A road perception method based on the integration of ETC technology and V2X technology

Abstract

The present invention discloses a road perception method based on the integration of ETC technology and V2X technology, belonging to the technical field of highway monitoring, and comprising the following steps: S1: A first perception system is set at the toll gantry; S2: Point matching of ETC gantry data and first perception system data is carried out in real time. After successful matching, the license plate information of the ETC toll system is given to the first perception system for perception target supplement or redundant judgment of the license plate; S3: After the toll gantry, a second perception system is set at every preset distance; S4: For the overlapping area between the radars, point algorithm matching is carried out to obtain a unique target ID, and the license plate data is synchronously inherited. The present invention realizes redundant processing of the license plate of the perception system, realizes the supplement of the license plate of the perception system at night or in poor visibility, solves the problem that it is difficult to identify the license plate of the perception system at night or in poor visibility, and saves economic costs.

Description

Technical Field

[0001] The present invention relates to the technical field of highway monitoring, and particularly to a road perception method based on the integration of ETC technology and V2X technology. Background Art

[0002] Highways, with their advantages of high vehicle speed, large traffic capacity, and low transportation costs, have become an important artery of the national economy on a par with aviation and railways. With the continuous growth of people's demand for convenient travel, the development of automobiles and highways has been quite rapid. License plate recognition has become an important part of the existing intelligent transportation system. License plate recognition technology requires being able to extract and recognize the license plate of a moving vehicle from a complex background and using the recognized license plate number as the basis for vehicle toll payment or electronic evidence collection.

[0003] Currently, in order to collect vehicle feature information and passing vehicle records, multiple ETC gantries are set up along highways, and a large number of intelligent monitoring and recording devices are installed on them. In order to clearly record and recognize vehicle license plates at night, supplementary lights are installed on the gantries. However, in order to take clear photos, the brightness of the supplementary lights is often relatively high, which will have a certain impact on the driving safety of drivers; and when other problems that affect visibility occur, such as rainy days, foggy days, and other bad weather, it will further increase the difficulty of license plate recognition; at the same time, there is a large amount of vehicle flow on highways, so data redundancy is also extremely large. How to reduce this part of data redundancy is also a very important task.

[0004] In the prior art, a high-precision highway license plate cloud recognition method disclosed in Patent Application No. CN201910822694.X, after using the vehicle-related information collected by the front-end site and setting up a cloud storage repository to save and manage the collected information data in real time for subsequent retrieval during recognition and judgment, has the beneficial technical effects of solving the problem of the original conventional single direct storage in the recognition hardware device, with a single source of recognition images and inability to make full use of the collected data, realizing timely and convenient data retrieval and use, and realizing data cloud storage management. However, this solution only aims to improve the highway entrance passing efficiency and the license plate recognition accuracy rate, but it cannot solve the problem of license plate recognition during the driving process of vehicles on highways.

[0005] In the method for anti-dazzle license plate capture on highways based on deep learning algorithms disclosed in patent application number CN202111093289.2, this solution sets up a speed measurement camera to enable the speed measurement camera to identify the position of the vehicle's headlights at night in real time, thereby controlling the supplementary light to swing with the movement of the vehicle, ensuring that the illumination range of the supplementary light is always below the vehicle's front windshield and above the license plate area, that is, ensuring that the license plate recognition camera can normally recognize the license plate, and at the same time avoiding the supplementary light directly shining into the driver's eyes, greatly improving the road traffic safety on highways. Especially for different types of vehicles, precise control of the supplementary light can be achieved, thus improving the driving safety of vehicles; however, in the case of a large traffic flow, this solution uses the method of adding multiple supplementary lights, which will undoubtedly increase a large amount of economic costs. At the same time, this solution only aims at the situation where the light is relatively dim at night to accurately recognize the license plate. When faced with the situation where the difficulty of license plate recognition is increased due to non-light brightness problems, accurate license plate recognition cannot be achieved. Summary of the Invention

[0006] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a road perception method based on the integration of ETC technology and V2X technology to solve the problems of poor visibility caused by weather conditions and difficult license plate recognition at night.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] A road perception method based on the integration of ETC technology and V2X technology includes the following steps:

[0009] Step S1: Set up a first perception system at the toll gantry;

[0010] Step S2: Perform point matching of ETC gantry data and first perception system data in real time. After successful matching, give the license plate information of the ETC toll system to the first perception system for perception target supplementation or redundant license plate judgment;

[0011] Step S3: After the toll gantry, set up a second perception system at every preset distance;

[0012] Step S4: Perform point algorithm matching on the overlapping area between radars to obtain a unique target ID and synchronously inherit the license plate data.

