Lane positioning method, apparatus and device

Abstract

The application discloses a lane positioning method, device and equipment, and relates to the fields of maps, navigation, intelligent traffic, Internet of Vehicles, intelligent vehicle control technology, automatic driving and advanced auxiliary driving. The method comprises the following steps: acquiring a road image of a target vehicle; acquiring lane information corresponding to a position of the target vehicle; performing visual recognition processing on the road image to obtain lane line attribute information; determining target prior information matched with the lane line attribute information; determining posterior probability information of the target vehicle based on the lane information and the target prior information; and determining a lane position of the target vehicle according to the posterior probability information. The technical scheme provided in the application embodiment can determine the posterior probability of the vehicle on each lane and then position the lane by recognizing the lane line attribute information in the road image and combining the prior probability knowledge related to the lane line attribute information, thereby reducing the dependence of lane positioning on equipment and high-precision map data and improving the efficiency of lane positioning.

Description

TECHNICAL FIELD

[0001] The present application relates to the fields of maps, navigation, intelligent transportation, Internet of Vehicles, intelligent vehicle control technology, automatic driving and advanced auxiliary driving, and particularly relates to a lane positioning method, device and equipment. BACKGROUND

[0002] Lane positioning is used to determine the lane number of a vehicle in a map. In the field of intelligent vehicle control technology, such as in automatic driving scenarios or advanced auxiliary driving scenarios, it becomes increasingly important to achieve lane-level positioning.

[0003] In related technologies, one way is to achieve vehicle positioning based on RTK (Real Time Kinematic) carrier phase difference technology; another way is to achieve vehicle position recognition based on sensor-laid methods, such as geomagnetic sensors; and still another way is to achieve accurate tracking of vehicle position based on laser radar ranging and three-dimensional point cloud feature scanning technology.

[0004] However, the above related technologies have high dependence on equipment and high-precision map data, high implementation difficulty and high implementation cost. SUMMARY

[0005] Embodiments of the present application provide a lane positioning method, device and equipment, which can reduce the dependence on equipment and high-precision map data during lane positioning, reduce the implementation difficulty and save the implementation cost.

[0006] According to an aspect of an embodiment of the present application, a lane positioning method is provided, and the method comprises:

[0007] obtaining a road image of a target vehicle;

[0008] obtaining lane information corresponding to a position of the target vehicle;

[0009] performing visual recognition processing on the road image to obtain lane line attribute information;

[0010] determining target prior information matched with the lane line attribute information, the target prior information being prior probability information for predicting a lane position under the condition of the lane line attribute information;

[0011] determining posterior probability information of the target vehicle based on the lane information and the target prior information;

[0012] determining a lane position of the target vehicle according to the posterior probability information.

[0013] According to an aspect of an embodiment of the present application, a lane positioning device is provided, and the device comprises:

[0014] a road image acquisition module configured to acquire a road image of a target vehicle;

[0015] a lane information acquisition module configured to acquire lane information corresponding to a position of the target vehicle;

[0016] a visual recognition module configured to perform visual recognition processing on the road image to obtain lane line attribute information;

[0017] a prior information determination module configured to determine target prior information matched with the lane line attribute information, the target prior information being prior probability information of predicting a lane position under a condition of the lane line attribute information;

[0018] a posterior information determination module configured to determine posterior probability information of the target vehicle based on the lane information and the target prior information;

[0019] a lane positioning module configured to determine a lane position of the target vehicle according to the posterior probability information.

[0020] According to an aspect of an embodiment of the present application, a computer device is provided, which comprises a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the lane positioning method.

[0021] According to an aspect of an embodiment of the present application, a computer readable storage medium is provided, which stores at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by a processor to implement the lane positioning method.

[0022] According to an aspect of an embodiment of the present application, a computer program product or a computer program is provided, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to enable the computer device to perform the lane positioning method.

[0023] The technical solution provided by the embodiments of the present application can bring the following beneficial effects:

[0024] The lane line recognition result in the road image is obtained through visual recognition of the road image collected by the vehicle, without relying on other sensor data, and then the prior probability knowledge related to the lane line recognition result and capable of predicting the lane position under the condition of the attribute recognition result is determined from the prior information of the lane line, only the prior probability knowledge and the lane information of the road where the vehicle is located are combined, without relying on high-precision map data and complex vehicle-mounted equipment, the posterior probability of the vehicle on the real lane can be determined, the lane where the vehicle is located is positioned, the scheme implementation difficulty is effectively reduced, the scheme implementation cost is saved, and the lane positioning efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0026] Figure 1 is a schematic diagram of an implementation environment provided by an embodiment of the present application;

[0027] Figure 2 A schematic diagram of a vehicle coordinate system is exemplarily shown;

[0028] Figure 3 A schematic diagram of lane line recognition information is exemplarily shown;

[0029] Figure 4 A schematic diagram of a virtual observation lane is exemplarily shown;

[0030] Figure 5 is a flowchart of a lane positioning method provided by an embodiment of the present application;

[0031] Figure 6 is a flowchart of a lane positioning method provided by an embodiment of the present application;

[0032] Figure 7 is a flowchart of a lane positioning method provided by another embodiment of the present application;

[0033] Figure 8 A technical architecture schematic diagram of a lane-level positioning scheme is exemplarily shown;

[0034] Figure 9 is a block diagram of a lane positioning device provided by an embodiment of the present application;

[0035] Figure 10 is a structural block diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0036] The lane positioning method provided by the embodiments of the present application relates to the fields of map, navigation, intelligent transportation, Internet of Vehicles, intelligent vehicle control technology, automatic driving and advanced auxiliary driving, and the following is briefly described taking a map navigation application as an example.

[0037] With the continuous increase in the number of vehicles, the application of map navigation is also more and more extensive. In the map navigation application scenario, lane-level positioning of vehicles is very important, and lane-level positioning is of great significance for vehicles to determine their own lateral position and formulate navigation strategies. In addition, based on the results of lane-level positioning, lane-level path planning and guidance of vehicles can also be carried out.

[0038] Lane-level positioning is also conducive to improving the vehicle passing rate of the existing road network to alleviate traffic congestion, and on the other hand, it can improve the safety of automobile driving, reduce the rate of traffic accidents, improve traffic safety, reduce energy consumption and environmental pollution, and has great significance for energy transformation and pollution reduction, and relieving traffic congestion in China.

[0039] On the other hand, the lane-level positioning scheme based on vision reduces the technical cost and can better support mass production.

[0040] The lane positioning method provided by the embodiments of the present application relates to artificial intelligence technology, and the following is briefly described to facilitate understanding by those skilled in the art.

[0041] Artificial intelligence (AI) is to use digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which tries to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making.

[0042] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. Artificial intelligence basic technologies generally include technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning, etc.

[0043] Machine Learning (ML) is a multi-disciplinary subject, involving probability theory, statistics, approximation theory, convex analysis, algorithmic complexity theory, etc. It is a specialized study of how computers simulate or implement human learning behavior to acquire new knowledge or skills, reorganize existing knowledge structure, and continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental approach to making computers intelligent. Its applications are widespread in various fields of artificial intelligence. Machine learning and deep learning generally include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and rule-based learning.

[0044] Computer Vision (CV) Computer vision is a science that studies how to make machines "see". More specifically, it refers to using cameras and computers to replace human eyes to identify, track, and measure targets, and further process images to make them more suitable for human observation or transmission to instruments for detection. As a scientific discipline, computer vision researches related theories and technologies, and attempts to establish artificial intelligence systems that can obtain information from images or multidimensional data. Computer vision technology generally includes image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content / behavior recognition, three-dimensional object reconstruction, 3D technology, virtual reality, augmented reality, simultaneous localization and mapping, and other technologies. It also includes common face recognition, fingerprint recognition, and other biometric identification technologies.

[0045] To make the purpose, technical solutions and advantages of the present application clearer, the embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

[0046] Reference is made to Figure 1 which shows a schematic diagram of an implementation environment provided by an embodiment of the present application. The implementation environment can be implemented as an auxiliary driving or automatic driving task system. The implementation environment can include a terminal 10 and a server 20.

[0047] The terminal 10 can be an electronic device such as a mobile phone, a tablet computer, a game console, an e-book reader, a multimedia playback device, a wearable device, a PC (Personal Computer), etc. It can also be an intelligent vehicle-mounted terminal loaded in an unmanned vehicle, an intelligent vehicle, or a general vehicle. It can also be any terminal related to image processing. The terminal 10 can be configured or connected with a camera to capture images. Optionally, a client of an application program can be installed in the terminal 10.

[0048] In the embodiments of the present application, the application program can be any application program capable of image processing. Typically, the application program is a map navigation type application program or a driving type application program. Of course, image processing can also be performed in other types of application programs in addition to the map navigation type application program and the driving type application program. For example, a driving record application program, an interactive entertainment type application program, a game type application program, a virtual reality (VR) type application program, an augmented reality (AR) type application program, and the like, are not limited in the embodiments of the present application.

