A method and system for monitoring and managing overloaded vehicles based on multi-source data fusion

Abstract

This invention relates to the field of overloaded vehicle monitoring technology, specifically a method and system for monitoring and managing overloaded vehicles based on multi-source data fusion. The method includes acquiring target vehicle video and total mass data; determining whether the target vehicle is a suspected overloaded vehicle; if the target vehicle is determined to be a suspected overloaded vehicle, extracting the target vehicle's license plate information and calculating the overload rate; if the overload rate is greater than a pre-set overload threshold, marking the target vehicle as an overloaded vehicle and generating a management case; predicting the time window for the overloaded vehicle to arrive at a downstream overload control station; within the time window, if the total mass data of the same overloaded vehicle is obtained again from the downstream overload control station, comparing and identifying the same overload behavior. This invention, by utilizing multi-source data for vehicle identification, reduces the computational and communication resource consumption of frequently calling external systems, thereby improving the efficiency of overloaded vehicle monitoring and management.

Description

Technical Field

[0001] This invention relates to the field of overloaded vehicle monitoring technology, specifically a method and system for monitoring and managing overloaded vehicles based on multi-source data fusion. Background Technology

[0002] Managing overloaded and oversized transport on highways is crucial for ensuring road safety and maintaining transportation infrastructure. Among these measures, addressing overweight is the most challenging aspect. "Overweight" refers to a situation where the difference between the total mass of the target vehicle and its maximum permissible total mass, divided by the maximum permissible total mass, exceeds the overweight threshold, meaning the overweight rate exceeds the threshold. Current technical steps for managing overweight vehicles involve: first, measuring the total mass of the target vehicle at pre-designated monitoring stations on key road sections; then, extracting the vehicle's license plate information from captured video footage, and retrieving the maximum permissible total mass from the vehicle information query system based on this license plate information; finally, calculating the overweight rate using the total mass and the maximum permissible total mass, and then determining if the vehicle is overweight.

[0003] However, existing technologies have the following drawbacks. First, to accurately calculate the overload rate, it is necessary to obtain the maximum permissible gross weight of each vehicle passing through the overload control station. To obtain this maximum permissible gross weight, it is necessary to retrieve the target vehicle's maximum permissible gross weight from an external vehicle information query system based on the vehicle license plate information extracted from vehicle videos. Performing this operation for every passing vehicle would frequently call external systems, consuming significant computing and communication resources and increasing data security risks. Second, when overloaded vehicles travel on the highway network, they may be repeatedly detected by different downstream overload control stations. Existing technologies lack effective prediction of vehicle travel trends, leading to the potential generation of repeated enforcement cases for the same overload behavior, resulting in a waste of law enforcement resources.

[0004] Therefore, it is necessary to design a more convenient method that can utilize multi-source data from local overload control stations for preliminary identification of overloaded vehicles in the initial stage of overload identification. This would enable preliminary screening of suspected overloaded vehicles, reducing the frequency of data retrieval from external vehicle information query systems, thus minimizing communication resource consumption and data security risks. Simultaneously, the method could predict the driving trajectory of overloaded vehicles and collaborate with downstream overload control stations to achieve rapid identification and merging of the same overload behavior, thereby improving governance efficiency while ensuring governance effectiveness. Summary of the Invention

[0005] (1) Technical problems to be solved The purpose of this invention is to provide a method and system for monitoring and managing overloaded vehicles based on multi-source data fusion, so as to improve the efficiency of monitoring and managing overloaded vehicles and reduce the computing power and communication consumption of calling external systems.

[0006] (2) Technical solution To achieve the above objectives, the present invention provides a method for monitoring and managing overloaded vehicles based on multi-source data fusion, the method comprising the following steps: S1: Obtain the target vehicle video captured by the fixed camera at the current overload control station (pre-set) and the target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain the target vehicle model information; perform preliminary overload identification based on the target vehicle model information, the target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, steps S2 to S4 are not executed, and the process exits directly.

[0007] S2, if the target vehicle is determined to be a suspected overloaded vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overloaded vehicle and generate a management case; otherwise, skip steps S3 to S4 and exit directly.

[0008] S3. Based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, a trajectory prediction method is used to deduce the driving trend of the overloaded vehicle and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, a unique monitoring task instruction for the overloaded vehicle is generated and sent to the downstream overload control station.

[0009] S4. Within the time window, if the total mass data of the same overloaded vehicle is obtained again from the downstream overload control station, the total mass data of the target vehicle obtained again is compared with the total mass data of the target vehicle obtained in step S1. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overloaded behavior, and the control case will not be generated again. Only data association and recording will be performed.

[0010] Furthermore, the method for identifying the target vehicle video to obtain the target vehicle model information includes: The target vehicle's lane and image area are extracted from the target vehicle video when the target vehicle enters a pre-defined sampling area. If the target vehicle is found to have changed lanes in the pre-defined sampling area, the target vehicle image area is calculated based on the target vehicle before the lane change.

[0011] Area correction is performed based on the target vehicle image area and the lane where the target vehicle is located at the start of the sampling area to obtain the equivalent pixel area at the standard position; the equivalent pixel area at the standard position is then matched with a pre-set area-vehicle model mapping table to obtain the target vehicle model information.

