V2x-based station vehicle perception assisted positioning method

Abstract

The application discloses a kind of based on V2X's yard vehicle perception auxiliary positioning method, including laying auxiliary positioning system, which includes multiple RSUs, multiple yard cameras and multiple MEC nodes;High-precision map of parking lot is obtained, objects in parking lot are selected as markers, markers, RSUs, acquisition units and yard end MEC nodes are marked on high-precision map according to actual position, to obtain marked map, and the marked map is stored in cloud;Vehicle enters, vehicle obtains marked map through RSU;Sub-area positioning;Reference point selection in camera range;The relative distance of reference point and vehicle is calculated, and is mapped in marked map, to obtain the actual position of vehicle on marked map.The application proposes a new vehicle real-time positioning method, which can accurately assist in positioning the vehicle in the underground parking lot when the GPS signal is weak, and the positioning accuracy is not affected by the displacement change of the reference point.

Description

TECHNICAL FIELD

[0001] The application relates to a vehicle positioning method in an underground parking lot, in particular to a station vehicle perception assisted positioning method based on V2X. BACKGROUND

[0002] With the development of society, indoor parking lots are becoming more and more popular. At present, when a vehicle drives in an underground parking lot, the vehicle needs to be positioned for path planning or real-time map updating. Common positioning methods at present include GPS positioning, video positioning in a parking lot, Bluetooth positioning and uwb positioning.

[0003] GPS positioning is most commonly used in automobile positioning devices, but it has certain requirements for the environment. For example, in an open place outdoors, the GPS satellite signal is good and the positioning is accurate, but various obstructions exist in an underground parking lot, so that the satellite signal positioning in the underground parking lot is not good.

[0004] For video positioning in a parking lot, a camera arranged in the parking lot is generally relied on. In order to ensure the accuracy of positioning, the camera coordinates are generally marked on a high-precision map of the parking lot. However, the camera may be displaced in a long-term use process, so that the video positioning is inaccurate.

[0005] The positioning accuracy of Bluetooth is more than 1 meter, which is relatively low. The positioning accuracy of uwb can reach about 40 cm, but the investment cost is high.

[0006] At present, with the continuous development and maturity of automatic driving technology, more and more vehicles support V2X communication. V2X is English for Vehicle-to-Everything Communication, which is Chinese for vehicle-to-everything communication. The application is to position the vehicle by using V2X communication.

[0007] Nomenclature:

[0008] RSU is the English abbreviation of Road Side Unit, which is directly translated as road side unit. It is installed on the roadside in the ETC system, uses the DSRC (Dedicated Short Range Communication) technology, communicates with the on-board unit (OBU, On Board Unit), realizes vehicle identity recognition and electronic deduction device. The detection range of RSU is a circular area with a radius of about 800 meters with the arrangement point as the center.

[0009] OBU, the English abbreviation of On Board Unit, is a Chinese vehicle unit. OBU is placed on the car, using DSRC (Dedicated Short Range Communication) technology, and RSU communication microwave device.

[0010] MEC, the English abbreviation of Mobile Edge Computing, is a Chinese mobile edge computing. The technology can improve user experience and save bandwidth resources on the one hand, and provide third-party application integration by sinking computing power to the mobile edge node, providing unlimited possibilities for mobile edge entry service innovation. The technology effectively integrates wireless network and Internet technologies together, and adds computing, storage, processing and other functions on the wireless network side. The application sets up an MEC node in the parking lot for computing and processing data obtained in the parking lot. SUMMARY

[0011] The purpose of the application is to provide a V2X-based yard vehicle perception auxiliary positioning method that can accurately position vehicles when the GPS signal is weak.

