A video monitoring method for underground coal mines

Abstract

The present invention discloses a method for video monitoring underground in a coal mine, which relates to the field of image communication technology. The method comprises the following steps: acquiring historical video data and extracting historical key feature data for analysis, establishing a feature baseline range for a normal operating state of an underground coal mine, constructing an abnormal rule library for different abnormality types, establishing a mapping relationship between historical key feature data and abnormal rules, acquiring real-time video data and audio data underground in the coal mine and extracting real-time key feature data, comparing the data with the feature baseline range, matching corresponding rules in the abnormal rule library based on the established mapping relationship, outputting abnormality analysis results, and triggering corresponding alarms according to the abnormal rules if an abnormality is determined to exist, and associating an abnormality handling process, extracting features by collecting underground videos of multiple time periods, establishing baselines and differentiated judgment conditions, constructing an abnormality library and a triggering mechanism, comparing real-time data to determine abnormalities, performing hierarchical response and handling, and realizing intelligent safety monitoring.

Description

Technical Field

[0001] The present invention relates to the field of image communication technology, and in particular to a video monitoring method for underground coal mines. Background Art

[0002] In the field of coal mine production safety, the application of video surveillance technology has become an important means to ensure operational safety and improve production efficiency. Through remote monitoring of personnel behavior, equipment status and environmental changes in coal mines, this technology plays a key role in the following aspects: In personnel management, video surveillance can be used to identify workers' illegal operating behaviors, and real-time reminders and corrections can be made, effectively reducing safety risks caused by human factors; in equipment operation monitoring, it can track the operating status of coal mining machines, belt conveyors, ventilators and gas sensors in real time, capture equipment failure hazards in advance, and avoid production accidents caused by equipment abnormalities; at the same time, video surveillance can also be used to monitor underground environmental conditions, such as tunnel ventilation efficiency, gas concentration fluctuations, etc., providing intuitive and real-time data basis for environmental safety early warning.

[0003] However, existing technical solutions still have many problems that need to be solved urgently. On the one hand, the underground environment of coal mines is complex and harsh. Objective factors such as poor lighting conditions, high dust concentration, and strong electromagnetic interference seriously affect the quality of video images, making it difficult for traditional video surveillance methods to accurately extract key feature data. Problems such as equipment recognition errors and misjudgment of personnel behavior are common, which greatly reduces the accuracy and reliability of monitoring. On the other hand, current monitoring systems lack the ability to dynamically analyze the normal operating status of underground coal mines. They usually use fixed thresholds and rules to judge anomalies, which cannot adapt to the differentiated changes in normal status in different operating periods and different production scenarios. They are very likely to generate a large number of false alarms or missed alarms, disrupting normal safe production order. In the abnormality handling link, most existing technologies can only issue simple alarms, lack a complete abnormality rule library and an automated handling process association mechanism, and cannot provide accurate handling guidance for different types and levels of abnormal situations. It is difficult to achieve rapid response and effective control of safety risks, and it is difficult to meet the needs of coal mine safety production for intelligent and refined monitoring. Summary of the Invention

[0004] The technical problem solved by the present invention is that the existing technical solutions still have many problems that need to be solved urgently: on the one hand, the underground environment of coal mines is complex and harsh, and objective factors such as poor lighting conditions, high dust concentration, and strong electromagnetic interference seriously affect the quality of video images, making it difficult for traditional video monitoring methods to accurately extract key feature data. Problems such as equipment recognition errors and misjudgment of personnel behavior are common, which greatly reduces the accuracy and reliability of monitoring; on the other hand, the current monitoring system lacks the ability to dynamically analyze the normal operating status of underground coal mines. It usually uses fixed thresholds and rules to judge anomalies, which cannot adapt to the differentiated changes in normal status in different operating periods and different production scenarios. It is very easy to generate a large number of false alarms or missed alarms, which interfere with normal safe production order; in the abnormality handling link, most existing technologies can only issue simple alarms, lack a complete abnormality rule library and an automated handling process association mechanism, and cannot provide accurate handling guidance for different types and levels of abnormal situations. It is difficult to achieve rapid response and effective control of safety risks, and it is difficult to meet the needs of coal mine safety production for intelligent and refined monitoring.

[0005] In order to solve the above technical problems, the present invention provides the following technical solution, comprising the following steps:

[0006] Step S1, acquiring historical video data, and extracting historical key feature data from historical video data of early morning shift operation scenes, historical video data of mid-shift operation scenes, and historical video data of night shift operation scenes in the historical video data;

[0007] Step S2: Analyze the historical key feature data and establish a feature baseline range for normal operation in the coal mine. For different types of anomalies, establish a rule system according to the anomaly level to build an anomaly rule library, and establish a mapping relationship between the historical key feature data and the anomaly rules.

[0008] Step S3, acquiring real-time video data and audio data from underground coal mines, and extracting real-time key feature data from the real-time video data;

[0009] Step S4: comparing the real-time key feature data with the feature baseline range, combining the real-time audio data analysis results, and matching the corresponding rules in the anomaly rule library based on the established mapping relationship between historical key feature data and anomaly rules, and outputting the anomaly analysis results;

[0010] Step S5: If it is determined that an anomaly exists based on the anomaly analysis result, a corresponding alarm is triggered according to the anomaly rule and an anomaly handling process is associated.

[0011] As a preferred embodiment of the method for video surveillance of an underground coal mine according to the present invention, the historical video data includes historical video data of operation scenes during the morning shift, historical video data of operation scenes during the mid-shift, and historical video data of operation scenes during the night shift in the underground coal mine;

[0012] The historical key feature data is extracted through image recognition technology, specifically including:

[0013] Decomposing the historical video data into a set of image frames according to a time sequence;

[0014] A fixed template matching method is used to identify the features of equipment, including a coal mining machine, a belt conveyor, a ventilator, and a gas sensor;

[0015] Preset reflective marking points in the lane and determine the equipment displacement deviation by calculating the pixel coordinate offset of the reflective marking points;

[0016] Based on grayscale threshold segmentation to extract environmental features, the area of ​​the image frame with a grayscale value higher than a first threshold is identified as bright light, and the area below a second threshold is identified as shadow;

[0017] The inter-frame difference method is used to capture dynamic features. Two adjacent image frames are selected from the image frame to perform pixel difference calculation. Areas where the grayscale values ​​of pixels between adjacent frames change and the number exceeds a set threshold are marked as moving targets. Contour analysis is used to distinguish between people and equipment.

[0018] The image frame includes a frame image in a video image sequence in historical video data.