[0049] The server 20 is configured to provide background services for the terminal 10. For example, the server 20 can be a background server of the application program. The server 20 can be a stand-alone physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and basic cloud computing services such as big data and artificial intelligence platforms. Alternatively, the server 20 simultaneously provides background services for application programs in multiple terminals 10.

[0050] Optionally, the terminal 10 and the server 20 can communicate with each other through the network 30. The terminal 10 and the server 20 can be directly or indirectly connected through wireless communication, which is not limited in the present application.

[0051] Before introducing the method embodiments provided by the present application, the application scenarios, related terms or names that can be involved in the method embodiments of the present application are briefly introduced, so as to facilitate the understanding of the present application by those skilled in the art.

[0052] ADAS (Advanced Driving Assistance System) is a system that uses various sensors (millimeter wave radar, laser radar, monocular / dual camera, and satellite navigation) installed on a vehicle to sense the environment around the vehicle at any time during driving, collect data, identify, detect and track static and dynamic objects, and combine navigation map data to perform system calculation and analysis, so as to pre-alert the driver of potential dangers and effectively increase the comfort and safety of vehicle driving. In the embodiments of the present application, road image data can be collected by a monocular camera installed on a vehicle.

[0053] ADAS data: Traditional ordinary road data is SD (Standard Definition) road data, which mainly records basic road attributes such as road length, number of lanes, direction, and topology. The most information-rich high-definition road data is HD (High Definition) road data, which records very precise and abundant road information, including lane line equations / crosspoint coordinates, utility pole coordinates, road sign locations, and camera / traffic light locations. However, due to the very high production cost of HD road data and its lack of universality, a transitional form of data between SD and HD road data has been proposed, called ADAS data (which can also be named differently, such as SD+ data). Its information richness and accuracy fall between SD and HD road data. Compared to SD road data, ADAS data mainly adds information such as lane line type, color, lane attributes, lane number change points, dashed / solid line change points, and lane topology information. In this application embodiment, the main information used is the total number of lanes in the road data mentioned above. Therefore, the lane positioning technology solution provided in this application embodiment has a very low dependence on road data, and it is even possible to simply add the total number of lanes information to the traditional SD road data.

[0054] GPS (Global Positioning System) is a high-precision radio navigation positioning system based on artificial Earth satellites. It can provide accurate geographical location, vehicle speed, and precise time information anywhere in the world and in near-Earth space. The GPS signal in this application refers to the positioning signal obtained by the vehicle's GPS module through the Global Navigation Satellite System, including coordinates, speed, and time. Current GPS positioning systems can only locate the road / highway where the vehicle is located and the direction of the road, but cannot accurately locate the vehicle's lane. The technical solution provided in this application can accurately locate the vehicle's lane.

[0055] Vehicle Coordinate System (VCS): The vehicle coordinate system is a special three-dimensional moving coordinate system O-xyz used to describe the motion of a vehicle; its origin O is fixed relative to the vehicle's position, and is generally taken as the vehicle's center of mass. For example... Figure 2 As shown, Figure 2 An exemplary schematic diagram of a vehicle coordinate system is shown. When the vehicle is stationary on a level road, the X-axis is parallel to the ground and points forward, the Y-axis points to the driver's left, and the Z-axis passes through the vehicle's center of mass O and points vertically upward. Of course, Figure 2This is merely an illustration; there are various ways to establish a vehicle coordinate system, such as a left-handed system or a right-handed system. The choice of the coordinate system origin also varies, such as the midpoint of the front axle, the midpoint of the vehicle's front, or the midpoint of the rear axle. This application does not limit the choice of origin. It should be noted that the lane line equations mentioned in this application are based on the vehicle coordinate system.

[0056] Lane intercept: The distance from the vehicle's center of gravity to the lane lines on the left and right sides of the vehicle. Vehicles obtain road information ahead by visually recognizing road images, including the lane line equations in the vehicle's coordinate system. The lane line equations may be expressed as quadratic polynomials, cubic polynomials, or other forms, for example: y = d + ax + bx 2 +c*x 3 Or y = d + ax + bx 2 And so on. Where a, b, c, and d are the fitting coefficients of the polynomial.

[0057] In one example, such as Figure 3 As shown, Figure 3 An exemplary schematic diagram of lane line recognition information is shown. Figure 3 In the image, two lane lines, L1 and L2, are identified to the left of the current vehicle; and two lane lines, R1 and R2, are identified to the right of the current vehicle. Through visual recognition processing of the road image, the lane line equations corresponding to L1, L2, R1, and R2 can be determined. Using these lane line equations, the lane intercepts for L1, L2, R1, and R2 can be determined. For example, by taking the coordinate x=0 corresponding to the X-axis in the vehicle coordinate system and substituting it into the lane line equations for L1 and R1, the lane intercept d of L1 can be obtained. L1 and the lane intercept d of R1 R1 .

[0058] Virtual observation lane: A logically abstract lane set up for ease of calculation. The width of the virtual observation lane can be set to different values ​​according to the actual situation, for example, 3.75 meters for urban expressways, 3.5 meters for general urban roads, and so on. Optionally, there is an offset in the correspondence between the virtual observation lane and the real lane. For example, if the offset is selected as 4, then virtual observation lane 7 corresponds to real lane 3, and virtual observation lane 6 corresponds to real lane 2. This application does not limit the width D of a single virtual observation lane or the offset between the virtual observation lane and the real lane.

[0059] In one example, such as Figure 4 As shown, Figure 4An example diagram of a virtual observation lane is shown. It is assumed here that there are 5 virtual observation lanes (the number of virtual observation lanes is optional). Figure 4 (This example uses five virtual observation lanes, including left lane 1, left lane 2, middle lane, right lane 2, and right lane 1.) Figure 3 The double solid lines on the left and right sides represent the road edges, and there are a total of 5 virtual observation lanes. It should be noted that in other possible embodiments, the number of virtual observation lanes can be set to different values, and this application does not limit the selection of the number of virtual observation lanes or the method of representing virtual observation lanes.

[0060] Please refer to Figure 5 The diagram illustrates a flowchart of a lane positioning method according to an embodiment of this application. This method can be applied to a computer device, which refers to an electronic device capable of data calculation and processing. The method may include the following steps (510-560).

[0061] Step 510: Obtain the road image of the target vehicle.

[0062] In some applications, images of the road are captured by a monocular camera installed on the vehicle (such as on the windshield or roof) to obtain road images of the vehicle.

[0063] Optionally, the aforementioned road image can be a road image in front of the vehicle or a panoramic image around the vehicle.

[0064] Step 520: Obtain the lane information corresponding to the location of the target vehicle.

[0065] The vehicle's location can be determined based on its GPS signal, such as its latitude and longitude coordinates.

[0066] Based on the vehicle's location, the local map data corresponding to that location is determined, and from the local map data, the number of lanes on the road where the vehicle is located at the current location can be determined.

[0067] In a possible implementation, historical state information collected by the vehicle in a historical positioning period is tracked, where the historical state information includes, but is not limited to, global positioning system (GPS) information, vehicle control information, vehicle visual perception information, inertial measurement unit (IMU) information, and the like. Then, the positioning information of the vehicle at the current moment is output through certain algorithms and rules. According to the positioning information, a corresponding road position is matched, and then the local map information of the current position is obtained. Optionally, the obtained road information is mainly the total number of lanes at the position of the current vehicle, and therefore, the scheme has low dependence on map data, and is convenient for traditional vehicles to be directly upgraded to lane-level positioning capability.

[0068] Optionally, the local map data can be stored locally in the vehicle terminal, or can be requested and obtained from a server by the vehicle terminal.

[0069] In step 530, visual recognition processing is performed on the road image to obtain lane line attribute information.

[0070] The lane line attribute information refers to data information used to describe the attribute of the lane line. The lane line attribute includes, but is not limited to, the lane line type, the lane line color, and the lane line equation.

[0071] The lane line color includes, but is not limited to, yellow, white, blue, green, gray, black, and other colors. The lane line type includes, but is not limited to, a single solid line, a single dashed line, a double solid line, a double dashed line, a left dashed right solid line, a left solid right dashed line, a guardrail, a curbstone, a road curb, a road edge, and the like.

[0072] In some application scenarios, the AI model can be used to perform image recognition segmentation on the road image to obtain the left and right lane lines of the lane in which the current vehicle is located, and the machine learning technology can be used to obtain the type and color attribute of each lane line in the road image, and the lane line equation information.

[0073] In a possible implementation, the lane line attribute information includes the lane line type and the lane line color of at least one lane line. The at least one lane line can be all or part of the lane lines identified from the road image.

[0074] Optionally, each of the two lane lines on the left and right sides of the vehicle in the road image is identified.

[0075] In a possible implementation, the lane line attribute information further includes the lane line equation of at least one lane line.

[0076] In step 540, target prior information matched with the lane line attribute information is determined.

[0077] The target prior information refers to prior probability information of predicting a lane position under the condition of lane line attribute information.

[0078] The target prior information includes, but is not limited to, prior probability of predicting a lane position under the condition of a lane line being a target lane line type, prior probability of predicting a lane position under the condition of a lane line being a target lane line color, and prior probability of predicting a lane position under the condition of a detected lane line spacing.

