CN117499600A - An intelligent analysis method, platform and storage medium for mine surveillance videos - Google Patents

Abstract

The invention provides an intelligent analysis method, platform and storage medium for mine monitoring videos. The method of the present invention includes: S1. Entering and maintaining user personal information; S2. Configuring monitoring equipment and algorithms based on the user's personal information; S3. Based on the processing results of steps S1 and S2, dynamically analyzing the monitoring screen and reporting abnormalities The event is pushed intelligently; S4. Visually display the analysis results in step S3 on the cloud. The invention can realize one-click query of original monitoring pictures, intelligent analysis pictures and intelligent alarm of abnormal personnel behavior, which can greatly avoid the problems of traditional manual video observation methods such as heavy workload and error-prone, and further improve the safety of mine personnel activities. . On the basis of the above, the intelligent transformation of monitoring equipment can also be realized, thereby empowering the intelligent construction of mines.

Description

An intelligent analysis method, platform and storage medium for mine surveillance videos

Technical field

The present invention relates to the technical field of intelligent mining of mining engineering video surveillance information and cloud management and control. Specifically, it relates to an intelligent analysis method, platform and storage medium for mine surveillance video.

Background technique

Video surveillance is the most traditional method of mine production safety supervision. It is often used to monitor production activities in various areas of the mine in real time to facilitate timely discovery of potential risks and safety hazards to ensure the safety of workers. However, traditional video surveillance has the following shortcomings:

(1) Traditional video surveillance requires a dedicated person to watch the surveillance screen for a long time and always pay attention to whether there are any abnormalities. This working mode requires a lot of manpower and material resources. At the same time, the personnel operating the monitoring system also need to be highly concentrated for a long time, which is easy to cause Fatigue leads to omissions;

(2) The effect of manual surveillance depends on the experience and quality of the staff, and individual differences are large, making it difficult to ensure consistency.

(3) For traceback of abnormal events, manual viewing is required segment by segment, which is inefficient. Important videos may also be overwritten due to loop recording.

(4) For emergencies, manual monitoring is difficult to quickly identify and respond.

In recent years, the rapid development in the field of machine vision has greatly enriched the connotation of video surveillance, transforming its functions from traditional visual viewing to intelligent recognition. However, many industrial and mining enterprises have already deployed a large number of monitoring equipment. It will be a huge expense to introduce AI cameras. Therefore, it is necessary to design a highly compatible intelligent monitoring video mining platform to realize the intelligent upgrade of traditional monitoring equipment. Significantly.

Contents of the invention

Based on the technical issues raised above, an intelligent analysis method, platform and storage medium for mine surveillance video are provided. The invention realizes one-click query of original monitoring pictures, intelligent analysis pictures and intelligent alarm of abnormal personnel behavior, which can greatly avoid the problems of traditional manual video observation methods such as heavy workload and error-proneness, and further improve the safety of mine personnel activities.

The technical means adopted in the present invention are as follows:

An intelligent analysis method for mine surveillance videos, including:

S1. Enter and maintain user personal information;

S2. Configure monitoring equipment and algorithms based on user personal information;

S3. Based on the processing results of steps S1 and S2, dynamically analyze the monitoring screen and intelligently push abnormal events;

S4. Visually display the analysis results in step S3 on the cloud.

Further, step S1 includes:

S11. Enter the user's personal information, including name, contact number, email address and personal photo. At the same time, it is necessary to ensure that the face in the photo is clearly visible;

S12. After the photo is uploaded, the face detection module based on CompreFace is used to verify the photo. If the verification fails, the photo needs to be uploaded again.

Further, step S2 includes:

S21. Equipped with a mine environment surveillance camera. Compatible brands include Hikvision, Huawei, and Dahua. If the brand requirements are not met, use a hard disk video recorder that supports RTSP communication protocol to manage the cameras in a unified manner;

S22. Configure the surveillance video intelligent analysis network model, the compatible architecture is YOLO, and the model file suffix is pt;

S23. Configure the basic information of the camera in the cloud and upload the IP information, port number information, account information, and password information. After the upload is completed, the RTSP communication protocol is used to verify whether the uploaded information is correct. If the verification fails, it cannot be persisted;

S24. Configure the algorithm in the cloud and upload the algorithm name information, encoding information, function introduction information, and model file information. After the upload is complete, verify the architecture and file type of the uploaded model. If the verification fails, it cannot be persisted;

S25. Assign a fully configured intelligent analysis algorithm to the camera that has been entered. One camera can be assigned multiple intelligent analysis algorithms at the same time.

