Bulk coal in-transit packaging transportation safety supervision method based on video AI identification

Abstract

The invention discloses a bulk coal in-transit packaging transportation safety supervision method based on video AI identification, and relates to the technical field of data processing. According to the application, a carriage video, an electric tarpaulin state and vehicle positioning and motion data can be collected at the same time, and multi-modal fusion analysis is carried out when a vehicle is static and the tarpaulin is opened; therefore, the technical problems that an existing physical lead seal is prone to being damaged, static abnormity cannot be recognized through track monitoring, and video monitoring depends on manual work, so that report missing and false report are caused can be solved. The technical effects of improving the abnormity identification accuracy, reducing the false report and missing report rate, shortening the response time, realizing automatic evidence solidification and improving the intelligent level of coal transportation supervision are achieved.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a method for safety supervision of the in-transit packaged transportation of bulk coal based on video AI recognition. Background Technology

[0002] Coal transportation, as a crucial link in the energy supply chain, is widely used in thermal power generation, industrial production, and other fields. With the development of IoT and AI technologies, traditional anti-cheating systems have gradually evolved from physical seals to electronic monitoring. Currently, anti-cheating systems employ a combination of physical seals and trajectory monitoring. This seal management method suffers from drawbacks such as susceptibility to damage and lack of traceability. While trajectory monitoring can acquire location information, it cannot identify abnormal activity when the vehicle is stationary. Existing video surveillance systems often use a fixed view from the driver's cab, or, although they possess cabin monitoring capabilities, rely on manual interpretation, resulting in a high false negative rate. Therefore, existing technologies suffer from systemic problems such as high false alarm rates, delayed response times, and difficulties in evidence consolidation when dealing with multi-factor collaborative fraud in complex transportation scenarios, severely hindering the intelligent supervision process in the coal transportation industry. Summary of the Invention

[0003] This disclosure provides a method for safety supervision of bulk coal transported in transit using video AI recognition. Its main purpose is to address systemic problems such as high false alarm rates, delayed response times, and difficulties in evidence consolidation, which severely hinder the progress of intelligent supervision in the coal transportation industry.

[0004] According to a first aspect of this disclosure, a method for safety supervision of bulk coal during transit and packaged transportation based on video AI recognition is provided, comprising: Real-time acquisition of video monitoring data of coal transport vehicles' cargo compartments, opening and closing status signals of electric tarpaulins, and vehicle positioning and movement status data; Determine whether the vehicle is in a stopped state based on the vehicle's motion status data and positioning information; Once it is determined that the vehicle is in a stopped state, the open / closed status signal of the electric tarpaulin is detected; If the tarpaulin is detected to be open, the monitoring video stream of the carriage is analyzed in real time to identify whether there is a human-shaped target or a significantly moving object, and multimodal information fusion judgment is performed by combining the vehicle's stationary state and the tarpaulin's open state.

[0005] Optionally, the real-time acquisition of video monitoring data of the coal transport vehicle's cargo compartment, the opening and closing status signals of the electric tarpaulin, and the vehicle's positioning and movement status data includes: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

[0006] Optionally, detecting the opening / closing status signal of the electric tarpaulin after determining that the vehicle is in a stopped state includes: Illegal opening behavior is identified by analyzing the current waveform characteristics of the tarpaulin motor. When a sudden current change is detected that exceeds a preset threshold and the duration is less than a preset time, it is determined to be an abnormal opening. Record the timestamp of the tarpaulin opening event and verify it by associating it with the timestamp of the vehicle's stopped state.

[0007] Optionally, if the tarpaulin is detected to be open, the monitoring video stream of the carriage is analyzed in real time to identify whether there is a human-shaped target or a significantly moving object, and a multimodal information fusion judgment is performed by combining the vehicle's stationary state and the tarpaulin's open state, including: A weighted logic threshold is adopted, with the vehicle's stationary state as the first weight, the tarpaulin opening state as the second weight, and the detection of carriage movement as the third weight. When the overall score exceeds the preset score threshold, a first-level alarm is triggered. Centered on the alarm trigger time, a video segment of preset duration is extracted, and the vehicle's acceleration curve, positioning trajectory points, and tarpaulin status change records within that time period are stored simultaneously.