[0012] To achieve the above purpose, the technical scheme adopted by the application is as follows: a V2X-based yard vehicle perception auxiliary positioning method applied to a parking lot capable of V2X communication with the vehicle's infotainment system, comprising the following steps:

[0013] (1) Laying an auxiliary positioning system, the auxiliary positioning system comprising a plurality of RSUs, a plurality of in-yard cameras and a plurality of MEC nodes;

[0014] Among them, the RSU and the in-yard camera are laid in the parking lot according to the size of the parking lot, the detection range of the plurality of RSUs completely covers the parking lot, the camera range of each in-yard camera corresponds to a sub-area in the parking lot, all sub-areas completely cover the parking lot, and the in-yard cameras are numbered;

[0015] The vehicle is provided with a vehicle-mounted camera;

[0016] The MEC node is laid in the parking lot or on the vehicle, used for acquiring and processing video data of the in-yard camera or the vehicle-mounted camera, and communicating with the cloud through the RSU;

[0017] (2) Artificial marking;

[0018] Obtain a high-precision map of the parking lot, select an object in the parking lot as a marker, mark the marker, RSU, acquisition unit and field end MEC node on the high-precision map according to the actual position, obtain a marked map, and store the marked map in the cloud;

[0019] (3) Vehicle enters, vehicle system obtains marker map through RSU;

[0020] (4) Sub-area positioning;

[0021] Intra-field camera works, when a vehicle is identified to enter its sub-area, video information is sent to the MEC node, and the MEC node obtains the sub-area of the vehicle in the marker map according to the number of the intra-field camera;

[0022] (5) Reference point selection in the camera range;

[0023] (51) Obtain video information collected by a camera, the camera is an intra-field camera corresponding to the sub-area or a vehicle-mounted camera of the vehicle;

[0024] (52) Image recognition is performed on the video information, the vehicle and a plurality of markers are identified, the identified markers are taken as to-be-measured markers, the coordinates of the to-be-measured markers in manual marking are obtained, and the theoretical relative distances of each to-be-measured marker and other to-be-measured markers are calculated;

[0025] (53) A sub-coordinate system is established with the camera as the center, a video positioning method is used on the video information, the positions of the vehicle and each to-be-measured marker in the sub-coordinate system are obtained, and the actual relative distances of each to-be-measured marker and other to-be-measured markers in the sub-coordinate system are calculated;

[0026] (54) For each to-be-measured marker, the distance change sum is calculated according to the theoretical relative distance and the actual relative distance, and the to-be-measured marker with the smallest distance change sum is taken as the reference point in the camera range;

[0027] (6) The relative distance between the reference point and the vehicle is calculated and mapped in the marker map to obtain the actual position of the vehicle on the marker map.

[0028] As preferred: It further includes step (7) sending the actual position of the vehicle to the cloud through the RSU.

[0029] As preferred: In step (1), if the MEC node is arranged in the parking lot, 1-3 intra-field cameras are corresponded according to the computing power, and a plurality of MEC nodes can cover all intra-field cameras; if the MEC node is arranged on the vehicle, one vehicle is arranged with one.

[0030] As preferred: In step (2), the objects serving as markers include marking lines, stop lines, speed reduction belts, columns and / or other objects that do not move without active intervention.

[0031] As preferred: In step (52), image recognition is performed on the video information;

[0032] If the video information is collected by the intra-field camera, the vehicle is directly identified from the video information;

[0033] If the video information is collected by the vehicle-mounted camera, the position of the vehicle-mounted camera is marked as the vehicle.

[0034] As preferred: in the step (53), the video positioning method comprises the object video ranging and direction positioning method in front of the motor vehicle.

[0035] As preferred: in the step (54), the distance variation sum calculation method of the to-be-measured markers is:

[0036] (a1) suppose that the to-be-measured markers are m in total, and are sequentially marked as D1 to Dm;

[0037] (a2) the theoretical relative distance of D1 and D2 is L12, and the actual relative distance is S12, then the variation of D1 and D2 is B12 = |L12-S12|;

[0038] (a3) the variations of D1 and D3-DM are sequentially calculated as B13-B1m;

[0039] (a4) the distance variation sums B12-B1m of D1 are added to obtain the distance variation sum B1 of D1;

[0040] (a5) the distance variation sums B2-Bm of D2-Dm are sequentially calculated according to the steps (a2)-(a4).

[0041] As preferred: the step (55) further comprises: if the distance variation sums of at least two to-be-measured markers are the same and are the minimum, then any one of the to-be-measured markers is selected as the reference point.

[0042] The overall idea of the present application is:

[0043] After the auxiliary positioning system is laid out, the markers are selected in the parking lot, and the markers, RSUs, collection units and field end MEC nodes are marked on the high-precision map according to the actual positions to obtain a marker map.

