A municipal road roadside parking charging statistics method

Abstract

This invention relates to the field of smart parking technology and provides a method for calculating and counting parking fees on municipal roads, as follows: A license plate recognition camera and a video camera are installed above the roadside parking spaces; the roadside parking spaces are divided into multiple groups in an adjacent order; the video cameras capture video images, and trigger zones are set. When the trigger zone is triggered and the vehicle is identified as entering a parking space, the license plate recognition camera captures an image of the vehicle; after recognizing the license plate number, once the vehicle has fully entered the parking space group, the license plate number is bound to the corresponding parking space group; when the trigger zone is triggered and the vehicle is identified as leaving a parking space, the vehicle is captured, and after recognizing the license plate number, the license plate number is unbound from the parking space group; parking fees are calculated for the current vehicle based on its entry and exit times. This roadside parking fee calculation method requires no manual intervention and can complete the parking fee monitoring of an entire row of roadside parking spaces with just one or a few groups, significantly reducing labor costs and improving fee calculation efficiency.

Description

Technical Field

[0001] This invention belongs to the field of smart parking technology, and in particular relates to a method for calculating and counting roadside parking fees on municipal roads. Background Technology

[0002] With the rise of cutting-edge technologies such as artificial intelligence, the Internet of Things, big data, cloud computing, fintech, and contactless payment, the transformation of parking lots into smart parking facilities has accelerated, and the deployment of smart parking lots has achieved varying degrees of success. Unmanned parking lots are widely used; license plates are automatically recognized when drivers enter, and payment is made by scanning a code when exiting, significantly reducing labor costs. However, for roadside parking on municipal roads, the open nature of these spaces presents challenges for smart parking billing. To meet the parking needs of citizens, many cities have opened up a large number of inconvenient roadside parking spaces, implementing smart parking payment functions through the transformation of these spaces.

[0003] Initially, video-based parking sensors were used. These were vertical poles installed near parking spaces, similar to the slow-charging stations near existing roadside parking spaces for new energy vehicles. They used cameras on the sensors to automatically acquire vehicle parking information, automatically identify license plate numbers, and transmit the data to a central system. However, video-based parking sensors had several drawbacks, including encroaching on municipal space, obstructing pedestrian traffic, disrupting municipal road planning, and incurring high construction costs. Coupled with factors such as installation and operating costs, susceptibility to vandalism by third parties, and technological maturity, many places ultimately abandoned the video-based parking sensor solution and opted for the more mature, simpler to install, and cheaper geomagnetic technology.

[0004] Geomagnetic technology has been used for many years and is unaffected by weather, temperature, or other environmental factors. It primarily relies on precise radar to detect vehicles. Geomagnetic sensors have a long battery life, and their single detection function reduces maintenance costs. By excavating and installing geomagnetic sensors at parking spaces, data on whether a car is parked in the space is transmitted via a repeater to a backend computer, which then feeds this information back to the terminal. While geomagnetic technology is currently the most widely used technology for roadside parking, its drawbacks are also significant. It can only track the entry and exit times of vehicles in parking spaces, not the license plate numbers. Therefore, dedicated road toll collectors are needed to manually input license plate information for each parking space using handheld terminals, which obviously increases labor costs. Another approach uses shoulder-mounted sensors, which are vehicle recognition devices installed on the shoulder of the parking space. These devices have built-in batteries and solar panels, with one device for each parking space. However, these also have significant problems, such as being easily damaged by pedestrians, and the high cost of individual devices. Some road shoulder sensing devices use cameras to identify vehicles, but the cameras are easily contaminated by the accumulation of dust on the road, so their application is relatively limited.

