A management system for real-time mining area headcount based on video streaming

Abstract

The present invention provides a real-time mining area headcount management system based on video streams. The system comprises physical counting devices equal in number to the number of mine entrances and exits, optical counting devices equal in number to the number of mine tunnel nodes, a processor, and a feature recognition device. The physical counting devices are located at the mine entrances and exits to count the actual number of people entering and exiting the mine; the optical counting devices are located at the top of the mine tunnel intersection and within the mine tunnel to capture real-time video streams from various areas within the mine tunnel; and the feature recognition devices are located at the mine entrances and exits to capture optical images of people entering in the standard direction. The processor receives various data, uses the physical counting information as a benchmark for the total number of people in the mine, and uses the video stream to determine the movement of people between the optical counting devices and the number of people who can be tracked within the mine tunnel per unit time. This technical solution uses node segmentation and video streaming to determine the activity of people within the mine.

Description

Technical Field

[0001] The present invention relates to the field of information and communication technology specifically suitable for administrative, commercial, financial, management or supervisory purposes, and in particular to a management system for real-time mining area headcount based on video streaming. Background Art

[0002] Coal mine safety has always threatened the health and life of miners. At present, the underground conditions of coal mines are harsh, and the mining process is threatened by natural disasters such as gas, roof collapse, and water seepage.

[0003] Mining personnel management involves pre-entry registration, real-time monitoring of personnel information, and uploading of their status information to monitoring terminals. These methods, leveraging the internet, enable real-time monitoring of miners. However, they lack detailed management of the number of miners entering the mining area, their information, and the personnel requirements for the current mining process. This results in lax management and the potential for monitoring loopholes, which can lead to mining accidents.

[0004] The existing patent CN110821565B discloses a coal mine safety production area personnel management system and method based on coal mine personnel position monitoring. The mining area monitoring terminal is set in each mining area to detect the status information of the coal mine personnel, the coal mine environmental status information and the coal mine personnel's mine entry information, and upload the above information to the monitoring cloud; the monitoring portable terminal is set on each coal mine personnel to obtain the coal mine personnel's current location information, identity information and the mining area to be worked in real time, and upload the above information to the monitoring cloud; the monitoring cloud is used to configure the coal mine work information, coal mine personnel information and coal mine equipment information operation port; the various types of coal mine information received are also heterogeneous and aggregated, and a unified access port is configured to provide users with a coal mine information visualization operation interface. The mining process is inspected through the mining area monitoring terminal and the monitoring portable terminal, the mining process environment is monitored, and the coal mine personnel attendance status is monitored.

[0005] This approach uses real-time monitoring, which, while sufficient for managing personnel in the mines, severely impacts their privacy. Furthermore, this solution relies heavily on location transmission technology, making the system less stable in the event of equipment failure or extreme conditions.

[0006] Therefore, it is necessary to provide a management system based on real-time mining area headcount counting of video streams to solve the above technical problems. Summary of the Invention

[0007] The present invention aims to solve the problem that the existing technology adopts a real-time monitoring method. Although this method meets the implementation management of miners, it seriously affects the privacy of miners. At the same time, this technical solution relies heavily on position transmission technology. When equipment fails or under extreme conditions in the mine, the system stability is poor. A management system based on real-time mining area population statistics based on video streams is provided. The number of people is calculated by physical and optical directions, and non-corresponding data in optical detection is searched based on physical detection, thereby solving the above problems.

[0008] The present invention provides a management system for real-time mining area headcount based on video streams, comprising a physical counting device having the same number of mining area entrances and exits, a plurality of optical counting devices having the same number of mining area roadway nodes, a processor and a feature recognition device;

[0009] The physical counting device is installed at the entrance and exit of the mining area to count the actual number of people entering and leaving the mining area and use it as an adjustment base for actual statistics. The physical counting device increases or decreases the total number of people in the mine based on the two situations of entry and exit.