[0012] Furthermore, the method for extracting the lane and image area of ​​the target vehicle when it enters a pre-defined sampling area from the target vehicle video, and if it is identified that the target vehicle has changed lanes within the pre-defined sampling area, then the method for converting the target vehicle image area based on the target vehicle before the lane change includes: Extract the target vehicle video snapshots when the target vehicle travels into the pre-defined sampling area from the target vehicle video, and record them as the first image to the next. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval.

[0013] For the first image to the... N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... N The closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first...M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the N Original area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

[0014] Furthermore, the method for obtaining the equivalent pixel area at the standard position by performing area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area includes: Based on the distance between the fixed cameras at the current overload control stations and the lanes, the lanes are marked from closest to furthest as lane one to lane number two. H Lanes; among them, H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. H The geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained.

[0015] Identify the lane where the target vehicle was located at the start of the sampling area, and denot it as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ; in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

[0016] Furthermore, the method of using trajectory prediction to deduce the driving trend of the overloaded vehicle and predict the time window for the overloaded vehicle to arrive at the downstream overload control station based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station includes: Based on pre-acquired road network topology data and the location data of the current overload control station, the path distance from the current overload control station to the downstream overload control station is obtained; the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station are obtained; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, respectively, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speedometer and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station is higher than the maximum speed limit, then the minimum speed limit and the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued.

[0017] The second time starting point is obtained by subtracting a pre-set time margin from the first time starting point; the second time ending point is obtained by adding a pre-set time margin to the first time ending point; the time period from the second time starting point to the second time ending point is recorded as the time window for overloaded vehicles to arrive at the downstream overload control station.

[0018] Based on the same inventive concept, this invention also provides an overloaded vehicle monitoring and control system based on multi-source data fusion, the system comprising: The preliminary overload identification module is used to acquire target vehicle videos captured by fixed cameras at the current overload control station and target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain target vehicle model information; perform preliminary overload identification based on the target vehicle model information, target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, the precise overload identification module, trajectory prediction module, and overload behavior merging module are not executed, and the module exits directly.

[0019] The overload precision identification module, connected to the overload preliminary identification module, is used to perform the following actions: if the target vehicle is determined to be a suspected overload vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overload vehicle and generate a management case; otherwise, do not execute the trajectory prediction module and the overload behavior merging module and exit directly.

[0020] The trajectory prediction module, connected to the overloaded vehicle precise identification module, is used to deduce the driving trend of the overloaded vehicle based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, it generates a unique monitoring task instruction for the overloaded vehicle and sends it to the downstream overload control station.

[0021] The overload behavior merging module, connected to the trajectory prediction module, is used to compare the total vehicle mass data of the same overloaded vehicle obtained from the downstream overload control station again within the time window. If the total vehicle mass data of the target vehicle obtained again is obtained from the downstream overload control station, the difference between the obtained target vehicle total mass data and the target vehicle total mass data obtained from the overload preliminary identification module is compared. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overload behavior, and no treatment case is generated again. Only data association and recording are performed.

[0022] Furthermore, the preliminary over-limit identification module includes: The target vehicle image area calculation module is used to extract the lane and target vehicle image area of ​​the target vehicle when it travels into a pre-set sampling area from the target vehicle video. If it is identified that the target vehicle has changed lanes in the pre-set sampling area, the target vehicle image area is converted based on the target vehicle before changing lanes.

[0023] The target vehicle model information recognition module is connected to the target vehicle image area calculation module. It is used to perform area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area to obtain the equivalent pixel area at the standard position. The equivalent pixel area at the standard position is matched with a pre-set area-model mapping table to obtain the target vehicle model information.

[0024] Furthermore, the target vehicle image area calculation module includes: The sampling module is used to extract snapshot images of the target vehicle video when the target vehicle travels into a pre-defined sampling area, denoted as the first image to the second. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval.

[0025] The area calculation module, connected to the sampling module, is used to calculate the area of ​​the first image up to the second image. N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... N The closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first... M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the NOriginal area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

[0026] Furthermore, the target vehicle model information identification module includes: The lane marking module is used to mark the lanes as lane one through lane two, from closest to farthest, based on the distance between the fixed camera at the current overload control station and the lane. H Lanes; among them, H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. H The geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained.

[0027] The area correction module, connected to the lane marking module, is used to identify the lane where the target vehicle is located at the start of the sampling area, denoted as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ; in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

[0028] Furthermore, the trajectory prediction module includes: The time window preliminary calculation module is used to obtain the path distance from the current overload control station to the downstream overload control station based on pre-acquired road network topology data and the location data of the current overload control station; obtain the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speed measuring device at the overload control station and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station is higher than the maximum speed limit, then the minimum speed limit and the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued.

[0029] The time margin overlay module, connected to the time window preliminary calculation module, is used to subtract a pre-set time margin from the first time starting point to obtain the second time starting point; add a pre-set time margin to the first time ending point to obtain the second time ending point; and record the time period from the second time starting point to the second time ending point as the time window for overloaded vehicles to arrive at the downstream overload control station.

