A method for monitoring the operation of a grain depot grain transport vehicle

Abstract

The application relates to the technical field of intelligent monitoring of grain transport vehicles, and discloses a grain depot grain transport vehicle operation supervision method, which binds the card number of an IC card with a grain transport vehicle track ID; obtains the grain transport vehicle track from the monitoring video data in the grain depot area in real time; extracts the grain transport vehicle stationary point and stationary time interval from the grain transport vehicle track; matches the stationary point with the grain depot operation point area, and if the time point of operation report at the stationary point is not in the stationary time interval, an alarm is sent; the grain transport vehicle track data is grouped and grain transport vehicle track anomaly detection and stationary point anomaly detection are carried out, and when an anomaly is detected, an alarm is sent. The application binds the IC card with the grain transport vehicle track, can guarantee the one-to-one correspondence relationship between the IC card and the grain transport vehicle, simultaneously supervises the operation time of the grain transport vehicle at each operation point and the grain transport vehicle track in real time, and ensures that the whole operation process of the grain transport vehicle is supervised.

Description

Technical Field

[0001] This invention belongs to the field of intelligent monitoring technology for grain transport vehicles, specifically relating to a method for supervising the operation of grain transport vehicles in grain depots. Background Technology

[0002] Currently, my country's policy-oriented grain management has achieved full information coverage, ensuring that information on acquisition, sales, and rotation is searchable and traceable. Similarly, effective tracking and management of grain transport vehicles entering and leaving grain depots are also necessary. Currently, grain depots mostly issue IC cards or attach / affix disposable tags for temporary management of grain transport vehicles. The problem with using IC cards or tags for temporary management is that the correspondence between the IC card / tag and the grain transport vehicle cannot be guaranteed. During grain transport operations, disposable tags are prone to falling off or being lost, as well as being dismantled, moved, or damaged by unauthorized personnel. Using IC cards, on the other hand, is prone to unauthorized swapping and cheating, resulting in loopholes in the supervision of grain depot entry and exit operations.

[0003] Chinese invention patent CN201610849841.9 proposes an electronic tag that can be attached to a measured object, such as a vehicle, turnover box, or freight container, via an internal permanent magnet. This tag stores electronic information about the vehicle, personnel, and goods, and actively detects the association between the measured object and the tag. However, it requires a power module and network support. The power module results in a short lifespan, reduced reliability, and increased maintenance costs. Network support issues prevent it from functioning properly in areas with poor network connectivity. In contrast, IC cards offer advantages such as long lifespan, reusability, small size for portability, high reliability, and strong security. Therefore, this electronic tag cannot effectively replace IC cards.

[0004] Chinese invention patent CN202010233879.X proposes a method for compressing grain depot video data based on target detection and trajectory analysis. This method involves collecting real-time video monitoring data from grain depots using surveillance cameras, analyzing the acquired video data using a target detection algorithm to extract valid frames containing moving targets, as well as the target's location and sub-image information; employing a convolutional neural network model to perform target recognition on the target sub-images to obtain target category information; and performing target matching and tracking based on target information collected from multiple consecutive frames from the same camera to obtain the target's motion trajectory data in the current scene. Based on the information collected from each surveillance camera, a storage architecture for image data, tag data, and trajectory data is constructed to achieve efficient data storage. This invention can extract and store valid data related to moving targets from massive amounts of surveillance video data, improving the utilization rate of storage resources. However, this invention focuses on the compressed storage of surveillance video data and does not further utilize the extracted moving target trajectory data.

[0005] Chinese invention patent CN202210344357.6 proposes a vehicle detection method, system, and storage medium. The vehicle detection method includes: acquiring video surveillance information within a grain depot area; determining whether there are moving vehicles in the video surveillance information, and if so, determining the type of the moving vehicle; if the moving vehicle is a grain transport vehicle, determining whether the grain transport vehicle is loaded; if so, acquiring the movement trajectory of the loaded vehicle and determining whether the loaded vehicle is entering or leaving the depot. This saves time and reduces labor costs associated with manual measurement. However, this invention only determines whether a grain transport vehicle is loaded and whether it is entering or leaving the depot, and does not fully utilize the extracted grain transport vehicle trajectory. Therefore, the provided vehicle detection system has limited practical application value in the supervision of grain transport vehicle operations in grain depots. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method for supervising the operation of grain transport vehicles in grain depots, so as to supervise the entire operation process of grain transport vehicles in grain depots.