[0019] As a preferred solution of the method for video monitoring underground coal mines described in the present invention, the historical key feature data is analyzed and a feature baseline range of a normal operating state of the coal mine is established, specifically including:

[0020] Establishing a baseline library by equipment type, including shearers, belt conveyors, ventilators, and gas sensors;

[0021] The coal mining machine includes a drum and a cutting motor, and a first baseline range is determined based on the drum height offset and the cutting motor temperature;

[0022] The belt conveyor includes a conveyor belt and rollers, and a second baseline range is determined based on the deviation of the conveyor belt and the fluctuation of the roller speed;

[0023] The ventilator includes blades, and a third baseline range is determined based on the wind pressure fluctuation magnitude and the vibration frequency of the blades;

[0024] The gas sensor determines a fourth baseline range based on a response time and a zero drift of the gas sensor;

[0025] The characteristic baseline range serves as the basic data for differentiated judgment conditions of equipment type and operation period.

[0026] As a preferred solution of the method for underground video monitoring in a coal mine according to the present invention, differentiated judgment conditions are set according to equipment type and normal operating period, specifically including:

[0027] During the morning shift, set the criteria for determining the number of moving targets in areas with dense human activity;

[0028] During the mid-shift period, set the equipment operation noise judgment conditions;

[0029] During the night shift, set the conditions for determining the brightness of the laneway lighting.

[0030] As a preferred solution of the method for underground video monitoring in a coal mine described in the present invention, a rule system is established according to the abnormality level for different abnormality types to construct an abnormality rule library, and a mapping relationship between historical key feature data and abnormality rules is established, specifically including:

[0031] Build an exception rule library and establish a rule system according to the exception level, including first-level exception rules, second-level exception rules and third-level exception rules;

[0032] The first-level abnormal rule automatically cuts off the power supply to the area and initiates the emergency evacuation process when triggered;

[0033] The secondary abnormality rule, when triggered, suspends the delivery and notifies the maintenance team process;

[0034] The three-level abnormal rules will trigger a voice warning and record the information of the violator;

[0035] Establish a mapping relationship between historical key feature data and anomaly rules, including single feature trigger rules and multi-feature association trigger rules. The multi-feature association trigger rules determine the anomaly level based on a combination of logical AND and logical OR.

[0036] Based on the statistical analysis of historical key feature data, an abnormal feature decision table is constructed. The abnormal feature decision table includes normal state judgment conditions for equipment displacement deviation, gray value fluctuation, and moving target duration, as well as corresponding first-level abnormality, second-level abnormality, and third-level abnormality judgment conditions;

[0037] Divide the equipment status into normal status, warning status, abnormal status and fault status;

[0038] The device status corresponding to the device normal status determination condition and the device abnormality determination condition specifically includes:

[0039] When the displacement deviation exceeds the normal state judgment condition, it enters the warning state; when the displacement deviation exceeds the warning state judgment condition, it enters the abnormal state; when the displacement deviation exceeds the abnormal state judgment condition, it enters the fault state;

[0040] When the grayscale value fluctuation exceeds the normal state judgment condition, it enters the warning state; when the grayscale value fluctuation exceeds the warning state judgment condition, it enters the abnormal state; when the grayscale value fluctuation exceeds the abnormal state judgment condition, it enters the fault state;

[0041] When the duration of the moving target exceeds the normal state judgment conditions, it enters the warning state;

[0042] Based on the equipment normal state determination conditions, equipment abnormality determination conditions and abnormality rules, a state transition table is established to record the correspondence between equipment state transitions and abnormality rules, specifically including:

[0043] When the equipment status changes from the warning state to the abnormal state, an audible and visual alarm is triggered and the inspection personnel are notified;

[0044] When the equipment status changes from an abnormal state to a fault state, the power supply of the equipment is automatically cut off and the emergency shutdown process is initiated;

[0045] For multi-feature association triggering rules, if two or more feature data exceed the normal state judgment conditions and reach the fault state judgment conditions at the same time, the first-level abnormality rule is triggered; if only the abnormal state judgment conditions are reached, the second-level abnormality rule is triggered; if a single feature data exceeds the normal state judgment conditions but does not reach the warning state judgment conditions, it is marked as normal state; if it exceeds the warning state judgment conditions but does not reach the abnormal state judgment conditions, the third-level abnormality rule is triggered.

[0046] As a preferred solution of the video monitoring method for underground coal mines described in the present invention, wherein: the real-time video data includes equipment operation data, personnel activity data and environmental status data;

[0047] The real-time key feature data in the video data is extracted using image recognition technology, specifically including:

[0048] Using the preset device shape template, perform template matching on the image frame;

[0049] When the matching degree exceeds the threshold set based on the mean or standard deviation of historical key feature data, the device is identified and its location information is recorded;

[0050] Grayscale processing is performed on the image frame, and areas with grayscale values ​​below a threshold set according to historical data of normal lighting in the lane are marked as shadows, and areas above the threshold are marked as highlights, so as to identify abnormal lighting in the lane;

[0051] Detect moving targets by calculating the changes in pixel positions between two adjacent image frames;

[0052] When the area of ​​the moving target exceeds a set threshold pixel, distinguishing between people and equipment by their contour shapes;

[0053] De-duplicate the extracted real-time key feature data to remove repeated occurrences of the same real-time key feature data;

[0054] The real-time key feature data are numbered in chronological order, a corresponding relationship between the real-time key feature data and the acquisition time and acquisition location is established, and the data are stored in a database.

[0055] As a preferred solution of the underground video monitoring method of a coal mine according to the present invention, the real-time key feature data is compared with the feature baseline range, and based on the established mapping relationship between historical key feature data and abnormal rules, corresponding rules are matched in the abnormal rule library, and abnormal analysis results are output, which specifically includes:

[0056] When comparing the real-time key feature data with the feature baseline range of the normal operating state, the differentiated judgment conditions corresponding to the operating period are preferentially adopted;

[0057] The device displacement deviation, grayscale value fluctuation and moving target duration are compared with the baseline differentiation judgment conditions of the device type and operation period.

[0058] As a preferred solution of the method for underground video monitoring of a coal mine according to the present invention, the rule matching logic includes:

[0059] If a single feature data exceeds the normal state judgment condition but does not meet the warning state judgment condition, it is marked as normal state; if it exceeds the warning state judgment condition but does not meet the abnormal state judgment condition, the third-level abnormality rule is triggered and it is marked as warning state;

[0060] If a single feature data reaches or exceeds the abnormal state judgment condition, the secondary abnormal rule is triggered and marked as abnormal state;

[0061] If two or more characteristic data exceed the normal state judgment conditions and meet the corresponding fault state judgment conditions at the same time, the first-level abnormality rule is triggered and marked as a fault state;

[0062] The anomaly analysis result includes the anomaly type, anomaly location and anomaly level, and the anomaly occurrence timestamp in the anomaly analysis result is marked;

[0063] The characteristic data includes device displacement deviation, gray value fluctuation, and moving target duration. The judgment conditions of the characteristic data are determined according to the abnormal characteristic decision table and state transition rules;

[0064] The moving target duration is the cumulative duration that the moving target appears in consecutive video frames.