[0013] Furthermore, the specific steps of step S2 include the following sub-steps:

[0014] Step S201: The first perception system receives data from the ETC gantry;

[0015] Step S202: Circularly extract longitude and latitude, license plate, and timestamp information, and match them with the targets perceived by the first perception system in turn;

[0016] Step S203: Determine whether the difference between the extracted timestamp and the timestamp of the perceived target in the first perception system is within the set threshold range; if so, calculate the intersection density ratio between the extracted longitude and latitude and the longitude and latitude of the perceived target in the first perception system according to the Hungarian algorithm; if not, discard the data and return to step S201;

[0017] Step S204: Determine whether the obtained intersection density ratio is within the set threshold range; if so, the matching is successful, and the perception data inherits the license plate data in the ETC gantry; if not, discard the data and return to step S201.

[0018] Furthermore, the RSU in the ETC gantry and the first perception system achieve time synchronization through an ntp server.

[0019] Furthermore, the first perception system includes a camera, a radar, and an edge computing unit.

[0020] Furthermore, after the radar and the camera collect data, the radar points, pixels, and longitude and latitude are calibrated and bound, and conversions from radar points to pixels, radar points to longitude and latitude, and pixels to longitude and latitude are performed.

[0021] Furthermore, the second perception system includes a radar and an edge computing unit.

[0022] Furthermore, the edge computing unit of the second perception system is deployed according to the computing power of the edge computing unit.

[0023] Furthermore, step S4 specifically includes the following sub-steps:

[0024] Step S401: Receive and process data in real time;

[0025] Step S402: Loop to determine whether the target ID recognized by the radar exists in the track_list; if so, update the target track in the track_list and return to step S401; if not, calculate the distance between the target and the vehicles in the track_list;

[0026] Step S403: Loop to determine whether the distance is greater than the set threshold; if so, create a new target track and return to step S401; if not, calculate the difference in the heading angles between the target and the vehicles in the track_list;

[0027] Step S404: Loop to determine whether the difference in the heading angles is greater than the threshold; if so, create a new target track and return to step S401; if not, calculate the intersection density ratio between the target and the vehicles in the track_list using the Hungarian algorithm;

[0028] Step S405: Continuously determine whether the intersection encryption ratio is within the threshold; if not, create a new target track and return to step S401; if so, the matching is successful, and the original track is updated and maintained.

[0029] Further, when the radar in step S4 performs detection, for the blind area detected by the radar, the data of three or more frames before the target disappears is used to predict the target's travel based on longitude and latitude, average speed, average acceleration, and average heading angle to complete the target travel data.

[0030] Further, the radar is a millimeter-wave radar.

[0031] Advantages of the present invention: The present invention realizes the processing of license plate redundancy in the perception system, realizes the supplement of the license plate in the perception system at night or in poor visibility, solves the pain point problem of difficult recognition of the license plate in the perception system at night or in poor visibility, and at the same time reduces the cost of license plate recognition on highways. Description of the Drawings

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on the structures shown in these drawings without creative efforts.

[0033] Figure 1 is the flowchart of the method steps of the present invention.

[0034] Figure 2 is the fusion flowchart of the gantry ETC and the perception system of the present invention.

[0035] Figure 3 is the fusion flowchart of the edge computing unit of the perception system of the present invention. Detailed Embodiments

[0036] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0037] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0038] In this embodiment, as Figure 1As shown in the figure, a road perception method based on the integration of ETC technology and V2X technology includes the following steps:

[0039] Step S1: Set up a first perception system at the toll gantry.