[0030] (3) Beneficial effects Compared with the prior art, the beneficial effects of the present invention are: 1. Based on local multi-source data from overload control stations, the system automatically identifies the vehicle model information of the target vehicle and performs preliminary identification of overload based on the target vehicle model information and the target vehicle total mass data. This allows for the screening of suspected overloaded vehicles in the early stages of identification, avoiding the need to retrieve the maximum permissible total mass for suspected overloaded vehicles. This reduces the consumption of computing power and data retrieval communication resources, and also reduces the data security risks caused by frequent data retrieval from external vehicle information query systems.

[0031] 2. By using trajectory prediction to generate time windows and issuing unique monitoring task instructions to downstream overload control stations, collaborative monitoring and data comparison of the same overloaded vehicle can be achieved. Once the same overload behavior is determined, no more governance cases will be generated, thus avoiding the waste of law enforcement resources. Attached Figure Description

[0032] Figure 1 This is a flowchart of a method for monitoring and managing overloaded vehicles based on multi-source data fusion, according to Embodiment 1 of the present invention. Figure 2 This is a schematic diagram of the module composition of an overloaded vehicle monitoring and management system based on multi-source data fusion according to Embodiment 2 of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Before providing examples, it is necessary to describe the application scenario of this invention. This invention is applied to identifying overloaded vehicles at overload control stations and generating enforcement cases. Several overload control stations are distributed throughout the road network topology. These stations are located at predetermined positions on one-way roads. Each station is equipped with a fixed camera positioned to the side of the one-way road, its field of view fully covering the pre-defined sampling area of ​​the road. The road surface is clearly marked with road dividing lines, clearly dividing the road surface into several lanes.

[0035] Example 1: As Figure 1 As shown in the figure, this embodiment provides a method for monitoring and managing overloaded vehicles based on multi-source data fusion. The method includes the following steps: S1: Obtain the target vehicle video captured by the fixed camera at the current overload control station (pre-set) and the target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain the target vehicle model information; perform preliminary overload identification based on the target vehicle model information, the target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, steps S2 to S4 are not executed, and the process exits directly.

[0036] S2, if the target vehicle is determined to be a suspected overloaded vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overloaded vehicle and generate a management case; otherwise, skip steps S3 to S4 and exit directly.

[0037] S3. Based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, a trajectory prediction method is used to deduce the driving trend of the overloaded vehicle and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, a unique monitoring task instruction for the overloaded vehicle is generated and sent to the downstream overload control station.

[0038] S4. Within the time window, if the total mass data of the same overloaded vehicle is obtained again from the downstream overload control station, the total mass data of the target vehicle obtained again is compared with the total mass data of the target vehicle obtained in step S1. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overloaded behavior, and the control case will not be generated again. Only data association and recording will be performed.

[0039] For example, firstly, based on the target vehicle video captured by a fixed camera at the current overload control station and the target vehicle's total mass data obtained from the monitoring equipment, an image recognition method is used to analyze the vehicle video to obtain the target vehicle's model information. The target vehicle model information is one of the following: "two-axle truck," "three-axle truck," "four-axle truck," "five-axle truck," or "six-axle truck." Then, based on this model information and the total mass data, a pre-set target vehicle model and total mass range mapping table is consulted to perform preliminary overload identification, in order to determine whether the target vehicle is a suspected overloaded vehicle. For example, in the target vehicle model and total mass range mapping table, the total mass range corresponding to the "two-axle truck" model is "below 18,000 kg"; and the total mass range corresponding to the "six-axle truck" model is "below 46,000 kg." It is worth noting that, according to current traffic regulations, the weight limit for trucks is determined not only by the number of axles but also by factors such as axle arrangement and vehicle type. Therefore, the total vehicle mass range is a value comprehensively determined after considering all common scenarios for that vehicle type. Furthermore, to avoid overlooking suspected overloaded vehicles, the allowable range for the total vehicle mass range in the target vehicle type-to-total vehicle mass range mapping table is set to the smaller interval. If the target vehicle's total mass data falls within the total vehicle mass range corresponding to the target vehicle type information, the target vehicle is determined not to be a suspected overloaded vehicle; otherwise, it is determined to be a suspected overloaded vehicle. The core of this step lies in using locally available multi-source data for preliminary screening. If the vehicle is determined not to be a suspected overloaded vehicle, all subsequent processes are terminated, and the process exits directly.

[0040] Next, if the target vehicle is preliminarily identified as a suspected overloaded vehicle, its license plate information is extracted from the target vehicle video. Based on this license plate information, the system sends a request to the remote vehicle information query system (i.e., the traffic safety integrated service management platform) to retrieve the legally permitted maximum gross vehicle weight, which is 49,000 kg. It is worth noting that while the maximum permissible gross vehicle weight can be initially determined by the number of axles, it is also related to factors such as axle arrangement and vehicle type. Therefore, the accurate maximum permissible gross vehicle weight can only be obtained by querying the official legal information of the vehicle from the vehicle information query system. Subsequently, based on the target vehicle's gross vehicle weight data obtained from local monitoring and the retrieved maximum permissible gross vehicle weight, the overload rate of 43.12% is calculated. Since the overload rate corresponding to this target vehicle is greater than the preset overload threshold of 5%, the target vehicle is marked as an overloaded vehicle, and a management case is generated.