[0079] In an example embodiment, as shown in FIG. 5, a flowchart of a lane positioning method provided by an embodiment of the present application is shown. Before step 540, the method further includes step 570. Figure 6

[0080] Step 570: obtaining prior information.

[0081] The prior information is used to represent the association between the attributes of the lane line and the virtual observation lane. The prior information refers to prior probability information of predicting a lane position based on historical road images. In an example embodiment, the prior information is represented by a prior probability matrix, and the association between various attributes of the lane line and the virtual observation lane is represented by the prior probability in the prior probability matrix. For example, when a vehicle is in an edge lane, the probability that the lane line on one side of the vehicle belongs to the road edge is relatively high, which is usually higher than the probability that the lane line belongs to the road edge when the vehicle is in a middle lane.

[0082] In an example embodiment, the prior information includes at least one of a type observation matrix, a color observation matrix, and a spacing observation matrix of the lane line.

[0083] The type observation matrix is used to represent the association between the lane line type and the virtual observation lane.

[0084] The rows of the type observation matrix correspond to the lane line type, the columns of the type observation matrix correspond to the virtual observation lane, and the type observation matrix includes type prior probabilities determined according to the virtual observation lane and the lane line type. The type prior probability refers to the prior probability that the lane line belongs to a certain type when the vehicle is in a certain virtual observation lane. Optionally, the type observation matrix includes type prior probabilities that the lane line belongs to different lane line types under different virtual observation lanes.

[0085] For example, the type prior probability in the mth row and the nth column of the type observation matrix can represent the probability that the lane line belongs to the lane line type corresponding to the mth row when the vehicle is in the virtual observation lane n.

[0086] ​The rows of the color observation matrix correspond to lane line colors, and the columns correspond to virtual observation lanes. The color observation matrix includes color prior probabilities determined based on the virtual observation lanes and lane line colors. A color prior probability refers to the prior probability that a lane line in a given virtual observation lane belongs to a particular color. Optionally, the color observation matrix includes color prior probabilities that lane lines in different virtual observation lanes belong to different lane line colors.

[0087] For example, the prior probability of color in the m-th row and n-th column of the color observation matrix can represent the probability that the lane line is the color of the lane line corresponding to the m-th row if the vehicle is in the virtual observation lane n.

[0088] The rows of the spacing observation matrix correspond to the lane spacing, and the columns correspond to the virtual observation lanes. The spacing observation matrix includes prior probabilities of spacing determined based on the virtual observation lanes and lane spacing. Optionally, the spacing observation matrix includes prior probabilities of spacing used to characterize the accuracy of lane spacing under different virtual observation lanes.

[0089] Correspondingly, such as Figure 6 As shown, step 540 above can be replaced by step 541 below.

[0090] Step 541: Determine the target prior information that matches the lane line attribute information in the prior information.

[0091] In one possible implementation, the lane line attribute information includes the lane line type and lane line color of at least one lane line, and the target prior information includes the type prior probability and color prior probability corresponding to each of the at least one lane line.

[0092] Each lane line corresponds to a type observation matrix and a color observation matrix. The type observation matrix and color observation matrix corresponding to different lane lines can be the same or different.

[0093] In the above embodiments, one implementation of step 541 includes the following process:

[0094] For a target lane line in at least one lane line, determine the type prior probability corresponding to the target type of the target lane line under each virtual observation lane in the prior information. The type prior probability refers to the prior probability that the target lane line belongs to the target type when the target vehicle is in the virtual observation lane. The aforementioned target type is the lane line type of the target lane line. Optionally, based on the target type of the target lane line, determine the type prior probability corresponding to the target type of the target lane line under each virtual observation lane from the type observation matrix corresponding to the target lane line.

[0095] The color prior probability corresponding to the target color of the target lane line under each virtual observation lane in the prior information is determined. The color prior probability refers to the prior probability that the target lane line is of the target color under the virtual observation lane of the target vehicle. Optionally, the type prior probability corresponding to the target color of the target lane line under each virtual observation lane is determined from the color observation matrix corresponding to the target lane line according to the target color of the target lane line.

[0096] The type prior probability and the color prior probability corresponding to each of the at least one lane line are obtained.

[0097] For any lane line, there is a vehicle under each virtual observation lane, and the type prior probability and the color prior probability corresponding to the lane line.

[0098] In another possible implementation, the lane line attribute information further includes a lane line equation of the at least one lane line, and the prior information includes lane line spacing prior information.

[0099] In this implementation, an implementation of the step 570 includes the following process:

[0100] The lane line spacing information is determined according to the lane line equations of the at least two lane lines, and the lane line spacing prior information is determined based on the lane line spacing information.

[0101] The lane line spacing information can represent information of lane line spacing between each lane line, information of lane line spacing on both sides of the vehicle, or information of spacing between fixed lane lines, which is not limited in the embodiments of the application.

[0102] Optionally, the lane line spacing prior information is represented by a spacing prior matrix. Optionally, different lane line spacing corresponds to different spacing prior matrices.

[0103] Optionally, the target prior information includes a spacing prior probability corresponding to the at least one lane line. The spacing prior probability is used to represent the authenticity of the lane line spacing under each virtual observation lane of the vehicle. The spacing prior probability can be represented by a coefficient positively correlated with the true probability, and the relative possibility of the vehicle under each virtual observation lane can be determined by the coefficient.

[0104] Correspondingly, in this implementation, an implementation of the step 541 further includes the following process:

[0105] In a case where the lane line distance information comprises first lane line distance, first distance prior probability corresponding to each virtual observation lane is determined. The first lane line distance is used to represent the distance information of the lane line on the left side of the target vehicle, and the first distance prior probability is used to represent the authenticity of the first lane line distance under the virtual observation lane. Optionally, the first distance prior probability corresponding to each virtual observation lane is determined from the distance prior matrix corresponding to the first lane line distance.

[0106] In a case where the lane line distance information comprises second lane line distance, second distance prior probability corresponding to each virtual observation lane is determined. The second lane line distance is used to represent the distance information of the lane line on the right side of the target vehicle, and the second distance prior probability is used to represent the authenticity of the second lane line distance under the virtual observation lane. Optionally, the second distance prior probability corresponding to each virtual observation lane is determined from the distance prior matrix corresponding to the second lane line distance.

[0107] At step 550, posterior probability information of the target vehicle is determined based on the lane information and the target prior information.

[0108] In an exemplary embodiment, the lane information comprises a total number of lanes, and the posterior probability information comprises posterior probability of the target vehicle on a real lane.

[0109] In an exemplary embodiment, as shown in FIG. 5B, the step 550 comprises the following sub-steps (551-552). Figure 6

[0110] At step 551, a weight coefficient of the target vehicle on each virtual observation lane is determined based on the target prior information.

[0111] The weight coefficient is positively correlated with the probability of the target vehicle being located on each virtual observation lane.

[0112] At step 552, the posterior probability is determined according to the total number of lanes and the weight coefficient.

[0113] In an exemplary embodiment, one implementation of the step 552 comprises the following process:

[0114] A quantity relationship between the total number of lanes and the number of each virtual observation lane is determined.

[0115] According to the quantity relationship, the weight coefficient corresponding to each virtual observation lane is mapped to a real lane to obtain the posterior probability.

[0116] ​In some application scenarios, the total number of real lanes, the number of detected lanes, and the total number of virtual observation lanes are not necessarily the same, and there can be some offset between the three. For example, the total number of lanes is 6, the total number of virtual observation lanes is 5, and the offset is 1. Therefore, different probability mapping rules can be determined according to the number relationship, and then the posterior probability conforming to the situation is determined. The specific probability mapping scheme can refer to the description of step 715 in the next embodiment.

[0117] In step 560, the lane position of the target vehicle is determined according to the posterior probability information.

[0118] In an exemplary embodiment, as shown in FIG. 6, step 560 can be replaced by the following steps. Figure 6

[0119] In step 561, the real lane corresponding to the maximum posterior probability is determined as the lane where the target vehicle is located.

[0120] In summary, the technical scheme provided by the embodiments of the present application obtains the lane line recognition result in the road image by visual recognition of the road image collected by the vehicle, without relying on other sensor data. Then, the prior probability knowledge related to the lane line recognition result and capable of predicting the lane position under the condition of the attribute recognition result is determined from the prior information of the lane line. Only by combining the prior probability knowledge with the lane information of the road where the vehicle is located, without relying on high-precision map data and complex vehicle-mounted equipment, the posterior probability of the vehicle on the real lane can be determined, and the lane where the vehicle is located is positioned. Moreover, the road does not need to be modified, which effectively reduces the difficulty of scheme implementation, saves the cost of scheme implementation, and improves the efficiency of lane positioning.