[0005] One foreseeable and feasible solution is to directly use surveillance cameras installed on high-mounted crossbars to identify vehicles and record their information. While these cameras have a long shooting distance, even reaching 200 meters, they only have license plate recognition capabilities and lack complex computing power. If the vehicle is too far away, they cannot accurately and quickly trigger license plate recognition, nor can they pinpoint the vehicle's current parking space. Furthermore, there is a significant problem: if the car is far away, and the front and rear parking spaces are occupied while the middle space is empty, the surveillance camera may not be able to identify the empty middle space due to obstructions.

[0006] Currently, the technology of fusing LiDAR and cameras to track people or vehicles is relatively mature. However, applying this technology to roadside parking also presents some challenges: LiDAR hardware and software development costs are high; LiDAR is prone to errors when tracking vehicles at long distances; and the presence of heavy traffic and interference from non-parked vehicles makes it impossible to achieve constant online vehicle tracking. Many problems still need to be solved. Summary of the Invention

[0007] In view of the above problems, the purpose of this invention is to provide a method for calculating and counting roadside parking fees on municipal roads, in order to solve the above-mentioned technical problems.

[0008] The present invention adopts the following technical solution:

[0009] The method for calculating and counting parking fees on municipal roads includes the following steps:

[0010] Step S1: Install a recognition mechanism on the high horizontal bar above the roadside parking space. The recognition mechanism includes a license plate recognition camera and a video camera, both of which are set in the same direction.

[0011] Step S2: Divide the roadside parking spaces into multiple groups according to their adjacent order to form parking space groups, each group having a number of parking spaces;

[0012] Step S3: The video camera captures video images. First, the outer edge of the parking space group is located. Then, a trigger area is set near the outer edge of the parking space. When the trigger area of ​​each parking space group is triggered and the vehicle is identified as entering the parking space, the license plate recognition camera captures the vehicle.

[0013] Step S4: After identifying the license plate number from the captured image, once the vehicle has fully entered the parking space group, bind the license plate number to the corresponding parking space group and record the vehicle's entry time.

[0014] Step S5: When the trigger area of ​​each parking space group is triggered and it is identified as leaving the parking space, the license plate recognition camera captures the vehicle, identifies the license plate number, unbinds the license plate number from the parking space group, and records the time the vehicle leaves.

[0015] Step S6: Complete the parking fee calculation for the current vehicle based on the vehicle's entry and exit times.

[0016] Furthermore, in step S2, when dividing parking spaces into groups, the parking space groups closer to the identification mechanism have fewer parking spaces, while the parking space groups farther away from the identification mechanism have more parking spaces.

[0017] Furthermore, in step S3, the triggering area consists of two parallel detection lines, namely the first detection line and the second detection line from the outside to the inside. The triggering area is activated and identified as having entered a parking space in the following way:

[0018] When a vehicle target touches the first detection line, the outer rectangle of the vehicle target is identified and tracked.

[0019] When the aspect ratio of the circumscribed rectangle changes significantly, the license plate recognition camera captures the vehicle target for the first time. When the circumscribed rectangle touches the second detection line, the license plate recognition camera captures the vehicle target for the second time.

[0020] Furthermore, each parking space group has a list of license plate information. The specific process of step S4 is as follows:

[0021] The two captured images are used for license plate recognition to obtain the license plate number of the current vehicle target, and the two captured images are saved at the same time.

[0022] When a vehicle leaves the first detection line and then the second detection line, it is determined that the vehicle has entered the parking space. The license plate number is added to the corresponding license plate information list, and the current time is recorded as the vehicle's entry time.

[0023] Furthermore, the specific process of step S5 is as follows:

[0024] When a vehicle in a parking space touches the second detection line, the outermost pixel coordinates of the vehicle are tracked. When the vehicle crosses the first detection line, the license plate recognition camera captures the vehicle for the first time.

[0025] When the vehicle target leaves the second detection line first and then leaves the first detection line, the license plate recognition camera completes a second capture of the vehicle target and determines that the vehicle target has left the parking space;

[0026] The two captured images are used for license plate recognition to obtain the license plate number of the current vehicle target, and the two captured images are saved at the same time.