[0008] Optionally, the method further includes: Human posture recognition is performed on the video footage from the carriage, and a secondary alarm signal is generated when a human posture is detected. The secondary alarm signal is logically associated with the primary alarm signal. If the secondary alarm signal exists after the primary alarm signal is triggered, a remote voice warning is activated and the warning timestamp is recorded.

[0009] According to a second aspect of this disclosure, a safety monitoring device for the in-transit packaged transportation of bulk coal based on video AI recognition is provided, comprising: The data acquisition unit is used to collect real-time monitoring video data of the coal transport vehicle's cargo compartment, the opening and closing status signals of the electric tarpaulin, and the vehicle's positioning and movement status data. The determining unit is used to determine whether the vehicle is in a stopped state based on the vehicle's motion state data and positioning information; The detection unit is used to detect the opening and closing status signal of the electric tarpaulin when it is determined that the vehicle is in a stopped state; The analysis unit is used to perform real-time analysis of the monitoring video stream of the carriage if the tarpaulin is detected to be open, to identify whether there is a human-shaped target or a significantly moving object, and to perform multimodal information fusion judgment by combining the vehicle's stationary state and the tarpaulin's open state.

[0010] Optionally, the acquisition unit is further configured to: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

[0011] Optionally, the detection unit is further configured to: Illegal opening behavior is identified by analyzing the current waveform characteristics of the tarpaulin motor. When a sudden current change is detected that exceeds a preset threshold and the duration is less than a preset time, it is determined to be an abnormal opening. Record the timestamp of the tarpaulin opening event and verify it by associating it with the timestamp of the vehicle's stopped state.

[0012] Optionally, the analysis unit is further configured to: A weighted logic threshold is adopted, with the vehicle's stationary state as the first weight, the tarpaulin opening state as the second weight, and the detection of carriage movement as the third weight. When the overall score exceeds the preset score threshold, a first-level alarm is triggered. Centered on the alarm trigger time, a video segment of preset duration is extracted, and the vehicle's acceleration curve, positioning trajectory points, and tarpaulin status change records within that time period are stored simultaneously.

[0013] Optionally, the device further includes: The recognition unit is used to perform human posture recognition on the video of the carriage, and generates a secondary alarm signal when a human posture is detected. The alarm unit is used to logically associate the secondary alarm signal with the primary alarm signal. If the secondary alarm signal exists after the primary alarm signal is triggered, a remote voice warning is activated and the warning timestamp is recorded.

[0014] According to a third aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0015] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0016] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0017] This disclosure provides a method for safety supervision of bulk coal transported in transit based on video AI recognition. By simultaneously collecting video data from the vehicle compartment, the status of the electric tarpaulin, and vehicle positioning and movement data, and performing multimodal fusion analysis when the vehicle is stationary and the tarpaulin is open, it can identify abnormal activities in real time. Therefore, it solves the technical problems of existing methods such as the ease of damage to physical seals, the inability of trajectory monitoring to identify stationary anomalies, and the reliance on manual video monitoring leading to false alarms and missed alarms. This method achieves the technical effects of improving the accuracy of anomaly identification, reducing false alarm and missed alarm rates, shortening response time, automatically solidifying evidence, and enhancing the level of intelligent supervision of coal transportation.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0019] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 A flowchart illustrating a method for safety supervision of bulk coal in transit based on video AI recognition, provided in an embodiment of this disclosure; Figure 2 A schematic diagram of the structure of a bulk coal packaging and transportation safety monitoring device based on video AI recognition provided in this embodiment of the present disclosure; Figure 3 A schematic diagram of the structure of a bulk coal packaging and transportation safety monitoring device based on video AI recognition provided in this embodiment of the present disclosure; Figure 4 A schematic block diagram of an example electronic device 600 provided for embodiments of this disclosure. Detailed Implementation

[0020] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0021] The following describes a method for safety supervision of bulk coal packaging and transportation in transit based on video AI recognition, according to embodiments of the present disclosure, with reference to the accompanying drawings.

