A service area vehicle parking state monitoring system and a monitoring method

Abstract

A service area vehicle parking state monitoring method, comprising the following steps: step 1, reading the license plate information of the vehicle entering the service area, and binding the license plate information with the formed geomagnetic induction curve; step 2, after the vehicle enters the service area through the service area entrance, judging whether each lane geomagnetic sensor passed by the vehicle in the driving process is similar one by one, and binding the geomagnetic induction curve with the highest similarity meeting the similarity standard with the current read geomagnetic induction curve; step 3, repeating step 2 until no geomagnetic induction curve meeting the similarity standard appears; step 4, after determining the parking position of the vehicle, monitoring the parking time of the vehicle through the monitoring geomagnetic quantity, and triggering the alarm when the parking time of the vehicle is found to be longer than the alarm duration. The present application can help the service area managers to find the vehicle with too long parking time in time, and reduce the probability of accidental events.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology and relates to vehicle parking status monitoring, specifically to a service area vehicle parking status monitoring system and monitoring method. Background Technology

[0002] Highway service areas are places specifically designed for passengers and drivers to stop and rest. They are equipped with facilities such as parking lots, public restrooms, gas stations, vehicle repair shops, restaurants, and convenience stores. The service targets are usually drivers, passengers, and highway law enforcement personnel traveling on the highway. They typically make short stops and rest to allow drivers and passengers to rest and replenish their energy during long-distance driving. The mobility of drivers and passengers passing through highway service areas is very high, and they generally only need to eat and rest briefly, rarely staying overnight.

[0003] Currently, highway service areas do not monitor the status of vehicles entering and exiting. If a vehicle enters a service area and stops for an extended period, it is difficult for service area management personnel to detect it. Furthermore, if long-distance bus and freight truck drivers and others stop at highway service areas and spend the night in their vehicles, it is difficult to detect if they suffer sudden death due to fatigue or other health issues caused by long-distance driving. There have been several cases of truck drivers suffering sudden illnesses while spending the night at service areas, resulting in accidents that were not discovered until several days or even a week later. There is an urgent need for a monitoring system that can detect abnormally long-term vehicle stops in advance. Summary of the Invention

[0004] To overcome the technical defects of the existing technology, this invention discloses a service area vehicle parking status monitoring system and monitoring method.

[0005] The service area vehicle parking status monitoring method of the present invention includes the following steps: Step 1. Read the license plate information of vehicles entering the service area, and read the geomagnetic induction curve formed when the vehicle enters the entrance by the entrance geomagnetic sensor set on the entrance road of the service area. The data processing system binds the license plate information with the formed geomagnetic induction curve. Step 2. For each lane's geomagnetic sensor, the data processing system reads the geomagnetic induction curve formed in real time and determines whether it meets the triggering conditions. If it meets the triggering conditions, it retrieves the geomagnetic induction curves of all adjacent lanes' geomagnetic sensors in the previous time period, and determines whether they are similar to the current lane's geomagnetic induction curve. The geomagnetic induction curve that meets the similarity criteria and has the highest similarity is then bound to the currently read geomagnetic induction curve. Step 3. Repeat step 2 until no geomagnetic induction curves that meet similar criteria appear; The geomagnetic sensor corresponding to the last similar geomagnetic induction curve is considered to be the position before parking. The parking space geomagnetic sensor curve of the adjacent geomagnetic sensors is retrieved, and the parking space geomagnetic sensor that only rises and does not fall in the next time period is found. The position corresponding to the parking space geomagnetic sensor is the parking space of the vehicle, and the license plate number read in step 1 is bound to the parking space. Step 4. After determining the parking space of the vehicle, the data processing system continuously monitors the geomagnetic field of the parking space and monitors the parking time of the vehicle by monitoring the geomagnetic field. When the parking time of the vehicle exceeds the alarm time limit, an alarm is triggered. Once the vehicle has left the parking space normally, the binding relationship between the parking space and the license plate will be removed.

[0006] Preferably, step 2 further includes determining the vehicle speed. When the current vehicle speed is lower than a set speed threshold, the similarity of the two curves is determined using the following method: The geomagnetic induction curves are preprocessed, and the geomagnetic induction curves (VE) of all adjacent geomagnetic sensors of the current lane are read for the current time period. j j=1,2,…,H; H is the total number of adjacent geomagnetic sensors, and the subscript j indicates different adjacent geomagnetic sensors; The adjacent geomagnetic induction curves VE j The current time period is divided into L points, and the ordinate value of each point is VE. j1 VE j2 ,…VE jL , Calculate the adjacent sensor perturbation sequence VE k1 VE k2 ,…VE kL ; Among them VE ki =(VE 1i +VE 2i +…+VE Hi ) / H,i=1,2,…L; Divide the current lane's geomagnetic sensor geomagnetic induction curve VE0 into L points within the current time period, with each point having a ordinate value of VE. 01 VE 02 ,…VE 0L , The current lane geomagnetic sensor geomagnetic induction curve VE0 is corrected, and the corrected ordinate values ​​are: VES 01 VES 02 ,…VES 0L ; VES 0i =VE 0i - VEki i = 1, 2, ..., L; VES 01 VES 02 ,…VES 0L The corrected geomagnetic induction curve VES was fitted. After both geomagnetic induction curves have undergone the above corrections, the corrected geomagnetic induction curve VES or the corrected sequence point set (VES) is used. 01 VES 02 ,…VES 0L Perform similarity judgment.

