Memory parking mapping method, vehicle-mounted system and computer readable storage medium

Abstract

The invention belongs to the technical field of automatic driving, and particularly relates to a memory parking mapping method, a vehicle-mounted system and a computer readable storage medium. According to the method, the local coordinate system and the satellite global coordinate system are established through the vehicle-mounted sensor, during mapping, the two coordinate systems are adopted at the same time for coordinate recording, and the two coordinate systems are bound in the mapping process. Further setting the confidence coefficient for judging the global coordinates, so that when the global coordinates with low confidence coefficient are obtained, the global coordinates are calculated through the local coordinates, and the global coordinate path of mapping is corrected. According to the method, when the global coordinates detected by the satellite are inaccurate or discrete, calculation is performed by using the local coordinates, so that the discretization of the path is reduced, the inaccurate global coordinates can be corrected, and the mapping is more accurate and accords with the reality.

Description

Technical Field

[0001] The present technology belongs to the technical field of autonomous driving, and in particular relates to a memory parking mapping method, a vehicle-mounted system, and a computer-readable storage medium. Background Art

[0002] In the field of autonomous driving, a mapping process is often set up so that the vehicle can record a path map of a certain section of the route, and then perform automatic parking driving according to the mapped path during the second driving.

[0003] In the prior art, a method of mapping includes using a global satellite GNSS method to locate coordinates, and in the process of mapping, recording the overall path in the form of recording coordinates at a certain distance. This method is called global mapping.

[0004] However, when using this method, GNSS judgment is often inaccurate or coordinates are lost, such as between high-rise buildings, in underground garages, in mountainous areas, etc. In this case, GNSS positioning will be discrete or deviated, resulting in inaccurate positioning. Summary of the invention

[0005] Based on the above problems, we introduce local mapping while building the map, that is, coordinate mapping is performed based on the movement and rotation of the vehicle itself. When the global map is built, the local coordinates are recorded while the global coordinates are recorded. The map is built using these two methods, and confidence is introduced at the global coordinates to determine whether the global coordinates are reliable. When the global coordinates are unreliable, the global coordinates are inferred from the local coordinates, and then the global coordinates are corrected, so that the overall path is more realistic.

[0006] The content of the present invention is achieved through the following technical solutions:

[0007] In a first aspect, the present invention provides a memory parking mapping method, comprising:

[0008] Establish a local coordinate system and a global coordinate system, obtain the local coordinate information of the local coordinate system in real time through the vehicle-mounted sensor, obtain the global coordinate information of the global coordinate system in real time through the satellite positioning technology, and bind the local coordinate information with the global coordinate information;

[0009] Obtain vehicle driving environment variables in real time, and obtain real-time position confidence based on vehicle driving environment variables and global coordinate information;

[0010] Determine whether the real-time position confidence is within the preset confidence threshold range. If so, record the global coordinate information at the current moment as the target coordinate information. Otherwise, calculate the target coordinate information based on the local coordinate information.

[0011] When building a map in memory, two different coordinate systems, namely the global coordinate system and the local coordinate system, are set up. The coordinates of the global coordinate system and the local coordinate system are simultaneously acquired and bound to build the map. As a result, when there is a problem with the detection of the global coordinate system, the global coordinates can be reversed through the local coordinate system. This method can effectively achieve that when the global coordinates detected by the satellite have detection errors in some environments, they can be reversed through the local coordinates to complete the path, reduce the path discreteness in the mapping process, and also make the global coordinates corrected through the local coordinates.