[0022] Figure 1 This is a flowchart illustrating a method for safety supervision of bulk coal during transit based on video AI recognition, as provided in an embodiment of this disclosure.

[0023] like Figure 1 As shown, the method includes the following steps: Step 101: Real-time acquisition of monitoring video data of the coal transport vehicle's cargo compartment, opening and closing status signals of the electric tarpaulin, and vehicle positioning and movement status data. Three key types of data related to fraud monitoring during coal transportation are acquired synchronously and continuously. The monitoring video data of the coal truck compartment is collected through high-definition, image-stabilized cameras installed on the trucks. These cameras must be positioned facing inwards to clearly capture real-time dynamic images inside the compartment, ensuring a complete record of possible personnel activities, object movements, and other scenarios, providing a high-quality visual data source for subsequent AI video analysis modules.

[0024] The opening and closing status signal of the electric tarpaulin is collected by the tarpaulin status sensor. This sensor adopts a non-invasive design and is connected to the control circuit of the electric tarpaulin. It does not require major modifications to the original tarpaulin control structure. It can accurately identify the current opening, closing or stopping status of the electric tarpaulin by detecting the working signal of the tarpaulin motor or the current change in the circuit, thus avoiding the impact of the sensor installation method on the normal use of the tarpaulin.

[0025] Vehicle positioning and motion status data are collected collaboratively by a BeiDou positioning module and an inertial measurement unit (IMU). The BeiDou positioning module can acquire the vehicle's precise geographical location information in real time, while the IMU can simultaneously collect motion parameters such as real-time vehicle speed and acceleration. The combined positioning and motion status data comprehensively reflects the vehicle's trajectory and motion status, providing data support for subsequent determination of whether the vehicle is stationary. The real-time acquisition of these three types of data must maintain temporal synchronization to ensure that each set of data corresponds to the vehicle's state at the same moment, laying an accurate and consistent data foundation for subsequent multi-dimensional information fusion analysis.

[0026] Step 102: Determine whether the vehicle is in a stopped state based on the vehicle's motion state data and positioning information; By collaboratively analyzing vehicle motion status data and positioning information, it is possible to accurately identify whether a vehicle is stationary, providing a prerequisite for subsequent investigation of abnormal events. The vehicle's motion status data comes from a previously deployed inertial measurement unit (IMU), which can collect key motion parameters such as vehicle speed and acceleration in real time. These parameters directly reflect the vehicle's dynamic characteristics, such as whether the vehicle speed is zero and whether the acceleration remains stable near zero, providing fundamental data support for determining whether the vehicle has left the driving state.

[0027] The vehicle's positioning information is obtained by the Beidou positioning module, which continuously outputs the vehicle's real-time geographical coordinates. Its core value lies in assisting in judging the vehicle's status by monitoring changes in coordinates. If the positioning coordinates do not show significant displacement within a period of time, combined with the characteristics of vehicle speed and acceleration in the motion state data, it can effectively eliminate misjudgments caused by non-stop states such as brief deceleration and idling.

[0028] In the specific judgment process, it is necessary to perform correlation analysis between motion status data and positioning information, rather than relying on a single data dimension. For example, when motion status data shows that the vehicle speed is zero and the acceleration remains stable, it is necessary to further combine the positioning information to confirm whether the vehicle coordinates are within a stable range, avoiding misjudging the vehicle as stationary due to brief stops during driving (such as temporary avoidance). If both motion status data and positioning information indicate that the vehicle has no displacement and no dynamic changes, then the vehicle can be determined to be stationary. If the vehicle is determined to be in motion, there is no need to trigger subsequent analysis processes; the real-time collection and monitoring of vehicle motion status data and positioning information should continue. If the vehicle is determined to be stationary, this judgment result should be used as the basis for triggering subsequent processes such as tarpaulin status detection, ensuring the orderly progress and accurate response of the monitoring process.