[0007] Preferably, the method for calculating vehicle speed is: V=S / ΔT, where S is the distance between the geomagnetic sensors corresponding to the two curves, and ΔT is the time difference between the rising edges of the two curves.

[0008] Preferably, in step 3, if no geomagnetic induction curve meeting the similarity standard appears, the following steps are performed: Retrieve the geomagnetic induction curves that meet the trigger conditions from all adjacent lane geomagnetic sensors, as well as the current geomagnetic induction curve; If the Y-axis induction curve of each of the above geomagnetic induction curves is lower than the preset Y-axis induction threshold, the X-axis and Z-axis induction curves corresponding to the Y-axis induction curve are compared with the X-axis and Z-axis induction curves of another geomagnetic induction curve to determine whether they are similar, and then proceed to step 2. If no Y-axis sensing curve is lower than the preset Y-axis sensing threshold, it is considered that no geomagnetic sensing curve conforming to similar standards has appeared.

[0009] Preferably, step 2 also includes screening for false detections of the magnetic induction curves of the lane magnetic sensor, specifically: Step 21. If two adjacent lane geomagnetic sensors meet the matching trigger conditions within the same time period, it is initially judged as a suspected false detection. Assuming the geomagnetic induction curves of two adjacent lane geomagnetic sensors are VE1 and VE2, the total magnetic field disturbance sequence of curve VE1 is... X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; Total magnetic field perturbation sequence of curve VE2 X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; For curve VE1, calculate the parameter vector. : , Thr is the set trigger threshold, and P is the pulse width. , , in For magnetic field perturbation sequence The slope of the sliding window at point L is calculated using the following formula: ; And calculate the parameter vector of curve VE2 using the same method. ; Calculate using fuzzy membership function and eigenvectors These are the first similarity coefficient, the second similarity coefficient, and the third similarity coefficient, respectively. ; Where En_max is =max(p 11 p 21 C1 and C2 are the first and second membership parameters, respectively. The slope kk of the first similarity and the intercept b of the first similarity are respectively: ; The formula for calculating the second similarity comparison parameter EnV is as follows: ; The formula for calculating the third similarity comparison parameter, meankV, is as follows: ; Compare and Similarity F: Set a difference threshold. If the similarity F is greater than the difference threshold, the geomagnetic sensor that meets the trigger condition later is considered a false detection; otherwise, it is determined that there are vehicles driving near the geomagnetic sensors of both lanes.

[0010] Preferably, the false detection judgment of the geomagnetic sensors of two adjacent lanes also includes time interval detection. If the time interval between the two geomagnetic sensors of two adjacent lanes that meet the trigger condition is less than the set hysteresis time length, then the false detection judgment is performed.

[0011] Preferably, the specific method for determining similarity in step 2 is as follows: Calculate the waveform similarity parameter NCC for the two curves.

[0012] N represents N points uniformly distributed along the time axis on both curves, covering the entire time period, x n y n These are the ordinate values ​​corresponding to the nth point; Set an NCC threshold. When the calculated NCC value is greater than the NCC threshold, it means that the two are similar. The closer the NCC value is to 1, the higher the similarity.

[0013] Preferably, the specific method for determining similarity in step 2 is as follows: Calculate the average amplitude and pulse width of the two geomagnetic induction curves, and compare the ratio of the average amplitude and pulse width of the two curves respectively. If both ratios are within the set threshold range, the two curves are considered similar; otherwise, they are considered dissimilar. The closer the ratio is to 1, the higher the similarity.

[0014] A service area vehicle parking status monitoring system includes an entrance license plate camera installed at the service area vehicle entrance, an entrance geomagnetic sensor installed at the entrance license plate camera, lane geomagnetic sensors continuously distributed along the service area lanes, parking space geomagnetic sensors installed at each parking space in the service area, and a data processing system connected to the entrance license plate camera, lane geomagnetic sensors, and parking space geomagnetic sensors. The data processing system can execute the monitoring method described above.

[0015] Preferably, the lane geomagnetic sensor is a MEMS geomagnetic sensor located beside the lane and above the ground, and the parking space geomagnetic sensor is an underground geomagnetic induction coil.