[0121] The technical scheme provided by the embodiments of the present application will be further described in combination with some implementation scenarios. In some application scenarios, RTK (Real Time Kinematic, Real-Time Kinematic) carrier phase difference technology is used to realize vehicle positioning, and the scheme is to process the difference of two measurement station carrier phase observations in real time. The carrier phase collected by the reference station is sent to the user receiver, and the difference is solved to obtain coordinates, which can obtain centimeter / decimeter level positioning accuracy. The above positioning technology based on RTK has high dependence on equipment, and is not popular. In addition, it also needs to rely on high-precision map data for positioning matching, and the collection cost of high-precision map is also very high, which cannot be used in a large range. Through the technical scheme provided by the embodiments of the present application, high-precision map data and complex vehicle-mounted equipment are not needed, and only the lane number information of the road and the camera installed in the vehicle are needed to realize lane positioning.

[0122] ​In some scenarios, a method based on a paving sensor, such as a geomagnetic sensor, is used to achieve the sensing identification of the vehicle position. The above paving sensor-based solution has a high cost and is difficult to implement in road reconstruction, and does not have conditions for large-scale promotion. Through the technical solution provided in the embodiments of the present application, only the lane number information of the road and the camera installed in the vehicle are needed to perform lane positioning, without the need to modify the road, and the implementation cost is low.

[0123] In some scenarios, a method based on laser radar ranging and three-dimensional point cloud feature scanning technology is used to achieve accurate tracking of the vehicle position, but the cost is high and cannot be commercially applied in a short time. Through the technical solution provided in the embodiments of the present application, only the lane number information of the road and the camera installed in the vehicle are needed to perform lane positioning, with low cost and wide application prospect.

[0124] Please refer to Figure 7 which shows a flowchart of a lane positioning method provided by another embodiment of the present application. The method can be applied in a computer device, which refers to an electronic device with data calculation and processing capability. The method can include the following steps (701-716).

[0125] Step 701, obtaining a road image of a target vehicle.

[0126] Step 702, obtaining a total number of lanes corresponding to a position of the target vehicle.

[0127] Step 703, performing visual recognition processing on the road image to obtain a lane line type and a lane line color of at least one lane line, and lane line spacing information corresponding to the at least one lane line.

[0128] Determining the lane line spacing based on a lane line equation of the at least one lane line.

[0129] Step 704, obtaining a type observation matrix and a color observation matrix in prior information.

[0130] Step 705, determining a spacing observation matrix based on the lane line spacing information.

[0131] Step 706, for a target lane line in the at least one lane line, determining a target type observation matrix and a target color observation matrix corresponding to the target lane line.

[0132] Step 707, finding a type prior probability and a color prior probability corresponding to a target type of the target lane line in each virtual observation lane from the target type observation matrix and the target color observation matrix.

[0133] Step 708, obtaining a type prior probability and a color prior probability corresponding to each lane line.

[0134] Step 709, in the case that the lane line spacing information comprises the first lane line spacing, determining a first spacing observation matrix corresponding to the first lane line spacing.

[0135] Step 710, in the case that the lane line spacing information comprises the second lane line spacing, determining a second spacing observation matrix corresponding to the second lane line spacing.

[0136] Step 711, determining a first spacing prior probability corresponding to each virtual observation lane from the first spacing observation matrix.

[0137] Step 712, determining a second spacing prior probability corresponding to each virtual observation lane from the second spacing observation matrix.

[0138] Step 713, determining a weight coefficient of the target vehicle on each virtual observation lane based on the type prior probability and the color prior probability corresponding to each lane line, the first spacing prior probability and the second spacing prior probability.

[0139] Step 714, determining a quantity relationship between the total number of lanes and the number of each virtual observation lane.

[0140] Step 715, according to the quantity relationship, mapping the weight coefficient corresponding to each virtual observation lane to a real lane to obtain a posterior probability of the target vehicle on each real lane.

[0141] In an exemplary embodiment, the virtual observation lane comprises a first virtual lane, a second virtual lane, a third virtual lane, a fourth virtual lane and a fifth virtual lane.

[0142] If the total number of lanes N is equal to 1, a posterior probability 1 is determined, and N is a positive integer greater than or equal to 1;

[0143] If the total number of lanes N is equal to 2, a posterior probability of a first real lane is determined based on the weight coefficient of the first virtual lane, and a posterior probability of a second real lane is determined based on the weight coefficient of the fifth virtual lane;

[0144] If the total number of lanes N is equal to 3, a posterior probability of a first real lane is determined based on the weight coefficient of the first virtual lane, a posterior probability of a second real lane is determined based on the weight coefficients of the second virtual lane and the fourth virtual lane, and a posterior probability of a third real lane is determined based on the weight coefficient of the fifth virtual lane;

[0145] If the total number of lanes N equals 4, then the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the third real lane is determined based on the weight coefficient of the fourth virtual lane, and the posterior probability of the fourth real lane is determined based on the weight coefficient of the fifth virtual lane.

[0146] If the total number of lanes is equal to 5, then the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the third real lane is determined based on the weight coefficient of the third virtual lane, the posterior probability of the fourth real lane is determined based on the weight coefficient of the fourth virtual lane, and the posterior probability of the fifth real lane is determined based on the weight coefficient of the fifth virtual lane.

[0147] If the total number of lanes N is greater than or equal to 6, then the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the (N-1)th real lane is determined based on the weight coefficient of the fourth virtual lane, the posterior probability of the Nth real lane is determined based on the weight coefficient of the fifth virtual lane, and the posterior probability of the remaining real lanes is determined based on the weight coefficient of the third virtual lane.

[0148] Step 716: Determine the true lane corresponding to the maximum posterior probability as the lane where the target vehicle is located.

[0149] In summary, the technical solution provided in this application, by performing visual recognition on road images collected by vehicles, obtains lane line recognition results in the road images without relying on other sensor data. Then, prior probability knowledge related to the lane line recognition results and capable of predicting the virtual lane position can be determined from the prior information of the lane lines. Only by combining the above-mentioned prior probability knowledge with the lane information of the road where the vehicle is located, without relying on high-precision map data and complex vehicle equipment, the posterior probability of the vehicle on the real lane can be mapped, thereby locating the lane where the vehicle is located. There is no need to modify the road, effectively reducing the difficulty of implementing the solution, saving the implementation cost, and improving the efficiency of lane positioning.

[0150] In one example, such as Figure 8 As shown, it exemplarily illustrates a technical architecture diagram of a lane-level positioning scheme. Figure 8The illustrated lane-level positioning scheme is a lane positioning scheme based on a Bayesian model, and the technical architecture thereof includes a visual processing module, a vehicle positioning module, a base map data module, and a Bayesian observation model, which is further divided into two steps of virtual observation probability calculation and real probability mapping. The task functions of each part are described below.

[0151] 1. Visual processing module

[0152] The visual processing module is used to provide the identification results of lane line types, colors, and lane line equations. The specific implementation of this module is not limited in the present application, as long as the required output content is met.

[0153] The image acquisition unit in the visual processing module includes a monocular camera installed on the vehicle (installed on the windshield, or the roof, etc.) to obtain the road image in front of the vehicle.

[0154] The image processing unit in the visual processing module analyzes and processes the collected road image, and finally outputs the type and color information of the lane lines adjacent to the vehicle (left and right), as well as their confidence (quality).

[0155] The lane line color includes but is not limited to yellow, white, blue, green, gray, black, and other colors, etc. The lane line type includes but is not limited to single solid line, single dashed line, double solid line, double dashed line, left dashed right solid line, left solid right dashed line, guardrail, curbstone, road curb, road edge, and others.

[0156] A possible implementation is as follows:

[0157] (1) Obtain the road image in front of the vehicle through the monocular camera installed on the vehicle (installed on the windshield, or the roof, etc.)

[0158] (2) Element segmentation is performed on the image to identify the information of each lane line adjacent to the vehicle

[0159] (3) Extract the identified lane line information, and obtain the color and type attributes of the lane line through a machine learning model.

[0160] (4) Perform inverse perspective transformation on each adjacent lane line to convert each adjacent lane line in the image from the image coordinate system to the world coordinate system, and perform fitting reconstruction on the transformed lane line to obtain the lane line equation information of each lane line. The lane line equation can be a second-degree polynomial, a third-degree polynomial, or other forms, such as:

[0161] y = d + a * x + b * x 2 + c * x 3

[0162] or, as:

[0163] y = d + a * x + b * x 2

[0164] Optionally, the above-mentioned world coordinate system is a vehicle coordinate system, and the lane line equation is given based on the vehicle coordinate system. The vehicle coordinate system has been described above, and will not be described here.

[0165] In one possible implementation, the vision module only identifies information of at most 4 lane lines, i.e., two on the left side of the vehicle and two on the right side of the vehicle. The first lane line on the left side of the vehicle is L1, the second lane line on the left side of the vehicle is L2, the first lane line on the right side of the vehicle is R1, and the second lane line on the right side of the vehicle is R2. For reference, see Figure 3 If the leftmost side of the road image is a road edge, then the leftmost lane line represents the road edge / road edge / curb. When the vehicle is driving in the leftmost lane, L1 represents the left lane line of the leftmost lane, and L2 represents the road edge / road edge / curb.

