Positioning method, electronic equipment and vehicle

Abstract

The invention discloses a positioning method, electronic equipment and a vehicle, and the positioning method comprises the steps: correcting a current motion pose of motion equipment according to a semantic local map, so as to obtain a final motion pose of the motion equipment. According to the method, a global map or a high-precision map does not need to be established, the manufacturing cost is reduced, the matching time consumption is reduced, the real-time performance is high, and the positioning precision is higher.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a positioning method, electronic device, and vehicle. Background Technology

[0002] In related technologies, vehicle localization is a key issue for autonomous driving. Currently, localization methods based on semantic information mainly rely on the semantic information of the entire map or high-precision map to perform semantic matching with the local map. However, the production cost of global maps and high-precision maps is high, and their real-time performance is insufficient, which can easily lead to large errors. Furthermore, global maps require a large amount of storage and are time-consuming to match with local maps. Summary of the Invention

[0003] The present invention aims to at least solve one of the technical problems existing in the prior art. To this end, one object of the present invention is to propose a positioning method that does not require the creation of a global map or high-precision map, reducing production costs and matching time, and offering high real-time performance and higher positioning accuracy.

[0004] The second objective of this invention is to provide an electronic device.

[0005] The third objective of this invention is to provide a computer storage medium.

[0006] The fourth objective of this invention is to provide a computer program product.

[0007] The fifth objective of this invention is to provide a vehicle.

[0008] To address the aforementioned problems, a first aspect of the present invention proposes a positioning method, comprising: acquiring a historical frame semantic local map and sensor positioning information surrounding a vehicle; determining the current motion pose of the vehicle based on the sensor positioning information; and correcting the current motion pose based on the historical frame semantic local map to obtain the final motion pose of the vehicle.

[0009] The positioning method according to embodiments of the present invention improves positioning accuracy by using historical frame semantic local maps to correct the current motion pose of the moving device. At the same time, it eliminates the need to create global or high-precision maps, reducing production costs. Furthermore, it reduces matching time, has higher real-time performance, and avoids problems such as matching failure or increased positioning errors.

[0010] In some embodiments, the semantic local map includes a historical frame semantic local map and a current frame semantic local map. The step of correcting the current motion pose of the motion device based on the semantic local map to obtain the final motion pose of the motion device includes: determining a pose transformation matrix based on the historical pose information of the motion device and the current motion pose; obtaining a relative pose matrix based on the pose transformation matrix, the historical frame semantic local map, and the current frame semantic local map; and correcting the current motion pose based on the relative pose matrix to obtain the final motion pose.

[0011] In some embodiments, obtaining a relative pose matrix based on the pose transformation matrix, the historical frame semantic local map, and the current frame semantic local map includes: obtaining a current frame reference semantic local map based on the pose transformation matrix and the historical frame semantic local map; and determining the relative pose matrix based on the current frame reference semantic local map and the current frame semantic local map.

[0012] In some embodiments, obtaining the current frame reference semantic local map based on the pose transformation matrix and the historical frame semantic local map includes: performing coordinate system transformation on the historical frame semantic local map based on the pose transformation matrix to obtain the current frame reference semantic local map in the vehicle coordinate system at the current moment.

[0013] In some embodiments, correcting the current motion pose according to the relative pose matrix to obtain the final motion pose includes: determining the position information of the motion device according to the relative pose matrix; and correcting the current motion pose using the position information as an observation to obtain the final motion pose.

[0014] In some embodiments, the positioning method further includes: determining the current motion posture based on the sensing positioning information of the motion device.

[0015] In some embodiments, the sensing and positioning information includes at least one or more of the following: inertial sensing data collected by an inertial sensor, wheel speed sensing data collected by a wheel speedometer, and navigation and positioning data collected by a GNSS sensor.

[0016] In some embodiments, determining the current motion posture based on the sensing and positioning information of the motion device includes: fusing the inertial sensing data, the wheel speed sensing data, and the navigation and positioning data to determine the current motion posture.