[0007] To address the aforementioned technical problems, this invention provides a method for supervising the operation of grain transport vehicles in grain depots, comprising the following steps:

[0008] S100. Bind the IC card number to the grain transport vehicle trajectory ID;

[0009] S200: Obtain the real-time trajectory of grain transport vehicles from the monitoring video data within the grain depot area;

[0010] S300. Extract the stationary point and stationary time interval of the grain transport vehicle from the trajectory of the grain transport vehicle;

[0011] S400. Match the stationary point with the grain depot operation point area, and use the stationary time interval of the matched stationary point as the first reference value of the operation time.

[0012] S500: The time when the grain transport vehicle at the stationary point completes the operation is taken as the second reference value of the operation time. If the difference between the second reference value of the operation time and the time interval of the first reference value of the operation time exceeds the preset time difference threshold, an alarm is issued.

[0013] S600: Group the grain transport vehicle trajectory data and perform grain transport vehicle trajectory anomaly detection and stationary point anomaly detection. When an anomaly is detected, an alarm prompt is issued.

[0014] As an improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0015] The process of obtaining the trajectory of the grain transport vehicle is as follows:

[0016] The surveillance video stream data from the same surveillance camera is input into the target detection algorithm model to detect vehicles in the surveillance video stream data and obtain the target detection box and confidence of the grain transport vehicle. The target detection box of the grain transport vehicle is then processed by the multi-target tracking algorithm model to extract the trajectory data of the grain transport vehicle under a single camera. Then, the cross-camera trajectory association algorithm is used to obtain the associated trajectory data of the grain transport vehicle.

[0017] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0018] The cross-camera trajectory association algorithm is specifically as follows:

[0019] The trajectory data of grain trucks under a single camera is converted into relative coordinates within a grain depot setting. Then, the coordinate points of multiple trajectories are iterated based on time sequence for the trajectory data of the same grain truck from multiple single cameras.

[0020] (1) When there are overlapping coordinate points in different trajectory data at the same time, merge the multiple overlapping coordinate points into one coordinate point;

[0021] (2) When there are inconsistent coordinate points in different trajectory data at the same time, the coordinate points with high confidence are spliced ​​together and the coordinate points with low confidence are discarded.

[0022] (3) In addition to the above two cases, the coordinate points are directly spliced ​​to obtain the associated grain transport vehicle trajectory data.

[0023] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0024] The target detection algorithm model is a YOLOv8 convolutional neural network, and the multi-target tracking algorithm model is the BoT-SORT algorithm.

[0025] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0026] The stationary point and stationary time interval of the grain transport vehicle are obtained from the associated grain transport vehicle trajectory using the Traj-DBSCAN algorithm.

[0027] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0028] The process for obtaining the first reference value of the operation time is as follows:

[0029] The grain depot operation point area includes the areas for tare weight operation point, gross weight measurement operation point, pre-inspection and sampling operation point, and grain loading and unloading operation point. The stationary point is compared with the operation point area of ​​each grain depot in turn:

[0030] (1) When the stationary point is a single coordinate point, when the minimum Euclidean distance between the coordinates of the stationary point and the boundary of the grain depot operation point location area is less than the preset distance threshold, the stationary time interval of the current operation point is used as the first reference value of the operation time, and the grain depot operation point location area that matches the stationary point is recorded.

[0031] (2) When the stationary point is a cluster of coordinate points, when the coordinates of the stationary point that exceed the proportional threshold are in the location area of ​​the grain depot operation point, the stationary time interval of the current operation point is used as the first reference value of the operation time, and the location area of ​​the grain depot operation point that matches the stationary point is recorded.