[0041] Then, based on the speed of the marked overloaded vehicle measured by the speed camera at the current overload control station and the geographical location data of the current station, a trajectory prediction method is used to extrapolate the future travel trend of the overloaded vehicle and calculate the expected time range, i.e., the time window, for the overloaded vehicle to arrive at the downstream overload control station. The downstream overload control station refers to the next necessary overload control station along the vehicle's travel direction in the road network topology. Based on the prediction results, a unique monitoring task instruction specifically for the overloaded vehicle is generated and sent in advance to the corresponding downstream overload control station to guide it in conducting targeted monitoring.

[0042] Finally, at the downstream overload control station, if the same overloaded vehicle is captured again within the predicted time window and its total vehicle mass data is obtained, the difference between this data and the target vehicle's total vehicle mass data first obtained at the current station is compared. If the absolute value of the difference does not exceed a pre-set change threshold of 10%, then the two detections are determined to belong to the same overloaded transport behavior. In this case, the system will not generate a new management case for this detection at the downstream station, but will instead associate and uniformly record the two detection data, thereby completing the closed-loop management of a single overloaded behavior. The change threshold is determined based on the accuracy of the quality monitoring equipment at the overload control station. If the accuracy of the quality monitoring equipment at the overload control station is high, the change threshold can be smaller; otherwise, a larger value will be used.

[0043] Furthermore, the method for identifying the target vehicle video to obtain the target vehicle model information includes: The target vehicle's lane and image area are extracted from the target vehicle video when the target vehicle enters a pre-defined sampling area. If the target vehicle is found to have changed lanes in the pre-defined sampling area, the target vehicle image area is calculated based on the target vehicle before the lane change.

[0044] Area correction is performed based on the target vehicle image area and the lane where the target vehicle is located at the start of the sampling area to obtain the equivalent pixel area at the standard position; the equivalent pixel area at the standard position is then matched with a pre-set area-vehicle model mapping table to obtain the target vehicle model information.

[0045] For example, firstly, the lane and image area of ​​the target vehicle when it enters a pre-defined sampling area are extracted from the target vehicle video. The sampling area refers to a fixed segment defined in the video frame, set within a specific distance from the camera to ensure the target vehicle image is clear and its proportions are relatively stable. During the extraction process, if a lane change is detected within the pre-defined sampling area, the target vehicle image area is recalculated based on the area before the lane change. This is to avoid image area measurement distortion caused by lateral vehicle movement.

[0046] Next, based on the obtained target vehicle image area and the lane information of the target vehicle at the starting point of the sampling area, area correction processing is performed to obtain the equivalent pixel area at the standard position. The main purpose is to eliminate the perspective distortion caused by the difference in distance between the vehicle and the camera in different lanes, and to uniformly convert the original pixel areas measured at different lane positions to the equivalent value at the same reference lane position.

[0047] Finally, the equivalent pixel area at the corrected standard position is matched against a pre-defined area-vehicle model mapping table to obtain the target vehicle model information. The area-vehicle model mapping table stores the correspondence between different vehicle models and their typical pixel area ranges at the standard position. By comparing the actually measured equivalent pixel area with this mapping table, the best-matching vehicle model category can be determined. This method utilizes the stable feature of image area to achieve rapid and automated vehicle model identification. It's worth noting that the principle behind this step is that different vehicle models correspond to different body lengths and sizes; for example, the body length and size of a "six-axle truck" are much larger than those of a "two-axle truck." Therefore, by comparing the size of the image area, the target vehicle model information can be identified.

[0048] Furthermore, the method for extracting the lane and image area of ​​the target vehicle when it enters a pre-defined sampling area from the target vehicle video, and if it is identified that the target vehicle has changed lanes within the pre-defined sampling area, then the method for converting the target vehicle image area based on the target vehicle before the lane change includes: Extract the target vehicle video snapshots when the target vehicle travels into the pre-defined sampling area from the target vehicle video, and record them as the first image to the next. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval.

[0049] For the first image to the... N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... NThe closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first... M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the N Original area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

[0050] For example, multiple video snapshots are extracted from the target vehicle video when the target vehicle travels to a pre-defined sampling area, and these are sequentially denoted as the first image to the tenth image. The image acquisition follows a preset fixed time interval, and the number of target vehicle video snapshots depends on the total time required for the vehicle to pass through the entire sampling area and the set sampling time interval.

[0051] Next, the acquired first to tenth images are processed using contour recognition and segmentation techniques to obtain the corresponding first to tenth target vehicle contours in each frame. Then, based on the relative positional changes of the first to tenth target vehicle contours within the first to tenth images, the system automatically identifies whether the target vehicle has engaged in lane-changing behavior within the sampling area. If no lane-changing behavior is detected, the pixel area occupied by the closed plane enclosed by each of the first to tenth target vehicle contours in their respective images is calculated, yielding the first to tenth areas. Finally, the average of the first to tenth areas is calculated as the final target vehicle image area. If lane-changing behavior is detected within the sampling area, a special processing procedure for lane-changing is initiated. First, the contours acquired before the lane change, i.e., the first to fifth target vehicle contours, are identified as the contours of the vehicle on its original path, denoted as the first to fifth target vehicle original path contours. The pixel area occupied by each of the first to fifth target vehicle original path contours in their corresponding images is calculated, yielding the first to fifth original areas. Then, based on this, a fitting algorithm is used to analyze its changing trend, and the contour area that the vehicle should have in the sixth to tenth frames of the image if it had not changed lanes is predicted by extrapolation, i.e., the sixth to tenth original areas. Finally, the actual measured area before lane change and the extrapolated area after lane change are combined, and their average value is calculated to obtain the target vehicle image area for subsequent analysis. This method aims to eliminate the interference caused by lane change behavior on vehicle image area measurement and ensure the accuracy of preliminary vehicle type recognition.