[0012] In some implementations, a local coordinate system and a global coordinate system are established, local coordinate information of the local coordinate system is acquired in real time through a vehicle-mounted sensor, global coordinate information of the global coordinate system is acquired in real time through a satellite positioning technology, and the local coordinate information is bound to the global coordinate information; including:

[0013] With the front of the vehicle as the positive X-axis and the vertical upward as the positive Z-axis, a local coordinate system is established using the right-hand coordinate system, and the real-time vehicle translation component and rotation component are recorded and combined into local coordinates;

[0014] Establish a global coordinate system based on the longitude, latitude and altitude detected by satellite, and record the real-time longitude, latitude and altitude to form a global coordinate system;

[0015] After the vehicle is powered on, the initial second local coordinates of the vehicle local coordinate system are recorded, and the second local coordinates are transformed according to the vehicle driving state to obtain the first local coordinates;

[0016] After obtaining the initial mapping point, the first global coordinate and confidence at that moment, as well as the first local coordinate, are obtained, and the first global coordinate and the first local coordinate are bound.

[0017] The local coordinate system is established through the vehicle's own moving posture, and the global coordinate system is established through the satellite. When building the map, the coordinate system recorded at a certain moment will record the local coordinate system and the global coordinate system and bind the coordinates. This allows the corresponding local coordinate system to be queried and the coordinates to be calculated when the global real-time position confidence is low.

[0018] In some implementations, the vehicle driving environment variables are acquired in real time, and the real-time position confidence is obtained according to the vehicle driving environment variables and global coordinate information; including:

[0019] Obtain real-time global coordinate information through the GNSS status flag;

[0020] Obtain vehicle driving environment variables through SLAM technology;

[0021] According to the global coordinate information and vehicle driving environment variables, the preset confidence list is queried to obtain the real-time position confidence.

[0022] When judging the confidence, a marker will be set on the GNSS satellite for confidence judgment. At the same time, in order to further improve the accuracy of the judgment, the surrounding environment is judged through SLAM technology to obtain the surrounding environmental variables and further analyze the confidence of the real-time position.

[0023] In some implementations, determining whether the real-time position confidence is within a preset confidence threshold range, if so, recording the global coordinate information at the current moment as the target coordinate information, otherwise calculating the target coordinate information based on the local coordinate information includes:

[0024] Obtaining the real-time location confidence, and determining whether the real-time location confidence is greater than a preset real-time location confidence threshold;

[0025] If it is greater than, record the current first global coordinate and set the first confidence state.

[0026] Otherwise, the second global coordinates are calculated according to the first local coordinates, the second global coordinates are recorded and the second confidence state is set to replace the first global coordinates, and the second global coordinates and the second confidence state are recorded in the map construction.

[0027] The reliability of the current global coordinates is determined by the relationship between the real-time position confidence and the confidence threshold. When the coordinates are unreliable, the global coordinates are calculated and corrected through the local coordinates, and then the entire path is corrected to further prevent the global coordinates from becoming discrete.

[0028] In some implementations, calculating a second global coordinate according to the first local coordinate, recording the second global coordinate and replacing the first global coordinate to set a second confidence state, and recording the second global coordinate and the second confidence state in the map construction includes:

[0029] Obtain the first local coordinate in the map, the third global coordinate in the first confidence state N moments before, and the third local coordinate in the first N moments before;

[0030] The third global coordinate is converted into a fourth local coordinate in the coordinate system of the northeastern sky station center, and the fifth local coordinate is calculated according to the first local coordinate, the fourth local coordinate and the third local coordinate;

[0031] The fifth local coordinate is converted into a sixth global coordinate in the satellite global coordinate system, the sixth global coordinate is assigned to the second global coordinate, and the second global coordinate is set to a third confidence state.

[0032] When extrapolating, follow the local coordinates corresponding to the current global coordinates, and query the global coordinates with high confidence closest to the current time in the previous N recorded moments and their corresponding local coordinates, and convert the global coordinates into coordinates in the Northeast Sky Station Center coordinate system that can be used for calculations with the local coordinates. Through the principle of rigid body change, the extrapolated coordinates in the Northeast Sky Station Center coordinate system are obtained, and then converted into global coordinates for recording. In this way, the calculation of local and global coordinates is realized.

