number. The front face image of the person is compared with the face recognition information stored in the personnel data table, and matched based on face recognition technology. When the face recognition result matches the personnel information in the personnel data table, the person is marked as green, when the person information in the personnel data table cannot be matched, the person is marked as yellow, and when the face recognition result matches the blacklist in the personnel data table, the person is marked as red. The personnel marked as green, yellow and red are counted to obtain the number of green personnel n1, the number of yellow personnel n2 and the number of red personnel n3. The personnel abnormality parameter A is obtained by the formula A=n3 / (n1+n2+n3).

[0034] For the same video group, the vehicle license plate numbers in the vehicle images in the same video group are obtained based on image processing technology and matched with the license plate numbers stored in the vehicle data table. When the vehicle license plate number result matches the vehicle information in the vehicle data table, the vehicle is marked as green. When the vehicle data in the vehicle data table cannot be matched, the vehicle is marked as yellow. When the vehicle license plate number recognition result matches the blacklist in the vehicle data table, the vehicle is marked as red. The vehicles marked as green, yellow and red are counted to obtain the number of green vehicles m1, the number of yellow vehicles m2 and the number of red vehicles m3. The vehicle abnormality parameter B is obtained by the formula B=m3 / (m1+m2+m3).

[0035] When the personnel abnormality parameter A is greater than the preset threshold value A', a personnel alarm signal is generated; when the vehicle abnormality parameter B is greater than the preset threshold value B', a vehicle alarm signal is generated. For the video recording group that generates a fire alarm signal, obtain the number k1 of its edge monitoring module and mark it as abnormal module one. For the video recording group that generates a personnel alarm signal, obtain the number k2 of its edge monitoring module and mark it as abnormal module two. For the video recording group that generates a vehicle alarm signal, obtain the number k3 of its edge monitoring module and mark it as abnormal module three.

[0036] The central monitoring module obtains the numbers of abnormal module one, abnormal module two and abnormal module three and the abnormal videos recorded by them and plays them through the display screen.

[0037] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0038] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms "a", "an", and "the" are intended to include plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations;

[0039] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A monitoring transmission system based on an edge computing gateway, comprising an edge computing unit and an edge gateway, characterized in that: The edge computing unit records the video surveillance recording and audio recording information obtained by the edge monitoring unit, and analyzes the objects in the video surveillance recording whose positions are different from the previous positions, suddenly appear or disappear based on the video motion analysis technology at every preset time t, and numbers the objects, and the number symbol is i; i=1, 2, ..., n; where n is the total number of objects with different positions, suddenly appear or disappear; obtain the moving distance di of each object and the projection area Si of the moving object in the video, and obtain the average speed vi of each object through the formula vi=di / t; When an object with a projection area of ​​Si suddenly disappears or suddenly appears compared with the previous moment, the average speed of the object is set to a preset value; the hour in the current time is obtained, the sound decibels in the current recording information are analyzed, the impact factor is calculated by operation, and the abnormal index of the current moment is obtained based on the value of the impact factor; Get the hour h in the current time t, analyze the sound decibel B in the current recording information, and use the formula Calculate the impact factor k1, where h2 is the preset hourly threshold, λ1 and λ2 are the preset weight factors, and k0 is the preset impact factor reference value; By formula Calculate the abnormal index et at the current moment; Record the abnormal index of the current time at every preset time, retrieve the abnormal index before the preset time and the abnormal index twice the preset time and form an abnormal index vector; obtain the overall abnormal factor and mutation factor through formula calculation; generate a start recording signal when the overall abnormal factor is greater than the preset maximum threshold and the mutation factor is greater than the preset maximum threshold; generate a stop recording signal when the overall abnormal factor is less than the preset minimum threshold and the mutation factor is less than the preset threshold; Obtain the time when the recording start signal is generated and the time when the recording end signal is generated, automatically edit the video surveillance video between the time when the recording start signal is generated and the time when the recording end signal is generated and record it as an abnormal video, and send the abnormal video to the edge gateway; count the data points about the temperature, count the data points about the smoke concentration, obtain the average temperature and the average smoke concentration through calculation, and obtain the correlation coefficient through calculation; when the average temperature is greater than the preset temperature threshold or the average smoke concentration is greater than the preset smoke concentration threshold, further judge the value of the correlation coefficient. If the correlation coefficient is greater than 0 at this time, generate a fire alarm signal and record the video surveillance information recorded in the previous preset monitoring base as an abnormal video; The edge gateway obtains and stores all abnormal videos, pre-processes the abnormal videos, obtains the occupied space of each abnormal video, compresses the abnormal videos whose occupied space is greater than the preset threshold value 2, and deletes the abnormal videos whose occupied space is less than the preset threshold value 1, and the preset threshold value 2 is greater than the preset threshold value 1; Each edge gateway sends the pre-processed abnormal video to the central database.