[0016] The service area vehicle parking status monitoring system and method described in this invention tracks vehicle trajectories and monitors parking duration through geomagnetic induction. This helps service area managers to promptly identify vehicles that have been parked for too long and provides the specific parking location of the vehicles. This facilitates timely detection and handling by service area managers, improves the informatization level of vehicle parking monitoring in service areas, and can reduce the probability of unexpected events. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a specific implementation of the service area vehicle parking status monitoring system described in this invention; Figure 2 This is a schematic diagram of a specific implementation of the service area vehicle parking status monitoring method of the present invention; Figure 3 A schematic diagram of the geomagnetic induction curve when a vehicle passes a geomagnetic sensor; Figure 4 A schematic diagram of the geomagnetic induction curve when a vehicle is moving and stopping in a parking space; Figure 5 This is a schematic diagram of the sensing threshold described in this invention; Figure 6 This is a schematic diagram of vehicle driving according to a specific embodiment of the present invention; Figure 7 for Figure 6 Schematic diagram of X, Y, and Z axis sensing curves corresponding to a specific embodiment; Figure 3 , Figure 4 , Figure 5 , Figure 7 The horizontal axis represents time, and the vertical axis represents the change in geomagnetic field.

[0018] The attached diagram is labeled as follows: 1-Entrance license plate camera, 2-Parking space geomagnetic sensor. Detailed Implementation

[0019] The following is in conjunction with the appendix Figures 1 to 3 The specific embodiments of the present invention will be further described in detail below.

[0020] The service area vehicle parking status monitoring system of the present invention includes an entrance license plate camera installed at the vehicle entrance of the service area, an entrance geomagnetic sensor installed at the entrance license plate camera, lane geomagnetic sensors continuously distributed on the lanes of the service area, parking space geomagnetic sensors installed in each parking space of the service area, and a data processing system connected to the entrance license plate camera, lane geomagnetic sensors and parking space geomagnetic sensors.

[0021] The service area vehicle parking status monitoring method of the present invention, such as Figure 2 As shown, it includes the following steps: Step 1. Read the license plate information of vehicles entering the service area, and read the geomagnetic induction curve formed when the vehicle enters the entrance by the entrance geomagnetic sensor set on the entrance road of the service area. The data processing system binds the license plate information with the formed geomagnetic induction curve. Step 2. After the vehicle enters the service area through the service area entrance, the data processing system reads the geomagnetic induction curve generated by the geomagnetic sensor of each lane the vehicle passes through during its journey, and retrieves the geomagnetic induction curves of all adjacent lane geomagnetic sensors in the previous time period, judges whether they are similar one by one, and binds the geomagnetic induction curve that meets the similarity standard and has the highest similarity to the currently read geomagnetic induction curve. Step 3. Repeat step 2 until no geomagnetic induction curves that meet similar criteria appear; The geomagnetic sensor corresponding to the last similar geomagnetic induction curve is considered to be the position before parking. The parking space geomagnetic sensor curve of the adjacent geomagnetic sensors is retrieved, and the parking space geomagnetic sensor that only rises and does not fall in the next time period is found. The position corresponding to the parking space geomagnetic sensor is the parking space of the vehicle, and the license plate number read in step 1 is bound to the parking space. Step 4. After determining the parking space of the vehicle, the data processing system continuously monitors the geomagnetic field of the parking space and monitors the parking time of the vehicle by monitoring the geomagnetic field. When the parking time of the vehicle exceeds the alarm time limit, an alarm is triggered. Once the vehicle has left the parking space normally, the binding relationship between the parking space and the license plate will be removed.

[0022] When a vehicle enters the service area, the license plate is read by the entrance license plate camera, and the geomagnetic induction curve formed when the vehicle enters the entrance is read by the entrance geomagnetic sensor on the road at the entrance of the service area. The data processing system binds the license plate information with the geomagnetic induction curve formed here.

[0023] The nearest few geomagnetic sensors connected to a road and to a given geomagnetic sensor A are defined as the adjacent geomagnetic sensors of lane geomagnetic sensor A. These adjacent geomagnetic sensors can be lane geomagnetic sensors or parking space geomagnetic sensors, such as... Figure 1 As shown, there are two lanes. The upper lane has magnetic sensors on one side for lanes 11, 12, 13, and 14, and the lower lane has magnetic sensors on one side for lanes 21, 22, 23, and 24. The upper lane is adjacent to parking spaces 31 and 32, and the lower lane is adjacent to parking spaces 33 and 34. For the entrance of the two lanes, magnetic sensors 11 and 12 are entrance magnetic sensors. The magnetic sensors adjacent to the magnetic sensor in lane 13 include the magnetic sensors in lanes 12, 14, 22, 23, and 24, as well as the magnetic sensor for parking space 31.

[0024] The parking space geomagnetic sensor can be a buried geomagnetic induction coil that is not affected by adjacent geomagnetic sensors, while the lane geomagnetic sensor is a small MEMS geomagnetic sensor that is located beside the lane and above the ground and is easy to install and maintain.