[0027] Find the corresponding license plate number from the parking space group corresponding to the vehicle target, then remove the license plate number from the license plate information list, and record the current time as the vehicle's departure time.

[0028] Furthermore, if the corresponding license plate number cannot be found in the current parking space group, the corresponding license plate number is searched for in other parking space groups. If found, the license plate number is removed from the recorded license plate information list.

[0029] Furthermore, in step S4, if the vehicle target leaves the first detection line but continues to touch the second detection line, and the distance from the first detection line is greater than k*D, where D is the pixel distance between the two detection lines at the current position of the vehicle target, and k is a scaling factor with a value of 0.80-0.90, then the vehicle target has entered the parking space.

[0030] The beneficial effects of this invention are at least four inventive points:

[0031] First, ordinary video cameras and high-precision license plate recognition cameras are used together. The license plate recognition camera only plays the role of capturing images and does not need to record continuously in real time, which can extend the life of the high-cost license plate recognition camera. Ordinary video cameras are low in cost, can work continuously, have low accuracy requirements, and do not need to recognize license plates, but only need to recognize and track vehicles.

[0032] Secondly, this invention divides roadside vehicles into groups. Due to inaccurate identification of specific parking spaces or obstruction, the specific parking space number cannot be identified. This invention does not need to determine the specific parking space number that the current vehicle is parked in. It only needs to record the parking space group that the vehicle enters. After grouping the parking spaces, the approximate parking range of the vehicle can be known, avoiding errors.

[0033] Third, the present invention has a reasonable setting of the trigger area and the judgment logic for vehicle entry and exit, which can efficiently and quickly trigger the license plate recognition camera to capture and recognize the license plate. Basically, there are at least two images for both entry and exit, which reduces the probability of license plate recognition errors.

[0034] Fourth, the roadside parking billing method of the present invention does not require manual intervention. One set or a few sets are sufficient to complete the parking billing monitoring of the entire roadside parking spaces, which can significantly reduce labor costs and improve billing efficiency. Attached Figure Description

[0035] Figure 1 This is a flowchart of the municipal roadside parking fee calculation and statistics method provided in the embodiments of the present invention;

[0036] Figure 2 This is a diagram showing the positional relationship between parking spaces and recognition mechanisms;

[0037] Figure 3 This is a schematic diagram of the vehicle's circumscribed rectangle. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0039] To illustrate the technical solution described in this invention, specific embodiments are described below.

[0040] like Figure 1 As shown in this embodiment, the method for calculating and counting parking fees on municipal roads includes the following steps:

[0041] Step S1: Install a recognition mechanism on the high horizontal bar above the roadside parking space. The recognition mechanism includes a license plate recognition camera and a video camera, both of which are set in the same direction.

[0042] like Figure 2 As shown, a pole is assumed in front of a roadside parking space on a municipal road. The pole can be an existing street light pole, utility pole, etc. A high-level horizontal bar is installed at a certain height above the ground on the pole, and an identification mechanism 101 is installed on the high-level horizontal bar. In this embodiment, the identification mechanism includes a license plate recognition camera and a video camera, both of which are set in the same direction and face the parking space.

[0043] The license plate recognition camera is a high-definition camera and can be equipped with a supplementary light. The cost of a license plate recognition camera is relatively higher than that of a regular video camera. The license plate recognition camera is only used for vehicle capture and does not record video, which extends the lifespan of the expensive camera and reduces data storage space. On the other hand, regular video cameras are low-cost, can work continuously, have lower accuracy requirements, and do not need to recognize license plates; they only need to identify and track vehicles. Therefore, this embodiment combines the two types of camera equipment, allowing each device to leverage its own advantages to work together to complete vehicle recognition and parking space entry / exit judgment.

[0044] Roadside parking spaces are usually arranged in a straight line. If there are a large number of roadside parking spaces in a row, two or more poles can be set up in different locations. Each pole has an identification mechanism, and the parking space areas processed by different identification structures are strictly divided.