[0052] Furthermore, the method for obtaining the equivalent pixel area at the standard position by performing area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area includes: Based on the distance between the fixed cameras at the current overload control stations and the lanes, the lanes are marked from closest to furthest as lane one to lane number two. H Lanes; among them, H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. HThe geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained.

[0053] Identify the lane where the target vehicle was located at the start of the sampling area, and denot it as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ; in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

[0054] For example, all lanes are marked according to the relative spatial relationship between the fixed cameras at the current overload control station and each lane. Specifically, the lanes are marked as lane one to lane three in order of increasing distance between the cameras and the lanes.

[0055] Then, to complete the conversion from the original pixel area to the equivalent pixel area at the standard position, an area correction coefficient needs to be assigned to each lane. Specifically, a pre-set first area correction coefficient to a third area correction coefficient is assigned to each of the first to third lanes. The first lane, being closest to the camera, is designated as the reference lane, and its first area correction coefficient is set to 1. Due to perspective effects, vehicles appear smaller in lanes further away; therefore, the values ​​of the second to third area correction coefficients are all greater than 1, and their values ​​increase sequentially with the lane number to compensate for the image size reduction caused by increased distance. The specific values ​​of the first to third area correction coefficients are determined as follows: the same test vehicle (with constant physical size) is placed precisely in sequence at the geometric center of the area overlapping with the sampling area in the first to third lanes, and the fixed camera at the current overload control station is used to take pictures. The area correction coefficient of each lane is obtained by calculating the ratio of the pixel area of ​​the closed plane enclosed by the outline of the test vehicle in each lane position to the pixel area in the reference lane (first lane). That is, the second area correction coefficient is 1.13 and the third area correction coefficient is 1.24.

[0056] The specific lane where the target vehicle is located at the start of the sampling area is identified as the second lane. Therefore, the second area correction coefficient corresponding to this lane is called. By multiplying the measured area of ​​the target vehicle image by the second area correction coefficient, perspective distortion correction can be completed, and the equivalent pixel area at the standard position can be obtained.

[0057] Furthermore, the method of using trajectory prediction to deduce the driving trend of the overloaded vehicle and predict the time window for the overloaded vehicle to arrive at the downstream overload control station based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station includes: Based on pre-acquired road network topology data and the location data of the current overload control station, the path distance from the current overload control station to the downstream overload control station is obtained; the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station are obtained; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, respectively, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speedometer and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station is higher than the maximum speed limit, then the minimum speed limit and the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued.

[0058] The second time starting point is obtained by subtracting a pre-set time margin from the first time starting point; the second time ending point is obtained by adding a pre-set time margin to the first time ending point; the time period from the second time starting point to the second time ending point is recorded as the time window for overloaded vehicles to arrive at the downstream overload control station.

[0059] For example, based on pre-acquired road network topology data including road connectivity, and combined with the specific location data of the current and downstream overload control stations, the most likely actual route for vehicles between the current and downstream stations is determined, and the precise distance of this route, i.e., the path distance, is obtained. Simultaneously, the legally mandated minimum and maximum speed limits on this route segment are obtained as a benchmark range for vehicle speed prediction.

[0060] Then, based on the real-time speed of overloaded vehicles measured by the speed camera at the current overload control station, it is compared with the obtained speed limit range, and different speed parameters are selected to calculate the arrival time. Specifically, there are three cases: if the overloaded vehicle's speed is between the minimum and maximum speed limits, indicating that its speed is compliant, then the minimum and maximum speed limits are used as calculation parameters respectively; if the speed is below the minimum speed limit, indicating low-speed driving, then the measured speed and the maximum speed limit are used as calculation parameters, and a speeding alarm is issued to indicate abnormal low speed; if the speed is above the maximum speed limit, indicating speeding, then the minimum and measured speeds are used as calculation parameters, and a speeding alarm is issued. Based on the selected speed parameters and the known path distance, two different travel times are calculated, and then superimposed with the obtained current time point to obtain the first time start and first time end, forming the basic predicted arrival time range. Based on the obtained basic predicted arrival time range, a pre-set time margin is introduced for expansion. Specifically, this time margin is subtracted from the first time starting point to obtain an earlier second time starting point; and this time margin is added to the first time ending point to obtain a later second time ending point. The time margin is used to compensate for prediction deviations caused by uncertainties in actual driving, such as traffic congestion and overtaking.

[0061] Finally, the entire time period between the extended second time start point and the second time end point will be used as the time window in which overloaded vehicles are expected to arrive at the downstream overload control station.