[0032] (3) If no grain depot operation point matches the stationary point, repeat steps S200, S300 and S400 until the first reference value of operation time is obtained.

[0033] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0034] The time point at which the grain transport vehicle at the stationary point completes its operation is when the grain transport vehicle uses an IC card to report operation data, including the operation type and the time point at which the operation is completed.

[0035] As a further improvement to the grain depot grain transport vehicle operation supervision method of the present invention:

[0036] The process of trajectory anomaly detection and stationary point anomaly detection is as follows:

[0037] The grain depot's planar map is gridded, and the relative coordinates of the associated grain transport vehicle trajectories and the stationary points of the grain transport vehicles in the grain depot scenario are converted into grid coordinates. Then, the grain transport vehicle trajectories and stationary points converted into grid coordinates are grouped according to the business type and grain depot number of the grain transport vehicles. Finally, an anomaly detection algorithm based on isolation is used to detect anomalies in the grouped grain transport vehicle trajectories and stationary points.

[0038] The beneficial effects of this invention are mainly reflected in:

[0039] 1. This invention ensures a one-to-one correspondence between IC cards and grain transport vehicles by binding the IC card to the grain transport vehicle's trajectory when issuing the IC card;

[0040] 2. This invention extracts the stationary time intervals of grain transport vehicles in the work area by tracking their trajectories. It reports the working time of the grain transport vehicles in each work area through IC cards. By comparing the working time and stationary time intervals of the grain transport vehicles in real time, the correspondence between the IC cards and the grain transport vehicles can be monitored, ensuring that the entire operation process of the grain transport vehicles is monitored. This helps to monitor cheating behaviors such as "changing vehicles and changing cards" during the operation of grain transport vehicles.

[0041] 3. This invention, through the detection of abnormal grain transport vehicle trajectories and abnormal stationary points, can effectively utilize grain transport vehicle trajectory data to monitor abnormal situations such as detours and abnormal stationary states that occur during the operation of grain transport vehicles. This enables systematic, intelligent, and information-based management of grain depots, which is conducive to the construction of smart grain depots. Attached Figure Description

[0042] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0043] Figure 1 This is a flowchart of a grain depot grain transport vehicle supervision method according to the present invention.

[0044] Figure 2 This is a flowchart illustrating the real-time acquisition of grain transport vehicle trajectory data in a grain depot transport vehicle monitoring method according to the present invention. Detailed Implementation

[0045] The present invention will be further described below with reference to specific embodiments, but the scope of protection of the present invention is not limited thereto:

[0046] Example 1: A method for supervising the operation of grain transport vehicles in grain depots, such as... Figure 1-2 As shown, it includes the following steps:

[0047] S100: Issue IC cards, create a grain transport vehicle trajectory field, and bind the IC card number to the trajectory ID.

[0048] When a grain truck enters the grain depot, it needs to be registered, its operation serial number written into an IC card and issued, and the IC card issuance time reported. Then, a grain truck trajectory field is created, binding the IC card number with the grain truck trajectory ID. Optionally, binding means that the corresponding grain truck trajectory ID can be directly queried through the IC card number, and the corresponding IC card number can also be directly queried through the grain truck trajectory ID. After the IC card is issued, a grain truck trajectory starting point is created, and the coordinates and time of the starting point are the coordinates of the IC card issuance point and the IC card issuance time, respectively.

[0049] S200 can acquire real-time trajectory data of grain transport vehicles.

[0050] The process of obtaining the trajectory of grain transport vehicles is as follows Figure 2 As shown, it specifically includes:

[0051] S201, acquire surveillance video stream data of the grain depot area.

[0052] Surveillance cameras should be installed within the grain depot area to acquire video streaming data. It should be noted that the surveillance within the grain depot should be comprehensive, and there may be overlap in the viewing angles between the cameras.

[0053] S202 uses object detection and multi-object tracking algorithms to extract the trajectory data of grain transport vehicles under a single camera.