[0065] As a preferred embodiment of the method for underground video surveillance of a coal mine according to the present invention, if an abnormality is determined to exist based on the abnormality analysis result, a corresponding alarm is triggered according to the abnormality rule, and an abnormality handling process is associated, specifically including:

[0066] When the abnormality analysis result is a level three abnormality, a voice warning device in a local area of ​​the coal mine is triggered, and the abnormal location is marked with a yellow mark on the ground monitoring center interface;

[0067] When the abnormality analysis result is a level 2 abnormality, the underground sound and light alarm system is activated, the light flashing frequency is set to c times / second, the alarm sound intensity reaches d decibels, and the abnormality information is pushed to the inspection personnel's handheld terminal;

[0068] When the abnormality analysis result is a level one abnormality, the system automatically cuts off the power supply to the abnormal area, activates the underground emergency evacuation sound and light indication system, sets the evacuation signal frequency to e times / second, sends an emergency alarm to the mine safety command center, and automatically associates the abnormality handling process;

[0069] The exception handling process is automatically associated, including:

[0070] If the abnormality type is equipment failure, the corresponding equipment maintenance manual is automatically retrieved and pushed to the maintenance team through the underground explosion-proof terminal of the coal mine;

[0071] If the exception type is a personnel violation, an electronic work order containing the time, location, and behavior photo of the violation will be automatically generated and sent to the safety management department;

[0072] If the abnormality type is environmental risk, immediately close the front and rear dampers of the abnormal area and switch the ventilation system to emergency mode.

[0073] As a preferred solution of the underground video monitoring method of a coal mine described in the present invention, after the abnormality handling process is initiated, the location of the handling personnel is tracked in real time through the RFID positioning system; if the unresponsive time exceeds the set time, a reminder is automatically sent to the immediate superior;

[0074] The RFID positioning system uses UHF frequency band technology to accurately locate personnel positions through positioning base stations deployed at intervals in underground tunnels;

[0075] After the disposal is completed, confirm the operation by scanning the device QR code. The system records the disposal completion time and archives the exception handling record.

[0076] The beneficial effects of the present invention are as follows: through multi-period historical video data collection and image recognition technology, accurate feature baselines and differentiation thresholds are established, and combined with a multi-level anomaly rule library and trigger mechanism, the problems of inaccurate feature extraction and high false alarm and missed alarm rates of traditional monitoring are effectively solved, and equipment, environment and personnel anomalies are accurately monitored, significantly reducing accident risks; at the same time, a three-level response system is constructed, and corresponding alarms are triggered for different levels of anomalies and associated handling processes, so as to achieve reasonable allocation of resources, shorten the average fault handling time, and greatly improve fault handling efficiency; in addition, full process automation reduces labor costs, and automatically recorded exception handling data provides support for management optimization. Combined with RFID positioning to track handling personnel and scan code archiving of processing records, it is convenient to review and optimize exception handling processes and rule thresholds, and ultimately realize intelligent, efficient and safe production in coal mines, which can reduce comprehensive management costs in the long term and comprehensively improve the overall management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic diagram of the basic flow of a method for underground video monitoring in a coal mine provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0078] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0079] Example 1, with reference to Figure 1 , as an embodiment of the present invention, provides a method for underground video monitoring in a coal mine, comprising the following steps:

[0080] Step S1, acquiring historical video data, and extracting historical key feature data from historical video data of the coal mine underground morning shift operation scene, historical video data of the mid-shift operation scene, and historical video data of the night shift operation scene;

[0081] Step S2: Analyze the historical key feature data and establish a feature baseline range for normal operation in the coal mine. For different anomaly types, establish a rule system according to the anomaly level to build an anomaly rule library, and establish a mapping relationship between the historical key feature data and the anomaly rules.

[0082] Step S3, acquiring real-time video data and audio data from underground coal mines, and extracting real-time key feature data from the real-time video data;

[0083] Step S4: compare the real-time key feature data with the feature baseline range, combine the real-time audio data analysis results, and match the corresponding rules in the anomaly rule library based on the established mapping relationship between historical key feature data and anomaly rules, and output the anomaly analysis results;

[0084] Step S5: Based on the abnormality analysis results, if it is determined that an abnormality exists, a corresponding alarm is triggered according to the abnormality rules and the abnormality handling process is associated.

[0085] In one embodiment, historical video data of the morning shift operation scene, the mid-shift operation scene and the night shift operation scene are collected, and image recognition technology is used to extract equipment features (including equipment type and displacement deviation), environmental features (including strong light, shadows) and dynamic features (including personnel movement targets and equipment movement targets), to construct historical key feature data, and then establish a baseline library according to the equipment type. The statistical process control method is used to determine the feature baseline range of the normal operation status of each device, and differentiated judgment conditions are set according to different operation time periods. At the same time, an abnormal rule library including first-level abnormal rules, second-level abnormal rules and third-level abnormal rules is constructed, a mapping relationship between feature data and abnormal rules is established, and the corresponding rules of equipment status division and status transfer are clarified. During real-time monitoring, real-time video of the underground well is obtained. The system uses image recognition technologies such as template matching and grayscale processing to extract real-time key feature data from video and audio data, and establishes a corresponding relationship between it and the collection time and location. It compares the real-time key feature data with the feature baseline range, combines it with real-time audio analysis, matches the rules in the abnormal rule library according to the rule matching logic, and outputs the analysis results containing the abnormal type, location, level and timestamp. If it is determined to be abnormal, the corresponding alarm will be triggered according to the abnormality level, including a local voice warning triggered by a three-level abnormality and a power cut-off and emergency evacuation initiated by a first-level abnormality. At the same time, it automatically associates the abnormality handling process, such as pushing the maintenance manual for equipment failure, generating electronic work orders for personnel violations, and switching the ventilation system for environmental risks. After the handling process is started, the RFID positioning system is used to track the handling personnel. If there is no response within the timeout, the superior will be reminded. After the handling is completed, the QR code is scanned for confirmation and the records are archived.

[0086] The historical video data includes historical video data of the morning shift operation scene, historical video data of the mid-shift operation scene, and historical video data of the night shift operation scene in the coal mine;

[0087] Historical key feature data is extracted through image recognition technology, including:

[0088] Decompose historical video data into a set of image frames according to time series;

[0089] Fixed template matching is used to identify equipment features, including coal mining machines, belt conveyors, ventilators, and gas sensors.