[0029] Step 103: After determining that the vehicle is in a stopped state, detect the opening and closing status signal of the electric tarpaulin; By accurately detecting the opening and closing status signals of the electric tarpaulin, potential fraudulent scenarios are identified. The core component for detecting these signals is the tarpaulin status sensor. This sensor uses a non-invasive design to connect to the electric tarpaulin's control circuit, requiring no structural modifications to the vehicle's existing tarpaulin control system. This avoids affecting the normal opening and closing operation of the tarpaulin while stably capturing the operating status information of the tarpaulin motor. Specifically, the sensor identifies the tarpaulin status through two key parameters: first, it directly detects whether the tarpaulin motor receives an opening or closing control command; second, it monitors the current changes in the tarpaulin control circuit. When the tarpaulin motor starts, the circuit current fluctuates to a specific amplitude. The sensor can identify these fluctuations to help determine whether the tarpaulin is currently open, closed, or stopped.

[0030] After the test is completed, the acquired opening and closing status signals need to be preliminarily judged: if the signal shows that the electric tarpaulin is in the closed state, it means that the current carriage is in a closed state, and there is no risk of fraud caused by the opening of the tarpaulin. At this time, there is no need to trigger the subsequent process, and the real-time monitoring of the vehicle status and the tarpaulin signal should continue. If the signal shows that the electric tarpaulin is in the open state, it means that the sealing of the carriage has been broken, and there is a potential possibility of fraud. It is necessary to trigger the subsequent further verification process or alarm mechanism to provide key evidence for accurately identifying fraud.

[0031] Step 104: If the tarpaulin is detected to be open, the monitoring video stream of the carriage is analyzed in real time to identify whether there is a human-shaped target or a significantly moving object, and multimodal information fusion judgment is performed by combining the vehicle's stationary state and the tarpaulin's open state.

[0032] By combining AI analysis and multi-dimensional state fusion of the train carriage surveillance video stream, suspected fraudulent activities were further identified. The video stream originated from previously installed high-definition, image-stabilized cameras in the carriage. The continuous real-time video stream output from these cameras provided ongoing visual data support for the AI ​​analysis, ensuring complete capture of every dynamic change within the carriage. Real-time analysis of the video stream relied on an AI video analysis module. This module incorporates a pre-trained deep learning model that efficiently processes frames in the video stream, accurately identifying human figures or significantly moving objects. Human figure identification can detect individuals potentially entering the carriage to steal or switch coal, while the identification of significantly moving objects covers the movement of objects related to fraud, such as coal-making tools. The combination of these two methods forms a comprehensive monitoring system for abnormal dynamics within the carriage, preventing potential fraudulent activities from being missed due to the identification of a single target.

[0033] While completing the video stream analysis, a multimodal information fusion judgment needs to be initiated. This involves associating and integrating the human-shaped target or significantly moving object identification results obtained in this step with the vehicle's stationary state and the tarpaulin's open state. This fusion judgment is not a simple superposition of single-dimensional results, but rather uses logical association to exclude non-fraudulent scenarios. For example, it avoids false judgments caused by situations where the vehicle is still moving when the tarpaulin is open, or where only the tarpaulin is open but there is no abnormal activity inside the vehicle. It ensures that only when the vehicle remains stationary, the tarpaulin remains open, and there is a human-shaped target or significantly moving object inside the vehicle is it considered a suspicious scenario with fraud risk, providing a reliable basis for generating accurate alarm signals in the future.

[0034] In some embodiments, the real-time acquisition of monitoring video data of the coal transport vehicle's cargo compartment, the opening and closing status signal of the electric tarpaulin, and the vehicle's positioning and movement status data includes: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

[0035] Based on pre-set, identical video data collected from the cargo compartments of coal transport vehicles, the core is to pre-define a unified video acquisition standard and on-site installation specifications for each coal transport vehicle. This ensures that the high-definition, image-stabilized cameras used for video acquisition are fixedly installed facing the interior of the cargo compartment at pre-set positions, and that acquisition parameters such as resolution, frame rate, and exposure remain consistent. This avoids problems such as incomplete video information and insufficient clarity caused by deviations in camera installation angles or differences in acquisition parameters. As a result, the video data from each vehicle's cargo compartment has a unified analytical benchmark, facilitating the subsequent stable identification of dynamics within the cargo compartment by the AI ​​video analysis module.