[0017] In some embodiments, before fusing the inertial sensing data, the wheel speed sensing data, and the navigation and positioning data, the method further includes: determining that the confidence level of the navigation and positioning data is greater than a preset confidence threshold.

[0018] In some embodiments, fusing the inertial sensing data, the wheel speed sensing data, and the navigation positioning data to determine the current motion pose includes: obtaining a predicted motion pose of the vehicle based on the inertial sensing data; using the wheel speed sensing data as an observation to correct the predicted motion pose to obtain an updated motion pose; and using the navigation positioning data as an observation to correct the updated motion pose to obtain the current motion pose.

[0019] A second aspect of the present invention provides an electronic device, including at least one processor; a memory communicatively connected to at least one of the processors; wherein the memory stores a computer program executable by at least one of the processors, and the positioning method of the above embodiment is implemented when the at least one processor executes the computer program.

[0020] According to an embodiment of the present invention, the electronic device implements a positioning method by executing a computer program stored in a memory through a processor. This method does not require the creation of a global map or a high-precision map, reduces production costs, and reduces matching time, and has high real-time performance, thereby achieving higher positioning accuracy.

[0021] A third aspect of the present invention provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the positioning method of the above embodiments.

[0022] A fourth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the positioning method described above.

[0023] A fifth aspect of the present invention provides a vehicle including the electronic equipment of the above embodiments; or, the vehicle includes a data acquisition module and a controller, the data acquisition module being used to acquire a semantic local map and a current motion pose, and the controller being used to execute the localization method of the above embodiments.

[0024] According to the vehicle of the present invention, the positioning method is executed by an electronic device, or by acquiring a semantic local map and the current motion pose by an acquisition module, and then the semantic local map and the current motion pose are sent to the controller, and the positioning method is executed by the controller. There is no need to build a global map or a high-precision map, which reduces production costs and matching time, and has high real-time performance and higher positioning accuracy.

[0025] In some embodiments, the acquisition module includes: a camera unit for acquiring a semantic local map; and a sensor unit for acquiring the current motion pose.

[0026] In some embodiments, the sensor unit includes one or more of an inertial sensor, a wheel speedometer, and a GNSS sensor.

[0027] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0028] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart of a positioning method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a positioning method according to an embodiment of the present invention; Figure 3 This is a flowchart of a positioning method according to another embodiment of the present invention; Figure 4 This is a flowchart of a positioning method according to another embodiment of the present invention; Figure 5 This is a structural block diagram of an electronic device according to an embodiment of the present invention.

[0029] Figure label: 10 electronic devices; Processor 1; Memory 2. Detailed Implementation

[0030] The embodiments of the present invention are described in detail below. The embodiments described with reference to the accompanying drawings are exemplary. The embodiments of the present invention are described in detail below.

[0031] Currently, most semantic matching-based positioning schemes rely on the pre-built high-precision maps or global maps, and then perform positioning by matching the current local map with the global map. However, global maps and high-precision maps are costly to produce, lack real-time performance, and are prone to matching failures, resulting in large positioning errors.

[0032] To address the aforementioned issues, the first aspect of this invention proposes a positioning method that eliminates the need for establishing a global map or high-precision map, reducing production costs and matching time, while offering high real-time performance and higher positioning accuracy.

[0033] The following is for reference. Figure 1 A positioning method according to an embodiment of the present invention is described. The positioning method includes step S1, and the specific steps are as follows.

[0034] Step S1: Correct the current motion pose of the motion device based on the semantic local map to obtain the final motion pose of the motion device.

[0035] The semantic local map refers to a map of a certain area collected by the moving device at each moment and incorporating semantic information. In other words, the map only contains environmental details near the current location of the moving device. Based on this, compared with global maps or high-precision maps, the positioning method of this application using the semantic local map can be updated in real time as the moving device moves, quickly responding to environmental changes, thereby improving the real-time positioning performance and having a low cost.