[0090] Preset reflective marking points in the lanes and determine the equipment displacement deviation by calculating the pixel coordinate offset of the reflective marking points;

[0091] Based on grayscale threshold segmentation to extract environmental features, the area of ​​the image frame with a grayscale value higher than a first threshold is identified as bright light, and the area below a second threshold is identified as shadow;

[0092] The inter-frame difference method is used to capture dynamic features. Two adjacent image frames are selected for pixel difference calculation. Areas where the grayscale value of pixels between adjacent frames changes and the number exceeds a set threshold are marked as moving targets. Contour analysis is used to distinguish between people and equipment.

[0093] The image frame includes a frame image in a video image sequence in the historical video data.

[0094] In one embodiment, when implementing the video surveillance method for underground coal mines, historical video data of the morning shift operation scene, historical video data of the mid-shift operation scene, and historical video data of the night shift operation scene are first collected. The morning shift period is 6:00-14:00, the mid-shift period is 14:00-22:00, and the night shift period is 22:00-6:00. The data are then decomposed into a set of image frames in time series, and a fixed template matching method is used to identify the equipment features of the coal mining machine and the belt conveyor. The equipment displacement deviation is calculated by the pixel coordinate offset of the preset reflective identification point in the lane. In terms of environmental feature extraction, the first threshold is set to 200, and based on the historical video data, the image frame is automatically generated. The grayscale histogram of the image frame during normal operation is distributed to the right side of the peak, and the grayscale value that can cover 90% of the normal lighting area is selected for determination; the second threshold is set to 30, and the grayscale value of the shadow area of ​​the historical video is statistically analyzed, and the mean is subtracted by 1.5 times the standard deviation. In this way, the area with the grayscale of the image frame higher than the first threshold is judged as strong light, and the area below 30 is identified as shadow. When capturing dynamic features, the threshold is set to 40 pixels. Based on the statistics of the grayscale changes of pixels in adjacent frames of the historical video, the 80% percentile value is taken, and the pixel difference between the two adjacent frames is calculated. The area where the grayscale value changes and the amount exceeds the set threshold is marked as a moving target, and then contour analysis is used to distinguish between people and equipment.

[0095] Analyze historical key characteristic data and establish a characteristic baseline range for normal underground coal mine operation conditions, including:

[0096] Establish a baseline library by equipment type, including coal mining machines, belt conveyors, ventilators, and gas sensors;

[0097] The coal mining machine includes a drum and a cutting motor, and a first baseline range is determined based on a drum height offset and a cutting motor temperature;

[0098] The belt conveyor includes a conveyor belt and rollers, and a second baseline range is determined based on the conveyor belt deviation and the roller speed fluctuation;

[0099] The ventilator includes blades, and a third baseline range is determined based on the blade wind pressure fluctuation magnitude and the blade vibration frequency;

[0100] A gas sensor, determining a fourth baseline range based on a response time and a zero drift of the gas sensor;

[0101] The characteristic baseline range serves as the basic data for differentiated judgment conditions based on equipment type and operating period.

[0102] In one of the embodiments, when establishing the characteristic baseline range of the normal operating status of a coal mine underground, differentiated settings need to be implemented according to the type of equipment: for coal mining machines, based on the statistical analysis of historical operating data in the past 6 months, the normal range of the roller height offset is set to ±12cm, the fluctuation threshold is determined with a 95% confidence interval, and the cutting motor temperature is set in the range of 45°C-75°C. The first baseline range is formed based on the temperature distribution of the equipment during the stable period of continuous operation; for belt conveyors, by analyzing historical operating data, the allowable range of conveyor belt deviation is set to ±8cm. Referring to the equipment safety operation standards and actual working conditions, the roller speed fluctuation is controlled within ±10% of the rated speed. The second baseline range is formed based on the speed monitoring data during daily load changes; the ventilator baseline library is Based on the historical data of wind pressure and blade vibration, the normal range of wind pressure fluctuation is set to ±150Pa. According to the statistics of wind pressure fluctuations in stable operation during different operating periods, the blade vibration frequency is between 22Hz and 28Hz. The vibration frequency distribution range during normal operation of the equipment is taken to form the third baseline range; the gas sensor baseline library sets the response time to greater than or equal to 12 seconds based on sensor performance tests and actual operation data to meet the timeliness requirements of safety monitoring. The zero point drift is controlled at ±0.03%FS. Based on the statistics of zero point offset tests during long-term operation, the fourth baseline range is formed. Each baseline range is based on the statistical analysis of historical key feature data, combined with equipment safety standards and actual operating conditions, to provide a quantitative basis for differentiated judgment of equipment status in subsequent operating periods.

[0103] Differentiated judgment conditions are set based on equipment type and normal operating hours, including:

[0104] During the morning shift, set the criteria for determining the number of moving targets in areas with dense human activity;

[0105] During the mid-shift period, set the equipment operation noise judgment conditions;

[0106] During the night shift, set the conditions for determining the brightness of the laneway lighting.

[0107] In one embodiment, when implementing differentiated judgment conditions based on equipment type and normal operating hours, the specific operations are as follows: During the morning shift, based on the patterns of human activity in historical video data, areas with dense human activity are delineated, such as working face entrances and shift handover points. A threshold for the number of moving targets in this area is set to greater than or equal to 15 people, determined based on the 90% confidence interval of the morning shift data for the past three months. Exceeding this threshold triggers an early warning. During the mid-shift period, when equipment is concentrated in operation, audio sensors are used to collect operating audio from coal mining machines and belt conveyors, extracting frequency and intensity characteristics, and comparing them with the mid-shift noise baseline range corresponding to the equipment type. For example, the normal operating noise range of a coal mining machine is 80-100 decibels. When the measured noise exceeds the baseline by ±10%, it is determined to be abnormal. During the night shift, the core is to ensure lighting safety. Based on historical tunnel lighting data, the average tunnel brightness is set to no less than 20 lux. Based on coal mine safety regulations and night operation lighting standards, if the measured brightness is lower than this threshold, it is determined to be a lighting abnormality. The judgment conditions for each time period are set in combination with historical data statistics and safety standards to ensure the targeted and effective monitoring of different operating hours.

[0108] For different anomaly types, a rule system is established according to the anomaly level to build an anomaly rule library, and a mapping relationship between historical key feature data and anomaly rules is established, including:

[0109] Build an exception rule library and establish a rule system according to the exception level, including first-level exception rules, second-level exception rules and third-level exception rules;

[0110] Level 1 exception rules, when triggered, automatically cut off the power supply to the area and initiate the emergency evacuation process;

[0111] Secondary abnormality rules: when triggered, transportation is suspended and the maintenance team is notified;

[0112] Level 3 exception rules: when triggered, voice warnings are issued and the violator's information flow is recorded;

[0113] Establish a mapping relationship between historical key feature data and anomaly rules, including single feature trigger rules and multi-feature association trigger rules. Multi-feature association trigger rules determine the anomaly level based on a combination of logical AND and logical OR.