[0036] The key to detecting current changes in the tarpaulin motor using a non-invasive current sensor and combining this with voltage signals for opening / closing status determination lies in leveraging the characteristics of the non-invasive current sensor to accurately capture current changes during motor operation without disrupting the original control circuit structure of the electric tarpaulin. When the motor receives an opening command, the current in the circuit will fluctuate upwards to a specific amplitude; when the motor receives a closing command and reverses, the current will exhibit different fluctuation characteristics compared to when it is on; when the motor stops working, the current returns to a stable zero value or standby current. Simultaneously, monitoring the voltage signal in the tarpaulin control circuit further verifies whether the motor is in a normal energized state, avoiding misreading of the current signal due to abnormal circuit voltage. For example, if the voltage signal indicates that the circuit is not energized, even a brief current fluctuation can be identified as interference, not an actual opening / closing action of the tarpaulin. This significantly improves the accuracy of determining the opening / closing status of the electric tarpaulin and reduces subsequent process mis-triggers caused by signal misinterpretation.

[0037] In some embodiments, detecting the opening / closing status signal of the electric tarpaulin after determining that the vehicle is in a stopped state includes: Illegal opening behavior is identified by analyzing the current waveform characteristics of the tarpaulin motor. When a sudden current change is detected that exceeds a preset threshold and the duration is less than a preset time, it is determined to be an abnormal opening. Record the timestamp of the tarpaulin opening event and verify it by associating it with the timestamp of the vehicle's stopped state.

[0038] Identifying illegal opening behavior by analyzing the current waveform characteristics of the tarpaulin motor relies on a non-intrusive current sensor previously connected to the tarpaulin control circuit. This sensor can collect the complete current waveform of the motor in real time. Under normal circumstances, the motor current will show a regular waveform of steady rise, stable operation, and slow decline when the tarpaulin is opened or closed. However, illegal opening (such as forcibly prying the tarpaulin, causing abnormal motor operation, or improperly connecting external circuits to trigger short-term motor action) will cause irregular sudden changes in the current waveform. At this time, the preset current threshold and duration parameters play a key role: when the detected current change amplitude exceeds the preset threshold and the duration of the change is less than the preset duration, it can be determined as an abnormal opening. This judgment logic can effectively exclude the slight current fluctuations during normal motor startup and accurately identify tarpaulin opening behavior caused by non-standard operation.

[0039] The verification of the timestamp of the tarpaulin opening event and its correlation with the timestamp of the vehicle's stationary state relies on the system's time synchronization mechanism. The timestamp of the tarpaulin opening event is generated in real time by the tarpaulin status sensor, accurately marking the moment the tarpaulin begins to open. The timestamp of the vehicle's stationary state is generated collaboratively by the BeiDou positioning module and the inertial measurement unit (IMU), recording the start time when the vehicle enters a stationary state. By comparing the two timestamps, it can be confirmed whether the tarpaulin opening behavior occurred within the time interval when the vehicle was stationary: if the tarpaulin opening timestamp falls after the vehicle stationary state timestamp, it indicates that the opening behavior occurred during the vehicle's stationary period, which matches the scenario characteristics of possible fraudulent behavior; if the tarpaulin opening timestamp and the vehicle stationary state timestamp do not overlap (e.g., the tarpaulin opens while the vehicle is moving), the association between this opening and fraud during the journey can be ruled out, further reducing invalid analysis and providing a more reliable basis for subsequent accurate triggering of the anomaly verification process.