[0036] Specifically, due to the characteristics of sensors, relying solely on sensor positioning information to locate moving devices will result in accumulated errors, and the longer the operation time, the greater the positioning error. To address this issue, this application, based on scenarios with clear semantic features, utilizes multiple sensors installed on the moving device to acquire semantic feature information, such as lane lines, parking space lines, and lane markings. A local map can be constructed based on this semantic feature information, and the current motion pose of the moving device can be corrected based on this semantic local map to obtain the final motion pose of the moving device. This improves positioning accuracy and avoids problems such as matching failures or increased positioning errors. Furthermore, this application does not require the creation of a global map or high-precision map, reducing production costs. The moving device can be a vehicle, robot, mobile device, etc., without specific limitations.

[0037] The positioning method according to embodiments of the present invention improves positioning accuracy by correcting the current motion pose of the moving device based on a semantic local map. It also eliminates the need to create a global map or a high-precision map, reducing production costs and shortening matching time, thus providing higher real-time performance and avoiding problems such as matching failure or increased positioning error.

[0038] In some embodiments, the semantic local map includes a historical frame semantic local map and a current frame semantic local map. The current motion pose of the motion device is corrected based on the semantic local map to obtain the final motion pose of the motion device, including: determining a pose transformation matrix based on the historical pose information and the current motion pose of the motion device; obtaining a relative pose matrix based on the pose transformation matrix, the historical frame semantic local map, and the current frame semantic local map; and correcting the current motion pose based on the relative pose matrix to obtain the final motion pose.

[0039] Among them, historical pose information can be the final motion pose of the motion device at a historical moment such as the previous moment, or historical pose information can be the motion pose determined based on the sensor positioning information of the motion device at a historical moment such as the previous moment.

[0040] Specifically, the change in vehicle pose from a historical moment to the current moment can be obtained by comparing historical pose information with the current motion pose. A pose transformation matrix is ​​then obtained based on the motion poses of the historical and current frames of the moving device. This matrix describes the change in the device's pose from the historical frame to the current frame. Furthermore, a matching algorithm is used to solve for the semantic features of the semantic local maps of the historical and current frames. The solution is combined with the pose transformation matrix to obtain a relative pose matrix. This relative pose matrix describes the change in the motion pose of the moving device within the semantic local maps of the historical and current frames. The current motion pose can then be corrected using the relative pose matrix to obtain a more accurate final motion pose. It should be noted that after solving for the semantic features, filtering techniques can be used for further optimization to improve matching accuracy. The relative pose matrix includes position and attitude information. The pose transformation matrix is ​​determined based on the historical pose information and the current motion pose, as shown in Formula 1. Here is the pose transformation matrix. For historical pose information, This is the current motion posture.

[0041] Formula 1 In some embodiments, obtaining a relative pose matrix based on a pose transformation matrix, a historical frame semantic local map, and a current frame semantic local map includes: obtaining a current frame reference semantic local map based on the pose transformation matrix and the historical frame semantic local map; and determining a relative pose matrix based on the current frame reference semantic local map and the current frame semantic local map.

[0042] Specifically, the pose of the moving device in the current frame is predicted based on the semantic local map of the historical frames and the pose transformation matrix, that is, the reference semantic local map of the current frame is obtained. Then, the reference semantic local map of the current frame is compared with the semantic local map of the current frame to obtain the difference between the reference semantic local map of the current frame and the semantic local map of the current frame. Based on the difference result, the relative pose matrix is ​​determined. The current motion pose can be corrected by the relative pose matrix to obtain a more accurate final motion pose.

[0043] In some embodiments, obtaining a current frame reference semantic local map based on a pose transformation matrix and a historical frame semantic local map includes: performing coordinate system transformation on the historical frame semantic local map based on the pose transformation matrix to obtain a current frame reference semantic local map in the vehicle coordinate system at the current moment.