[0040] Step S2: Perform real-time point matching of ETC gantry data and first perception system data. After successful matching, give the license plate information of the ETC toll system to the first perception system for perception target supplementation or redundant license plate judgment.

[0041] Step S3: After the toll gantry, set up a second perception system at every preset distance. In this embodiment, since the effective coverage radius of a single radar is usually more than 300 meters, the preset distance is 600 meters.

[0042] Step S4: Perform point algorithm matching on the overlapping area between radars to obtain a unique target ID and synchronously inherit the license plate data. At the same time, for the radar detection blind area, use a prediction algorithm to predict the driving of the target to complete the target driving data.

[0043] At the toll station, the ETC gantry is integrated with the perception system. When the gantry RSU and the in-vehicle ETC use DSRC for communication transactions at a specific location (fixed longitude and latitude), at the same time, the gantry RSU and the perception system use the ntp server to achieve time synchronization, and perform one-to-one matching of the ETC gantry and the perception system according to the longitude and latitude and timestamp (millisecond level), and synchronously inherit the license plate information of the etc gantry; as Figure 2 shown, specifically, step S2 includes the following sub-steps:

[0044] Step S201: The first perception system receives data from the ETC gantry.

[0045] Step S202: Circularly extract longitude and latitude, license plate, and timestamp information, and sequentially match them with the targets perceived by the first perception system.

[0046] Step S203: Judge whether the difference between the extracted timestamp and the timestamp of the perceived target in the first perception system is within the set threshold range; if so, calculate the cross-correlation ratio of the extracted longitude and latitude and the longitude and latitude of the perceived target in the first perception system according to the Hungarian algorithm; if not, discard the data and return to step S201.

[0047] Step S204: Judge whether the obtained cross-correlation ratio is within the set threshold range; if so, the matching is successful, and the perception data inherits the license plate data in the ETC gantry; if not, discard the data and return to step S201.

[0048] The first perception system includes a camera, a radar, and an edge computing unit, which are installed at the toll gantry. After the radar and the camera collect data, the radar points, pixels, and longitude and latitude are calibrated and bound, and conversions from radar points to pixels, from radar points to longitude and latitude, and from pixels to longitude and latitude are performed. Then, multiple second perception systems are installed at intervals (determined by the effective monitoring distance of the radar).

[0049] The second perception system includes a radar and an edge computing unit; among them, the obtained radar points can be converted from radar points to pixels and from radar points to longitude and latitude.

[0050] In a preferred embodiment, after the radar points of the present invention are calibrated, they can be converted to pixel points. For example: obtaining radar points (x, y), and through a 3×3 matrix obtained by calibration, at this time, to obtain the corresponding pixel points

[0051]

[0052] Among them, the 3×3 matrix is obtained by fitting multiple radar points and corresponding pixel points, and specifically includes three groups of fittings in total: radar points → pixel points, radar points → longitude and latitude, pixel points → longitude and latitude.

[0053] The edge computing unit is deployed according to the computing power of the edge computing unit, that is, one or more edge computing units are used between multiple poles to process the vehicle data detected by the radar; among them, the radar is a millimeter-wave radar.

[0054] Specifically, as Figure 3 shown, step S4 specifically includes the following sub-steps:

[0055] Step S401: Receive and process data in real time;

[0056] Step S402: Circularly determine whether the target ID recognized by the radar exists in track_list; if so, update the target track in track_list and return to step S401; if not, calculate the distance between the target and the vehicles in track_list;

[0057] Step S403: Circularly determine whether the distance is greater than the set threshold; if so, create a new target track and return to step S401; if not, calculate the difference in the heading angle between the target and the vehicles in track_list;

[0058] Step S404: Circularly determine whether the difference in the heading angle is greater than the threshold; if so, create a new target track and return to step S401; if not, use the Hungarian algorithm to calculate the intersection density ratio of the longitude and latitude of the target and the vehicles in track_list;

[0059] Step S405: Continuously determine whether the intersection encryption ratio is within the threshold; if not, create a new target track and return to step S401; if so, the matching is successful, and the original track is updated and maintained.