[0045] Step S2: Divide the roadside parking spaces into multiple groups according to their adjacent order to form parking space groups, with each group having a number of parking spaces.

[0046] In practical identification, although each parking space can be assigned a parking space number, conventional roadside parking methods require recording the association between the vehicle and the parking space number. However, in reality, car owners do not care about the specific parking space number, only the parking fee. Therefore, this embodiment does not record the parking space number separately. Furthermore, due to the absence of a separate geomagnetic sensor and the occlusion of parking spaces, accurate identification of the parking space number is quite difficult.

[0047] Simultaneously, to facilitate parking space management, this step divides parking spaces into multiple groups according to their adjacent order, forming parking space groups. When dividing parking space groups, the groups closer to the recognition mechanism have fewer parking spaces, while the groups farther away from the recognition mechanism have more parking spaces. Because the closer parking space groups have a closer field of view, and the farther parking space groups have a farther field of view, vehicle recognition is relatively more difficult compared to vehicles in the parking space groups in front. If there are fewer parking spaces in the farther parking space groups, the length of the parking space group will be shorter in the video image, which is not conducive to judging subsequent vehicles entering different parking space groups.

[0048] Step S3: The video camera captures video images. First, the outer edge of the parking space group is located. Then, a trigger area is set near the outer edge of the parking space. When the trigger area of ​​each parking space group is triggered and the vehicle is identified as having entered the parking space, the license plate recognition camera captures the vehicle.

[0049] The trigger zone is simply the area near the outer edge of the parking space in the image used to detect the vehicle's position and determine whether the vehicle has entered or exited the parking space. There are no physical markers for the trigger zone on the actual parking space surface. As one implementation, the trigger zone shown in the illustration consists of two parallel detection lines, namely the first detection line 102 and the second detection line 103, from the outside in. Entering or exiting the parking space requires passing through these detection lines in the image. Furthermore, the detection lines extend to the front or rear of the parking spaces at both ends, effectively enclosing all parking spaces within the two detection lines.

[0050] The trigger zone is activated and identified as having entered a parking space in the following ways:

[0051] S31. When a vehicle target touches the first detection line, identify the outer rectangle of the vehicle target and track the outer rectangle.

[0052] S32. When the aspect ratio of the circumscribed rectangle changes significantly, the license plate recognition camera captures the vehicle target for the first time. When the circumscribed rectangle touches the second detection line, the license plate recognition camera captures the vehicle target for the second time.

[0053] When a vehicle enters a parking space, it typically approaches the space first, then turns forward or reverses to enter. This embodiment uses two detection lines to accommodate different parking scenarios. The distance between the two lines should not be too far from the parking space to avoid triggering the detection lines by normally moving vehicles. However, they also cannot be too close, otherwise it will be difficult to detect and distinguish the two lines in the image. Furthermore, the second detection line should not be too close to the outer edge of the parking space, as improperly parked vehicles partially exposed within the parking space can easily cause false triggering. As an example, the first and second detection lines are 0.70 meters and 0.30 meters from the edge of the parking space, respectively.

[0054] Video cameras can monitor vehicles. When a vehicle approaches a parking space, it will first touch the first detection line. To facilitate image capture, the camera identifies and tracks the circumscribed rectangle of the vehicle target. For example... Figure 3 As shown, the dashed line represents the circumscribed rectangle of the vehicle. Due to variations in vehicle size and distance, the size and aspect ratio of the circumscribed rectangle in the image are not fixed. To identify when a vehicle is turning, the aspect ratio of the vehicle changes significantly during reversing or turning into a parking space. This feature allows for quick determination of whether the vehicle is turning. The detection method is simple to implement and highly efficient. When a significant change occurs in the circumscribed rectangle, the vehicle is captured.