[0010] Multiple optical counting devices are installed at the top of the mine tunnel intersection and inside the mine tunnel to obtain real-time video streams from various areas within the mine tunnel and perform video recognition of the number and movement of people in each area of ​​the mine tunnel;

[0011] Feature recognition devices are installed at the entrances and exits of the mining area to collect optical images of the standard direction of people entering;

[0012] The processor receives the physical counting information of each physical counting device, the video stream of each optical counting device, and the position mark of the corresponding optical counting device. The physical counting information is used as the benchmark for the total number of people in the mine, and the video stream is used to determine the movement of people between each optical counting device and the number of people who can be tracked in the mine tunnel per unit time.

[0013] The management system for real-time mining area headcount based on video stream of the present invention is preferably configured such that the processor includes a personnel movement module, a feature recognition module, a data correction module, an image preprocessing module, a missing personnel module, and a missing personnel judgment module.

[0014] The image preprocessing module enhances the acquired video image;

[0015] The feature recognition module generates an image set personnel database based on the personnel features obtained at the entrance and exit;

[0016] The personnel movement module identifies the personnel and their movements in the video stream based on the acquired personnel features;

[0017] The data correction module compares the results collected by the personnel movement module with the image set personnel database, adds the identified personnel features to the corresponding personnel information in the image set personnel database, and updates the personnel features;

[0018] The missing person module is used to identify people who have not appeared in any video stream within a unit of time, generate a list of people to be found, and track their last location and movement in the video stream;

[0019] The missing judgment module is used to determine the maximum range of the area where the search personnel should be located based on the last appearance time, location, movement and average moving speed, combined with the mine map.

[0020] The present invention provides a management system for real-time mining area headcount based on video streams. As an optimal method, the unit time corresponding to each person in the image set personnel library is inversely proportional to the maximum value of the preset time and the personnel density in the current area.

[0021] The management system for real-time mining area headcount based on video stream described in the present invention is, as an optimal method, that the feature recognition device and the optical counting device are both arranged at the top of the mine tunnel, and the height at which the feature recognition device is set is the same as the average height at which the optical counting device is set.

[0022] The management system for real-time mining area headcount based on video streams described in the present invention, as a preferred embodiment, further includes an authentication device carried by the mine entrants, and each authentication device can authenticate each other without duplication. The authentication device receives data from the missing person module's to-be-found person table and verifies whether the user identity corresponding to the authentication device exists. If so, a verification procedure is initiated. Otherwise, the to-be-found person table is received after the next unit of time.

[0023] When the verification procedure is started, the user needs to authenticate his or her own authentication device through other authentication devices, or appear in the video stream of the optical counting device within a threshold time. Otherwise, the missing judgment module will broadcast the missing person of the current user to the server.

[0024] The beneficial effects of the present invention are as follows:

[0025] This technical solution uses nodes and video surveillance to divide the mine area into several zones based on the longitudinal layout of the tunnels, dividing each tunnel into several zones at their intersections. Based on the image features of the video stream, the characteristics of individuals are analyzed to establish a personnel database. Based on this, the characteristics of the tunnels allow the approximate range of individuals' movements to be determined. This provides the most likely search instructions in the event of missing individuals. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A schematic diagram of a management system for real-time mining area headcount based on video streaming;

[0027] Figure 2 This is a processor architecture diagram for a management system that uses real-time mining area headcount counting based on video streaming.

[0028] Reference numerals:

[0029] 1. Physical counting device; 2. Optical counting device; 3. Processor; 31. Personnel movement module; 32. Feature recognition module; 33. Data correction module; 34. Image preprocessing module; 35. Missing personnel module; 36. Missing judgment module; 4. Feature recognition device. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Example 1

[0031] like Figure 1 As shown, a management system for real-time mining area headcount based on video streams includes a physical counting device 1 equal to the number of mining area entrances and exits, a plurality of optical counting devices 2 equal to the number of mining area roadway nodes, a processor 3 and a feature recognition device 4;

[0032] The physical counting device 1 is set at the entrance and exit of the mining area, and is used to physically count the actual number of people entering and leaving the mining area and use it as an adjustment base for actual statistics. The physical counting device 1 increases or decreases the total number of people in the mine according to the two situations of entry and exit;