[0025] When a vehicle enters the service area from the lane where the No. 11 geomagnetic sensor is located, the license plate is read by the entrance license plate camera, and the No. 11 geomagnetic sensor reads the geomagnetic curve formed by the vehicle's movement and binds it to the license plate information.

[0026] Assuming the time it takes for a vehicle to pass a lane's geomagnetic sensor is considered a time period, as the vehicle travels, the data processing system reads the geomagnetic curve from the previous time period and compares it with the geomagnetic sensors in adjacent lanes that monitored the geomagnetic change curve in the previous time period. If a geomagnetic change occurs in an adjacent lane's geomagnetic sensor in the current time period, the geomagnetic curve of the adjacent lane's geomagnetic sensor is compared with the geomagnetic curve from the previous time period. If the similarity criteria are met, the geomagnetic curve is then bound to the geomagnetic curve from the previous time period.

[0027] By binding each geomagnetic curve individually and connecting the corresponding lane geomagnetic sensors, the vehicle's driving trajectory can be obtained.

[0028] The length of the time period depends on the spacing of the geomagnetic sensors and the vehicle's speed. The spacing is generally 5-7 meters. When the vehicle first enters the service area, the speed is relatively fast, generally 4-6 meters per second. After entering the road with parking areas, the speed is reduced to 1-3 meters per second. The initial time period is shorter and the subsequent time period is longer.

[0029] The similarity judgment of the geomagnetic sensor curve can also be determined based on whether the trigger condition is met. A typical implementation is that the trigger condition for reading and comparing the geomagnetic induction curve is that the amount of geomagnetic change sensed by the geomagnetic sensor exceeds a preset sensing threshold. The sensing threshold includes a time threshold and a numerical threshold. If the time when the sensing curve detected by the geomagnetic sensor is higher than the numerical threshold is greater than the time threshold, it is considered to meet the trigger condition, that is, it is considered that a vehicle may pass through at this time.

[0030] like Figure 5 As shown, the trigger condition refers to the sensing change in the geomagnetic field exceeding the sensing threshold. The sensing threshold includes a time threshold VT and a numerical threshold VM. If the time at which the geomagnetic sensor detects a change higher than the numerical threshold VM is greater than the time threshold VT, then the sensing threshold is considered to have been exceeded.

[0031] Subsequently, in each time period, the data processing system monitors and compares all adjacent geomagnetic sensors of the geomagnetic sensor that showed a geomagnetic change curve in the previous time period to see if they meet the similarity criteria, and binds the geomagnetic curves that meet the similarity criteria.

[0032] The geomagnetic induction curve refers to the change in induced data caused by the change in geomagnetic induction when a vehicle passes a geomagnetic sensor. The induction curve is trapezoidal or has a sharp peak shape, which first rises and then falls. Figure 3 As shown; when the vehicle approaches the geomagnetic sensor and comes to a complete stop, the sensing curve rises but does not fall, instead remaining at a high level, as... Figure 4 As shown.

[0033] The changes in geomagnetic induction include induction curves in three directions: X, Y, and Z. The X-axis represents the vehicle's direction of travel, the Y-axis is parallel to the road surface and perpendicular to the X-axis, and the Z-axis is perpendicular to the road surface. The total changes in geomagnetic induction... ; The geomagnetic induction curve is usually defined as the curve showing the total change in geomagnetic induction T over time.

[0034] For vehicles traveling within a service area, the initial speed is typically higher when entering the open area, and lower when entering the parking area with parking spaces on both sides of the road. In the open area, there are few vehicles, but in the parking area, assuming the lane magnetic sensor is a MEMS magnetic sensor positioned beside the lane and above the ground, the parking area usually has many densely parked vehicles, and other vehicles are also traveling at low speeds on the road within the parking area, causing the distance between vehicles to shorten. The MEMS magnetic sensor is easily interfered with by the magnetic curves of nearby vehicles and parked vehicles. To overcome the interference from the parking area, step 2 also includes judging the vehicle's speed. For example, a speed threshold of 10 kilometers per hour is set. When the current vehicle speed is lower than the set speed threshold, it is considered that the vehicle is about to enter or has already entered the parking area. The similarity judgment of the two curves at this time uses the following method: The geomagnetic induction curves are preprocessed, and the geomagnetic induction curves (VE) of all adjacent geomagnetic sensors of the current lane are read for the current time period. j j=1,2,…,H; H is the total number of adjacent geomagnetic sensors, and the subscript j indicates different adjacent geomagnetic sensors; The adjacent geomagnetic induction curves VE j The current time period is divided into L points, and the ordinate value of each point is VE. j1 VE j2 ,…VE jL , Calculate the adjacent sensor perturbation sequence VE k1 VE k2 ,…VE kL ; Among them VE ki = (VE 1i +VE 2i +…+VE Hi ) / H,i=1,2,…L; Divide the current lane's geomagnetic sensor geomagnetic induction curve VE0 into L points within the current time period, with each point having a ordinate value of VE. 01 VE 02 ,…VE 0L , The current lane geomagnetic sensor geomagnetic induction curve VE0 is corrected, and the corrected ordinate values ​​are: VES 01 VES 02 ,…VES 0L ; VES 0i =VE 0i - VE ki i = 1, 2, ..., L; VES 01 VES 02 ,…VES 0L Fit the corrected curve VES; By using the geomagnetic induction curve of a nearby geomagnetic sensor to correct the current geomagnetic sensor's induction curve using the above method, interference from adjacent geomagnetic induction curves can be effectively eliminated.