[0166] According to the fitting result of the lane line in the world coordinate system, the lane line equation information of each lane line can be obtained, and the intercept information d of each lane line to the vehicle can be determined L1 , d L2 , d R1 , d R2 . Wherein, d L1 , d L2 , d R1 , d R2 are the intercepts of lane lines L1, L2, R1 and R2 respectively.

[0167] 2. Vehicle positioning module

[0168] The vehicle positioning module tracks the historical state information collected in the historical positioning period, wherein the historical state information includes but is not limited to global positioning system (GPS) information, vehicle control information, vehicle visual perception information, and inertial measurement unit (IMU) information. Finally, the module outputs a positioning point information P (such as vehicle position coordinates latitude and longitude) at a current time through certain algorithms and rules. The positioning point information is used to obtain local map data from the map data module, which can be used to obtain local map data information from the map data module and compare with the visual recognition result.

[0169] 3. Map data module

[0170] The map data module matches to a corresponding road position according to the positioning information of the vehicle positioning module, and then obtains local map information of the current position. Optionally, the obtained road information is mainly the total number of lanes at the position of the current vehicle, and therefore, the scheme has low dependence on map data, and is convenient for directly upgrading a traditional vehicle to a lane-level positioning capability.

[0171] 4. Bayesian observation model

[0172] (4.1) Virtual observation probability calculation

[0173] A. Obtain a prior probability matrix

[0174] In a possible implementation, the number of virtual observation lanes is 5, that is, there are 5 observation states of the virtual observation lane. Optionally, the number of virtual observation lanes is optional, and in this embodiment, only 5 is taken as an example for description. Specifically, the virtual observation lane includes a left 1 lane, a left 2 lane, a middle lane, a right 2 lane, and a right 1 lane.

[0175] The number of virtual observation lanes is determined to fix the dimension of the Bayesian observation matrix, so as to facilitate calculation.

[0176] As known from the above description of the visual recognition module, the real lane lines obtained by visual recognition include a total of 4 lane lines, which are L1, L2, R1, and R2.

[0177] An observation matrix is designed for the lane line type of each lane line, which is ObsProbL1Type, ObsProbL2Type, ObsProbR1Type, and ObsProbR2Type. Among them, ObsProbL1Type, ObsProbL2Type, ObsProbR1Type, and ObsProbR2Type are type observation matrices corresponding to L1, L2, R1, and R2. The number of rows of the four type observation matrices is the total number of identifiable lane line types, and the number of columns is 5, indicating 5 virtual observation lanes.

[0178] An observation matrix is designed for the lane line color of each lane line, which is ObsProbL1Color, ObsProbL2Color, ObsProbR1Color, and ObsProbR2Color. Among them, ObsProbL1Color, ObsProbL2Color, ObsProbR1Color, and ObsProbR2Color are color observation matrices corresponding to L1, L2, R1, and R2. The number of rows of the four color observation matrices is the total number of identifiable lane line colors, and the number of columns is 5, indicating 5 virtual observation lanes.

[0179] The distance interval of the two lane lines L1, L2 on the left side of the vehicle and the distance interval of the two lane lines R1, R2 on the right side of the vehicle are designed into two observation matrices, which are ObsProbLeftDist and ObsProbRightDist. Among them, ObsProbLeftDist is a distance observation matrix corresponding to the lane line interval of L1 and L2, and ObsProbRightDist is a distance observation matrix corresponding to the lane line interval of R1 and R2. The two distance observation matrices are 1 row and 5 columns, and the probabilities in the matrix are calculated through the lane line equations of L1, L2, R1 and R2 obtained in real time and certain rules.

[0180] For the 8 observation matrices of types and colors (4 type observation matrices and 4 color observation matrices), their values are fixed and obtained according to prior knowledge (empirical knowledge) or statistical results (through a large number of road videos and pictures). The matrix is normalized by column, the i-th row and j-th column of the matrix represents the probability of the lane line type / color corresponding to the i-th row under the condition that the vehicle is in the virtual observation lane j. Alternatively, it can also be non-normalized, and the relative probability size is represented by the value size of each column of the matrix, which has little effect on the calculation of probability and the final result.

[0181] In order to be general, ObsProbL1Type and ObsProbL2Type are taken as examples for brief description. In one possible implementation, the lane line types that can be recognized by the visual recognition module include single solid line, single dashed line and road edge, and the observation matrices ObsProbL1Type and ObsProbL2Type are both 3 rows and 5 columns. A feasible representation is as follows:

[0182] ObsProbL1Type:

[0183] Left 1 lane Left 2 lane Middle lane Right 2 lane Right 1 lane Single solid line 0.5 0.1 0.1 0.1 0.1 Single dashed line 0.1 0.8 0.8 0.8 0.8 Road edge 0.4 0.1 0.1 0.1 0.1

[0184] ObsProbL2Type:

[0185] Left 1 lane Left 2 lane Middle lane Right 2 lane Right 1 lane Single solid line 0.2 0.4 0.1 0.1 0.1 Single dashed line 0.1 0.1 0.8 0.8 0.8 Road edge 0.7 0.5 0.1 0.1 0.1

[0186] Note: When the vehicle is in the left 1 lane, the first lane line on the left side of the vehicle is generally a solid line, and the second lane line is a road edge marking such as a curbstone or a guardrail. Therefore, according to the above type observation matrix, when the vehicle is in the left 1 lane, the probability of L1 being a single solid line is the largest (0.5), and the probability of L2 being a road edge is the largest (0.7).

[0187] The distance observation matrix is also explained below through a specific example. The rows of the matrix ObsProbLeftDist and ObsProbRightDist are supposed to represent the distance between lane lines in theory, but the distance between lane lines is not a discrete value, but a continuous value, so it is difficult to represent the matrix with finite rows. Therefore, only one row is used here to represent the relative possibility of the distance between lane lines appearing a given distance between lane lines under different virtual observation lanes. The term relative possibility is used here because the value of the matrix is a coefficient representation, not a real probability, but this does not affect the final calculation.

[0188] In one possible implementation, let the distance between lane lines L1 and L2 be LDist = abs(d L1 - d L2 ), in meters (m), and abs represents the absolute value operation; the distance between lane lines R1 and R2 be RDist = abs(d R1 - d R2 ), and one possible design of the distance probability matrix is as follows:

[0189] For ObsProbLeftDist, ObsProbLeftDist[1], ObsProbLeftDist[2], ObsProbLeftDist[3], ObsProbLeftDist[4], and ObsProbLeftDist[5] are the values of each column of ObsProbLeftDist, respectively.

[0190]

[0191]

[0192]

[0193]

[0194]

[0195] For ObsProbRightDist, ObsProbRightDist[1], ObsProbRightDist[2], ObsProbRightDist[3], ObsProbRightDist[4], and ObsProbRightDist[5] are the values of each column of ObsProbRightDist, respectively.

[0196]

[0197]

[0198]

[0199]

[0200]

[0201] where Width is an optional parameter, representing the lane width, such as 3.5m, 3.75m. C is an optional coefficient adjustment parameter, such as 8.0, 10.0.

[0202] B. Virtual lane probability calculation

[0203] The optimization goal in this embodiment is to determine the lane with the maximum probability given the lane line color, type and equation. That is:

[0204]

[0205] where LaneIndex represents the lane with the maximum probability, i represents the virtual observation lane number, and the value is left 1 lane, left 2 lane, middle lane, right 2 lane, right 1 lane. For convenience of representation, we use i = 1, 2, 3, 4, 5 to represent them respectively. P(i|L1, L2, R1, R2) represents the conditional probability of the virtual observation lane where the vehicle is located given the lane lines L1, L2, R1, R2.

[0206] According to Bayes' formula, the formula can be transformed into:

[0207]

[0208] Here it is assumed that the initial probability of the vehicle in each virtual observation lane is equal, i.e. P(i) corresponding to each lane is equal, so the influence of P(i) can be removed, and the following is obtained:

[0209]

[0210] P(L1, L2, R1, R2) represents the probability of observing L1, L2, R1, R2 appearing in a given type, color, equation. P(L1, L2, R1, R2) is a constant, so the influence of the common denominator P(L1, L2, R1, R2) on the maximum value can be ignored, and the above formula is further transformed into:

[0211]

[0212] Then, in some application scenarios, the probabilities between each lane line can be considered as independent, and the joint probability P(L1, L2, R1, R2 | i) can be expressed as the product of each probability:

[0213]

[0214] Further, the color and type of each lane line can be considered as independent observation variables, and the above formula is further transformed as:

[0215]

[0216] wherein, L1 type is the type of lane line L1, L1 color is the color of lane line L1, L2 type is the type of lane line L2, L2 color is the color of lane line L2, R1 type is the type of lane line R1, R1 color is the color of lane line R1, R2 type is the type of lane line R2, R2 color is the color of lane line R2, and LDist is the lane line distance between L1 and L2, and RDist is the lane line distance between R1 and R2.

[0217] The final probability calculation becomes the product of the above 10 probabilities, and each of the 10 probabilities can be obtained from the prior probability matrix in Part A.

[0218] For example, the value of P(L1 type | i) is the value of the i-th virtual observation lane line type L1 type corresponding to the row in the matrix ObsProbL1Type.