[0044] The vehicle enters the parking lot to obtain the marker map, and then according to the positions of the in-field cameras capturing the vehicle, the approximate area of the vehicle in the parking lot, i.e. the sub-area positioning of the present application, is calculated.

[0045] Then a reference point is selected from the sub-area, and the relative distance between the reference point and the vehicle is calculated and mapped to the marker map, since the position of the reference point in the marker map is known, the position of the vehicle in the marker map can be accurately obtained.

[0046] The application proposes a new selection method in steps (51)-(54) when selecting reference points. The method only identifies the vehicle and several markers in the image recognition range through image recognition. Since the vehicle has been positioned to the sub-area, if the camera is an in-field camera, the actual image recognition range is the sub-area, and if it is a vehicle-mounted camera, it is a part of the sub-area with a small range, so the marker can be quickly identified. After identifying the marker, the coordinates of the marker can be known according to the artificial marker. However, since the marker may move, for example, due to in-field planning, the identification line of the left turn moves 10 meters forward, or due to the impact of the vehicle, the in-field camera moves, which leads to the difference between the actual position and the theoretical position. The application proposes a method for calculating the "total distance change", regardless of whether the marker and the camera move, we only select the marker with the smallest total distance change as the reference point.

[0047] Compared with the prior art, the application has the following advantages:

[0048] The application proposes a new real-time vehicle positioning method based on V2X communication and video positioning, which is used for accurate auxiliary positioning of vehicles in an underground parking lot when the GPS signal is weak.

[0049] The method first selects markers in the parking lot to mark on the high-precision map to obtain a marker map. After the vehicle enters the parking lot, the camera determines the sub-area where the vehicle is located, and then selects a marker in the sub-area as a reference point. The relative distance between the reference point and the vehicle is calculated and mapped to the marker map. Since the reference point is known in the marker map, the position of the vehicle in the marker map can be accurately obtained.

[0050] The method of the application is not affected by the position change of the marker. Without using the reference point selection method of the application, the relative distance between the vehicle and a certain marker is directly calculated for vehicle positioning. When the marker moves and is not updated in the map, the positioning of the vehicle will also be offset, thereby causing inaccurate positioning. The application selects the reference point to select the marker that is least likely to change, thereby improving the positioning accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 The flowchart of the application;

[0052] Figure 2 The principle diagram of the auxiliary positioning system of the application. EMBODIMENT

[0053] The application will be further described below with reference to the accompanying drawings.

[0054] Example 1: see Figure 1 and Figure 2A station vehicle perception assisted positioning method based on V2X is applied to a parking lot capable of V2X communication with a vehicle machine system, and includes the following steps:

[0055] (1) An auxiliary positioning system is laid out, and the auxiliary positioning system includes a plurality of RSUs, a plurality of in-lot cameras, and a plurality of MEC nodes;

[0056] The RSUs and the in-lot cameras are laid out in the parking lot according to the size of the parking lot, the detection range of the plurality of RSUs completely covers the parking lot, the camera range of each in-lot camera corresponds to a sub-area in the parking lot, all sub-areas completely cover the parking lot, and the in-lot cameras are numbered;

[0057] The vehicle is provided with a vehicle-mounted camera;

[0058] The MEC nodes are laid out in the parking lot or on the vehicle, are used to acquire video data of the in-lot cameras or the vehicle-mounted camera for processing, and communicate with the cloud through the RSUs;

[0059] (2) Artificial marking;

[0060] A high-precision map of the parking lot is acquired, objects in the parking lot are selected as markers, the markers, the RSUs, the acquisition units, and the in-lot MEC nodes are marked on the high-precision map according to actual positions to obtain a marked map, and the marked map is stored in the cloud;

[0061] (3) The vehicle enters the parking lot, and the machine system acquires the marked map through the RSU;

[0062] (4) Sub-area positioning;

[0063] The in-lot cameras work, and when a vehicle is identified to enter a sub-area thereof, video information is sent to the MEC node, the MEC node acquires the sub-area of the vehicle in the marked map according to the number of the in-lot cameras;

[0064] (5) Reference point selection in a camera range;