2. According to claim 1, a monitoring transmission system based on edge computing gateway is characterized in that: It also includes edge monitoring unit, central database and central monitoring module: The edge monitoring unit obtains video monitoring information of the monitoring area through a high-speed spherical camera, and records sound when the ambient sound is detected to be greater than a preset decibel threshold through an acoustic sensor. It obtains the temperature of the monitoring area through a temperature sensor at preset intervals to generate a set of data points, and obtains the smoke concentration of the monitoring area through a smoke sensor at preset intervals to generate a set of data points. The central database classifies the acquired abnormal videos according to the number of the edge computing module sent, and classifies the abnormal videos with the same edge computing module number into a video group; for the same video group, images of the front faces of people and vehicles are acquired based on live monitoring and image processing technology, and the central database is preset with a personnel data table and a vehicle data table; the personnel data table contains the name of the person and the face recognition information of the person, and the vehicle data table contains the vehicle number and license plate number; the front face image of the person is compared with the face recognition information stored in the personnel data table, and matched based on face recognition technology; when the face recognition result matches the personnel information in the personnel data table, the person is marked as green, when it cannot match the personnel information in the personnel data table, the person is marked as yellow, and when the face recognition result matches the blacklist in the personnel data table, the person is marked as red; the personnel marked as green, yellow and red are counted to obtain the number of green personnel, the number of yellow personnel and the number of red personnel; the personnel abnormality parameters are obtained through calculation; Based on image processing technology, the vehicle license plate number in the vehicle image of the same video group is obtained, and matched with the license plate number stored in the vehicle data table; when the vehicle license plate number result matches the vehicle information in the vehicle data table, the vehicle is marked as green, when it cannot match the vehicle data in the vehicle data table, the vehicle is marked as yellow, and when the vehicle license plate number recognition result matches the blacklist in the vehicle data table, the vehicle is marked as red; the vehicles marked as green, yellow and red are counted to obtain the number of green vehicles, the number of yellow vehicles and the number of red vehicles; the vehicle abnormality parameters are obtained through calculation; When the personnel abnormality parameter is greater than the preset threshold, a personnel alarm signal is generated; when the vehicle abnormality parameter is greater than the preset threshold, a vehicle alarm signal is generated; for the video group that generates a fire alarm signal, the number of its edge monitoring module is obtained and marked as abnormal module one; for the video group that generates a personnel alarm signal, the number of its edge monitoring module is obtained and marked as abnormal module two; for the video group that generates a vehicle alarm signal, the number of its edge monitoring module is obtained and marked as abnormal module three; The central monitoring module obtains the numbers of abnormal module one, abnormal module two and abnormal module three and the abnormal videos recorded by them and plays them through the display screen.

3. According to claim 1, a monitoring transmission system based on edge computing gateway is characterized in that: The specific process of obtaining the overall abnormal factor and mutation factor through formula calculation is: The abnormal index et of the current time is recorded every preset time t, and the abnormal index et-1 before time t and the abnormal index e2t-1 before time 2t are retrieved to form the abnormal index vector (et-2, et-1, et); through the formula Calculate the overall anomaly factor and mutation factor p.

4. A monitoring transmission system based on edge computing gateway according to claim 1, characterized in that: The specific process of compression operation is: Get the mutation factor p at each moment in the abnormal video, mark the moment when the mutation factor is less than the preset threshold pmin as a non-feature moment, and get the original resolution r1 (0), original bit rate r2 (0) and original frame rate r3 (0) of the non-feature moment. Calculate compression parameters: compressed resolution r1, compressed bit rate r2 and compressed frame rate r3, and compress the video at non-feature moments in the abnormal video according to the calculated compression parameter values, where c1, c2 and c3 are preset parameters.

5. The monitoring transmission system based on edge computing gateway according to claim 1 is characterized in that: The specific process of obtaining the average temperature and average smoke density through calculation and the correlation coefficient through calculation is as follows: A two-dimensional rectangular coordinate system is generated with time t as the horizontal axis and temperature T as the vertical axis, all data points about temperature are filled into the coordinate system, and then all data points (t, T) are connected with a smooth curve to generate a two-dimensional image of temperature T changing with time t, and a two-dimensional rectangular coordinate system is generated with time t as the horizontal axis and smoke concentration q as the vertical axis, all data points about smoke concentration are filled into the coordinate system, and then all data points (t, q) are connected with a smooth curve to generate a two-dimensional image of smoke concentration q changing with time t; The analytical expression of the image T=f(t1) is obtained in the two-dimensional image of temperature T changing with time t, where t1 is the time variable in the two-dimensional image, and the formula is used every preset monitoring base t0. Calculate the average temperature E(T); obtain the analytical expression q=g(t2) in the two-dimensional image of smoke concentration q changing with time t, where t2 is the time variable in the two-dimensional image, and calculate the average temperature E(T) by the formula every preset monitoring base t0 Calculate the average smoke concentration E (q); by the formula Construct a two-dimensional probability density function F(T, q) of temperature T and smoke concentration q. 0, t0); through the formula Calculate the correlation coefficient B(T,q).

Images