[0035] In subsequent similarity assessments, the corrected curve VES or the corrected sequence point set before fitting (VES) is used. 01 VES 02 ,…VES 0L To perform similarity judgment, when using the modified sequence point set for similarity judgment, the geomagnetic induction curves to be compared can be divided into L points in the current time period in the same way.

[0036] The inventors discovered that, when using a MEMS geomagnetic sensor positioned beside the lane and above the ground, when two vehicles travel in the same direction past both sides of the same lane's geomagnetic sensor, approximately equal in magnitude but opposite in direction geomagnetic changes occur in the Y-axis direction. The Y-axis direction refers to the direction perpendicular to the vehicle's travel direction. Figure 6 As shown, vehicles A and B simultaneously travel in the same direction past the MEMS geomagnetic sensor from both sides of the sensor, generating geomagnetic induction curves as follows. Figure 7 As shown, since the magnetic fields in the Y-axis direction are equal in magnitude but opposite in direction, they will be severely canceled out. Although the magnetic fields in the X-axis and Z-axis will be superimposed, they are not enough to offset the loss in the Y-axis direction. As a result, the overall geomagnetic induction curve after synthesis is severely distorted due to interference from adjacent vehicles and cannot be considered similar.

[0037] For ordinary roads, such as Figure 1 As shown, the MEMS geomagnetic sensors are installed on both sides of the road, preventing vehicles from traveling on both sides of the sensor. However, since service areas are generally paved plazas with only road markings, vehicles may still pass by on the outer side of the road. Therefore, it is possible for vehicles to pass by on either side of the sensor. Figure 6 The scenario shown depicts two different vehicles passing by the sensors on either side, causing a misjudgment.

[0038] To avoid misjudgment, when no geomagnetic induction curve meeting similar criteria appears, the following steps should be taken: The system retrieves all geomagnetic induction curves from adjacent lanes that meet the trigger conditions and detects the Y-axis induction curves. If a Y-axis induction curve is lower than a preset Y-axis induction threshold, but the curve still meets the trigger conditions, it indicates a significant change in the geomagnetic induction curve, suggesting a vehicle has passed. In this case, a trigger is determined. Figure 6 In the scenario where a vehicle simultaneously passes by the sensor from the other side in the same direction, the overall geomagnetic induction curve meets the triggering condition. However, the change in the Y-axis direction is canceled out, causing curve distortion. In this case, the overall geomagnetic induction curve is no longer used as the comparison object. Instead, the X-axis and Z-axis induction curves corresponding to the Y-axis induction curve are compared with the X-axis and Z-axis induction curves of another geomagnetic induction curve to determine their similarity. The sum of the similarity of the X-axis and Z-axis geomagnetic induction curves is used as the similarity ranking standard, and step 2 is continued. This avoids misjudgment and loss of target when vehicles pass by on both sides of the sensor.

[0039] Therefore, the data processing system compares the similarity of the geomagnetic induction curves of adjacent geomagnetic sensors within a continuous time period, binds each pair of similar geomagnetic induction curves, and thus obtains the vehicle's driving trajectory. Finally, when no similar curves can be found, the geomagnetic sensor corresponding to the last similar geomagnetic induction curve is considered to be the position of the vehicle before it stopped. The system then retrieves the parking space geomagnetic sensor curve from the adjacent geomagnetic sensors of that geomagnetic sensor, finds the parking space geomagnetic sensor that only rises and does not fall in the next time period, and the position corresponding to the parking space geomagnetic sensor is the vehicle's parking space. The initial license plate number is then bound to that parking space.

[0040] Due to variations in vehicle speed, the similarity of the geomagnetic induction curves from two adjacent geomagnetic sensors over a given time period is determined. This similarity assessment calculates the waveform similarity parameter NCC between the two curves.

[0041] N represents N points uniformly distributed along the time axis on both curves, covering the entire time period, x n y n These are the ordinate values ​​corresponding to the nth point.

[0042] Set an NCC threshold. The closer the NCC value is to 1, the more similar the two curves are. For example, you can set the NCC threshold to 0.5. When the calculated NCC value is greater than 0.5, it means that the two are similar.