[0065] (51) Video information acquired by an in-lot camera corresponding to the sub-area or a vehicle-mounted camera of the vehicle is acquired;

[0066] (52) Image recognition is performed on the video information, a vehicle and a plurality of markers are identified, the identified markers are taken as to-be-measured markers, coordinates of the to-be-measured markers at the time of artificial marking are acquired, and theoretical relative distances of each to-be-measured marker from other to-be-measured markers are calculated;

[0067] (53) Establishing a sub-coordinate system with the camera as the center, using a video positioning method to obtain the positions of the vehicle and each to-be-measured marker in the sub-coordinate system, and calculating the actual relative distances between each to-be-measured marker and other to-be-measured markers in the sub-coordinate system;

[0068] (54) For each to-be-measured marker, calculating the distance change sum according to the theoretical relative distance and the actual relative distance, and taking the to-be-measured marker with the smallest distance change sum as the reference point in the camera range;

[0069] (6) Calculating the relative distance between the reference point and the vehicle and mapping it in the marker map to obtain the actual position of the vehicle on the marker map.

[0070] In this embodiment, the MEC nodes are arranged in the parking lot, and each MEC node corresponds to 1-3 in-parking-lot cameras according to the computing power, and multiple MEC nodes can cover all in-parking-lot cameras.

[0071] In step (2), the objects serving as markers include marking lines, stop lines, speed bumps, columns and / or other objects that do not move without active intervention. The marking lines can be turning lines, no-entry lines, etc.

[0072] In step (52), image recognition is performed on the video information.

[0073] If the video information is collected by the in-parking-lot camera, the vehicle is directly identified from the video information.

[0074] If the video information is collected by the vehicle-mounted camera, the position of the vehicle-mounted camera is marked as the vehicle.

[0075] In the step (53), the video positioning method includes a motor vehicle front object video ranging and direction positioning method.

[0076] In step (54), the distance change sum calculation method of the to-be-measured marker is:

[0077] (a1) Let the to-be-measured markers be m, and mark them as D1 to Dm in turn.

[0078] (a2) The theoretical relative distance of D1 and D2 is L12, and the actual relative distance is S12, so the change of D1 and D2 is B12=|L12-S12|.

[0079] (a3) The changes B13~B1m of D1 and D3~DM are calculated in turn.

[0080] (a4) The distance change sum B1 of D1 is obtained by adding B12~B1m.

[0081] (a5) sequentially calculate the distance change sum B2-Bm of D2-Dm according to steps (a2)-(a4).

[0082] The step (55) further comprises: if the distance change sums of the at least two to-be-tested markers are same and are all minimum, then selecting any one of the at least two to-be-tested markers as the reference point.

[0083] Embodiment 2: refer to Figure 1 and Figure 2 In this embodiment, the MEC node is arranged on the vehicle, and one vehicle is arranged with one MEC node. The rest is the same as Embodiment 1. In this embodiment, the video information is sent to the MEC node on the vehicle, processed, and then sent to the cloud through the RSU.

[0084] Embodiment 3: refer to Figure 1 and Figure 2 On the basis of Embodiment 1 or Embodiment 2, this embodiment comprises steps (1)-(6), and further comprises step (7) of sending the actual position of the vehicle to the cloud through the RSU. The cloud is mainly used to record the driving track of the vehicle, and provide basis for subsequent data analysis.

[0085] In addition, in step (4) of the present application, the vehicle sub-area positioning can also not rely on the camera, but directly use the BSM message of the vehicle. BSM message: is the abbreviation of Basic Safety Message, is a basic message type in vehicle networking. BSM is the most widely used message in V2X communication, and all V2V applications are implemented based on BSM message. BSM message is the basis for vehicle-to-vehicle communication, which contains the basic information of vehicle position, speed, direction, etc., for traffic safety and traffic flow optimization between vehicles. BSM message can realize real-time communication between vehicles through vehicle networking technology, and improve traffic safety and efficiency.

[0086] The BSM message of the vehicle is sent to the MEC node, and the MEC node is compared with the marker map, and the approximate area of the vehicle can also be obtained.