Furthermore, the persistent storage of cameras, intelligent analysis algorithms, and camera-algorithm relationship information relies on three data tables, including:

Data table that stores camera information, including ID, IP, port number, account number, and password fields;

A data table that stores algorithm information, including ID, algorithm name, encoding, function introduction, and model file storage path;

Camera-algorithm association table, including camera ID and algorithm ID.

Further, step S3 includes:

S31. Use Python’s multiprocessing module to control multiple processes;

S32. Use RTSP communication protocol to obtain surveillance video in real time;

S33. Analyze the surveillance video obtained in real time. When analyzing any surveillance video, two independent processes will be opened for reading and processing of surveillance images, and the inter-process shared queue will be used to complete the transmission of the images, thereby avoiding the need for analysis. Blockage caused by long time;

S34. Push fully processed videos based on the zlm streaming service, and adopt an on-demand streaming strategy. The video stream push will be enabled only after the user accesses the specific algorithm of a specific camera. At the same time, no one will view this video for 5 minutes. When the video is streaming, it will automatically turn off.

Further, step S33 includes:

S331. Read the video mining algorithm data table information, obtain the ID, coding, and model file address information of the existing algorithm, globally load the model through the model file storage address, and use the coding as the key and the loaded model as the value to construct a surveillance video. analysis method object;

S332. Based on the camera and algorithm ID in the camera-algorithm association table, multi-table joint query obtains a matrix composed of camera IP, port, account, password, and algorithm encoding;

S333. Traverse the above matrix and create two processes for each piece of data. One of the processes is based on the camera IP, port, account, and password. It uses the RTSP protocol to complete the real-time storage of the monitoring screen, and uses the inter-process sharing queue method. Store each frame of monitoring screen; another process is based on the monitoring screen in the inter-process shared queue, combines algorithm coding and constructed monitoring video analysis method object, matches the corresponding global loading model, and uses the video screen in the global queue as input Perform analysis.

Further, step S4 includes:

S41. Based on the entered basic information of the camera, the RTSP communication protocol is used to regularly detect the operating status of the camera, and finally display the number of normal and abnormal operating cameras. The default periodic detection interval is 10 minutes, which can be adjusted according to user needs;

S42. Based on the entered camera information, algorithm information and camera-algorithm association information, use a tree structure to display the camera and its matching data analysis algorithm. Click on the camera or its corresponding analysis algorithm, and step S3 will be called. Through ZLM , obtain the real-time video stream and display it in the monitoring screen display module;

S43. Based on the camera-algorithm association information, use the switch method to control the start and stop of the camera intelligent analysis algorithm to achieve flexible control of the camera function. When the camera function is adjusted, step S3 will be called to update the configuration of the camera and algorithm. Information, dynamic processing of surveillance videos;

S44. Click on the camera function and any camera or its subordinate algorithm to display the corresponding picture;

S45. Perform daily statistics on the abnormal results identified by the intelligent analysis algorithm, and use a calendar to visually display whether there is abnormal information every day;

S46. When the specified date is selected, the alarm information captured within the specific date will be displayed in the form of a carousel. When no date is selected, all alarm information will be displayed in descending order of time.

The present invention also provides a mine monitoring video intelligent analysis platform, which is implemented based on the above mine monitoring video intelligent analysis method, including:

User personal information entry and face recognition unit, used to enter and maintain user personal information;

The monitoring equipment and algorithm cloud configuration unit is used to configure the monitoring equipment and algorithm cloud based on the user's personal information;

Surveillance video edge multi-channel, multi-process dynamic analysis and push unit is used to dynamically analyze surveillance images and intelligently push abnormal events;

The cloud visual display unit is used to visually display the analysis results on the cloud.