[0044] Specifically, since the historical frame semantic local map and the current frame semantic local map are at different times and their coordinate systems may be different, it is necessary to perform coordinate system transformation on the historical frame semantic local map. For example, ... Figure 2As shown, the pose transformation matrix S404 is determined based on the pose information of the moving device in the historical frame semantic local map S401 and the current frame semantic local map S402. Since the historical frame semantic local map S401 and the current frame semantic local map S402 are not in the same coordinate system, the pose transformation matrix S404 and coordinate system transformation convert the semantic features in the historical frame semantic local map S401 into the current frame reference semantic local map S403. The coordinate system-transformed current frame reference semantic local map S403 is then matched and optimized with the current frame semantic local map S402, and further optimized using filtering techniques. This allows the acquisition of the relative pose matrix in the vehicle coordinate system at the current moment. The relative pose matrix includes vehicle position information and vehicle posture information. Simultaneously, as... Figure 2 As shown, the semantic local map S403 of the current frame reference frame after coordinate system transformation has higher geometric information consistency with the semantic local map S402 of the current frame, thus enabling better matching of feature points.

[0045] In some embodiments, correcting the current motion pose based on the relative pose matrix to obtain the final motion pose includes: determining the position information of the motion device based on the relative pose matrix; and using the position information as an observation, correcting the current motion pose using the Kalman filter method to obtain the final motion pose.

[0046] Specifically, after obtaining the relative pose matrix, the position information of the moving device, such as translation or rotation information, can be obtained by decomposing the relative pose matrix. Using the KF or EKF method as an observation, the current motion pose is corrected to obtain the final motion pose of the moving device. This effectively utilizes information from the semantic local map of historical frames to adjust and optimize the current motion pose, improving positioning accuracy. After completing the coordinate system transformation, the relative pose matrix is ​​used again to correct the current motion pose. The correction process can be referred to in Equation 2, where T is the transformation matrix. The transformation matrix T is decomposed to obtain position information P, and the position information P is used as an observation. The observation matrix can be referred to in Equation 3, where... It is a measurement vector. It is the transformation matrix from the state vector to the measurement vector. The measurement noise can be corrected based on the degree of overlap and iteration count of the semantic features between the historical frame semantic local map and the current frame semantic local map.

[0047] Formula 2 Formula 3 In some embodiments, the positioning method further includes: determining the current motion posture based on the sensing positioning information of the motion device.

[0048] Specifically, by setting up multiple sensors on the motion equipment to acquire various types of data, such as IMU information, wheel speed and chassis information, camera image information, and GNSS information, the sensor positioning information acquired by the sensors can be processed by algorithms to accurately describe the current motion state of the motion equipment, i.e., determine the current motion pose of the motion equipment. For example, by using Kalman Filter (KF) or Error-State Kalman Filter (EKF), the various information acquired by the sensors can be fused and filtered to obtain the current motion pose of the vehicle. Historical pose information can also be obtained by processing historical sensor positioning information through algorithms. Historical pose information can include the pose information of the motion equipment in the previous frame or the pose information of the motion equipment in the previous few frames, without limitation. At the same time, by preprocessing the information collected by the sensors, such as noise reduction, filtering, or data format conversion, the required data format of sensor positioning information can be obtained. The sensor positioning information can include position, velocity, acceleration, etc., without specific limitations.

[0049] For example, such as Figure 2 As shown, S401 is a schematic diagram of the semantic local map of historical frames. The semantic local map of historical frames consists of multiple sampling points with corresponding coordinate positions, denoted as P1, P2...Pn. The information of sampling point P is as (x, y, z), where x is the horizontal coordinate value in the current vehicle coordinate system, y is the vertical coordinate value, and z is the altitude coordinate value.

[0050] In some embodiments, the sensing positioning information includes at least one or more of the following: inertial sensing data collected by an inertial sensor, wheel speed sensing data collected by a wheel speedometer, and navigation positioning data collected by a GNSS sensor.