[0040] In some embodiments, if the tarpaulin is detected to be open, the real-time analysis of the monitoring video stream of the carriage to identify whether there is a human-shaped target or a significantly moving object, and the multimodal information fusion judgment combining the vehicle's stationary state and the tarpaulin's open state, includes: A weighted logic threshold is adopted, with the vehicle's stationary state as the first weight, the tarpaulin opening state as the second weight, and the detection of carriage movement as the third weight. When the overall score exceeds the preset score threshold, a first-level alarm is triggered. Centered on the alarm trigger time, a video segment of preset duration is extracted, and the vehicle's acceleration curve, positioning trajectory points, and tarpaulin status change records within that time period are stored simultaneously.

[0041] When using a weighted logic threshold, corresponding weights must be assigned to the three key states: The first weight is for the vehicle being stopped, as this is the fundamental scenario where fraudulent behavior might occur. The weight setting must ensure the basic role of this state; only when the vehicle is actually stopped can subsequent anomalies be associated with fraud. The second weight is for the tarpaulin being open, which directly signifies that the vehicle's sealing has been broken, a necessary prerequisite for fraudulent behavior. The weight must match its key triggering significance for the fraud scenario. The third weight is for vehicle activity detection, which directly reflects whether there are any fraud-related dynamics such as the movement of people or objects inside the vehicle. It is the core basis for judging whether fraudulent behavior has actually occurred, and its weight is set relatively higher to highlight its decisive role. When calculating the comprehensive score, the weights corresponding to each state must be correlated with the state judgment results. When the final comprehensive score exceeds the preset score threshold, a level-one alarm can be accurately triggered. This weighted fusion method avoids invalid alarms caused by misjudgments of a single state, significantly improving alarm reliability.

[0042] After a Level 1 alarm is triggered, a video clip of a preset duration must be extracted, centered on the alarm trigger time. This clip must completely cover the dynamics of the vehicle compartment for a period of time before the alarm and the scene for a period of time after the alarm, ensuring that no key footage is missing. Simultaneously, the vehicle's acceleration curve, location trajectory points, and tarpaulin status change records for this period must be stored: the acceleration curve verifies whether the vehicle remained in a stable, stationary state during the alarm period, eliminating false alarms caused by brief shaking; the location trajectory points accurately mark the geographical location where the alarm occurred, providing locational evidence for subsequent tracing; the tarpaulin status change records reflect whether the tarpaulin remained open or underwent abnormal switching during the alarm period, further supplementing information about the fraud scenario. These data, together with the video clips, form a complete set of evidence, providing comprehensive support for post-incident verification and liability determination.

[0043] In some embodiments, the method further includes: Human posture recognition is performed on the video footage from the carriage, and a secondary alarm signal is generated when a human posture is detected. The secondary alarm signal is logically associated with the primary alarm signal. If the secondary alarm signal exists after the primary alarm signal is triggered, a remote voice warning is activated and the warning timestamp is recorded.

[0044] Human posture recognition in the train carriage video relies on an activated AI video analysis module. This module, equipped with a specially trained deep learning model, can not only identify human figures within the carriage but also more accurately capture specific human movements and postures. These include typical postures directly related to coal theft or substitution, such as bending over to move coal, climbing the carriage edge, or adjusting the arrangement of coal piles, rather than simply determining the presence of a human figure. When the model detects such human postures with clear signs of fraud, a secondary alarm signal is immediately generated. This signal, compared to the primary alarm, focuses more on abnormal details in human movements, providing more specific visual evidence for suspected fraudulent behavior and reducing potential misjudgments that might arise from relying solely on human figure recognition.

[0045] The secondary alarm signal is then logically correlated with the primary alarm signal. This correlation requires the primary alarm signal to have already been triggered. Only when the primary alarm signal is triggered first, and the secondary alarm signal appears within a preset time interval after the primary alarm signal is triggered, is it considered a genuine fraud risk scenario requiring intervention. If only the secondary alarm signal appears without the support of the primary alarm signal, it may indicate human activity within the carriage that is not a fraudulent scenario, and no further intervention is needed. If the secondary alarm signal is not detected after the primary alarm signal is triggered, continuous monitoring of the carriage dynamics is required. When the condition that the secondary alarm signal exists after the primary alarm signal is triggered is met, the system initiates a remote voice warning through the vehicle's 4G / 5G communication module, pushing the preset standard transportation warning voice to the vehicle's onboard terminal in real time, providing an immediate warning to the personnel inside the vehicle to prevent the continuation of fraudulent behavior. At the same time, the system automatically records the timestamp of the remote voice warning. This timestamp is synchronously correlated with previously stored carriage video clips, vehicle positioning data, tarpaulin status records, and other information to further supplement the evidence chain for fraud monitoring and provide complete time and event correspondence for subsequent accountability verification.