Further, the cloud visual display unit includes: camera operation status display module, camera function and picture query module, camera algorithm management module, monitoring picture display module, alarm statistics module and abnormal information display module, wherein:

The camera operating status display module is used to regularly detect the operating status of the camera based on the entered basic camera information using the RTSP communication protocol, and finally displays the number of normal and abnormal operating cameras. The default periodic detection time interval is 10 minutes, which can be based on Adjust to user needs;

The camera function and screen query module is used to display the camera and its matching data analysis algorithm in a tree structure based on the entered camera information, algorithm information and camera-algorithm related information. Click on the camera or its corresponding analysis The algorithm will establish a communication relationship with the multi-channel, multi-thread dynamic analysis and push module of the surveillance video edge, obtain the real-time video stream through ZLM, and display it in the surveillance screen display module;

The camera algorithm management module is used to control the start and stop of the camera's intelligent analysis algorithm based on the camera-algorithm association information using a switch to achieve flexible control of the camera function. When the camera function is adjusted, it will be closely related to the surveillance video edge. The channel and multi-process dynamic analysis establishes communication with the push unit, allowing it to update the configuration information of the camera and algorithm, and perform dynamic processing of the surveillance video;

The monitoring picture display module is used to click on the camera function and any camera or its subordinate algorithm to display the corresponding picture;

The alarm statistics module is used to perform daily statistics on the abnormal results identified by the intelligent analysis algorithm, and uses a calendar to visually display whether there is abnormal information every day.

The abnormal information display module has an interactive relationship with the alarm statistics module. When a specified date is selected in the alarm statistics module, the abnormal information display module displays the alarm information captured within the specific date in the form of a carousel. When selecting a date in the alarm statistics module, all alarm information will be displayed in descending order of time.

The present invention also provides a storage medium that includes a stored program, wherein when the program is run, the above intelligent analysis method for mine monitoring video is executed.

Compared with the prior art, the present invention has the following advantages:

1. The mine monitoring video intelligent analysis method provided by the present invention can realize one-click query of original monitoring pictures, intelligent analysis pictures and intelligent alarm of abnormal personnel behavior, which can greatly avoid the heavy workload and error-prone of traditional manual video observation methods. and other issues to further improve the safety of mine personnel activities.

2. The mine monitoring video intelligent analysis method provided by the present invention can realize cloud configuration and management of monitoring equipment and algorithms, has strong compatibility, and has good scalability.

3. The mine monitoring video intelligent analysis method provided by the present invention can upgrade traditional monitoring equipment to AI equipment at a lower cost.

Based on the above reasons, the present invention can be widely promoted in fields such as intelligent mining of mining engineering video surveillance information and cloud management and control.

Description of the drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that need to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting any creative effort.

Figure 1 is an interface diagram of a user personal information entry module provided by an embodiment of the present invention;

Figure 2 is a new camera interface provided by an embodiment of the present invention;

Figure 3 is a new analysis algorithm interface provided by the embodiment of the present invention;

Figure 4 is a camera algorithm configuration page provided by an embodiment of the present invention;

Figure 5 is an alarm email for not wearing a helmet provided by an embodiment of the present invention;

Figure 6 is a real-time push log provided by an embodiment of the present invention;

Figure 7 Intelligent mining of big data from surveillance video provided by an embodiment of the present invention.

Detailed ways

In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only These are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts should fall within the scope of protection of the present invention.

It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances so that the embodiments of the invention described herein are capable of being practiced in sequences other than those illustrated or described herein. Furthermore, the terms "include" and "having" and any variations thereof are intended to cover non-exclusive inclusions, e.g., a process, method, system, product, or apparatus that encompasses a series of steps or units and need not be limited to those explicitly listed. Those steps or elements may instead include other steps or elements not expressly listed or inherent to the process, method, product or apparatus.

The invention provides an intelligent analysis method for mine monitoring video, which includes:

S1. Enter and maintain user personal information;

S2. Configure monitoring equipment and algorithms based on user personal information;

S3. Based on the processing results of steps S1 and S2, dynamically analyze the monitoring screen and intelligently push abnormal events;

S4. Visually display the analysis results in step S3 on the cloud.

In specific implementation, as a preferred embodiment of the present invention, step S1 includes:

S11. Enter the user's personal information, including name, contact number, email address and personal photo. At the same time, it is necessary to ensure that the face in the photo is clearly visible; as shown in Figure 1, when the uploaded photo does not contain facial features, it cannot be saved;

S12. After the photo is uploaded, the face detection module based on CompreFace is used to verify the photo. If the verification fails, the photo needs to be uploaded again.