[0114] Based on the statistical analysis of historical key feature data, an abnormal feature decision table is constructed. The abnormal feature decision table includes the normal state judgment conditions of equipment displacement deviation, gray value fluctuation, and moving target duration, as well as the corresponding first-level abnormality, second-level abnormality, and third-level abnormality judgment conditions;

[0115] Divide the equipment status into normal status, warning status, abnormal status and fault status;

[0116] The device status is determined based on the normal status judgment conditions and abnormal status judgment conditions, including:

[0117] When the displacement deviation exceeds the normal state judgment condition, it enters the warning state; when the displacement deviation exceeds the warning state judgment condition, it enters the abnormal state; when the displacement deviation exceeds the abnormal state judgment condition, it enters the fault state;

[0118] When the grayscale value fluctuation exceeds the normal state judgment condition, it enters the warning state; when the grayscale value fluctuation exceeds the warning state judgment condition, it enters the abnormal state; when the grayscale value fluctuation exceeds the abnormal state judgment condition, it enters the fault state;

[0119] When the duration of the moving target exceeds the normal state judgment conditions, it enters the warning state;

[0120] Based on the equipment normal state determination conditions, equipment abnormality determination conditions and abnormality rules, a state transition table is established to record the correspondence between equipment state transitions and abnormality rules, specifically including:

[0121] When the equipment status changes from the warning state to the abnormal state, an audible and visual alarm is triggered and the inspection personnel are notified;

[0122] When the equipment status changes from an abnormal state to a fault state, the power supply of the equipment is automatically cut off and the emergency shutdown process is initiated;

[0123] For multi-feature association triggering rules, if two or more feature data exceed the normal state judgment conditions and reach the fault state judgment conditions at the same time, the first-level abnormality rule is triggered; if only the abnormal state judgment conditions are reached, the second-level abnormality rule is triggered; if a single feature data exceeds the normal state judgment conditions but does not reach the warning state judgment conditions, it is marked as normal state; if it exceeds the warning state judgment conditions but does not reach the abnormal state judgment conditions, the third-level abnormality rule is triggered.

[0124] In one embodiment, when constructing an abnormal rule library and establishing a mapping relationship between historical key feature data and abnormal rules, first, according to the severity of the abnormality, a first-level abnormal rule, a second-level abnormal rule, and a third-level abnormal rule are constructed. A first-level abnormality triggers automatic power cut-off of the area and emergency evacuation, a second-level abnormality triggers suspension of transportation and notification of the maintenance team, and a third-level abnormality triggers a voice warning and violation information record. Based on the statistical analysis of the historical key feature data of the past 6 months and with a 95% confidence interval as the benchmark, an abnormal feature decision table is constructed: the normal range of equipment displacement deviation is set to ±10cm. If it exceeds this range, it enters the warning state. If it reaches ±15cm, it is judged as abnormal and triggers the second-level abnormality rule. If it exceeds ±20cm, it enters the fault state and triggers the first-level abnormality rule; the normal range of gray value fluctuation is ± 15 gray units, exceeding it enters the warning stage, exceeding ±25 gray units triggers the second level abnormality, and exceeding ±35 gray units triggers the first level abnormality; the duration of the moving target normally does not exceed 60 seconds, exceeding it enters the warning stage, exceeding 120 seconds and not reaching the fault standard triggers the third level abnormality, and exceeding 180 seconds and meeting other fault conditions triggers the first level abnormality. At the same time, single and multi-feature association trigger rules are established, and the abnormality level is determined based on logical combination. The equipment status is divided into normal status, warning status, abnormal status and fault status, and the corresponding relationship of state transition is clarified. For example, from warning to abnormality, sound and light alarms and inspection notifications are triggered, and from abnormality to fault, emergency shutdown is initiated. All judgment values ​​are set in combination with historical data, coal mine safety standards and actual equipment operating conditions to ensure the scientific and operability of abnormal handling.

[0125] Real-time video data includes equipment operation data, personnel activity data, and environmental status data;

[0126] Real-time key feature data in video data is extracted through image recognition technology, including:

[0127] Using the preset device shape template, perform template matching on the image frame;

[0128] When the matching degree exceeds 85%, the device is identified and its location information is recorded;

[0129] Grayscale processing is performed on the image frame, and the areas with grayscale values ​​below the 91st threshold are marked as shadows, and the areas with grayscale values ​​above the 92nd threshold are marked as highlights, so as to identify abnormal lighting in the lane;

[0130] Detect moving targets by calculating the changes in pixel positions between two adjacent image frames;

[0131] When the area of ​​the moving target exceeds the ninety-third threshold pixel, distinguish between people and equipment by their contour shape;

[0132] De-duplicate the extracted real-time key feature data to remove repeated occurrences of the same real-time key feature data;

[0133] The real-time key feature data are numbered in chronological order, a corresponding relationship between the real-time key feature data and the collection time and collection location is established, and the data are stored in the database.

[0134] In one embodiment, when extracting key features from real-time video data, the implementation steps and specific numerical settings are as follows: First, for the equipment operation data, the preset equipment shape templates of the coal mining machine and the conveyor are used to perform image frame matching, and the matching threshold is set to 85%. Based on the statistics of historical data in the past three months, the correct recognition rate of more than 90% is ensured. If the threshold is exceeded, the equipment is judged to exist and the location information is recorded. For the environmental status data, after the image frame is gray-scaled, the gray-scale threshold is set to 30 (shadow) and 200 (highlight) according to the historical data of normal lighting in the tunnel (taking the night shift data for the past six months). Gray-scale values ​​below 30 are marked as Shadows are marked as highlighted if the value is above 200, and are used to detect lighting anomalies. In personnel activity data extraction, moving targets are detected by calculating pixel changes between adjacent frames. When the moving target area exceeds 800 pixels², and referring to historical data on the size of normal activity areas for personnel and equipment, and covering more than 95% of the normal activity target area, personnel and equipment are distinguished based on their silhouette aspect ratio (the silhouette aspect ratio of personnel is approximately 1.5-2.5, while the silhouette aspect ratio of equipment varies significantly). Finally, the extracted real-time key feature data is deduplicated, numbered in chronological order, and stored in a database in correspondence with the collection time and location to ensure data accuracy and traceability.

[0135] Compare real-time key feature data with the feature baseline range. Based on the established mapping relationship between historical key feature data and anomaly rules, match the corresponding rules in the anomaly rule library and output anomaly analysis results, including:

[0136] When comparing real-time key feature data with the feature baseline range of normal operating status, differentiated judgment conditions corresponding to the operating period are used first;

[0137] The equipment displacement deviation, grayscale value fluctuation and moving target duration are compared with the baseline differentiation judgment conditions of equipment type and working period.