[0051] Specifically, the current motion posture can be determined solely by inertial sensor data, wheel speed data, or navigation and positioning data. Alternatively, it can be determined by combining different sensor data, thus compensating for the limitations of a single sensor. Examples include combining inertial and wheel speed data, combining inertial and navigation and positioning data, or combining inertial, wheel speed, and navigation and positioning data. The inertial sensor data includes 3-axis accelerometer information and 3-axis angular velocity information; the wheel speed data includes position, velocity, and angle information; and the navigation and positioning data includes longitude, latitude, altitude, velocity, and signal confidence level information.

[0052] In some embodiments, determining the current motion posture based on the sensing and positioning information of the motion device includes fusing inertial sensing data, wheel speed sensing data, and navigation and positioning data to determine the current motion posture.

[0053] Specifically, data fusion can be performed using KF or EKF fusion filtering based on inertial sensor data, wheel speed sensor data, and navigation and positioning data, thereby further improving the accuracy of determining the current motion pose.

[0054] In some embodiments, before fusing inertial sensing data, wheel speed sensing data, and navigation positioning data, the method further includes: determining that the confidence level of the navigation positioning data is greater than a preset confidence threshold.

[0055] Specifically, in complex environments, signals measured by GNSS sensors may encounter problems such as obstruction, reflection, and interference, leading to inaccurate observation results. Therefore, a confidence threshold is set to filter the observation data to ensure that only reliable data is used for data fusion, thereby avoiding deviations or errors due to inaccurate data. If the confidence of the navigation and positioning data is determined to be greater than the preset confidence threshold, the navigation and positioning data is used for data fusion; conversely, if the confidence of the navigation and positioning data is determined to be less than or equal to the preset confidence threshold, this data is not used.

[0056] In some embodiments, inertial sensing data, wheel speed sensing data, and navigation positioning data are fused to determine the current motion pose, including: obtaining a predicted motion pose of the motion device based on the inertial sensing data; using the wheel speed sensing data as an observation to correct the predicted motion pose to obtain an updated motion pose; and using the navigation positioning data as an observation to correct the updated motion pose to obtain the current motion pose.

[0057] Specifically, a predictive model is built based on historical data collected by inertial sensors, thereby providing the optimal estimate based on current inertial sensing data and historical moments. The motion pose of the motion device is predicted by combining the prediction equation, as shown in Formula 5. Combined with Formula 4, the motion pose of the motion device can be predicted. P represents the current pose of the motion device; P represents the position. For speed; Indicates a gesture; This refers to the angular velocity deviation. For acceleration deviation; Here is the state transition matrix. It is a control variable matrix. It is the state control vector, i.e., inertial sensor data. It is noise from the control system.

[0058] = Formula 4 Formula 5 Subsequently, the prediction model can be updated based on the historical data collected by the wheel speed sensor. Wheel speed sensor data is used as an observation, and the motion pose is corrected after integration of the wheel speed sensor data, thereby improving the accuracy of the prediction model. The correction can be performed using the Kjeldahl-Frankenstein method, as shown in Equation 6. It should be noted that the observation matrix can be obtained using Equations 7 and 8, where... The speed observed by the wheel speed gauge on the chassis in the vehicle body coordinate system; The transformation matrix from the vehicle coordinate system to the body coordinate system; It is the current predicted state velocity vector; It is the matrix obtained from the current state attitude vector.

[0059] Formula 6 Formula 7 Formula 8 Furthermore, after correcting the predicted motion pose using wheel speed sensor data, navigation and positioning data can be used as observations to further correct the motion pose. The correction process is shown in Formula 8, and the observation matrix is ​​shown in Formula 10. Location information; This is speed information.

[0060] Formula 9 Formula 10 The following is for reference. Figure 3 Taking an example, the positioning method of this embodiment of the invention is described, and the specific steps are as follows.

[0061] Step S2: The information acquisition module acquires the semantic local map and the current motion pose.

[0062] Step S3: The information preprocessing module performs information preprocessing on the sensor positioning information.

[0063] In step S4, the algorithm processing module corrects the current motion pose using the algorithm relative pose matrix to obtain the final motion pose of the motion device.