[0219] According to the product of the 10 probabilities corresponding to the virtual observation lane in the above formula, the probability weight coefficients corresponding to each of the 5 virtual observation lanes can be obtained, including: the weight coefficient ObsProbVirtual[1] of the left 1 lane, the weight coefficient ObsProbVirtual[2] of the left 2 lane, the weight coefficient ObsProbVirtual[3] of the middle lane, the weight coefficient ObsProbVirtual[4] of the right 2 lane, and the weight coefficient ObsProbVirtual[5] of the right 1 lane.

[0220] Optionally, for the i-th virtual observation lane, the weight coefficient ObsProbVirtual[i] can be determined by the following formula.

[0221]

[0222] Next, ObsProbVirtual[i] is normalized by the following equation.

[0223]

[0224] Optionally, the normalized value is taken as the probability of the vehicle in the 5 virtual observation lanes (virtual observation probability).

[0225] (4.2) Real lane probability mapping

[0226] The probability of the vehicle in the 5 virtual observation lanes is obtained from (4.1), but the number of real lanes is variable, so it is necessary to map the virtual observation probability to the real lane probability.

[0227] The total number of lanes in which the current vehicle is located obtained from the local road data is N, N is a positive integer greater than or equal to 1. Next, different probability mapping schemes are performed according to the value of N, see the following several cases.

[0228] Case 1 (N = 1):

[0229] If N = 1, output LaneIndex = 1, and probability ObsProbReal[1] = 1.0. ObsProbReal[1] represents the posterior probability of the vehicle in real lane 1. When there is only one lane on the current road, there is no dispute about the lane positioning result, and the vehicle must be in that lane.

[0230] Case 2 (N = 2):

[0231] If N = 2, ObsProbReal[1] = ObsProbVirtual[1], and ObsProbReal[2] = ObsProbVirtual[5]. ObsProbReal[2] represents the posterior probability of the vehicle in real lane 2.

[0232] Optionally, the posterior probability ObsProbReal is normalized.

[0233] Correspondingly, the real lane LaneIndex in which the vehicle is located is determined by the following equation.

[0234]

[0235] Case 3 (N = 3):

[0236] If N=3, ObsProbReal[l]=ObsProbVirtual[l], ObsProbReal[2]=ObsProbVirtual[2]+ObsProbVirtual[4], and ObsProbReal[3]=ObsProbVirtual[5]. ObsProbReal[3] represents the posterior probability of the vehicle in the real lane 3.

[0237] Optionally, the posterior probability ObsProbReal is normalized.

[0238] Correspondingly, the real lane LaneIndex in which the vehicle with the maximum probability is determined by the following formula.

[0239]

[0240] Case 4 (N=4):

[0241] If N=4, ObsProbReal[l]=ObsProbVirtual[l], ObsProbReal[2]=ObsProbVirtual[2], ObsProbReal[3]=ObsProbVirtual[4], and ObsProbReal[4]=ObsProbVirtual[5]. ObsProbReal[4] represents the posterior probability of the vehicle in the real lane 4.

[0242] Optionally, the posterior probability ObsProbReal is normalized.

[0243] Correspondingly, the real lane LaneIndex in which the vehicle with the maximum probability is determined by the following formula.

[0244]

[0245] Case 5 (N=5):

[0246] If N=5, ObsProbReal[l]=ObsProbVirtual[l], ObsProbReal[2]=ObsProbVirtual[2], ObsProbReal[3]=ObsProbVirtual[3], ObsProbReal[4]=ObsProbVirtual[4], and ObsProbReal[5]=ObsProbVirtual[5]. ObsProbReal[5] represents the posterior probability of the vehicle in the real lane 5.

[0247] Optionally, the posterior probability ObsProbReal is normalized.

[0248] Correspondingly, the real lane in which the vehicle with the maximum probability is determined by the following formula: LaneIndex.

[0249]

[0250] Case 6 (N≥6):

[0251] If N≥6, ObsProbReal[1]=ObsProbVirtual[1], ObsProbReal[2]=ObsProbVirtual[2], ObsProbReal[N-1]=ObsProbVirtual[4], and ObsProbReal[N]=ObsProbVirtual[5]. ObsProbReal[N-1] and ObsProbReal[N] represent the posterior probabilities of the vehicle in the real lane N-1 and the real lane N, respectively.

[0252] Then, the probabilities of the middle lanes in the virtual observation lane are distributed to the remaining lanes. Optionally, the distribution is performed according to the principle that the middle lane is high and the two side lanes are low. A feasible manner is as follows:

[0253] For the lane j that is not assigned a weight, j is a lane other than the real lane 1, the real lane 2, the real lane N-1, and the real lane N, and the assigned weight of the lane j is

[0254] wherein, indicates the lane number of the middle lane, and the above rule aims to make the weight close to the middle lane large and the weight far from the middle lane small. The embodiments of the present application do not limit the weight distribution rule.

[0255] Then, normalization processing is performed on all weight(j).

[0256] Then, for all unassigned lanes j, the corresponding posterior probability ObsProbReal[j] is determined by the following formula.

[0257] ObsProbReal[j]=ObsProbVirtual[3]×weight(j)

[0258] At this point, all real lanes are assigned with probability values.

[0259] Optionally, normalization can also be performed on ObsProbReal.

[0260] The finally output LaneIndex is the lane-level positioning result output by the Bayesian model.

[0261]

[0262] It should be noted that the embodiment can be based on a single video frame to perform lane-level positioning. In actual application, context information can be used to track lane-level to obtain more accurate positioning results. On this basis, HMM (Hidden Markov Model) can be added to track positioning information to obtain more accurate positioning results.

[0263] In the example embodiment, other observation variables can also be added to perform probability calculation, and the observation variables are not limited to the type, color and equation of the lane line. For example, the number of left and right lane lines recalled is added to the probability of lane positioning. For another example, the number of lane lines identified is increased from two on each side to three or more on each side, and the embodiment of the present application is not limited in this regard.

[0264] As can be seen from the technical solutions provided in the embodiment, the technical solutions of the embodiment only rely on the road image collected by the monocular camera loaded on the vehicle in hardware, without panoramic image, other sensor data, and high-precision data as support, which is simple to use, low in landing cost, and high in precision. An efficient lane-level positioning solution is provided, which provides a technical prerequisite for lane-level navigation and advanced auxiliary driving.

[0265] The following is an apparatus embodiment of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the apparatus embodiments of the present application, please refer to the method embodiments of the present application.

[0266] Please refer to Figure 9 which shows a block diagram of a lane positioning apparatus provided by an embodiment of the present application. The apparatus has the functions of implementing the lane positioning method described above, which can be implemented by hardware or corresponding software executed by hardware. The apparatus can be a computer device or can be arranged in a computer device. The apparatus 900 can include a road image acquisition module 910, a lane information acquisition module 920, a visual recognition module 930, a prior information determination module 940, a posterior information determination module 950, and a lane positioning module 960.

[0267] The road image acquisition module 910 is configured to acquire a road image of a target vehicle.

[0268] The lane information acquisition module 920 is configured to acquire lane information corresponding to a position of the target vehicle.

[0269] The visual recognition module 930 is configured to perform visual recognition processing on the road image to obtain lane line attribute information.

[0270] The prior information determination module 940 is configured to determine target prior information matched with the lane line attribute information, where the target prior information refers to prior probability information of predicting a lane position under the condition of the lane line attribute information.

[0271] The posterior information determination module 950 is configured to determine posterior probability information of the target vehicle based on the lane information and the target prior information.

[0272] The lane positioning module 960 is configured to determine a lane position of the target vehicle according to the posterior probability information.

[0273] In an example embodiment, the apparatus 900 further includes a prior information acquisition module (not shown in the figure).

[0274] The prior information acquisition module is configured to acquire prior information, where the prior information is used to represent an association between attributes of lane lines and virtual observation lanes.

[0275] The prior information determination module 940 is further configured to:

[0276] determine target prior information matched with the lane line attribute information in the prior information.

[0277] In an example embodiment, the lane line attribute information includes lane line types and lane line colors of at least one lane line, and the target prior information includes type prior probabilities and color prior probabilities respectively corresponding to the at least one lane line.

[0278] The prior information determination module 940 includes a type probability determination unit, a color probability determination unit, and a probability acquisition unit.

[0279] The type probability determination unit is configured to, for a target lane line in the at least one lane line, determine a type prior probability corresponding to a target type of the target lane line under each virtual observation lane in the prior information, where the type prior probability refers to a prior probability that the target vehicle is in the virtual observation lane and the target lane line belongs to the target type.

[0280] The color probability determination unit is configured to determine a color prior probability corresponding to a target color of the target lane line under each virtual observation lane in the prior information, where the color prior probability refers to a prior probability that the target vehicle is in the virtual observation lane and the target lane line is of the target color.

[0281] The probability acquisition unit is configured to obtain the type prior probabilities and the color prior probabilities respectively corresponding to the at least one lane line.