[0054] In the host computer, the monitoring video stream data from the same surveillance camera is input into the target detection algorithm model to detect vehicles in the monitoring video stream data, obtain the target detection box and confidence of the grain transport vehicle, and then the target detection box of the grain transport vehicle is processed by the multi-target tracking algorithm model to obtain the trajectory data of the grain transport vehicle.

[0055] As an example, the object detection algorithm model can be a YOLOv8 convolutional neural network, and the multi-object tracking algorithm model can be the BoT-SORT algorithm. The core of the BoT-SORT algorithm is the grain truck ReID algorithm, the Kalman filter algorithm, and the Hungarian algorithm. The BoT-SORT algorithm uses the Kalman filter algorithm to predict the grain truck object detection box in the current frame based on the grain truck object detection box in the previous frame. It then uses the grain truck ReID algorithm to obtain the features of the grain truck object detection box. Based on the grain truck object detection box features and confidence levels, it matches the real object detection box of the current frame obtained by the object detection algorithm with the Kalman filter predicted detection box. The successfully matched grain truck object detection box is used as the trajectory segment of the current frame. Trajectory segments from multiple frames constitute the grain truck trajectory segment. By continuously extracting grain truck trajectory segments, the final grain truck trajectory data is obtained. The grain truck ReID algorithm is a grain truck feature re-identification algorithm, and the grain truck ReID algorithm can be a ResNest50 convolutional neural network.

[0056] As one implementation method for generating object detection algorithm models, an initial model and dataset for object detection algorithm are constructed, wherein the dataset includes images of grain trucks in a grain depot scene; the dataset is labeled to obtain a training set; the initial model of object detection algorithm is trained based on the training set to obtain a grain truck object detection model.

[0057] As one implementation method for generating the ReID algorithm model for grain transport vehicles, the initial model and dataset of the ReID algorithm for grain transport vehicles are obtained, wherein the dataset includes images of grain transport vehicles; the dataset is labeled to obtain a training set; the initial model of the ReID algorithm for grain transport vehicles is trained based on the training set to obtain the ReID model for grain transport vehicles.

[0058] In this embodiment, the YOLOv8 convolutional neural network and the ResNest50 convolutional neural network are trained separately using training and test sets. The training process of the convolutional neural network is divided into two stages. The first stage is the forward propagation stage, where data is propagated from lower to higher levels. The second stage is the backpropagation stage, where, if the result obtained from forward propagation does not match the expectation, the error is propagated from higher to lower levels during training. Specifically, the training process is as follows:

[0059] The first step is to initialize the weights of the network;

[0060] The second step involves the input data being propagated forward through a convolutional layer, a downsampling layer, and a fully connected layer to obtain the output value.

[0061] The input data consists of categorized and labeled images of loaded and empty vehicles.

[0062] The third step is to calculate the error between the network's output value and the target value.

[0063] Step 4: When the error is greater than the expected value, the error is fed back into the network, and the errors of the fully connected layer, downsampling layer, and convolutional layer are calculated in sequence. The error of each layer can be understood as how much of the total error of the network should be borne by the network. When the error is equal to or less than the expected value, the training ends.

[0064] Step 5: Update the weights based on the obtained error, and then proceed to step 2.

[0065] S203 uses a cross-camera trajectory association algorithm to obtain the trajectory data of the grain transport vehicle.

[0066] The target detection algorithm and multi-target tracking algorithm models obtained the trajectory data of grain trucks from a single camera by inputting video data. The grain depot's monitoring system has full coverage, but there is overlap between the monitoring images. Therefore, the trajectory data extracted from multiple cameras may overlap, and the target detection and multi-target tracking algorithms may also obtain incorrect trajectories for the grain trucks, meaning that the coordinates of multiple trajectories of the same grain truck at the same time may be inconsistent. Therefore, a cross-camera trajectory association algorithm is needed to perform trajectory association operations on the grain truck trajectories extracted from multiple cameras, merging or discarding multiple trajectories. The input of the cross-camera trajectory association algorithm is the trajectory data of the same grain truck extracted from multiple cameras, and the output is the associated grain truck trajectory data.