[0043] Similarity can also be determined in the following ways: Calculate the average amplitude and pulse width of the two geomagnetic induction curves, and compare their ratios. Since both ratios are within the threshold range, it indicates that the two curves are similar.

[0044] The horizontal axis of the geomagnetic induction curve represents time, and the vertical axis represents the geomagnetic induction quantity. The average amplitude is the average value of the vertical axis of the geomagnetic induction curve over a complete time period, and the pulse width is the time when the vertical axis value of the geomagnetic induction curve is greater than the average amplitude. For example, if the thresholds for the average amplitude ratio and pulse width ratio are both set within the range of 0.7 to 1 / 0.7 = 1.42, the average amplitude and pulse width of the vertical axis of the first geomagnetic induction curve over the complete time period are 120 and 0.40 milliseconds, respectively, and the average amplitude and pulse width of the vertical axis of the second geomagnetic induction curve over the complete time period are 145 and 0.44 milliseconds, respectively. Then, 120 / 145 = 0.827 and 0.40 / 0.44 = 0.909, both of which are within the threshold range, and the two geomagnetic induction curves are considered similar.

[0045] After the parking space of the vehicle is determined, the data processing system continuously monitors the geomagnetic field of the parking space. When the vehicle does not leave the parking space, the geomagnetic field of the parking space remains at a high value. By monitoring the geomagnetic field, the parking time of the vehicle can be monitored. When it is found that the parking time of the vehicle is too long, such as more than 12 hours or 24 hours, an alarm mechanism can be triggered to remind staff to check.

[0046] Once the vehicle has driven out normally, the geomagnetic field decreases, thus unbinding the parking space from the registered license plate.

[0047] for Figure 1 The service area shown commonly features two lanes or two lanes with parking spaces on both sides. It's possible that the movement of one vehicle could trigger the geomagnetic sensors in adjacent lanes, leading to false detections. To avoid false detections, a false detection screening process can be implemented, specifically: When the current lane's geomagnetic sensor detects the geomagnetic induction curve VE1, after a set lag time of 450ms, it communicates with the geomagnetic sensor in the adjacent lane. If the geomagnetic sensor in the adjacent lane also has a geomagnetic induction curve VE2 within these 450ms, and both curves VE1 and VE2 meet the triggering conditions for subsequent geomagnetic induction curve comparison, then it is initially judged as a suspected false detection.

[0048] Further investigation was then conducted to determine whether the suspected false alarm was a genuine false alarm or a case of vehicles parked side-by-side.

[0049] Assuming the geomagnetic induction curves of two adjacent lane geomagnetic sensors are VE1 and VE2, the total magnetic field disturbance sequence of curve VE1 is... X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; Total magnetic field perturbation sequence of curve VE2 X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; The disturbance sequence refers to dividing the curve into n uniformly distributed points along the time axis, with the ordinate of each point forming a sequence. In this invention, the geomagnetic induction curve is actually obtained by synthesizing the geomagnetic induction curves in the X, Y, and Z directions. The X-axis magnetic field disturbance sequence... Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; are the sequences in the X, Y, and Z directions before synthesis, respectively.

[0050] Calculate the parameter vector based on each perturbation sequence. : , Thr is the set trigger threshold used to determine if a vehicle has passed by. If the disturbing magnetic field strength is greater than... Thr If P is the pulse width, then it is assumed that a vehicle has passed by. , , in For magnetic field perturbation sequence The slope of the sliding window at point L is calculated using the following formula: ; And calculate the parameter vector of curve VE2 using the same method. .

[0051] Calculate using fuzzy membership function and eigenvectors These are the first similarity coefficient, the second similarity coefficient, and the third similarity coefficient, respectively. ; Where En_max is =max(p 11 p 21 C1 and C2 are the first and second membership parameters, respectively; a specific value is C1=384, C2=315.

[0052] The slope kk of the first similarity and the intercept b of the first similarity are respectively: ; The formula for calculating the second similarity comparison parameter EnV is as follows: ; The formula for calculating the third similarity comparison parameter, meankV, is as follows: ; Compare and Similarity F: ; For the same vehicle, the waveform characteristics detected by geomagnetic sensors A and B should be similar. By looking for the similarity between the characteristics, it can be determined whether the two sensors have detected the same vehicle.

[0053] To evaluate the relevance of features, the above method uses a fuzzy membership function for calculation. F1 is used to evaluate feature P. 11 and feature P 21 The degree of correlation, F2 is used to evaluate feature P 13 and feature P 23 The degree of correlation, F3 is used to evaluate feature P 14 and feature P 24 The degree of correlation, their values ​​are all between 0 and 1, and the larger the value, the more correlated they are.