[0138] In one embodiment, when comparing real-time key feature data with the baseline range of normal operating status features, the "time period priority" principle is strictly followed and combined with quantitative standards: during the morning shift, priority is given to areas with dense human activity, such as the number of moving targets at the entrance of the working face. The threshold is set to be greater than or equal to 15 people, based on the 90% confidence interval of the morning shift data for the past three months. If exceeded, an early warning is triggered. At the same time, the equipment displacement deviation, taking the coal mining machine drum height as an example, the normal range is ±10cm. Based on the 95% confidence interval of the historical data for the past six months, if exceeded, the warning state is entered; during the mid-shift, priority is given to analyzing the equipment operating noise. Taking the coal mining machine as an example, the normal noise baseline is 80-100 decibels. According to historical operating data statistics, the measured value exceeding the baseline by ±10%, that is, greater than 110 decibels or less than 70 decibels, is judged to be abnormal. The roller speed fluctuation is controlled within ±10% of the rated speed. Based on daily load monitoring data, if exceeded, the secondary abnormality rule is triggered; during the night shift, priority is given to checking the tunnel lighting brightness, and the average brightness is set to be greater than or equal to 20 lux. According to coal mine safety regulations and nighttime operation standards, if the measured brightness is less than 20 lux and the grayscale value fluctuates by more than ±35 grayscale units, and the historical shadow area data mean is -1.5 times the standard deviation, it is determined to be a lighting anomaly. The core data of equipment displacement deviation, grayscale value fluctuation, and moving target duration are compared with the equipment type and time period baseline. For example, the normal range of belt conveyor belt deviation is ±8 cm (refer to equipment safety standards). If the measured deviation during the mid-shift period is greater than 15 cm and accompanied by abnormal roller speed, and the fluctuation is greater than ±15%, the first-level anomaly rule is triggered. The moving target duration is normally less than or equal to 60 seconds. According to historical video data statistics, if a single target lasts for more than 180 seconds during the night shift and the grayscale value is abnormal, it is determined to be an environmental risk by combining the logical "AND" rule. After comparison, according to the abnormal feature decision table, for example, a displacement deviation of ±20 cm triggers an emergency shutdown and state transition rule. The output includes the analysis results of the anomaly type, location, level, such as the first-level anomaly and timestamp, providing an accurate basis for disposal.

[0139] Rule matching logic, including:

[0140] If a single feature data exceeds the normal state judgment condition but does not meet the warning state judgment condition, it is marked as normal state; if it exceeds the warning state judgment condition but does not meet the abnormal state judgment condition, the third-level abnormality rule is triggered and it is marked as warning state;

[0141] If a single feature data reaches or exceeds the abnormal state judgment condition, the secondary abnormal rule is triggered and marked as abnormal state;

[0142] If two or more characteristic data exceed the normal state judgment conditions and meet the corresponding fault state judgment conditions at the same time, the first-level abnormality rule is triggered and marked as a fault state;

[0143] The anomaly analysis results include the anomaly type, anomaly location, and anomaly level, and the anomaly occurrence timestamp in the anomaly analysis results is marked;

[0144] The characteristic data includes device displacement deviation, grayscale value fluctuation, and moving target duration. The judgment conditions of the characteristic data are determined based on the abnormal characteristic decision table and state transition rules.

[0145] The moving target duration is the cumulative duration that the moving target appears in consecutive video frames.

[0146] In one embodiment, the rule matching logic is implemented according to the following criteria: If a single feature data (device displacement deviation, grayscale value fluctuation, motion target duration) exceeds the normal state judgment condition but does not reach the warning threshold, it is marked as normal; if it exceeds the warning state judgment condition but does not reach the abnormal threshold, the third-level abnormality rule is triggered and the warning state is marked; if it meets or exceeds the abnormal state judgment condition, the second-level abnormality rule is triggered and the abnormal state is marked; if two or more feature data simultaneously exceed the normal state judgment condition and meet the fault state judgment condition (based on the abnormal feature decision table and state transition rules), the first-level abnormality rule is triggered and the fault state is marked. The abnormality analysis results must include the abnormality type (device, environment, or personnel related), the abnormality location (accurate to the acquisition location), the abnormality level (level 1, level 2, level 3), and the timestamp of the abnormality occurrence (accurate to seconds). The motion target duration is calculated based on the cumulative duration of its appearance in consecutive video frames. The judgment conditions for all feature data are determined based on the abnormal feature decision table and state transition rules to ensure a precise correspondence between abnormality classification and handling process.

[0147] Based on the anomaly analysis results, if an anomaly is determined, the corresponding alarm is triggered according to the anomaly rules and the anomaly handling process is associated, including:

[0148] When the abnormality analysis result is level three abnormality, the voice warning device in the local area of ​​the coal mine is triggered, and the abnormal location is marked with a yellow mark on the ground monitoring center interface;

[0149] When the abnormality analysis result is a level 2 abnormality, the underground sound and light alarm system is activated, the light flashing frequency is set to c times / second, the alarm sound intensity reaches d decibels, and the abnormality information is pushed to the inspection personnel's handheld terminal;

[0150] When the abnormality analysis result is a level one abnormality, the system automatically cuts off the power supply to the abnormal area, activates the underground emergency evacuation sound and light indication system, sets the evacuation signal frequency to e times / second, and sends an emergency alarm to the mine safety command center, and automatically associates the abnormality handling process;

[0151] Automatic association of exception handling processes, including:

[0152] If the abnormality type is equipment failure, the corresponding equipment maintenance manual is automatically retrieved and pushed to the maintenance team through the underground explosion-proof terminal of the coal mine;

[0153] If the exception type is a personnel violation, an electronic work order containing the time, location, and behavior photo of the violation will be automatically generated and sent to the safety management department;

[0154] If the abnormality type is environmental risk, immediately close the front and rear dampers of the abnormal area and switch the ventilation system to emergency mode.

[0155] In one of the embodiments, when executing the abnormal alarm triggering and handling process, it is implemented according to the following standards: when the abnormal analysis result is determined to be a level 3 abnormality, the voice warning device of the local area of ​​the coal mine underground is immediately triggered, and the abnormal location is marked with a yellow mark on the ground monitoring center interface; if it is a level 2 abnormality, the underground sound and light alarm system is started, and the light flashing frequency is set to 3 times / second (c is set to 3, refer to the level 2 alarm visual prompt standard in the "General Technical Requirements for Coal Mine Safety Monitoring Systems"), the alarm sound intensity reaches 85 decibels (d is set to 85, based on the human body's response sensitivity to the alarm sound and the underground noise environment setting), and the abnormal information is pushed to the patrol personnel's handheld terminal at the same time. When the level 1 abnormality occurs, the system automatically cuts off the power supply of the abnormal area and starts the underground emergency evacuation sound. The light indication system sets the evacuation signal frequency to 5 times per second (e is set to 5, which meets the high recognition requirements of emergency evacuation signals), and sends an emergency alarm to the mine-wide safety command center. The abnormal handling process is automatically associated according to the abnormality type: if it is an equipment failure, the system automatically retrieves the corresponding equipment maintenance manual from the knowledge base and pushes it to the maintenance team through the underground explosion-proof terminal of the coal mine; if it is determined to be a personnel violation, an electronic work order including the time, place and behavior photos of the violation is automatically generated and sent to the safety management department; if it is an environmental risk, the front and rear air doors of the abnormal area are immediately closed, and the ventilation system is switched to emergency mode to ensure the effective isolation of harmful gases in the abnormal area. All handling processes strictly comply with the coal mine safety regulations and emergency plan requirements to ensure the timeliness and standardization of abnormal responses.