[0046] Corresponding to the aforementioned method for safety supervision of bulk coal in transit based on video AI recognition, this invention also proposes a device for safety supervision of bulk coal in transit based on video AI recognition. Since the device embodiment of this invention corresponds to the method embodiment described above, details not disclosed in the device embodiment can be referred to in the method embodiment, and will not be repeated here.

[0047] Figure 2 This is a schematic diagram of a safety monitoring device for the in-transit packaging and transportation of bulk coal based on video AI recognition, provided as an embodiment of this disclosure. Figure 2 As shown, it includes: The acquisition unit 21 is used to collect real-time monitoring video data of the coal transport vehicle's cargo compartment, the opening and closing status signals of the electric tarpaulin, and the vehicle's positioning and movement status data. Determining unit 22 is used to determine whether the vehicle is in a stopped state based on the vehicle's motion state data and positioning information; The detection unit 23 is used to detect the opening and closing status signal of the electric tarpaulin when it is determined that the vehicle is in a stopped state. The analysis unit 24 is used to perform real-time analysis on the monitoring video stream of the carriage if the tarpaulin is detected to be open, to identify whether there is a human-shaped target or a significantly moving object, and to perform multimodal information fusion judgment by combining the vehicle's stationary state and the tarpaulin's open state.

[0048] Furthermore, in one possible implementation of this disclosure, the acquisition unit 21 is further configured to: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

[0049] Furthermore, in one possible implementation of this disclosure, the detection unit 23 is further configured to: Illegal opening behavior is identified by analyzing the current waveform characteristics of the tarpaulin motor. When a sudden current change is detected that exceeds a preset threshold and the duration is less than a preset time, it is determined to be an abnormal opening. Record the timestamp of the tarpaulin opening event and verify it by associating it with the timestamp of the vehicle's stopped state.

[0050] Furthermore, in one possible implementation of this disclosure, the analysis unit 24 is further configured to: A weighted logic threshold is adopted, with the vehicle's stationary state as the first weight, the tarpaulin opening state as the second weight, and the detection of carriage movement as the third weight. When the overall score exceeds the preset score threshold, a first-level alarm is triggered. Centered on the alarm trigger time, a video segment of preset duration is extracted, and the vehicle's acceleration curve, positioning trajectory points, and tarpaulin status change records within that time period are stored simultaneously.

[0051] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 3 The apparatus shown also includes: The recognition unit 25 is used to perform human posture recognition on the video of the carriage and generate a secondary alarm signal when a human posture is detected. Alarm unit 26 is used to logically associate the secondary alarm signal with the primary alarm signal. If the secondary alarm signal exists after the primary alarm signal is triggered, a remote voice warning is activated and the warning timestamp is recorded.

[0052] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of the embodiments of this disclosure, and the principle is the same. Therefore, the embodiments of this disclosure are not limited thereto.

[0053] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0054] Figure 4 A schematic block diagram of an example electronic device 400 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0055] like Figure 4 As shown, device 400 includes a computing unit 401, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 402 or a computer program loaded from storage unit 408 into RAM (Random Access Memory) 403. RAM 403 may also store various programs and data required for the operation of device 400. The computing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. I / O (Input / Output) interface 405 is also connected to bus 404.

[0056] Multiple components in device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of monitors, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0057] The computing unit 401 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processors, controllers, microcontrollers, etc. The computing unit 401 performs the various methods and processes described above, such as a method for safety supervision of bulk coal in transit packaged transportation based on video AI recognition. For example, in some embodiments, the method for safety supervision of bulk coal in transit packaged transportation based on video AI recognition can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 408. In some embodiments, part or all of the computer program can be loaded and / or installed on device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by computing unit 401, one or more steps of the method described above can be performed. Alternatively, in other embodiments, computing unit 401 can be configured by any other suitable means (e.g., by means of firmware) to perform the aforementioned method for safety supervision of bulk coal in-transit packaged transportation based on video AI recognition.