[0282] In an example embodiment, the lane line attribute information further comprises a lane line equation of the at least one lane line, and the prior information comprises lane line spacing prior information, and the prior information acquisition module comprises a lane line spacing determination unit and a spacing information determination unit.

[0283] The lane line spacing determination unit is configured to determine lane line spacing information according to the lane line equations of the at least two lane lines.

[0284] The spacing information determination unit is configured to determine the lane line spacing prior information based on the lane line spacing information.

[0285] In an example embodiment, the target prior information comprises spacing prior probability corresponding to the at least one lane line, and the prior information determination module 940 further comprises a spacing probability determination unit.

[0286] The spacing probability determination unit is configured to determine first spacing prior probability corresponding to each virtual observation lane in a case where the lane line spacing information comprises first lane line spacing, the first lane line spacing being used to represent spacing information of a left lane line of the target vehicle, and the first spacing prior probability being used to represent authenticity of the first lane line spacing of the target vehicle in the virtual observation lane.

[0287] The spacing probability determination unit is further configured to determine second spacing prior probability corresponding to each virtual observation lane in a case where the lane line spacing information comprises second lane line spacing, the second lane line spacing being used to represent spacing information of a right lane line of the target vehicle, and the second spacing prior probability being used to represent authenticity of the second lane line spacing of the target vehicle in the virtual observation lane.

[0288] In an example embodiment, the prior information comprises at least one of a lane line type observation matrix, a color observation matrix, and a spacing observation matrix.

[0289] The rows of the type observation matrix correspond to lane line types, the columns of the type observation matrix correspond to the virtual observation lanes, and the type observation matrix comprises type prior probabilities determined according to the virtual observation lanes and the lane line types.

[0290] The rows of the color observation matrix correspond to lane line colors, the columns of the color observation matrix correspond to the virtual observation lanes, and the color observation matrix comprises color prior probabilities determined according to the virtual observation lanes and the lane line colors.

[0291] The rows of the spacing observation matrix correspond to lane line spacings, and the columns of the spacing observation matrix correspond to the virtual observation lanes, and the spacing observation matrix comprises spacing prior probabilities determined according to the virtual observation lanes and the lane line spacings.

[0292] In an example embodiment, the lane information comprises a total number of lanes, the posterior probability information comprises a posterior probability of the target vehicle being in a real lane, and the posterior information determination module 950 comprises a virtual probability determination unit and a posterior probability determination unit (not shown in the figure).

[0293] The virtual probability determination unit is configured to determine a weight coefficient of the target vehicle in each virtual observation lane based on the target prior information, the weight coefficient being positively correlated with a probability of the target vehicle being in the virtual observation lane.

[0294] The posterior probability determination unit is configured to determine the posterior probability according to the total number of lanes and the weight coefficient.

[0295] In an example embodiment, the posterior probability determination unit comprises a quantity relationship determination subunit and a probability mapping subunit.

[0296] The quantity relationship determination subunit is configured to determine a quantity relationship between the total number of lanes and a number of the virtual observation lanes.

[0297] The probability mapping subunit is configured to map the weight coefficient corresponding to each virtual observation lane to the real lane according to the quantity relationship to obtain the posterior probability.

[0298] In an example embodiment, the virtual observation lanes comprise a first virtual lane, a second virtual lane, a third virtual lane, a fourth virtual lane, and a fifth virtual lane, and the probability mapping subunit is specifically configured to:

[0299] If the total number of lanes N is equal to 1, the posterior probability is determined to be 1, and the N is a positive integer greater than or equal to 1;

[0300] If the total number of lanes N is equal to 2, a posterior probability of a first real lane is determined based on a weight coefficient of the first virtual lane, and a posterior probability of a second real lane is determined based on a weight coefficient of the fifth virtual lane;

[0301] If the total number of lanes N is equal to 3, a posterior probability of a first real lane is determined based on a weight coefficient of the first virtual lane, a posterior probability of a second real lane is determined based on weight coefficients of the second virtual lane and the fourth virtual lane, and a posterior probability of a third real lane is determined based on a weight coefficient of the fifth virtual lane;

[0302] If the total number of lanes N is equal to 4, the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the third real lane is determined based on the weight coefficient of the fourth virtual lane, and the posterior probability of the fourth real lane is determined based on the weight coefficient of the fifth virtual lane;

[0303] If the total number of lanes is equal to 5, the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the third real lane is determined based on the weight coefficient of the third virtual lane, the posterior probability of the fourth real lane is determined based on the weight coefficient of the fourth virtual lane, and the posterior probability of the fifth real lane is determined based on the weight coefficient of the fifth virtual lane.

[0304] If the total number of lanes N is greater than or equal to 6, the posterior probability of the first real lane is determined based on the weight coefficient of the first virtual lane, the posterior probability of the second real lane is determined based on the weight coefficient of the second virtual lane, the posterior probability of the N-1th real lane is determined based on the weight coefficient of the fourth virtual lane, the posterior probability of the Nth real lane is determined based on the weight coefficient of the fifth virtual lane, and the posterior probability of the remaining real lane is determined based on the weight coefficient of the third virtual lane.

[0305] In summary, the technical scheme provided by the embodiments of the present application obtains the lane line recognition result in the road image collected by the vehicle through visual recognition of the road image, without relying on other sensor data, and then determines the prior probability knowledge of the lane position related to the lane line recognition result under the condition of the attribute recognition result. Only the prior probability knowledge and the lane information of the road where the vehicle is located need to be combined, without relying on high-precision map data and complex vehicle-mounted equipment, the posterior probability of the vehicle in the real lane can be determined, the lane where the vehicle is located is positioned, and the road does not need to be modified, which effectively reduces the difficulty of scheme implementation, saves the cost of scheme implementation, and improves the efficiency of lane positioning.

[0306] It should be noted that the device provided in the above embodiments is only used as an example to divide the above functional modules in realizing its functions, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above described functions. In addition, the device and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process is described in detail in the method embodiments, which will not be described here.

[0307] Reference is made to Figure 10 which shows a structural block diagram of a computer device provided by an embodiment of the present application. The computer device can be a vehicle-mounted computing terminal. The computer device is used to implement the lane positioning method provided by the above-embodiments. Specifically,

[0308] Generally, the computer device 1000 includes a processor 1001 and a memory 1002.

[0309] The processor 1001 can include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 1001 can be implemented in at least one of a hardware form of a DSP (Digital Signal Processing), a FPGA (Field Programmable Gate Array), a PLA (Programmable Logic Array). The processor 1001 can also include a main processor and a coprocessor. The main processor is a processor for processing data in an awake state, also known as a CPU (Central Processing Unit). The coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 1001 can be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing the content required to be displayed by the display screen. In some embodiments, the processor 1001 can also include an AI (Artificial Intelligence) processor for processing machine learning-related computing operations.

[0310] The memory 1002 can include one or more computer-readable storage media, which can be non-transitory. The memory 1002 can also include a high-speed random access memory, and a non-volatile memory such as one or more disk storage devices, flash storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1002 is used to store at least one instruction, at least one program, a code set or an instruction set, which is configured to be executed by one or more processors to implement the lane positioning method described above.

[0311] In some embodiments, the computer device 1000 can further optionally include a peripheral device interface 1003 and at least one peripheral device. The processor 1001, the memory 1002, and the peripheral device interface 1003 can be connected through a bus or a signal line. Each peripheral device can be connected to the peripheral device interface 1003 through a bus, a signal line, or a circuit board. Specifically, the peripheral device includes at least one of a radio frequency circuit 1004, a touch display screen 1005, a camera assembly 1006, an audio circuit 1007, a positioning assembly 1008, and a power supply 1009.

[0312] Those skilled in the art can understand that the structure shown in the foregoing embodiments does not constitute a limitation on the computer device 1000, and the computer device 1000 can include more or fewer components than those shown in the drawings, or combine certain components, or adopt a different arrangement of components. Figure 10

[0313] In an example embodiment, a computer readable storage medium is also provided, and the storage medium stores at least one instruction, at least one program, a code set, or an instruction set. When the at least one instruction, the at least one program, the code set, or the instruction set is executed by a processor, the lane positioning method described above is implemented.

[0314] Optionally, the computer readable storage medium can include a ROM (Read Only Memory), a RAM (Random Access Memory), a SSD (Solid State Drives), or an optical disc, etc. The RAM can include a ReRAM (Resistance Random Access Memory) and a DRAM (Dynamic Random Access Memory).

[0315] In an example embodiment, a computer program product or a computer program is also provided, and the computer program product or the computer program includes computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the lane positioning method described above.

[0316] ​It should be understood that "multiple" mentioned herein refers to two or more than two. The "and / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents that the associated objects before and after it are in an "or" relationship. In addition, the step numbers described herein only exemplarily show a possible execution order between steps, and in some other embodiments, the above steps can also be executed in a non-numbered order, such as two different numbered steps being executed at the same time, or two different numbered steps being executed in an order opposite to that shown in the figure, and the embodiments of the present application are not limited in this regard.