[0067] As an example, the cross-camera trajectory association algorithm specifically involves first converting the coordinates of the grain truck trajectory data obtained in step S202 under a single camera into relative coordinates within the grain depot scene. When the same grain truck has trajectory data under multiple single cameras, trajectory association is required. During association, the coordinate points of multiple trajectories are traversed based on time order.

[0068] (1) When there are overlapping coordinate points between multiple trajectory data of grain transport vehicles at the same time, the overlapping coordinate points are merged into one coordinate point;

[0069] (2) When there are inconsistent coordinate points in multiple trajectory data of grain transport vehicles at the same time, the coordinate points are selected based on the confidence of the trajectory segment corresponding to the coordinate point. The confidence of the trajectory segment is the confidence of the target detection box of the grain transport vehicle corresponding to the trajectory segment. The coordinate points with high confidence are spliced ​​together, and the coordinate points with low confidence are discarded.

[0070] (3) Except for the two cases (1) and (2) above, the coordinate points are directly spliced ​​together. Finally, the associated grain transport vehicle trajectory data is obtained.

[0071] S300 extracts the stationary point and stationary time interval of the grain transport vehicle from the associated grain transport vehicle trajectory.

[0072] The stationary point of a grain transport vehicle refers to the location where the vehicle stops during operational activities within the grain depot area. This includes points such as the tare weight weighing point and the gross weight weighing point, as well as pre-inspection sampling points and grain loading / unloading points. The stationary time interval of a grain transport vehicle refers to the time interval during which the vehicle remains at each stationary point. For example, if the coordinates of the stationary point are (x, y), the corresponding stationary time interval is [x...y]. t y t That is, the grain transport vehicle departs from time x. t At that time y t It is stationary at coordinates (x, y).

[0073] As an example, the algorithm for extracting stationary points and stationary time intervals of grain transport vehicles can be the Traj-DBSCAN algorithm (《Robust and Hierarchical Stop Discovery in Sparse and Diverse Trajectories》, Tran, Le Hung; Nguyen, Quoc Viet Hung; Do, Ngoc Hoan; Yan, Zhixian; 2011). In the Traj-DBSCAN algorithm, stationary points in the trajectory are defined as clusters of trajectory points where the maximum distance between trajectory points is less than the maximum distance threshold ∈ and the time interval between trajectory points is greater than the minimum time threshold minTime. A trajectory point cluster is a set of trajectory points, and the time interval between trajectory points is the difference between the maximum and minimum time of a trajectory point in the cluster.

[0074] As an example, the maximum distance threshold ∈ is set to 10m, and the minimum time threshold minTime is set to 30s.

[0075] The pseudocode for the Traj-DBSCAN algorithm is shown in Table 1. The process is as follows:

[0076] Table 1. Pseudocode of the Traj-DBSCAN algorithm

[0077]

[0078] S400 matches the stationary point with the grain depot operation area and obtains the first reference value of the operation time based on the stationary time interval.

[0079] S401, Obtain the location and area information of each operation point in the grain depot;

[0080] The grain depot operation points include tare weight measurement operation points, gross weight measurement operation points, pre-inspection and sampling operation points, and grain loading and unloading operation points. The grain depot operation point location area is a square parking area for grain transport vehicles when they are operating at the tare weight measurement operation points, gross weight measurement operation points, pre-inspection and sampling operation points, and grain loading and unloading operation points, which includes four boundaries: upper, lower, left, and right.

[0081] As an example, the location information of grain depot operation points includes the coordinate values ​​of the location areas of each operation point.

[0082] S402, compare the location information of the stationary point with the location information of the grain depot operation point in sequence. When the stationary point is a single coordinate point, when the minimum Euclidean distance between the coordinates of the stationary point and the boundary of the grain depot operation point location area is less than the preset distance threshold, record the grain depot operation point location area that matches the stationary point, and take the stationary time interval of the current operation point as the first reference value of the operation time, and stop the comparison.