In specific implementation, as a preferred embodiment of the present invention, step S2 includes:

S21. Configure a mine environment monitoring camera. Compatible brands include Hikvision, Huawei, and Dahua. If the brand requirements are not met, use a hard disk video recorder that supports RTSP communication protocol to manage the cameras in a unified manner. In this embodiment, Hikvision is selected. Nuctech's camera uses an intelligent algorithm for helmet recognition, and the algorithm model meets the requirements with the suffix .pt;

S22. Configure the surveillance video intelligent analysis network model, the compatible architecture is YOLO, and the model file suffix is pt;

S23. Configure the basic information of the camera in the cloud, upload IP information, port number information, account information, and password information. After the upload is completed, use the RTSP communication protocol to verify whether the uploaded information is correct. If the verification fails, it cannot be persisted; as shown in Figure 2 Show.

S24. Configure the algorithm in the cloud and upload the algorithm name information, encoding information, function introduction information, and model file information. After the upload is complete, verify the architecture and file type of the uploaded model. If the verification fails, it cannot be persisted; as shown in Figure 3 As shown in the figure, after the upload is complete, the platform will store the model to the specified path and verify whether the model is a model generated by the YOLO architecture. Only when the verification is successful can the algorithm be added;

S25. Assign a fully configured intelligent analysis algorithm to the camera that has been entered. One camera can be assigned multiple intelligent analysis algorithms at the same time. Upgrade traditional monitoring equipment to AI equipment, as shown in Figure 4.

In specific implementation, as a preferred embodiment of the present invention, the persistent storage of cameras, intelligent analysis algorithms, and camera-algorithm correlation information relies on three data tables, including:

Data table that stores camera information, including ID, IP, port number, account number, and password fields;

A data table that stores algorithm information, including ID, algorithm name, encoding, function introduction, and model file storage path;

Camera-algorithm association table, including camera ID and algorithm ID.

In specific implementation, as a preferred embodiment of the present invention, step S3 includes:

S31. Use Python’s multiprocessing module to control multiple processes;

S32. Use RTSP communication protocol to obtain surveillance video in real time;

S33. Analyze the surveillance video obtained in real time. When analyzing any surveillance video, two independent processes will be opened for reading and processing of surveillance images, and the inter-process shared queue will be used to complete the transmission of the images, thereby avoiding the need for analysis. Blockage caused by long time; the step S33 includes:

S331. Read the video mining algorithm data table information, obtain the ID, coding, and model file address information of the existing algorithm, load the model globally through the model file storage address, and use the coding as the key and the loaded model as the value to construct a surveillance video analysis method object;

S332. Based on the camera and algorithm ID in the camera-algorithm association table, multi-table joint query obtains a matrix composed of camera IP, port, account, password, and algorithm encoding;

S333. Traverse the above matrix and create two processes for each piece of data. One of the processes is based on the camera IP, port, account, and password. It uses the RTSP protocol to complete the real-time storage of the monitoring screen, and uses the inter-process sharing queue method. Store each frame of monitoring screen; another process is based on the monitoring screen in the inter-process shared queue, combines algorithm coding and constructed monitoring video analysis method object, matches the corresponding global loading model, and uses the video screen in the global queue as input Perform analysis.

S334. When a person who is not wearing a helmet is detected, the platform will determine the specific information of the person who is not wearing a helmet based on the face database entered by the user, and notify and remind them through broadcasts, text messages, emails, etc. to avoid accidents. occurs, as shown in Figure 5. At the same time, the image, occurrence event, and exception type will be persisted to facilitate subsequent accountability and police information review. At the same time, the person will not be tested again within 5 minutes to see if he is wearing a safe Cap is designed to provide a certain processing time, and after more than 5 minutes, it will enter the alarm process again;

S34. Push the fully processed video based on the zlm streaming service, and adopt the on-demand streaming strategy. When and only when the user accesses the specific algorithm of a specific camera, the video stream push will be enabled, as shown in Figure 6. At the same time When no one views this video stream for 5 minutes, it will be automatically closed.