[0317] The above only describes exemplary embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A lane positioning method characterized by, The method comprises: acquiring a road image of a target vehicle; acquiring lane information corresponding to a position of the target vehicle; performing visual recognition processing on the road image to obtain lane line attribute information; the lane line attribute information comprises a lane line type and a lane line color of at least one lane line; acquiring prior information, the prior information being used to represent an association between a lane line attribute and a virtual observation lane; determining target prior information in the prior information that matches the lane line attribute information, the target prior information being prior probability information for predicting a lane position under the condition of the lane line attribute information; based on the lane information and the target prior information, determining posterior probability information of the target vehicle; determining a lane position of the target vehicle according to the posterior probability information; wherein the determination of the target prior information in the prior information that matches the lane line attribute information comprises: for a target lane line in the at least one lane line, determining a type prior probability corresponding to a target type of the target lane line under each virtual observation lane in the prior information, the type prior probability being a prior probability that the target lane line belongs to the target type under the virtual observation lane of the target vehicle; determining a color prior probability corresponding to a target color of the target lane line under each virtual observation lane in the prior information, the color prior probability being a prior probability that the target lane line is of the target color under the virtual observation lane of the target vehicle; obtaining the type prior probability and the color prior probability corresponding to each of the at least one lane line to determine the target prior information in the prior information that matches the lane line attribute information, the target prior information being prior probability information for predicting a lane position under the condition of the lane line attribute information, the target prior information comprising the type prior probability and the color prior probability corresponding to each of the at least one lane line.

2. The method of claim 1, wherein, The lane line attribute information further comprises a lane line equation of the at least one lane line, and the prior information comprises lane line spacing prior information, and the acquisition of the prior information comprises: determining lane line spacing information according to lane line equations of at least two lane lines; based on the lane line spacing information, determining the lane line spacing prior information.

3. The method of claim 2, wherein, The target prior information comprises a spacing prior probability corresponding to the at least one lane line, and the determination of the target prior information in the prior information that matches the lane line attribute information comprises: in a case where the lane line spacing information comprises a first lane line spacing, determining a first spacing prior probability corresponding to each virtual observation lane, the first lane line spacing being used to represent spacing information of a lane line on a left side of the target vehicle, and the first spacing prior probability being used to represent authenticity of the first lane line spacing under the virtual observation lane of the target vehicle. In a case where the lane line distance information comprises second lane line distance, a second distance prior probability corresponding to each virtual observation lane is determined, the second lane line distance is used to represent distance information of a right lane line of the target vehicle, and the second distance prior probability is used to represent authenticity of the second lane line distance of the target vehicle in the virtual observation lane.

4. The method according to any one of claims 1 to 3, characterized in that, The prior information comprises at least one of a type observation matrix, a color observation matrix and a distance observation matrix of a lane line. The rows of the type observation matrix correspond to lane line types, the columns of the type observation matrix correspond to the virtual observation lanes, and the type observation matrix comprises type prior probabilities determined according to the virtual observation lanes and the lane line types. The rows of the color observation matrix correspond to lane line colors, the columns of the color observation matrix correspond to the virtual observation lanes, and the color observation matrix comprises color prior probabilities determined according to the virtual observation lanes and the lane line colors. The rows of the distance observation matrix correspond to lane line distances, the columns of the distance observation matrix correspond to the virtual observation lanes, and the distance observation matrix comprises distance prior probabilities determined according to the virtual observation lanes and the lane line distances.

5. The method of claim 4, wherein, The lane information comprises a total number of lanes, the posterior probability information comprises a posterior probability of the target vehicle in a real lane, and the posterior probability information of the target vehicle is determined based on the lane information and the target prior information, comprising: Based on the target prior information, a weight coefficient of the target vehicle in each virtual observation lane is determined, the weight coefficient is positively correlated with a probability that the target vehicle is located in each virtual observation lane; The posterior probability is determined according to the total number of lanes and the weight coefficient.

6. The method of claim 5, wherein, The posterior probability is determined according to the total number of lanes and the weight coefficient, comprising: A quantity relationship between the total number of lanes and a number of the virtual observation lanes is determined; According to the quantity relationship, the weight coefficient corresponding to each virtual observation lane is mapped to the real lane to obtain the posterior probability.

7. A lane positioning device, characterized by, The device comprises: a road image acquisition module configured to acquire a road image of a target vehicle; a lane information acquisition module configured to acquire lane information corresponding to a position of the target vehicle; a visual recognition module configured to perform visual recognition processing on the road image to obtain lane line attribute information, the lane line attribute information comprising a lane line type and a lane line color of at least one lane line; a prior information acquisition module configured to acquire prior information, the prior information being used to represent an association relationship between attributes of a lane line and a virtual observation lane; a prior information determination module configured to determine target prior information matched with the lane line attribute information in the prior information, the target prior information being prior probability information of a predicted lane position under a condition of the lane line attribute information; a posterior information determination module configured to determine posterior probability information of the target vehicle based on the lane information and the target prior information. The lane positioning module is configured to determine a lane position of the target vehicle according to the posterior probability information. The prior information determination module comprises a type probability determination unit, a color probability determination unit, and a probability acquisition unit. The type probability determination unit is configured to determine, for a target lane line in the at least one lane line, a type prior probability corresponding to a target type of the target lane line in each virtual observation lane in the prior information, the type prior probability being a prior probability that the target lane line belongs to the target type in the virtual observation lane of the target vehicle. The color probability determination unit is configured to determine a color prior probability corresponding to a target color of the target lane line in each virtual observation lane in the prior information, the color prior probability being a prior probability that the target lane line is of the target color in the virtual observation lane of the target vehicle. The probability acquisition unit is configured to obtain target prior information in the prior information that matches the lane line attribute information according to the type prior probability and the color prior probability corresponding to each of the at least one lane line, the target prior information being prior probability information for predicting a lane position under the condition of the lane line attribute information, and the target prior information comprising the type prior probability and the color prior probability corresponding to each of the at least one lane line.

8. The apparatus of claim 7, wherein, The lane line attribute information further comprises lane line equations of the at least one lane line, and the prior information acquisition module is configured to: determine lane line spacing information according to the lane line equations of the at least two lane lines; and determine lane line spacing prior information based on the lane line spacing information.

9. The apparatus of claim 8, wherein, The target prior information comprises spacing prior probability corresponding to the at least one lane line, and the prior information determination module 940 comprises a spacing probability determination unit, which is configured to: in a case where the lane line spacing information comprises a first lane line spacing, determine a first spacing prior probability corresponding to each virtual observation lane, the first lane line spacing being used to represent spacing information of a lane line on a left side of the target vehicle, and the first spacing prior probability being used to represent authenticity of the first lane line spacing in the virtual observation lane of the target vehicle; and in a case where the lane line spacing information comprises a second lane line spacing, determine a second spacing prior probability corresponding to each virtual observation lane, the second lane line spacing being used to represent spacing information of a lane line on a right side of the target vehicle, and the second spacing prior probability being used to represent authenticity of the second lane line spacing in the virtual observation lane of the target vehicle.

10. The apparatus of any one of claims 7 to 9, wherein, The prior information comprises at least one of a type observation matrix, a color observation matrix, and a spacing observation matrix of a lane line. The rows of the type observation matrix correspond to lane line types, the columns of the type observation matrix correspond to the virtual observation lanes, and the type observation matrix comprises type prior probabilities determined according to the virtual observation lanes and the lane line types. The rows of the color observation matrix correspond to lane line colors, the columns of the color observation matrix correspond to the virtual observation lanes, and the color observation matrix comprises color prior probabilities determined according to the virtual observation lanes and the lane line colors; The rows of the distance observation matrix correspond to lane line distances, the columns of the distance observation matrix correspond to the virtual observation lanes, and the distance observation matrix comprises distance prior probabilities determined according to the virtual observation lanes and the lane line distances.

11. The apparatus of claim 10, wherein, The lane information comprises a total number of lanes, the posterior probability information comprises a posterior probability of the target vehicle being in a real lane, and the posterior information determination module comprises a virtual probability determination unit and a posterior probability determination unit: The virtual probability determination unit is configured to determine weight coefficients of the target vehicle in the virtual observation lanes based on the target prior information, the weight coefficients being positively correlated with probabilities of the target vehicle being in the virtual observation lanes; The posterior probability determination unit is configured to determine the posterior probability based on the total number of lanes and the weight coefficients.

12. The apparatus of claim 11, wherein, The posterior information determination unit comprises a number relationship determination subunit and a probability mapping subunit; The number relationship determination subunit is configured to determine a number relationship between the total number of lanes and a number of the virtual observation lanes; The probability mapping subunit is configured to map the weight coefficients corresponding to the virtual observation lanes to the real lane according to the number relationship to obtain the posterior probability.

13. A computer device, comprising: The computer device comprises a processor and a memory, and the memory stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by the processor to implement the lane positioning method according to any one of claims 1 to 6.

14. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by the processor to implement the lane positioning method according to any one of claims 1 to 6.

15. A computer program product, characterised in that, The computer program product or computer program comprises computer instructions stored in a computer device readable storage medium, a processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to implement the lane positioning method according to any one of claims 1 to 6.

Images