[0054] Finally, the average correlation is obtained by adding F1, F2, and F3 together and dividing by 3. This average correlation is then compared with the set difference threshold. If the difference threshold is greater than the threshold, the waveform characteristics obtained by the two sensors are considered to be similar, indicating that one sensor is a false detection. Otherwise, it is not a false detection, but rather a case of vehicles side by side.

[0055] The difference threshold can be set to 0.58. If F is greater than 0.58, it is a true false detection, and the geomagnetic sensor in the adjacent lane is considered to be a vehicle in the adjacent lane rather than a vehicle in the current lane; otherwise, it is judged to be two vehicles driving side by side.

[0056] After applying the above false detection filtering method, vehicle inspection data was collected over two hours within the two lanes of the service area. The original vehicle inspection results were compared with the vehicle inspection results obtained by applying the false detection algorithm, as shown below: Comparison table of original vehicle inspection results and vehicle inspection results using the false detection algorithm

[0057] It can be seen that this can reduce vehicle false detections by 75.64% and improve vehicle inspection accuracy to 97.97%.

[0058] The service area vehicle parking status monitoring system and method described in this invention tracks vehicle trajectories and monitors parking duration through geomagnetic induction. This helps service area managers to promptly identify vehicles that have been parked for too long and provides the specific parking location of the vehicles. This facilitates timely detection and handling by service area managers, improves the informatization level of vehicle parking monitoring in service areas, and can reduce the probability of unexpected events.

[0059] The foregoing descriptions are preferred embodiments of the present invention. Unless there is a clear contradiction between the preferred embodiments or a prerequisite for a particular preferred embodiment, the preferred embodiments can be arbitrarily combined and used. The embodiments and specific parameters described are only for clearly illustrating the inventor's invention verification process and are not intended to limit the scope of patent protection of the present invention. The scope of patent protection of the present invention shall still be determined by its claims. Similarly, any equivalent structural changes made based on the description and drawings of the present invention shall also be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the parking status of vehicles in a service area, characterized in that, Includes the following steps: Step 1. Read the license plate information of vehicles entering the service area, and read the geomagnetic induction curve formed when the vehicle enters the entrance by the entrance geomagnetic sensor set on the entrance road of the service area. The data processing system binds the license plate information with the formed geomagnetic induction curve. Step 2. For each lane's geomagnetic sensor, the data processing system reads the geomagnetic induction curve formed in real time and determines whether it meets the triggering conditions. If it meets the triggering conditions, it retrieves the geomagnetic induction curves of all adjacent lanes' geomagnetic sensors in the previous time period, and determines whether they are similar to the current lane's geomagnetic induction curve. The geomagnetic induction curve that meets the similarity criteria and has the highest similarity is then bound to the currently read geomagnetic induction curve. Step 3. Repeat step 2 until no geomagnetic induction curves that meet similar criteria appear; The geomagnetic sensor corresponding to the last similar geomagnetic induction curve is considered to be the position before parking. The parking space geomagnetic sensor curve of the adjacent geomagnetic sensors is retrieved, and the parking space geomagnetic sensor that only rises and does not fall in the next time period is found. The position corresponding to the parking space geomagnetic sensor is the parking space of the vehicle, and the license plate number read in step 1 is bound to the parking space. Step 4. After determining the parking space of the vehicle, the data processing system continuously monitors the geomagnetic field of the parking space and monitors the parking time of the vehicle by monitoring the geomagnetic field. When the parking time of the vehicle exceeds the alarm time limit, an alarm is triggered. Once the vehicle has left the parking space normally, the binding relationship between the parking space and the license plate will be removed. Step 2 also includes screening for false detections of the magnetic induction curves of the lane magnetic sensors, specifically: Step 21. If two adjacent lane geomagnetic sensors meet the matching trigger conditions within the same time period, it is initially judged as a suspected false detection. Assuming the geomagnetic induction curves of two adjacent lane geomagnetic sensors are VE1 and VE2, the total magnetic field disturbance sequence of curve VE1 is... X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; Total magnetic field perturbation sequence of curve VE2 X-axis magnetic field perturbation sequence Y-axis magnetic field perturbation sequence Z-axis magnetic field perturbation sequence ; For curve VE1, calculate the parameter vector. : ， Thr is the set trigger threshold, and P is the pulse width. , ， in For magnetic field perturbation sequence The slope of the sliding window at point L is calculated using the following formula: ； And calculate the parameter vector of curve VE2 using the same method. ; Calculate using fuzzy membership function and eigenvectors These are the first similarity coefficient, the second similarity coefficient, and the third similarity coefficient, respectively. ； Where En_max is =max(p 11 p 21 C1 and C2 are the first and second membership parameters, respectively. The slope kk of the first similarity and the intercept b of the first similarity are respectively: ； The formula for calculating the second similarity comparison parameter EnV is as follows: ； The formula for calculating the third similarity comparison parameter, meankV, is as follows: ； Compare and Similarity F: ； If the similarity F is greater than the difference threshold, the geomagnetic sensor that meets the trigger condition later is considered a false detection; otherwise, it is determined that there are vehicles driving near the geomagnetic sensors of both lanes.