[0087] The above only describes the preferred embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A V2X-based vehicle perception-assisted positioning method for parking lots, applied to parking lots capable of V2X communication with vehicle infotainment systems, characterized in that: The method comprises the following steps: (1) Laying an auxiliary positioning system, which comprises a plurality of RSUs, a plurality of in-field cameras and a plurality of MEC nodes; Wherein, the RSUs and in-field cameras are laid in the parking lot according to the size of the parking lot, the detection range of the plurality of RSUs completely covers the parking lot, the camera range of each in-field camera corresponds to a sub-area in the parking lot, all sub-areas completely cover the parking lot, and the in-field cameras are numbered; The vehicle is provided with a vehicle-mounted camera; The MEC nodes are laid in the parking lot or on the vehicle, used for acquiring and processing video data of the in-field cameras or the vehicle-mounted camera, and communicating with the cloud through the RSUs; (2) Artificial marking; Acquiring a high-precision map of the parking lot, selecting objects in the parking lot as markers, marking the markers, RSUs, acquisition units and field-end MEC nodes on the high-precision map according to the actual positions to obtain a marked map, and storing the marked map in the cloud; (3) Vehicle entry, the vehicle system acquires the marked map through the RSU; (4) Sub-area positioning; The in-field cameras work, and when a vehicle is identified to enter a sub-area thereof, video information is sent to the MEC node, and the MEC node obtains the sub-area of the vehicle in the marked map according to the number of the in-field cameras; (5) Reference point selection in the camera range; (51) Acquiring video information collected by a camera, the camera being the in-field camera corresponding to the sub-area or the vehicle-mounted camera of the vehicle; (52) Image recognition is performed on the video information, the vehicle and a plurality of markers are identified, the identified markers are taken as to-be-measured markers, the coordinates of the to-be-measured markers at the time of artificial marking are acquired, and the theoretical relative distances between each to-be-measured marker and other to-be-measured markers are calculated; (53) A sub-coordinate system is established with the camera as the center, a video positioning method is used on the video information, the positions of the vehicle and each to-be-measured marker in the sub-coordinate system are obtained, and the actual relative distances between each to-be-measured marker and other to-be-measured markers in the sub-coordinate system are calculated; (54) For each to-be-measured marker, the distance change sum is calculated according to the theoretical relative distance and the actual relative distance, and the to-be-measured marker with the smallest distance change sum is taken as the reference point in the camera range; (6) The relative distance between the reference point and the vehicle is calculated and mapped in the marked map to obtain the actual position of the vehicle on the marked map. 2.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: It further comprises step (7) of sending the actual position of the vehicle to the cloud through the RSU. 3.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: In step (1), if the MEC nodes are laid in the parking lot, 1-3 in-field cameras correspond to one MEC node according to the computing power, and a plurality of MEC nodes can cover all in-field cameras; if the MEC nodes are laid on the vehicle, one MEC node is laid on one vehicle. 4.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: In step (2), the objects taken as markers include marking lines, parking lines, speed bumps, columns and / or other objects that do not move without active intervention. 5.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: In step (52), image recognition is performed on the video information; If the video information is collected by the in-field camera, the vehicle is directly identified from the video information; If the video information is collected by the vehicle-mounted camera, the position of the vehicle-mounted camera is marked as the vehicle. 6.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: The video positioning method includes a motor vehicle front object video ranging and direction positioning method in the step (53). 7.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: In step (54), the distance change sum calculation method of the to-be-measured markers is as follows: (a1) Let the to-be-measured markers be m, and mark them as D1 to Dm in sequence; (a2) The theoretical relative distance of D1 and D2 is L12, and the actual relative distance is S12. Then, the change amount of D1 and D2 is B12 = |L12-S12|; (a3) The change amounts B13 to B1m of D1 and D3 to DM are calculated in sequence; (a4) The distance change sums B12 to B1m are added to obtain the distance change sum B1 of D1; (a5) The distance change sums B2 to Bm of D2 to Dm are calculated in sequence according to steps (a2) to (a4). 8.The V2X-based yard vehicle perception aided positioning method according to claim 1, characterized in that: The step (54) further includes that if the distance change sums of at least two to-be-measured markers are the same and are the minimum, then any one of the to-be-measured markers is selected as a reference point.

Images