[0055] The most obvious advantage of this method is that it filters out vehicles that are very close to the parking space or are temporarily parked, avoiding the saving of incorrect vehicle images. If the first detection line is triggered to take a picture, many vehicles will be seen touching the detection line but not stopping.

[0056] When the vehicle continues to turn and move, it will touch the second detection line, indicating that the vehicle is about to enter the parking space. At this time, the license plate recognition camera will capture the vehicle again.

[0057] Step S4: After identifying the license plate number from the captured image, once the vehicle has fully entered the parking space group, bind the license plate number to the corresponding parking space group and record the vehicle's entry time.

[0058] The advantage of capturing images at least twice during parking is that it allows for two license plate recognition checks, improving accuracy. After a vehicle enters a parking space, it needs to be linked to a parking space group. Each parking space group has a list of license plate information. The specific process of step S4 is as follows:

[0059] S41. Perform license plate recognition on the two captured images to obtain the license plate number of the current vehicle target, and save the two captured images at the same time.

[0060] S42. When a vehicle target leaves the first detection line and then leaves the second detection line, it is determined that the vehicle target has entered the parking space. The license plate number is added to the corresponding license plate information list, and the current time is recorded as the vehicle's entry time.

[0061] If there are discrepancies in license plate recognition between two captured images, the preferred approach is to capture another image when the vehicle leaves the second detection line for a comprehensive assessment, aiming to identify the license plate more clearly. If the license plate is still not captured and identified, the information needs to be sent to the backend for intervention.

[0062] As the vehicle continues its journey into the parking space, it will first leave the first inspection line and then the second inspection line. If, during the parking process, the driver stops and drives away directly, several situations may occur. For example, the vehicle may leave the second inspection line first and then the first inspection line, indicating that the vehicle left without stopping. In all these cases, the vehicle can be considered not to have stopped, and the license plate number cannot be added to the corresponding license plate information list.

[0063] Step S5: When the trigger zone of each parking space group is triggered and it is identified as leaving the parking space, the license plate recognition camera captures the vehicle, identifies the license plate number, unbinds the license plate number from the parking space group, and records the time the vehicle leaves.

[0064] The specific process for this step is as follows:

[0065] S51. When a vehicle target in a parking space group touches the second detection line first, the outermost pixel coordinate of the vehicle target is tracked. When it crosses the first detection line, the license plate recognition camera completes the first capture of the vehicle target.

[0066] S52. When the vehicle target leaves the second detection line first and then leaves the first detection line, the license plate recognition camera completes a second capture of the vehicle target and determines that the vehicle target has left the parking space.

[0067] S53. Perform license plate recognition on the two captured images to obtain the license plate number of the current vehicle target, and save the two captured images at the same time.

[0068] S54. Find the corresponding license plate number from the parking space group corresponding to the vehicle target, then remove the license plate number from the license plate information list, and record the current time as the vehicle's departure time.

[0069] When a vehicle leaves, the second detection line is triggered first, followed by the first detection line. When the vehicle first touches the second detection line, no capture is needed; the outermost pixel coordinates of the vehicle are located. When it crosses the first detection line, it indicates that most of the vehicle's front is visible, and the first capture is performed. Then, the vehicle sequentially leaves the second and first detection lines, indicating it has completely left, and a second capture is performed. This two-capture process allows for more accurate license plate identification. The license plate number is then removed from the license plate information list of the current parking space group. If the corresponding license plate number cannot be found in the current parking space group, it may be due to an error in classifying the parking space group during entry. Therefore, the corresponding license plate number can be searched for in other parking space groups, starting with adjacent groups and then other groups. If found, the license plate number is removed from the license plate information list of the corresponding parking space group.

[0070] Step S6: Complete the parking fee calculation for the current vehicle based on the vehicle's entry and exit times.

[0071] By recording the vehicle's entry and exit times, time statistics can be completed, ensuring billing is completed as designed.