[0083] When the stationary point is a cluster of coordinate points, if the coordinates of the stationary point that exceed the proportional threshold are in the location area, record the location area of ​​the grain depot operation point that matches the stationary point, and take the stationary time interval of the current operation point as the first reference value of the operation time, and stop the comparison.

[0084] As an example, when the stationary point is a single coordinate point, the preset distance threshold is 1m; when the stationary point is a cluster of coordinate points, the proportion threshold is 60%, that is, when more than 60% of the coordinates of the stationary point are in the work point location area.

[0085] S403, the grain transport vehicle has not completed all operations (returned from the vehicle and exited the gate), and continues to generate operation trajectory data. Therefore, trajectory extraction, stationary point extraction, and anomaly detection steps need to be continuously performed. If no stationary point matches the location information in the grain depot operation point location area information, repeat steps S200, S300, and S400 until the first reference value of the operation time and the grain depot operation point location area matching the stationary point are obtained.

[0086] S500 queries the operation data reported by the IC card of the corresponding work point matching the stationary point to obtain the second reference value of the grain transport vehicle operation time. It compares the first reference value of the grain transport vehicle operation time with the second reference value. When the comparison is abnormal, an alarm prompt is issued.

[0087] When performing operations, grain transport vehicles use IC cards to report operational data. Each time a grain transport vehicle completes an operation at a work point, it uses the IC card to report the operational data. The operational data reported by the IC card includes the operation type (tare weight measurement operation, gross weight measurement operation, pre-inspection sampling operation, and grain loading and unloading operation) and the operation time. Based on the grain depot operation point area matched with the stationary point obtained in step S402, the operation type corresponding to the stationary point is obtained. The time point of completion of the operation reported by the IC card is retrieved based on the operation type as the second reference value for the grain transport vehicle's operation time.

[0088] Specifically, the first reference value for the grain transport vehicle's operating time is a time interval, and the second reference value is a specific moment. The system determines whether the second reference value falls within the time interval of the first reference value. If not, it checks whether the difference between the second reference value and the upper or lower limit of the first reference value exceeds a preset time difference threshold. If so, an alarm is triggered to indicate an anomaly. This invention obtains the second reference value for the grain transport vehicle's operating time through data reported by an IC card. By comparing the first and second reference values ​​and issuing an alarm when an anomaly occurs, the system can monitor the correspondence between the IC card and the grain transport vehicle in real time.

[0089] As an example, the preset time difference threshold can be 1 minute.

[0090] The S600 groups the grain transport vehicle trajectory data and performs abnormal detection of grain transport vehicle trajectory and abnormal detection of stationary points. When an abnormality is detected, an alarm is issued.

[0091] S601, obtain the type of business being performed by the grain transport vehicle; optionally, the business type includes grain warehousing business and grain outbound business.

[0092] S602, obtain the grain warehouse number where the grain transport vehicle is performing its business;

[0093] S603: Based on the type of grain depot business performed by the grain transport vehicle and the grain depot number involved, the trajectory data of the grain transport vehicle is grouped, with trajectories of the same type of grain depot business and the grain depot number involved being grouped together.

[0094] S604, perform anomaly detection for grain transport vehicle trajectories and stationary points. When an anomaly is detected, issue an alarm. Specifically, obtain a grain depot planar map and grid it. Gridding involves dividing the grain depot planar map into grids. For example, the grid size can be 5m×5m. Based on the gridded grain depot map, perform coordinate transformation on the associated grain transport vehicle trajectories obtained in step 200 and the stationary points of the grain transport vehicles extracted in step 300. Coordinate transformation refers to converting the relative coordinates in the grain depot scene into grid coordinates. Then, for the grain transport vehicle trajectories and stationary points converted to grid coordinates, group them according to steps S601-S603, and use an isolation-based trajectory anomaly detection algorithm (iBOAT: Isolation-Based Online Anomalous Trajectory Detection, Chao Chen; Daqing Zhang; Pablo Samuel Castro; Nan Li; Lin Sun; Shijian Li; Zonghui Wang, 2013) for anomaly detection.