[0062] Example 2: Based on the same inventive concept, such as Figure 2 As shown in the figure, this embodiment also provides an overloaded vehicle monitoring and management system based on multi-source data fusion, the system comprising: The preliminary overload identification module is used to acquire target vehicle videos captured by fixed cameras at the current overload control station and target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain target vehicle model information; perform preliminary overload identification based on the target vehicle model information, target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, the precise overload identification module, trajectory prediction module, and overload behavior merging module are not executed, and the module exits directly.

[0063] The overload precision identification module, connected to the overload preliminary identification module, is used to perform the following actions: if the target vehicle is determined to be a suspected overload vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overload vehicle and generate a management case; otherwise, do not execute the trajectory prediction module and the overload behavior merging module and exit directly.

[0064] The trajectory prediction module, connected to the overloaded vehicle precise identification module, is used to deduce the driving trend of the overloaded vehicle based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, it generates a unique monitoring task instruction for the overloaded vehicle and sends it to the downstream overload control station.

[0065] The overload behavior merging module, connected to the trajectory prediction module, is used to compare the total vehicle mass data of the same overloaded vehicle obtained from the downstream overload control station again within the time window. If the total vehicle mass data of the target vehicle obtained again is obtained from the downstream overload control station, the difference between the obtained target vehicle total mass data and the target vehicle total mass data obtained from the overload preliminary identification module is compared. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overload behavior, and no treatment case is generated again. Only data association and recording are performed.

[0066] Furthermore, the preliminary over-limit identification module includes: The target vehicle image area calculation module is used to extract the lane and target vehicle image area of ​​the target vehicle when it travels into a pre-set sampling area from the target vehicle video. If it is identified that the target vehicle has changed lanes in the pre-set sampling area, the target vehicle image area is converted based on the target vehicle before changing lanes.

[0067] The target vehicle model information recognition module is connected to the target vehicle image area calculation module. It is used to perform area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area to obtain the equivalent pixel area at the standard position. The equivalent pixel area at the standard position is matched with a pre-set area-model mapping table to obtain the target vehicle model information.

[0068] Furthermore, the target vehicle image area calculation module includes: The sampling module is used to extract snapshot images of the target vehicle video when the target vehicle travels into a pre-defined sampling area, denoted as the first image to the second. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval.

[0069] The area calculation module, connected to the sampling module, is used to calculate the area of ​​the first image up to the second image. N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... N The closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first... M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the N Original area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

[0070] Furthermore, the target vehicle model information identification module includes: The lane marking module is used to mark the lanes as lane one through lane two, from closest to farthest, based on the distance between the fixed camera at the current overload control station and the lane. H Lanes; among them,H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. H The geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained.

[0071] The area correction module, connected to the lane marking module, is used to identify the lane where the target vehicle is located at the start of the sampling area, denoted as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ; in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

[0072] Furthermore, the trajectory prediction module includes: The time window preliminary calculation module is used to obtain the path distance from the current overload control station to the downstream overload control station based on pre-acquired road network topology data and the location data of the current overload control station; obtain the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speed measuring device at the overload control station and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station is higher than the maximum speed limit, then the minimum speed limit and the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to these values ​​and used as the first time start and first time end, respectively, and an overspeed alarm is issued.

[0073] The time margin overlay module, connected to the time window preliminary calculation module, is used to subtract a pre-set time margin from the first time starting point to obtain the second time starting point; add a pre-set time margin to the first time ending point to obtain the second time ending point; and record the time period from the second time starting point to the second time ending point as the time window for overloaded vehicles to arrive at the downstream overload control station.

[0074] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0075] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring and managing overloaded vehicles based on multi-source data fusion, characterized in that, The method includes the following steps: S1: Obtain the target vehicle video captured by the fixed camera at the current overload control station (pre-set) and the target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain the target vehicle model information; perform preliminary overload identification based on the target vehicle model information, the target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, steps S2 to S4 are not executed, and the process exits directly. S2, if the target vehicle is determined to be a suspected overloaded vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overloaded vehicle and generate a management case; otherwise, skip steps S3 to S4 and exit directly. S3. Based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, the trajectory prediction method is used to deduce the driving trend of the overloaded vehicle and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, a unique monitoring task instruction for the overloaded vehicle is generated and sent to the downstream overload control station. S4. Within the time window, if the total mass data of the same overloaded vehicle is obtained again from the downstream overload control station, the total mass data of the target vehicle obtained again is compared with the total mass data of the target vehicle obtained in step S1. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overloaded behavior, and the control case will not be generated again. Only data association and recording will be performed.

2. The method for monitoring and managing overloaded vehicles based on multi-source data fusion as described in claim 1, characterized in that, The method for identifying the target vehicle video to obtain the target vehicle model information includes: The target vehicle's lane and image area are extracted from the target vehicle video when the target vehicle enters a pre-defined sampling area. If the target vehicle is found to have changed lanes in the pre-defined sampling area, the image area of ​​the target vehicle is calculated based on the target vehicle before changing lanes. Area correction is performed based on the target vehicle image area and the lane where the target vehicle is located at the start of the sampling area to obtain the equivalent pixel area at the standard position; the equivalent pixel area at the standard position is then matched with a pre-set area-vehicle model mapping table to obtain the target vehicle model information.