[0064] Step S5: The positioning output module outputs the final motion pose of the motion device.

[0065] For example, such as Figure 4As shown, the current motion pose is determined by combining inertial sensing data, wheel speed sensing data, and navigation and positioning data. The current motion pose is then combined with the semantic local map of the previous frame to obtain the reference semantic local map of the current frame. The semantic local map of the current frame is then used to match and optimize the reference semantic local map of the current frame, and the current motion pose is corrected and optimized to obtain the final motion pose of the motion device.

[0066] A second aspect of the present invention provides an electronic device 10, such as... Figure 3 As shown, the electronic device 10 includes at least one processor 1; a memory 2 communicatively connected to the at least one processor 1; wherein the memory 2 stores a computer program that can be executed by the at least one processor 1, and the at least one processor 1 executes the computer program to implement the positioning method of the above embodiment.

[0067] According to an embodiment of the present invention, the electronic device 10 executes a computer program stored in the memory 2 through the processor 1 to implement a positioning method. This method does not require the creation of a global map or a high-precision map, reduces production costs, and reduces matching time, and has high real-time performance, thereby achieving higher positioning accuracy.

[0068] A third aspect of the present invention provides a computer storage medium on which a computer program is stored, wherein the computer program, when executed by a processor, implements the positioning method of the above embodiments.

[0069] A fourth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the positioning method described above.

[0070] A fifth aspect of the present invention provides a vehicle including the electronic equipment of the above embodiments; or, the vehicle includes a data acquisition module and a controller, the data acquisition module being used to acquire a semantic local map and the current motion pose, and the controller being used to execute the localization method of the above embodiments.

[0071] Specifically, the vehicle executes the positioning method through electronic device 10, or after acquiring semantic local map and current motion pose through acquisition module, it sends the semantic local map and current motion pose to controller, and then uses controller to execute positioning method to correct the vehicle's current motion pose, thereby improving positioning accuracy. At the same time, there is no need to create global map or high-precision map, reducing production costs, and the matching time is reduced, which has higher real-time performance and can avoid problems such as matching failure or increased positioning error.

[0072] According to the vehicle of the present invention, the positioning method is executed by the electronic device 10, or the semantic local map and the current motion pose are collected by the acquisition module and then sent to the controller, and the positioning method is executed by the controller. There is no need to build a global map or a high-precision map, which reduces the production cost and the matching time. It has high real-time performance and higher positioning accuracy.

[0073] In some embodiments, the acquisition module includes a camera unit and a sensor unit.

[0074] The camera unit is used to acquire semantic local maps; the sensor unit is used to acquire current motion poses.

[0075] Specifically, the camera unit acquires image information, which is then processed by models to generate a semantic local map based on semantic features. Sensor unit-acquired positioning information is processed by algorithms to accurately describe the vehicle's motion state, i.e., determine the vehicle's current pose. For example, Kalman Filter (KF) or Error-State Kalman Filter (EKF) can be used to fuse and filter various information acquired by the sensor unit to obtain the vehicle's current pose. In some embodiments, the sensor unit includes one or more of an inertial sensor, a wheel speedometer, and a GNSS sensor.

[0076] Specifically, the current motion posture can be determined solely by inertial sensor data, wheel speed data, or navigation and positioning data. Alternatively, it can be determined by combining different sensor data, thus compensating for the limitations of a single sensor. Examples include combining inertial and wheel speed data, combining inertial and navigation and positioning data, or combining inertial, wheel speed, and navigation and positioning data. The inertial sensor data includes 3-axis accelerometer information and 3-axis angular velocity information; the wheel speed data includes position, velocity, and angle information; and the navigation and positioning data includes longitude, latitude, altitude, velocity, and signal confidence level information.

[0077] In the description of this specification, any process or method described in the flowcharts or otherwise herein may be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order according to the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0078] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0079] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0080] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware, and the program can be stored in a computer-readable storage medium. When executed, the program includes one or a combination of the steps of the method embodiments.