In specific implementation, as a preferred embodiment of the present invention, step S4 includes:

S41. Based on the entered basic information of the camera, the RTSP communication protocol is used to regularly detect the operating status of the camera, and finally display the number of normal and abnormal operating cameras. The default periodic detection interval is 10 minutes, which can be adjusted according to user needs;

S42. Based on the entered camera information, algorithm information and camera-algorithm association information, use a tree structure to display the camera and its matching data analysis algorithm. Click on the camera or its corresponding analysis algorithm, and step S3 will be called. Through ZLM , obtain the real-time video stream and display it in the monitoring screen display module;

S43. Based on the camera-algorithm association information, use the switch method to control the start and stop of the camera intelligent analysis algorithm to achieve flexible control of the camera function. When the camera function is adjusted, step S3 will be called to update the configuration of the camera and algorithm. Information, dynamic processing of surveillance videos;

S44. Click on the camera function and any camera or its subordinate algorithm to display the corresponding picture;

S45. Perform daily statistics on the abnormal results identified by the intelligent analysis algorithm, and use a calendar to visually display whether there is abnormal information every day;

S46. When the specified date is selected, the alarm information captured within the specific date will be displayed in the form of a carousel. When no date is selected, all alarm information will be displayed in descending order of time.

Embodiments of the present invention also provide a mine monitoring video intelligent analysis platform, which is implemented based on the mine monitoring video intelligent analysis method, including:

User personal information entry and face recognition unit, used to enter and maintain user personal information;

The monitoring equipment and algorithm cloud configuration unit is used to configure the monitoring equipment and algorithm cloud based on the user's personal information;

Surveillance video edge multi-channel, multi-process dynamic analysis and push unit is used to dynamically analyze surveillance images and intelligently push abnormal events;

The cloud visual display unit is used to visually display the analysis results on the cloud.

During specific implementation, as a preferred embodiment of the present invention, the cloud visual display unit includes: camera operation status display module, camera function and picture query module, camera algorithm management module, monitoring picture display module, alarm statistics module and abnormal information display. module, where:

The camera operation status display module is used to regularly detect the operation status of the camera based on the basic information of the camera entered in Figure 2, using the RTSP communication protocol, to achieve statistics and visualization of the number of normal and abnormal operating cameras, as shown in ① in Figure 7 shown, where the camera running status detection interval is 10 minutes;

The camera function and screen query module is based on the camera information entered in Figure 2, the algorithm information entered in Figure 3, and the camera-algorithm association information configured in Figure 4, and uses a tree structure to display the cameras integrated in the current instance. and its corresponding analysis functions, as shown in ② in Figure 7. Click on the camera or its corresponding analysis algorithm in the tree structure to establish a communication relationship with the multi-channel, multi-process dynamic analysis and push unit at the edge of the surveillance video. Through ZLM, Obtain real-time video streams and display them in the monitoring screen display module;

The camera algorithm management module is based on the camera-algorithm association information configured in Figure 4, and uses a switch to control the start and stop of the camera's intelligent analysis algorithm to achieve flexible control of the camera function, as shown in ③ in Figure 7 . After the camera function is adjusted, it will also establish communication with the multi-channel, multi-process dynamic analysis and push module of the surveillance video edge, instructing it to re-read the configuration information and perform dynamic processing of the surveillance video;

The monitoring screen display module is a subordinate module of the camera function and screen query module. Clicking on any camera or its subordinate algorithm in the camera function and screen query module will display its corresponding screen in this module, as shown in Figure 7 Shown in ④;

The alarm statistics module is based on the alarm results of not wearing a safety helmet, and uses a calendar to visually display whether there is abnormal information every day, as shown in ⑤ in Figure 7;

The abnormal information display module has an interactive relationship with the alarm statistics module. When a specified date is selected in the alarm statistics module, this module will display the alarm information captured within the specific date in the form of a carousel. , and when no date is selected in the alarm statistics module, the module will display all alarm information in descending order of time, as shown in ⑥ in Figure 7.

In the above-mentioned embodiments of the present invention, each embodiment is described with its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units may be a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or may be Integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the units or modules may be in electrical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software functional units.

If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or contributes to the existing technology or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions to cause a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present invention. The aforementioned storage media include: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program code. .