[0095] The core idea of ​​the isolation-based trajectory anomaly detection algorithm is inspired by the Isolation Forest algorithm, and the pseudocode of the algorithm is shown in Table 2. In the grain depot scenario, the normal trajectories of grain transport vehicles are numerous and similar, while the abnormal trajectories are few and unique. Compared to normal trajectories, abnormal trajectories are easier to isolate. The core basis of the algorithm is that coordinate points that do not exist in normal trajectories often exist in abnormal trajectories. The anomaly score of a trajectory depends on the average number of times that trajectory is isolated from the trajectory set.

[0096] Table 2. Pseudocode examples of isolation-based trajectory anomaly detection algorithms

[0097]

[0098] The formula for calculating the anomaly score s of trajectory t is shown in Equation (1), where N is the total number of trajectories in a trajectory group. c(N) is the average number of isolations of the trajectories after a given number of trajectories N, which is used to regularize the anomaly detection score. The calculation method of c(N) is shown in Equation (2), where H(N-1) is the harmonic number, which can be estimated as shown in Equation (3).

[0099]

[0100] H(N-1)≈ln(N-1)+0.57721566 (3)

[0101] When the average number of trajectory isolations approaches c(N), the trajectory anomaly score approaches 0.5; when the average number of trajectory isolations approaches 0, the trajectory anomaly score approaches 1; when the average number of trajectory isolations approaches N-1, the trajectory anomaly score approaches 0; when the trajectory anomaly score obtained from the detection of grain transport vehicle trajectory anomalies exceeds the trajectory anomaly threshold, the trajectory is considered abnormal, and an anomaly alarm is issued.

[0102] As an example, the number of runs m is set to 20, the subsample size ψ is set to 80% of the number of trajectories in the entire trajectory set T, and the maximum isolation count is set to 50. The anomaly threshold for stationary point anomaly detection is set to 0.5, and the anomaly threshold for trajectory anomaly detection is set to 0.4.

[0103] The S700 stores data in a database.

[0104] The data includes trajectory data, trajectory point data, IC reporting business flow data, extracted trajectory stationary point data and stationary time interval data, and grain transport vehicle operation anomaly detection result data.

[0105] The trajectory data includes the trajectory ID and the sequence of trajectory points. The trajectory point data includes coordinate data and coordinate point time data.

[0106] The implementation principle of this invention is as follows:

[0107] By binding IC cards to the grain transport vehicle's trajectory when issuing them, the correspondence between the IC card and the grain transport vehicle can be guaranteed. By acquiring the grain transport vehicle's trajectory in real time and extracting data on the vehicle's stationary points and stationary time intervals, a first reference value for the grain transport vehicle's operating time is obtained. A second reference value for the grain transport vehicle's operating time is obtained through data reported by the IC card. By comparing the first and second reference values ​​for the grain transport vehicle's operating time and issuing alarms when an anomaly occurs, the correspondence between the IC card and the grain transport vehicle can be monitored in real time. This process is repeated until the grain transport vehicle completes all its operations, ensuring monitoring of the entire operation process. Through abnormal detection of grain transport vehicle trajectories and abnormal detection of abnormal stationary points, the grain transport vehicle trajectory data can be effectively utilized, thereby enabling systematic, intelligent, and information-based management of the grain depot, which is conducive to the construction of a smart grain depot.