[0033] Multiple optical counting devices 2 are set up at the top of the mine tunnel intersection and inside the mine tunnel to obtain real-time video streams of various areas in the mine tunnel and perform video recognition of the number and movement of people in each area of ​​the mine tunnel;

[0034] The feature recognition device 4 is set at the entrance and exit of the mining area to collect optical images of the standard direction of the entering personnel;

[0035] The processor 3 receives the physical counting information of each physical counting device 1, the video stream of each optical counting device 2, and the position mark of the corresponding optical counting device 2. The physical counting information is used as the basis for the total number of people in the mine, and the personnel movement between each optical counting device 2 and the number of people who can be tracked in the mine tunnel per unit time are judged through the video stream.

[0036] In order to avoid unnecessary waste of resources, the feature recognition device 4 should be set in the mine tunnel inside the physical counting device 1.

[0037] like Figure 2As shown, the processor 3 includes a personnel movement module 31, a feature recognition module 32, a data correction module 33, an image pre-processing module 34, a missing personnel module 35 and a missing judgment module 36;

[0038] The image pre-processing module 34 enhances the acquired video image;

[0039] The feature recognition module 32 generates an image set personnel database based on the personnel features obtained at the entrance and exit;

[0040] The personnel movement module 31 identifies the personnel and their movements in the video stream based on the acquired personnel features;

[0041] The data correction module 33 compares the results collected by the personnel movement module 31 with the image set personnel database, adds the identified personnel features to the corresponding personnel information in the image set personnel database, and updates the personnel features;

[0042] The missing person module 35 is used to identify persons who have not appeared in any video stream within a unit time and generate a list of persons to be found, and track their last position and movement in the video stream;

[0043] The missing judgment module 36 is used to judge the maximum range of the area where the search agent should be located based on the last appearance time, location, movement and average moving speed, combined with the mine tunnel map.

[0044] The unit time corresponding to each person in the image set personnel database is inversely proportional to the maximum value of the preset time and the personnel density in the current area.

[0045] The feature recognition device 4 and the optical counting device 2 are both installed at the top of the mine tunnel, and the feature recognition device 4 is installed at the same height as the average height of the optical counting device 2. In this embodiment, the feature recognition device 4 and the optical counting device 2 are both cameras.

[0046] The specific method of obtaining the person information of the video stream based on the image set person library includes the following steps:

[0047] S1. Select the target image features to be retrieved and obtain the video to be retrieved at the same time;

[0048] S2. Acquire all person target images in each frame of the video through target detection, form a person image set based on the person target images, and assign an ID to each person target image;

[0049] S3, extracting feature vectors of the target image and the images in the person image set;

[0050] S4, clustering the image feature vectors of the people in the video to obtain at least one cluster, and selecting the corresponding cluster center;

[0051] S5. Calculate the similarity between the target person's feature vector and the cluster center, and calculate the similarity between the target person's feature vector and all feature vectors in similar classes to obtain similar persons.

[0052] This embodiment is used in practice. Through the authentication device carried by the mining personnel, each authentication device can authenticate each other without duplication. The authentication device receives the data in the missing person module 35 to be found person table, verifies whether the user identity corresponding to the authentication device exists, and if so, starts the verification procedure. Otherwise, it waits for the next unit time to receive the to-be-found person table;

[0053] When the verification procedure is started, the user needs to authenticate his own authentication device through other authentication devices, or appear in the video stream of the optical counting device 2 within a threshold time. Otherwise, the absence judgment module 36 broadcasts the current user's absence to the server.

[0054] In this embodiment, the physical counting device 1 includes a chassis and a door body rotatably connected to the chassis. A rotary cylinder is provided at the bottom of the chassis, and a rotating shaft is provided on one side of the door body. The bottom end of the rotating shaft is connected to the output end of the rotary cylinder. The upper end of the rotating shaft passes through the chassis and extends to the outside. A sensor rotating bracket is provided at the top of the rotating shaft.