[0060] Preferably, in this embodiment, for the blind area detected by the radar, the data of the previous 5 frames before the target disappears is used, and the target driving is predicted according to the longitude and latitude, average speed, average acceleration, and average heading angle to complete the target driving data.

[0061] The present invention realizes the processing of license plate redundancy in the perception system, realizes the supplement of the license plate in the perception system at night or in poor visibility, solves the pain point problem that the license plate in the perception system is difficult to identify at night or in poor visibility, and at the same time reduces the cost of license plate recognition on highways.

[0062] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, some steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and units involved are not necessarily essential to this application.

[0063] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0064] Those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a ROM, a RAM, etc.

[0065] The above-disclosed are only the preferred embodiments of the present invention. Of course, the scope of the rights of the present invention cannot be limited thereby. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A road perception method based on the integration of ETC technology and V2X technology, characterized in that, It includes the following steps: Step S1: Set up a first sensing system at the toll gantry; the first sensing system includes a camera, a radar, and an edge computing unit; Step S2: Perform real-time point matching of ETC gantry data and first sensing system data. After successful matching, give the license plate information of the ETC toll system to the first sensing system for sensing target supplementation or redundant license plate judgment; the specific steps of Step S2 include the following sub-steps: Step S201: The first sensing system receives data from the ETC gantry; Step S202: Cyclically extract longitude and latitude, license plate, and timestamp information, and match them with the targets sensed by the first sensing system in turn; Step S203: Judge whether the difference between the extracted timestamp and the timestamp of the sensed target in the first sensing system is within the set threshold range; If so, calculate the cross-density ratio of the extracted longitude and latitude and the longitude and latitude of the sensed target in the first sensing system according to the Hungarian algorithm; if not, discard the data and return to Step S201; Step S204: Judge whether the obtained cross-density ratio is within the set threshold range; if so, the matching is successful, and the sensed data inherits the license plate data in the ETC gantry; if not, discard the data and return to Step S201; Step S3: After the toll gantry, set up a second sensing system at every preset distance; the second sensing system includes a radar and an edge computing unit; Step S4: Perform point algorithm matching on the overlapping area between radars to obtain a unique target ID and synchronously inherit the license plate data; The specific steps of Step S4 include the following sub-steps: Step S401: Receive and process data in real time; Step S402: Cyclically judge whether the target ID recognized by the radar exists in the track_list; if so, update the target track in the track_list and return to Step S401; if not, calculate the distance between the target and the vehicles in the track_list; Step S403: Cyclically judge whether the distance is greater than the set threshold; if so, create a new target track and return to Step S401; If not, calculate the heading angle difference between the target and the vehicles in the track_list; Step S404: Cyclically judge whether the heading angle difference is greater than the threshold; if so, create a new target track and return to Step S401; if not, calculate the cross-density ratio of the target and the longitude and latitude of the vehicles in the track_list using the Hungarian algorithm; Step S405: Cyclically judge whether the cross-density ratio is within the threshold; if not, create a new target track and return to Step S401; if so, the matching is successful, and the original track is updated and maintained; When the radar in Step S4 detects, for the blind area detected by the radar, use the data of three or more frames before the target disappears, and predict the target driving according to longitude and latitude, average speed, average acceleration, and average heading angle to complete the target driving data.

2. The road perception method based on the integration of ETC technology and V2X technology according to claim 1, characterized in that The RSU in the ETC gantry and the first sensing system achieve time synchronization through the ntp server.

3. The road perception method based on the integration of ETC technology and V2X technology according to claim 1, characterized in that, After the radar and the camera collect data, the radar points, pixels, and longitude and latitude are calibrated and bound, and conversions from radar points to pixels, from radar points to longitude and latitude, and from pixels to longitude and latitude are performed.

4. The road perception method based on the integration of ETC technology and V2X technology according to claim 1, wherein, The edge computing unit of the second perception system is deployed according to the computing power of the edge computing unit.

5. A road perception method based on the integration of ETC technology and V2X technology according to claim 1, characterized in that, The radar is a millimeter-wave radar.

Images