[0072] Finally, as an extreme case, if a vehicle leaves the first detection line but continues to touch the second detection line while entering a parking space—specifically, if the distance between the vehicle and the first detection line is greater than k*D, where D is the pixel distance between the two detection lines at the current position of the vehicle and k is a scaling factor with a value of 0.80-0.90—then the vehicle is considered to have entered the parking space. For example, if many car bodies remain outside the parking space after parking, or if there is insufficient space and the front of the car is still outside, this irregular parking is dangerous. This embodiment can push this information to the background for further processing, such as sending a text message to the car owner to move the car, or, if the car remains unmoved for an extended period, sending the message to the traffic police's handheld information terminal.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements 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 calculating and recording roadside parking fees on municipal roads, characterized in that, The method includes the following steps: Step S1: Install a recognition mechanism on the high horizontal bar above the roadside parking space. The recognition mechanism includes a license plate recognition camera and a video camera, both of which are set in the same direction. Step S2: Divide the roadside parking spaces into multiple groups according to their adjacent order to form parking space groups, each group having a number of parking spaces; Step S3: The video camera captures video images. First, the outer edge of the parking space group is located. Then, a trigger area is set near the outer edge of the parking space. When the trigger area of ​​each parking space group is triggered and the vehicle is identified as entering the parking space, the license plate recognition camera captures the vehicle. Step S4: After identifying the license plate number from the captured image, once the vehicle has fully entered the parking space group, bind the license plate number to the corresponding parking space group and record the vehicle's entry time. Step S5: When the trigger area of ​​each parking space group is triggered and it is identified as leaving the parking space, the license plate recognition camera captures the vehicle, identifies the license plate number, unbinds the license plate number from the parking space group, and records the time the vehicle leaves. Step S6: Complete the parking fee calculation for the current vehicle based on the vehicle's entry and exit times; In step S2, when dividing parking spaces into groups, the parking space groups closer to the identification mechanism have fewer parking spaces, while the parking space groups farther away from the identification mechanism have more parking spaces. In step S3, the triggering area consists of two parallel detection lines, namely the first detection line and the second detection line from the outside to the inside. The triggering area is activated and identified as having entered a parking space in the following way: When a vehicle target touches the first detection line, the outer rectangle of the vehicle target is identified and tracked. When the aspect ratio of the circumscribed rectangle changes significantly, the license plate recognition camera captures the vehicle target for the first time. When the circumscribed rectangle touches the second detection line, the license plate recognition camera captures the vehicle target for the second time. Each parking space group has a list of license plate information. The specific process of step S4 is as follows: The two captured images are used for license plate recognition to obtain the license plate number of the current vehicle target, and the two captured images are saved at the same time. When a vehicle leaves the first detection line and then the second detection line, it is determined that the vehicle has entered the parking space. The license plate number is added to the corresponding license plate information list, and the current time is recorded as the vehicle's entry time. The specific process of step S5 is as follows: When a vehicle in a parking space touches the second detection line, the outermost pixel coordinates of the vehicle are tracked. When the vehicle crosses the first detection line, the license plate recognition camera captures the vehicle for the first time. When the vehicle target leaves the second detection line first and then leaves the first detection line, the license plate recognition camera completes a second capture of the vehicle target and determines that the vehicle target has left the parking space; The two captured images are used for license plate recognition to obtain the license plate number of the current vehicle target, and the two captured images are saved at the same time. Find the corresponding license plate number from the parking space group corresponding to the vehicle target, then remove the license plate number from the license plate information list, and record the current time as the vehicle's departure time; If the corresponding license plate number cannot be found in the current parking space group, the corresponding license plate number will be searched for in other parking space groups. If found, the license plate number will be removed from the list of recorded license plate information. In step S4, if the vehicle target leaves the first detection line first, but continues to touch the second detection line, and the distance from the first detection line is greater than k*D, where D is the pixel distance between the two detection lines at the current position of the vehicle target, and k is a scaling factor with a value of 0.80-0.90, then the vehicle target has entered the parking space.

Images