[0058] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0059] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0060] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0061] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0062] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0063] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0064] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0065] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0066] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for safety supervision of bulk coal during transit packaging and transportation based on video AI recognition, characterized in that, include: Real-time acquisition of video monitoring data of coal transport vehicles' cargo compartments, opening and closing status signals of electric tarpaulins, and vehicle positioning and movement status data; Determine whether the vehicle is in a stopped state based on the vehicle's motion status data and positioning information; Once it is determined that the vehicle is in a stopped state, the open / closed status signal of the electric tarpaulin is detected; If the tarpaulin is detected to be open, the monitoring video stream of the carriage is analyzed in real time to identify whether there is a human-shaped target or a significantly moving object, and multimodal information fusion judgment is performed by combining the vehicle's stationary state and the tarpaulin's open state.

2. The method according to claim 1, characterized in that, The real-time acquisition of video monitoring data of the coal transport vehicle's cargo compartment, the opening and closing status signals of the electric tarpaulin, and the vehicle's positioning and movement status data includes: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

3. The method according to claim 1, characterized in that, The step of detecting the opening / closing status signal of the electric tarpaulin after determining that the vehicle is in a stopped state includes: Illegal opening behavior is identified by analyzing the current waveform characteristics of the tarpaulin motor. When a sudden current change is detected that exceeds a preset threshold and the duration is less than a preset time, it is determined to be an abnormal opening. Record the timestamp of the tarpaulin opening event and verify it by associating it with the timestamp of the vehicle's stopped state.

4. The method according to claim 1, characterized in that, If the tarpaulin is detected to be open, the monitoring video stream of the carriage is analyzed in real time to identify whether there is a human-shaped target or a significantly moving object, and a multimodal information fusion judgment is performed by combining the vehicle's stationary state and the tarpaulin's open state, including: A weighted logic threshold is adopted, with the vehicle's stationary state as the first weight, the tarpaulin opening state as the second weight, and the detection of carriage movement as the third weight. When the overall score exceeds the preset score threshold, a first-level alarm is triggered. Centered on the alarm trigger time, a video segment of preset duration is extracted, and the vehicle's acceleration curve, positioning trajectory points, and tarpaulin status change records within that time period are stored simultaneously.

5. The method according to claim 1, characterized in that, The method further includes: Human posture recognition is performed on the video footage from the carriage, and a secondary alarm signal is generated when a human posture is detected. The secondary alarm signal is logically associated with the primary alarm signal. If the secondary alarm signal exists after the primary alarm signal is triggered, a remote voice warning is activated and the warning timestamp is recorded.

6. A safety monitoring device for the in-transit packaging and transportation of bulk coal based on video AI recognition, characterized in that, include: The data acquisition unit is used to collect real-time monitoring video data of the coal transport vehicle's cargo compartment, the opening and closing status signals of the electric tarpaulin, and the vehicle's positioning and movement status data. The determining unit is used to determine whether the vehicle is in a stopped state based on the vehicle's motion state data and positioning information; The detection unit is used to detect the opening and closing status signal of the electric tarpaulin when it is determined that the vehicle is in a stopped state. The analysis unit is used to perform real-time analysis of the monitoring video stream of the carriage if the tarpaulin is detected to be open, to identify whether there is a human-shaped target or a significantly moving object, and to perform multimodal information fusion judgment by combining the vehicle's stationary state and the tarpaulin's open state.

7. The apparatus according to claim 6, characterized in that, The acquisition unit is also used for: Based on the preset on-site identical collection of monitoring video data of the coal transport vehicle's compartment; The current change characteristics of the tarpaulin motor are detected by a non-invasive current sensor, and the opening and closing status is determined by combining the voltage signal.

8. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-5.

Images