[0108] Finally, it should be noted that the above examples are merely some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments and many variations are possible. All variations that can be directly derived or conceived by those skilled in the art from the disclosure of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for monitoring the operation of a grain depot grain transport vehicle, characterized in that: The method comprises the following steps: S100, binding the card number of the IC card with the grain cart track ID; S200, obtaining the grain cart track from the monitoring video data in the grain depot area in real time; S300, extracting the grain cart stationary point and stationary time interval from the grain cart track; S400, matching the stationary point with the grain depot operation point area, and taking the stationary time interval of the matched stationary point as the first reference value of the operation time; S500, querying the operation data reported by the IC card of the corresponding operation point matched with the stationary point to obtain the second reference value of the grain cart operation time, the operation data reported by the IC card including the operation type and the operation time, obtaining the operation type corresponding to the stationary point according to the grain depot operation point area matched with the stationary point, querying the time point of completing the operation reported by the IC card as the second reference value of the grain cart operation time according to the operation type; if the difference between the second reference value of the operation time and the range time interval of the first reference value of the operation time exceeds the preset time difference threshold, an alarm prompt is issued; S600, grouping the grain cart track data and detecting the grain cart track anomaly and the stationary point anomaly, and issuing an alarm prompt when an anomaly is detected.

2. The grain depot grain cart operation monitoring method according to claim 1, wherein: the process of obtaining the grain cart track is: inputting the monitoring video stream data of the same monitoring camera into a target detection algorithm model to detect the vehicles in the monitoring video stream data to obtain a grain cart target detection frame and a confidence level, and converting the grain cart target detection frame through a multi-target tracking algorithm model to obtain the extracted grain cart track data under a single camera; then using a cross-camera track association algorithm to obtain the associated grain cart track data.

3. The grain depot grain cart operation monitoring method according to claim 2, wherein: the cross-camera track association algorithm specifically comprises, converting the grain cart track data under a single camera into relative coordinates in the grain depot scene, and then traversing the coordinate points of multiple grain cart track data under multiple single cameras of the same grain cart based on time sequence: (1) when there are overlapping coordinate points in different track data at the same time, merging the overlapping coordinate points into one coordinate point; (2) when there are inconsistent coordinate points in different track data at the same time, splicing the coordinate points with high confidence level and discarding the coordinate points with low confidence level; (3) directly splicing the coordinate points to obtain the associated grain cart track data except for the two cases.

4. The grain depot grain cart operation monitoring method according to claim 3, wherein: the target detection algorithm model is a YOLOv8 convolutional neural network, and the multi-target tracking algorithm model is a BoT-SORT algorithm.

5. The grain depot grain cart operation monitoring method according to claim 4, wherein: the grain cart stationary point and the stationary time interval are obtained from the associated grain cart track through a Traj-DBSCAN algorithm.

6. The grain depot grain cart operation monitoring method according to claim 5, wherein: The obtaining process of the first reference value of the operation time is: The grain depot operation point position area includes the areas of the tare operation point, the metering gross weight operation point, the pre-inspection sampling operation point, and the grain loading and unloading operation point. The stationary point is compared with each grain depot operation point position area in turn: (1) When the stationary point is a single coordinate point, if the minimum Euclidean distance between the stationary point coordinate and the boundary of the grain depot operation point position area is less than the preset distance threshold, the stationary time interval of the current operation point is taken as the first reference value of the operation time, and the grain depot operation point position area matched with the stationary point is recorded; (2) When the stationary point is a cluster of coordinate points, if more than a proportion threshold of the coordinates in the stationary point are in the grain depot operation point position area, the stationary time interval of the current operation point is taken as the first reference value of the operation time, and the grain depot operation point position area matched with the stationary point is recorded; (3) If no grain depot operation point is matched with the stationary point, steps S200, S300, and S400 are repeated until the first reference value of the operation time is obtained.

7. The grain depot grain transport vehicle operation monitoring method according to claim 6, characterized in that: The process of the trajectory anomaly detection and the stationary point anomaly detection is: The grain depot plane map is gridded, and the relative coordinates of the associated grain transport vehicle trajectory and the stationary point of the grain transport vehicle in the grain depot scene are converted into grid coordinates; then the grain transport vehicle trajectory and the stationary point converted into network coordinates are grouped according to the business type and the grain warehouse number executed by the grain transport vehicle, and the grouped grain transport vehicle trajectory and stationary point are respectively subjected to anomaly detection using an isolation-based trajectory anomaly detection algorithm.

Images