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions described in the foregoing embodiments can still be modified, or some or all of the technical features can be equivalently replaced; and these modifications or substitutions do not deviate from the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention. scope.

Claims (10)

1. The intelligent analysis method for the mine monitoring video is characterized by comprising the following steps of:

s1, inputting and maintaining personal information of a user;

s2, configuring monitoring equipment and algorithms based on personal information of a user;

s3, dynamically analyzing the monitoring picture based on the processing results of the step S1 and the step S2, and intelligently pushing the abnormal event;

and S4, carrying out cloud visualization display on the analysis result in the step S3.

2. The intelligent analysis method for mine monitoring video according to claim 1, wherein the step S1 comprises:

s11, inputting personal information of a user, including names, contact phones, mailboxes and personal photos, and ensuring that faces in the photos are clearly visible;

and S12, after the photo uploading is complete, checking the photo by adopting a face detection module based on CompreFace, and if the check fails, re-uploading the photo is needed.

3. The intelligent analysis method of mine monitoring video according to claim 1, wherein the step S2 comprises:

s21, configuring a mine environment monitoring camera, wherein the compatible brands comprise Haikang Wei view, hua Ji Dahua, if the brand requirements are not met, adopting a hard disk video recorder supporting an RTSP communication protocol to uniformly manage the camera;

s22, configuring a monitoring video intelligent analysis network model, wherein a compatible architecture is YOLO, and a model file suffix is pt;

s23, configuring basic information of a camera by a cloud, uploading IP information, port number information, account information and password information, and checking whether the uploaded information is correct or not by adopting an RTSP communication protocol after the uploading is finished, wherein if the checking fails, the information cannot be stored in a lasting mode;

s24, a cloud configuration algorithm is used for uploading algorithm name information, coding information, function introduction information and model file information, after uploading is completed, the architecture and the file type of an uploading model are checked, and if the checking fails, the persistence can not be maintained;

s25, distributing the well-configured intelligent analysis algorithms for the recorded cameras, wherein one camera can distribute a plurality of intelligent analysis algorithms at the same time.

4. The intelligent analysis method for mine monitoring video according to claim 3, wherein the persistence of the camera, the intelligent analysis algorithm and the camera-algorithm association relation information depends on three data tables, and the method comprises the following steps:

the data table for storing the camera information comprises an ID, an IP, a port number, an account number and a password field;

a data table for storing algorithm information, comprising an ID, an algorithm name, a code, a function introduction and a model file storage path;

the camera-algorithm association relation table comprises camera IDs and algorithm IDs.

5. The intelligent analysis method of mine monitoring video according to claim 1, wherein the step S3 comprises:

s31, performing multiprocessing control by adopting a multiprocessing module of Python;

s32, acquiring a monitoring video in real time by adopting an RTSP communication protocol;

s33, analyzing the monitoring video acquired in real time, opening up two independent processes for respectively reading and processing the monitoring picture when analyzing any monitoring video, and completing the transmission of the picture by utilizing an inter-process sharing queue, thereby avoiding the blocking caused by long analysis time;

s34, pushing the complete video based on zlm streaming media service, adopting a strategy of pushing the video according to the need, and starting pushing the video stream if and only after a user accesses a specific algorithm of a specific camera, and simultaneously automatically closing the video stream when no one views the video stream for 5 minutes.

6. The intelligent analysis method for mine monitoring video according to claim 5, wherein the step S33 includes:

s331, reading data table information of a video mining algorithm, obtaining ID, code and model file address information of the existing algorithm, globally loading a model through a model file storage address, and constructing a monitoring video analysis method object by taking the code as a key and the loaded model as a value;

s332, based on the camera and the algorithm ID in the camera-algorithm association relation table, the multi-table joint check obtains a matrix formed by the camera IP, the port, the account number, the password and the algorithm code;

s333, traversing the matrix, respectively creating 2 processes for each piece of data, wherein one process is based on the IP, the port, the account number and the password of the camera, and the real-time storage of the monitoring picture is completed by adopting an RTSP protocol mode, and each frame of monitoring picture is stored by adopting an inter-process sharing queue mode; the other process is based on the monitoring picture in the inter-process shared queue, combines the algorithm coding and the constructed monitoring video analysis method object, matches to obtain a corresponding global loading model, and takes the video picture in the global queue as input for analysis.