3. The method for monitoring and managing overloaded vehicles based on multi-source data fusion as described in claim 2, characterized in that, The method for extracting the lane and image area of ​​the target vehicle when it enters a pre-defined sampling area from the target vehicle video, and if it is identified that the target vehicle has changed lanes within the pre-defined sampling area, then the method for converting the target vehicle image area based on the target vehicle before the lane change includes: Extract the target vehicle video snapshots when the target vehicle travels into the pre-defined sampling area from the target vehicle video, and record them as the first image to the next. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval. For the first image to the... N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... N The closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first... M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the N Original area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

4. The method for monitoring and managing overloaded vehicles based on multi-source data fusion as described in claim 3, characterized in that, The method for obtaining the equivalent pixel area at a standard position by performing area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area includes: Based on the distance between the fixed cameras at the current overload control stations and the lanes, the lanes are marked from closest to furthest as lane one to lane number two. H Lanes; among them, H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. H The geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained. Identify the lane where the target vehicle was located at the start of the sampling area, and denot it as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ； in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

5. The method for monitoring and managing overloaded vehicles based on multi-source data fusion as described in claim 4, characterized in that, The method for predicting the travel trend of overloaded vehicles and the time window for their arrival at downstream overloaded vehicle checkpoints, based on the speed of overloaded vehicles measured by speed measuring instruments installed at the current overloaded vehicle checkpoints and the location data of the current overloaded vehicle checkpoints, using a trajectory prediction method, includes: Based on pre-acquired road network topology data and the location data of the current overload control station, the path distance from the current overload control station to the downstream overload control station is obtained; the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station are obtained; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, respectively, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speedometer and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to the calculated travel time and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station is higher than the maximum speed limit, the minimum speed limit and the speed of the overloaded vehicle measured by the speedometer installed at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to the calculated travel time and used as the first time start and first time end, respectively, and an overspeed alarm is issued. The second time starting point is obtained by subtracting a pre-set time margin from the first time starting point; the second time ending point is obtained by adding a pre-set time margin to the first time ending point; the time period from the second time starting point to the second time ending point is recorded as the time window for overloaded vehicles to arrive at the downstream overload control station.

6. A monitoring and management system for overloaded vehicles based on multi-source data fusion, characterized in that, The system includes: The preliminary overload identification module is used to acquire target vehicle videos captured by fixed cameras at the current overload control station and target vehicle total mass data obtained from monitoring; identify the target vehicle video to obtain target vehicle model information; perform preliminary overload identification based on the target vehicle model information, target vehicle total mass data, and a pre-obtained target vehicle model and total mass range mapping table to determine whether the target vehicle is a suspected overload vehicle; if the target vehicle is determined not to be a suspected overload vehicle, the precise overload identification module, trajectory prediction module, and overload behavior merging module are not executed, and the module exits directly; The overload precision identification module, connected to the overload preliminary identification module, is used to perform the following actions: if the target vehicle is determined to be a suspected overload vehicle, extract the target vehicle's license plate information from the target vehicle video; retrieve the target vehicle's maximum permissible gross weight from the vehicle information query system based on the target vehicle's license plate information; calculate the overload rate based on the target vehicle's gross weight data and the target vehicle's maximum permissible gross weight; if the overload rate is greater than a preset overload threshold, mark the target vehicle as an overload vehicle and generate a management case; otherwise, do not execute the trajectory prediction module and the overload behavior merging module and exit directly. The trajectory prediction module, connected to the overloaded vehicle precise identification module, is used to deduce the driving trend of the overloaded vehicle based on the speed of the overloaded vehicle measured by the speed measuring instrument installed at the current overload control station and the location data of the current overload control station, and predict the time window for the overloaded vehicle to arrive at the downstream overload control station; based on the time window, it generates a unique monitoring task instruction for the overloaded vehicle and sends it to the downstream overload control station. The overload behavior merging module, connected to the trajectory prediction module, is used to compare the total vehicle mass data of the same overloaded vehicle obtained from the downstream overload control station again within the time window. If the total vehicle mass data of the target vehicle obtained again is obtained from the downstream overload control station, the difference between the obtained target vehicle total mass data and the target vehicle total mass data obtained from the overload preliminary identification module is compared. If the absolute value of the difference does not exceed the preset change threshold, it is determined to be the same overload behavior, and no treatment case is generated again. Only data association and recording are performed.

7. The overload vehicle monitoring and control system based on multi-source data fusion as described in claim 6, characterized in that, The preliminary over-limit identification module includes: The target vehicle image area calculation module is used to extract the lane and target vehicle image area of ​​the target vehicle when it travels into a pre-set sampling area from the target vehicle video. If it is identified that the target vehicle has changed lanes in the pre-set sampling area, the target vehicle image area is converted based on the target vehicle before changing lanes. The target vehicle model information recognition module is connected to the target vehicle image area calculation module. It is used to perform area correction based on the target vehicle image area and the lane where the target vehicle is located at the starting point of the sampling area to obtain the equivalent pixel area at the standard position. The equivalent pixel area at the standard position is matched with a pre-set area-model mapping table to obtain the target vehicle model information.