2. The service area vehicle parking status monitoring method as described in claim 1, characterized in that, Step 2 also includes determining the vehicle's speed. When the current vehicle speed is lower than a set speed threshold, the similarity of the two curves is determined using the following method: The geomagnetic induction curves are preprocessed, and the geomagnetic induction curves (VE) of all adjacent geomagnetic sensors of the current lane are read for the current time period. j j=1,2,…,H; H is the total number of adjacent geomagnetic sensors, and the subscript j indicates different adjacent geomagnetic sensors; The adjacent geomagnetic induction curves VE j The current time period is divided into L points, and the ordinate value of each point is VE. j1 ,VE j2 ,…VE jL , Calculate the adjacent sensor perturbation sequence VE k1 VE k2 ,…VE kL ; where VE ki =(VE 1i +VE 2i +…+VE Hi ) / H, i=1,2,…L; Divide the current lane's geomagnetic sensor geomagnetic induction curve VE0 into L points within the current time period, with each point having a ordinate value of VE. 01 VE 02 ,…VE 0L , The current lane geomagnetic sensor geomagnetic induction curve VE0 is corrected, and the corrected ordinate values ​​are: VES 01 VES 02 ,…VES 0L ; VES 0i =VE 0i - VE ki i = 1, 2, ..., L; VES 01 VES 02 ,…VES 0L The corrected geomagnetic induction curve VES was fitted. After both geomagnetic induction curves have undergone the above corrections, the corrected geomagnetic induction curve VES or the corrected sequence point set (VES) is used. 01 VES 02 ,…VES 0L Perform similarity judgment.

3. The service area vehicle parking status monitoring method as described in claim 2, characterized in that, The method for calculating vehicle speed is: V=S / ΔT, where S is the distance between the geomagnetic sensors corresponding to the two curves, and ΔT is the time difference between the rising edges of the two curves.

4. The service area vehicle parking status monitoring method as described in claim 1, characterized in that, If no geomagnetic induction curve meeting the similarity criteria appears in step 3, the following steps shall be performed: Retrieve the geomagnetic induction curves that meet the trigger conditions from all adjacent lane geomagnetic sensors, as well as the current geomagnetic induction curve; Detect the Y-axis induction curve among the above geomagnetic induction curves. If a certain Y-axis induction curve is lower than the preset Y-axis induction threshold, then determine whether the X-axis and Z-axis induction curves corresponding to the Y-axis induction curve are similar to the X-axis and Z-axis induction curves of another geomagnetic induction curve, and continue to step 2. If no Y-axis sensing curve is lower than the preset Y-axis sensing threshold, it is considered that no geomagnetic sensing curve conforming to similar standards has appeared.

5. The service area vehicle parking status monitoring method as described in claim 1, characterized in that, The false detection judgment for two adjacent lane geomagnetic sensors also includes time interval detection. If the time interval between two adjacent lane geomagnetic sensors that meet the trigger condition is less than the set hysteresis time length, then a false detection judgment is made.

6. The service area vehicle parking status monitoring method as described in claim 1, characterized in that, The specific method for determining similarity in step 2 is as follows: Calculate the waveform similarity parameter NCC for the two curves. ； N represents N points uniformly distributed along the time axis on both curves, covering the entire time period, x n y n These are the ordinate values ​​corresponding to the nth point; Set an NCC threshold. When the calculated NCC value is greater than the NCC threshold, it means that the two are similar. The closer the NCC value is to 1, the higher the similarity.

7. The service area vehicle parking status monitoring method as described in claim 1, characterized in that, The specific method for determining similarity in step 2 is as follows: Calculate the average amplitude and pulse width of the two geomagnetic induction curves, obtain the ratio of the average amplitude of the two curves, and obtain the ratio of the pulse width of the two curves. If both ratios are within the set threshold range, it means that the two curves are similar; otherwise, they are considered dissimilar. The closer the ratio is to 1, the higher the similarity.

8. A service area vehicle parking status monitoring system, characterized in that, The system includes an entrance license plate camera installed at the vehicle entrance of the service area, an entrance geomagnetic sensor installed at the entrance license plate camera, lane geomagnetic sensors continuously distributed along the lanes of the service area, parking space geomagnetic sensors installed at each parking space in the service area, and a data processing system connected to the entrance license plate camera, lane geomagnetic sensors, and parking space geomagnetic sensors. The data processing system can execute the monitoring method as described in any one of claims 1 to 7.

9. The service area vehicle parking status monitoring system as described in claim 8, characterized in that, The lane geomagnetic sensor is a MEMS geomagnetic sensor located beside the lane and above the ground, while the parking space geomagnetic sensor is an underground geomagnetic induction coil.

Images