[0055] The sensor rotary bracket seat is set in an I-shape, and the two sides of the sensor rotary bracket seat are respectively rotatably connected with the sensor rotary bracket, and the inner ends of the sensor rotary bracket are connected to the sensor rotary bracket seat through reset hinges;

[0056] The sensor rotating bracket includes an exit personnel counting sensor bracket and an entry personnel counting sensor bracket. The lower end surfaces of the exit personnel counting sensor bracket and the entry personnel counting sensor bracket are respectively provided with independent counting personnel counting sensors, and the end of the sensor rotating bracket is provided with a pulley;

[0057] A slide bracket is provided on the outer side of the sensor rotary bracket, and a slot is provided on the side wall of the slide bracket for slidingly cooperating with the pulley.

[0058] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A management system for real-time mining area headcount based on video stream, characterized by: It includes a physical counting device (1) having a number equal to the number of the entrances and exits of the mining area, a plurality of optical counting devices (2) having a number equal to the number of the nodes in the mining area lanes, a processor (3) and a feature recognition device (4); The physical counting device (1) is arranged at the entrance and exit of the mining area, and is used to physically count the actual number of people entering and leaving the mining area and use it as an adjustment base for actual statistics. The physical counting device (1) increases or decreases the total number of people in the mine according to the two situations of entry and exit; A plurality of optical counting devices (2) are arranged at the top of the mine tunnel intersection and inside the mine tunnel, and are used to obtain real-time video streams of various areas in the mine tunnel, and perform video recognition of the number and movement of people in each area in the mine tunnel; A feature recognition device (4) is provided at the entrance and exit of the mining area to collect optical images of the standard direction of the entering personnel; The processor (3) receives the physical counting information of each of the physical counting devices (1) and the video stream of each of the optical counting devices (2) and the position mark of the corresponding optical counting device (2), uses the physical counting information as a benchmark for the total number of people in the mine, and judges the movement of people between each of the optical counting devices (2) and the number of people who can be tracked in the mine tunnel per unit time through the video stream; The processor (3) includes a personnel movement module (31), a feature recognition module (32), a data correction module (33), an image pre-processing module (34), a missing personnel module (35) and a missing personnel judgment module (36); The image pre-processing module (34) enhances the acquired video image; The feature recognition module (32) generates an image set personnel library based on the personnel features acquired at the entrance and exit; The personnel movement module (31) identifies personnel and movements in the video stream based on the acquired personnel features; The data correction module (33) compares the results collected by the personnel movement module (31) with the image set personnel database, adds the identified personnel features to the corresponding personnel information in the image set personnel database, and updates the personnel features; The missing person module (35) is used to identify persons who do not appear in any video stream within a unit time and generate a list of persons to be found, and to track their last position and movement in the video stream; The missing judgment module (36) is used to judge the maximum range of the area where the search personnel should be located based on the last appearance time, location, movement and average moving speed, combined with the mine road map; The management system also includes an authentication device carried by the personnel entering the mine, and each of the authentication devices can authenticate each other without duplication. The authentication device receives the data in the waiting-for-person table of the missing personnel module (35), verifies whether the user identity corresponding to the authentication device exists, and if so, starts the verification procedure, otherwise waits for the next unit time to receive the waiting-for-person table; When the verification program is started, the user needs to authenticate his own authentication device through other authentication devices, or appear in the video stream of the optical counting device (2) within a threshold time, otherwise the absence judgment module (36) broadcasts the current user's absence to the server.

2. A management system for real-time mining area headcount based on video stream according to claim 1, characterized in that: The unit time corresponding to each person in the image set person database is inversely proportional to the maximum value of the preset time and the density of people in the current area.

3. The management system for real-time mining area headcount based on video stream according to claim 1 is characterized by: The feature recognition device (4) and the optical counting device (2) are both arranged at the top of the mine tunnel, and the height at which the feature recognition device (4) is arranged is equal to the average height at which the optical counting device (2) is arranged.

Images