[0156] After the abnormal handling process is initiated, the location of the handling personnel is tracked in real time through the RFID positioning system; if the failure to respond exceeds the set time, a reminder is automatically sent to the person's immediate superior;

[0157] The RFID positioning system uses UHF frequency band technology to accurately locate personnel positions through positioning base stations deployed at intervals in underground tunnels;

[0158] After the disposal is completed, confirm the operation by scanning the device QR code. The system records the disposal completion time and archives the exception handling record.

[0159] In one of the embodiments, after the abnormal handling process is started, the system tracks the location of the handling personnel in real time through the RFID positioning system. The specific implementation is as follows: UHF frequency band technology is used to build an RFID positioning system, and positioning base stations are deployed at intervals in underground tunnels to achieve accurate positioning of personnel. When the handling task is issued, the system automatically obtains the RFID tag information of the handling personnel and monitors their movement trajectory in real time. If the handling personnel does not respond for more than 15 minutes (based on the coal mine emergency response standards and underground traffic efficiency settings), the system will automatically send a reminder message to their immediate superior. After the handling is completed, the handling personnel needs to confirm the operation by scanning the QR code of the device. The system will automatically record the completion time of the handling and archive the information of the abnormality type, handling process and time node to the database to form a complete abnormal handling record for subsequent review and tracing.

[0160] The present invention establishes accurate feature baselines and differentiation thresholds through multi-period historical video data collection and image recognition technology, and combines a multi-level anomaly rule library and trigger mechanism to effectively solve the problems of inaccurate feature extraction and high false alarm and missed reporting rates in traditional monitoring, accurately monitor equipment, environment and personnel anomalies, and significantly reduce accident risks; at the same time, a three-level response system is constructed to trigger corresponding alarms for different levels of anomalies and associate them with handling processes, so as to achieve rational allocation of resources, shorten the average fault handling time, and greatly improve fault handling efficiency; in addition, full process automation reduces labor costs, and automatically recorded exception handling data provides support for management optimization. Combined with RFID positioning to track handling personnel and scan code archiving of processing records, it is convenient to review and optimize exception handling processes and rule thresholds, and ultimately realize intelligent, efficient and safe production in coal mines, which can reduce comprehensive management costs in the long term and comprehensively improve the overall management level.

[0161] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium may be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0162] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for underground video monitoring in a coal mine, characterized in that: The following steps are involved: Step S1, acquiring historical video data, and extracting historical key feature data from historical video data of early morning shift operation scenes, historical video data of mid-shift operation scenes, and historical video data of night shift operation scenes in the historical video data; Step S2: Analyze the historical key feature data and establish a feature baseline range for normal operation in coal mines. For different types of anomalies, establish a rule system according to the anomaly level to build an anomaly rule library, and establish a mapping relationship between the historical key feature data and the anomaly rules, including: Build an exception rule library and establish a rule system according to the exception level, including first-level exception rules, second-level exception rules and third-level exception rules; The first-level abnormal rule automatically cuts off the power supply to the area and initiates the emergency evacuation process when triggered; The secondary abnormality rule, when triggered, suspends the delivery and notifies the maintenance team process; The three-level abnormal rules will trigger a voice warning and record the information of the violator; Establish a mapping relationship between historical key feature data and anomaly rules, including single feature trigger rules and multi-feature association trigger rules. The multi-feature association trigger rules determine the anomaly level based on a combination of logical AND and logical OR. Based on the statistical analysis of historical key feature data, an abnormal feature decision table is constructed. The abnormal feature decision table includes normal state judgment conditions for equipment displacement deviation, gray value fluctuation, and moving target duration, as well as corresponding first-level abnormality, second-level abnormality, and third-level abnormality judgment conditions; Step S3, acquiring real-time video data and audio data from underground coal mines, and extracting real-time key feature data from the real-time video data; Step S4: comparing the real-time key feature data with the feature baseline range, combining the real-time audio data analysis results, and matching the corresponding rules in the anomaly rule library based on the established mapping relationship between historical key feature data and anomaly rules, and outputting the anomaly analysis results; Step S5: Based on the abnormality analysis result, if it is determined that an abnormality exists, a corresponding alarm is triggered according to the abnormality rule and an abnormality handling process is associated.

2. The method for underground video monitoring in a coal mine according to claim 1, wherein: The historical key feature data is extracted through image recognition technology, specifically including: Decomposing the historical video data into a set of image frames according to a time sequence; A fixed template matching method is used to identify the features of equipment, including a coal mining machine, a belt conveyor, a ventilator, and a gas sensor; Preset reflective marking points in the lane and determine the equipment displacement deviation by calculating the pixel coordinate offset of the reflective marking points; Based on grayscale threshold segmentation to extract environmental features, the area of ​​the image frame with a grayscale value higher than a first threshold is identified as bright light, and the area below a second threshold is identified as shadow; The inter-frame difference method is used to capture dynamic features. Two adjacent image frames are selected from the image frame to perform pixel difference calculation. Areas where the grayscale values ​​of pixels between adjacent frames change and the number exceeds a set threshold are marked as moving targets. Contour analysis is used to distinguish between people and equipment. The image frame includes a frame image in a video image sequence in historical video data.

3. The method for underground video monitoring in a coal mine according to claim 2, wherein: Analyze the historical key characteristic data and establish a characteristic baseline range for normal underground coal mine operation conditions, specifically including: Establishing a baseline library by equipment type, including shearers, belt conveyors, ventilators, and gas sensors; The coal mining machine includes a drum and a cutting motor, and a first baseline range is determined based on the drum height offset and the cutting motor temperature; The belt conveyor includes a conveyor belt and rollers, and a second baseline range is determined based on the deviation of the conveyor belt and the fluctuation of the roller speed; The ventilator includes blades, and a third baseline range is determined based on the wind pressure fluctuation magnitude and the vibration frequency of the blades; The gas sensor determines a fourth baseline range based on a response time and a zero drift of the gas sensor; The characteristic baseline range serves as the basic data for differentiated judgment conditions of equipment type and operation period.