8. The overload vehicle monitoring and control system based on multi-source data fusion as described in claim 7, characterized in that, The target vehicle image area calculation module includes: The sampling module is used to extract snapshot images of the target vehicle video when the target vehicle travels into a pre-defined sampling area, denoted as the first image to the second. N Image; the extraction time interval for the target vehicle video snapshot image is a preset sampling time interval; N This indicates the number of video snapshot images of the target vehicle, which is related to the time the target vehicle passes through the sampling area and the sampling time interval. The area calculation module, connected to the sampling module, is used to calculate the area of ​​the first image up to the second image. N The target vehicle in the image is subjected to contour recognition and segmentation to obtain the contour of the first target vehicle up to the contour of the second target vehicle. N Target vehicle outline; based on the first target vehicle outline to the... N The outline of the target vehicle in the first image to the second image N The relative position in the image is used to determine whether the target vehicle has changed lanes within the sampling area; if no lane change is found, the outline of the first target vehicle is calculated up to the first... N The closed plane enclosed by the outline of the target vehicle in the first image to the second image N The pixel area occupied in the image is denoted as the first area to the second area. N Area, calculate the first area to the second area. N The average area is used to obtain the target vehicle image area; if a lane change occurs, the outline of the first target vehicle is moved to the next area. N The target vehicle profile before the target vehicle changes lanes is denoted as the first target vehicle's original path profile, up to the second. M The original path outline of the target vehicle, where... M greater than 2 and less than N Integers, respectively calculate the original path contour of the first target vehicle up to the first... M The closed plane enclosed by the original path contour of the target vehicle in the first image to the second image M The pixel area occupied in the image is denoted as the first original area to the second. M Original area, from the first original area to the... M Based on the trend of change of the original area, the fitting algorithm is used to extrapolate to obtain the first... Original area to the N Original area, calculate the first original area up to the second. N The average value of the original area is used to obtain the target vehicle image area.

9. The overload vehicle monitoring and control system based on multi-source data fusion as described in claim 8, characterized in that, The target vehicle model information recognition module includes: The lane marking module is used to mark the lanes as lane one through lane two, from closest to farthest, based on the distance between the fixed camera at the current overload control station and the lane. H Lanes; among them, H The number of lanes; respectively, the first lane to the second lane. H The lane is assigned a pre-set first area correction factor up to the [missing value]. H Area correction coefficients; wherein, the first area correction coefficient is set to 1, and the second area correction coefficient is set to the third. H The area correction factor is greater than 1, and the first area correction factor is up to the second... H The area correction coefficients increase sequentially; from the first area correction coefficient to the second... H The area correction factor was determined by placing the same pre-prepared test vehicle in lanes one through one of the lanes. H The geometric center of the overlapping area between the lane and the sampling area was determined, and the same test vehicle was captured by a fixed camera at the current overload control station, placed in the first lane to the second lane respectively. H The proportional relationship between the pixel areas of the closed plane enclosed by the profile of the test vehicle at the geometric center of the overlapping area between the lane and the sampling area is obtained. The area correction module, connected to the lane marking module, is used to identify the lane where the target vehicle is located at the start of the sampling area, denoted as the lane number. i Lane; i The value range is 1 to H The equivalent pixel area is calculated; the formula for calculating the equivalent pixel area is: ； in, Indicates the equivalent pixel area. This represents the area of ​​the target vehicle image. Indicates the first i Area correction factor.

10. The overload vehicle monitoring and control system based on multi-source data fusion as described in claim 9, characterized in that, The trajectory prediction module includes: The time window preliminary calculation module is used to obtain the path distance from the current overload control station to the downstream overload control station based on pre-acquired road network topology data and the location data of the current overload control station; obtain the minimum speed limit and maximum speed limit from the current overload control station to the downstream overload control station; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is between the minimum speed limit and the maximum speed limit, then the minimum speed limit and the maximum speed limit are used as the driving speed, and the travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station, and then superimposed with the obtained current time as the first time start and the first time end, respectively; if the speed of the overloaded vehicle measured by the speed measuring device installed at the current overload control station is lower than the minimum speed limit, then the speed measured by the speed measuring device installed at the current overload control station is used as the first time start and the first time end, respectively. The speed of the overloaded vehicle measured by the speed measuring device at the overload control station and the maximum speed limit are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to the calculated travel time and used as the first time start and first time end, respectively, and an overspeed alarm is issued. If the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station is higher than the maximum speed limit, then the minimum speed limit and the speed of the overloaded vehicle measured by the speed measuring device at the current overload control station are used as the driving speed. The travel time of the overloaded vehicle to the downstream overload control station is calculated based on the path distance from the current overload control station to the downstream overload control station. The current time is then added to the calculated travel time and used as the first time start and first time end, respectively, and an overspeed alarm is issued. The time margin overlay module, connected to the time window preliminary calculation module, is used to subtract a pre-set time margin from the first time starting point to obtain the second time starting point; add a pre-set time margin to the first time ending point to obtain the second time ending point; and record the time period from the second time starting point to the second time ending point as the time window for overloaded vehicles to arrive at the downstream overload control station.

Images