7. The intelligent analysis method of mine monitoring video according to claim 1, wherein the step S4 comprises:

s41, based on recorded basic information of the cameras, carrying out timing detection on the running states of the cameras by adopting an RTSP communication protocol, and finally displaying the number of the cameras which are normally and abnormally operated, wherein the default periodic detection time interval is 10min, and the adjustment can be carried out according to the requirements of users;

s42, based on the recorded camera information, algorithm information and camera-algorithm association information, displaying the cameras and the data analysis algorithms matched with the cameras in a tree structure mode, clicking the cameras or the corresponding analysis algorithms, calling the step S3, acquiring a real-time video stream through ZLM, and displaying the real-time video stream by a monitoring picture display module;

s43, based on camera-algorithm association information, a switch mode is adopted to control the start and stop of a camera intelligent analysis algorithm, flexible control of camera functions is achieved, and after the camera functions are adjusted, step S3 is called to update configuration information of the camera and the algorithm, and dynamic processing of a monitoring video is carried out;

s44, clicking the camera function and any camera or subordinate algorithm thereof to display a corresponding picture;

s45, carrying out daily statistics on the abnormal result identified by the intelligent analysis algorithm, and visually displaying whether abnormal information exists every day in a calendar form;

s46, after the appointed date is selected, displaying the alarm information captured in the specific date in a carousel form, and when the date is not selected, displaying all the alarm information in a descending order of time.

8. An intelligent analysis platform for mine monitoring video, which is characterized in that the intelligent analysis platform is realized based on the intelligent analysis method for mine monitoring video according to any one of claims 1-7, and comprises the following steps:

the personal information input and face recognition unit is used for inputting and maintaining personal information of the user;

the monitoring equipment and algorithm cloud configuration unit is used for configuring the monitoring equipment and algorithm cloud based on personal information of the user;

the monitoring video edge multichannel and multiprocess dynamic analysis and pushing unit is used for carrying out dynamic analysis on the monitoring picture and carrying out intelligent pushing on the abnormal event;

and the cloud visual display unit is used for carrying out cloud visual display on the analysis result.

9. The mine monitoring video intelligent analysis platform of claim 8, wherein the cloud visualization presentation unit comprises: the camera operation condition display module, the camera function and picture inquiry module, the camera algorithm management module, the monitoring picture display module, the alarm statistics module and the abnormal information display module, wherein:

the camera running condition display module is used for detecting the running state of the cameras at regular time by adopting an RTSP communication protocol based on the recorded basic information of the cameras, and finally displaying the quantity of the cameras which are normally and abnormally operated, wherein the default periodic detection time interval is 10min, and the camera running condition display module can be adjusted according to the user requirements;

the camera function and picture inquiry module is used for displaying the cameras and the data analysis algorithms matched with the cameras in a tree structure mode based on the recorded camera information, algorithm information and camera-algorithm association information, clicking the cameras or the corresponding analysis algorithms, establishing a communication relation with the monitoring video edge multichannel and multithread dynamic analysis and pushing module, acquiring real-time video streams through ZLM, and displaying the real-time video streams by the monitoring picture display module;

the camera algorithm management module is used for controlling the starting and stopping of the intelligent analysis algorithm of the camera based on the camera-algorithm association information in a switching mode to realize flexible control of the camera function, and after the camera function is adjusted, communication connection is established with the monitoring video edge multichannel and multiprocess dynamic analysis and pushing unit so as to update configuration information of the camera and the algorithm and perform dynamic processing of the monitoring video;

the monitoring picture display module is used for clicking the camera function and any camera or subordinate algorithm thereof to display corresponding pictures;

the alarm statistics module is used for carrying out daily statistics on the abnormal result identified by the intelligent analysis algorithm, and visually displaying whether abnormal information exists every day in a calendar mode.

The abnormal information display module has an interactive relation with the alarm statistics module, when a specified date is selected in the alarm statistics module, the abnormal information display module displays the alarm information captured in the specified date in a carousel form, and when the date is not selected in the alarm statistics module, all the alarm information is displayed in a descending order of time.

10. A storage medium comprising a stored program, wherein the program, when run, performs the mine monitoring video intelligent analysis method of any one of claims 1 to 7.