4. The method for underground video monitoring in a coal mine according to claim 3, wherein: Differentiated judgment conditions are set based on equipment type and normal operating hours, including: During the morning shift, set the criteria for determining the number of moving targets in areas with dense human activity; During the mid-shift period, set the equipment operation noise judgment conditions; During the night shift, set the conditions for determining the brightness of the laneway lighting.

5. The method for underground video monitoring in a coal mine according to claim 4, wherein: For different anomaly types, a rule system is established according to the anomaly level to build an anomaly rule library, and a mapping relationship between historical key feature data and anomaly rules is established, including: Divide the equipment status into normal status, warning status, abnormal status and fault status; The device status corresponding to the device normal status determination condition and the device abnormality determination condition specifically includes: When the displacement deviation exceeds the normal state judgment condition, it enters the warning state; when the displacement deviation exceeds the warning state judgment condition, it enters the abnormal state; when the displacement deviation exceeds the abnormal state judgment condition, it enters the fault state; When the grayscale value fluctuation exceeds the normal state judgment condition, it enters the warning state; when the grayscale value fluctuation exceeds the warning state judgment condition, it enters the abnormal state; when the grayscale value fluctuation exceeds the abnormal state judgment condition, it enters the fault state; When the duration of the moving target exceeds the normal state judgment conditions, it enters the warning state; Based on the equipment normal state determination conditions, equipment abnormality determination conditions and abnormality rules, a state transition table is established to record the correspondence between equipment state transitions and abnormality rules, specifically including: When the equipment status changes from the warning state to the abnormal state, an audible and visual alarm is triggered and the inspection personnel are notified; When the equipment status changes from an abnormal state to a fault state, the power supply to the equipment is automatically cut off and the emergency shutdown process is initiated; For multi-feature association triggering rules, if two or more feature data exceed the normal state judgment conditions and reach the fault state judgment conditions at the same time, the first-level abnormality rule is triggered; if only the abnormal state judgment conditions are reached, the second-level abnormality rule is triggered; if a single feature data exceeds the normal state judgment conditions but does not reach the warning state judgment conditions, it is marked as normal state; if it exceeds the warning state judgment conditions but does not reach the abnormal state judgment conditions, the third-level abnormality rule is triggered.

6. The method for underground video monitoring in a coal mine according to claim 5, wherein: The real-time video data includes equipment operation data, personnel activity data and environmental status data; The real-time key feature data in the video data is extracted using image recognition technology, specifically including: Using the preset device shape template, perform template matching on the image frame; When the matching degree exceeds the threshold set based on the mean or standard deviation of historical key feature data, the device is identified and its location information is recorded; Grayscale processing is performed on the image frame, and areas with grayscale values ​​below a threshold set according to historical data of normal lighting in the lane are marked as shadows, and areas above the threshold are marked as highlights, so as to identify abnormal lighting in the lane; Detect moving targets by calculating the changes in pixel positions between two adjacent image frames; When the area of ​​the moving target exceeds a set threshold pixel, distinguishing between people and equipment by their contour shapes; De-duplicate the extracted real-time key feature data to remove repeated occurrences of the same real-time key feature data; The real-time key feature data are numbered in chronological order, a corresponding relationship between the real-time key feature data and the acquisition time and acquisition location is established, and the data are stored in a database.

7. The method for underground video monitoring in a coal mine according to claim 6, wherein: Compare the real-time key feature data with the feature baseline range, match the corresponding rules in the anomaly rule library based on the established mapping relationship between historical key feature data and anomaly rules, and output anomaly analysis results, specifically including: When comparing the real-time key feature data with the feature baseline range of the normal operating state, the differentiated judgment conditions corresponding to the operating period are preferentially adopted; The device displacement deviation, grayscale value fluctuation and moving target duration are compared with the baseline differentiation judgment conditions of the device type and operation period.

8. The method for underground video monitoring in a coal mine according to claim 7, wherein: The rule matching logic includes: If a single feature data exceeds the normal state judgment condition but does not meet the warning state judgment condition, it is marked as normal state; if it exceeds the warning state judgment condition but does not meet the abnormal state judgment condition, the third-level abnormality rule is triggered and it is marked as warning state; If a single feature data reaches or exceeds the abnormal state judgment condition, the secondary abnormality rule is triggered and marked as abnormal state; If two or more characteristic data exceed the normal state judgment conditions and meet the corresponding fault state judgment conditions at the same time, the first-level abnormality rule is triggered and marked as a fault state; The anomaly analysis result includes the anomaly type, anomaly location and anomaly level, and the anomaly occurrence timestamp in the anomaly analysis result is marked; The characteristic data includes device displacement deviation, gray value fluctuation, and moving target duration. The judgment conditions of the characteristic data are determined according to the abnormal characteristic decision table and state transition rules; The moving target duration is the cumulative duration that the moving target appears in consecutive video frames.

9. The method for underground video monitoring in a coal mine according to claim 8, wherein: Based on the anomaly analysis results, if an anomaly is determined to exist, a corresponding alarm is triggered according to the anomaly rules, and an anomaly handling process is associated, specifically including: When the abnormality analysis result is a level three abnormality, a voice warning device in a local area of ​​the coal mine is triggered, and the abnormal location is marked with a yellow mark on the ground monitoring center interface; When the abnormality analysis result is a level 2 abnormality, the underground sound and light alarm system is activated, the light flashing frequency is set to c times / second, the alarm sound intensity reaches d decibels, and the abnormality information is pushed to the inspection personnel's handheld terminal; When the abnormality analysis result is a level one abnormality, the system automatically cuts off the power supply to the abnormal area, activates the underground emergency evacuation sound and light indication system, sets the evacuation signal frequency to e times / second, sends an emergency alarm to the mine safety command center, and automatically associates the abnormality handling process; The exception handling process is automatically associated, including: If the abnormality type is equipment failure, the corresponding equipment maintenance manual is automatically retrieved and pushed to the maintenance team through the underground explosion-proof terminal of the coal mine; If the exception type is a personnel violation, an electronic work order containing the time, location, and behavior photo of the violation will be automatically generated and sent to the safety management department; If the abnormality type is environmental risk, immediately close the front and rear dampers of the abnormal area and switch the ventilation system to emergency mode.

10. The method for underground video monitoring in a coal mine according to claim 9, wherein: After the abnormal handling process is initiated, the location of the handling personnel is tracked in real time through the RFID positioning system; if the failure to respond exceeds the set time, a reminder is automatically sent to the person's immediate superior; The RFID positioning system uses UHF frequency band technology to accurately locate personnel positions through positioning base stations deployed at intervals in underground tunnels; After the disposal is completed, confirm the operation by scanning the device QR code. The system records the disposal completion time and archives the exception handling record.

Images