[0081] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0082] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

[0083] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example.

[0084] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A positioning method, characterized in that, include: The current motion pose of the motion device is corrected based on the semantic local map to obtain the final motion pose of the motion device.

2. The positioning method according to claim 1, characterized in that, The semantic local map includes historical frame semantic local maps and current frame semantic local maps. The step of correcting the current motion pose of the motion device based on the semantic local map to obtain the final motion pose of the motion device includes: Determine the pose transformation matrix based on the historical pose information of the motion device and the current motion pose; The relative pose matrix is ​​obtained based on the pose transformation matrix, the semantic local map of the historical frames, and the semantic local map of the current frame. The current motion pose is corrected based on the relative pose matrix to obtain the final motion pose.

3. The positioning method according to claim 2, characterized in that, The relative pose matrix is ​​obtained based on the pose transformation matrix, the historical frame semantic local map, and the current frame semantic local map, including: The current frame reference semantic local map is obtained based on the pose transformation matrix and the historical frame semantic local map. The relative pose matrix is ​​determined based on the current frame reference semantic local map and the current frame semantic local map.

4. The positioning method according to claim 3, characterized in that, The step of obtaining the current frame reference semantic local map based on the pose transformation matrix and the historical frame semantic local map includes: Based on the pose transformation matrix, the semantic local map of the historical frame is transformed to obtain the reference semantic local map of the current frame in the vehicle coordinate system at the current moment.

5. The positioning method according to claim 2, characterized in that, Correcting the current motion pose based on the relative pose matrix to obtain the final motion pose includes: The position information of the motion device is determined based on the relative pose matrix; Using the position information as an observation, the current motion pose is corrected to obtain the final motion pose.

6. The positioning method according to any one of claims 2-5, characterized in that, The positioning method further includes: The current motion posture is determined based on the sensing and positioning information of the motion device.

7. The positioning method according to claim 6, characterized in that, The sensing and positioning information includes at least one or more of the following: inertial sensing data collected by inertial sensors, wheel speed sensing data collected by wheel speedometers, and navigation and positioning data collected by GNSS sensors.

8. The positioning method according to claim 7, characterized in that, Determining the current motion posture based on the sensing and positioning information of the motion device includes: The inertial sensing data, the wheel speed sensing data, and the navigation and positioning data are fused together to determine the current motion posture.

9. The positioning method according to claim 8, characterized in that, Before fusing the inertial sensing data, the wheel speed sensing data, and the navigation and positioning data, the method further includes: The confidence level of the navigation and positioning data is determined to be greater than a preset confidence threshold.

10. The positioning method according to claim 8 or 9, characterized in that, The inertial sensing data, wheel speed sensing data, and navigation positioning data are fused to determine the current motion pose, including: The predicted motion pose of the vehicle is obtained based on the inertial sensing data; The wheel speed sensing data is used as an observation to correct the predicted motion pose and obtain an updated motion pose. The navigation and positioning data is used as an observation to correct the updated motion pose to obtain the current motion pose.

11. An electronic device, characterized in that, include: At least one processor; A memory that is communicatively connected to at least one of the processors; The memory stores a computer program that can be executed by at least one of the processors, and when the at least one processor executes the computer program, it implements the positioning method according to any one of claims 1-10.

12. A computer storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the positioning method according to any one of claims 1-10.

13. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the positioning method according to any one of claims 1-10.

14. A vehicle, characterized in that, Including the electronic device as described in claim 10; Alternatively, the vehicle may include a data acquisition module and a controller, wherein the data acquisition module is used to acquire a semantic local map and the current motion pose, and the controller is used to execute the localization method according to any one of claims 1-10.

15. The vehicle according to claim 14, characterized in that, The acquisition module includes: A camera unit is used to acquire semantic local maps; A sensor unit, which is used to acquire the current motion pose.

16. The vehicle according to claim 15, characterized in that, The sensor unit includes one or more of an inertial sensor, a wheel speedometer, and a GNSS sensor.

Images