[0033] In some implementations, calculating the second global coordinates according to the first local coordinates, recording the second global coordinates and setting a second confidence state, and recording the second global coordinates and the second confidence state in the map construction further includes:

[0034] After the second global coordinates are calculated using the first local coordinates, the first calculation time is recorded;

[0035] It is determined whether the first estimated time is greater than the estimated time threshold. If so, a second confidence state is set for the second global coordinate; otherwise, a third confidence state is set.

[0036] During the calculation, the calculation time is recorded to determine the validity of the calculation. When the calculation time is too long, even if the global coordinates are calculated, the coordinates themselves are still discrete and the calculated coordinates cannot be recorded.

[0037] In some embodiments, if the calculated second global coordinate is in the second confidence state, the first local coordinate in the mapping, the seventh global coordinate in the first confidence state N-1 moments before, and the fourth local coordinate in the first N-1 moments before are reacquired;

[0038] and recalculating the second global coordinates according to the above method, and recording the second calculation time, and determining whether the second calculation time is greater than the calculation time threshold;

[0039] If so, perform the calculation and move forward one moment until the calculated second calculation time is less than the calculation time threshold, set the second global coordinate at this moment to the third confidence state, and record it in the map.

[0040] When the estimated time is too long, recalculation will be performed twice or more to make the estimated time less than the estimated time threshold, thereby making the data valid and preventing the data from being discrete.

[0041] In some embodiments, after the engine is turned off and the power is turned off, the coordinates with a confidence level higher than a confidence level threshold recorded most recently at the time of the engine being turned off are recorded.

[0042] The purpose is to prevent the initial coordinates from being discrete during the second startup due to changes in the environment or being placed in an environment with low confidence.

[0043] In a second aspect, this embodiment provides a vehicle-mounted system, comprising:

[0044] A mapping module is used to establish a local coordinate system and a global coordinate system, obtain the local coordinate information of the local coordinate system in real time through the vehicle-mounted sensor, obtain the global coordinate information of the global coordinate system in real time through the satellite positioning technology, and bind the local coordinate information with the global coordinate information;

[0045] The coordinate calculation module is used to determine whether the real-time position confidence is within a preset confidence threshold. If so, the global coordinate information at the current moment is recorded as the target coordinate information. Otherwise, the target coordinate information is calculated based on the local coordinate information.

[0046] The confidence judgment module is used to obtain vehicle driving environment variables in real time, and obtain real-time position confidence based on the vehicle driving environment variables and global coordinate information.

[0047] Through this system, the memory parking mapping method of the first aspect can be implemented on the system.

[0048] In a third aspect, the present invention proposes a computer-readable storage medium, comprising at least one execution instruction stored in the storage medium, and when the executable instruction is executed on the vehicle-mounted system as claimed in claim 9, the device executes the memory parking mapping method as claimed in any one of claims 1-8.

[0049] Beneficial effects of the memory parking map construction method, vehicle-mounted system and computer-readable storage medium of the present invention:

[0050] The present invention establishes a local coordinate system and a satellite global coordinate system through vehicle-mounted sensors. When building a map, two coordinate systems are used to record coordinates at the same time, and the two coordinate systems are bound during the process of building the map. It is further provided with a confidence level for judging the global coordinates, so that when obtaining global coordinates with low confidence, the global coordinates are extrapolated through the local coordinates to achieve the correction of the global coordinate path for building the map. Through the above method, when the global coordinates detected by the satellite are inaccurate or discrete, the local coordinates are used for extrapolation, which reduces the discreteness of the path and can also correct the inaccurate global coordinates, thereby making the map building more accurate and in line with reality. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present invention. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings:

[0052] Figure 1 It is the main flow chart of the method of the present invention;

[0053] Figure 2This is a flowchart of local mapping and global mapping of the present invention;

[0054] Figure 3 This is the confidence judgment process of the present invention;

[0055] Figure 4 A logical flow chart of coordinate calculation of the present invention;

[0056] Figure 5 It is a flow chart of the present invention for calculating global coordinates from local coordinates;

[0057] Figure 6 This is a flow chart of the present invention for judging whether the estimated global coordinates meet the requirements according to the estimated time;

[0058] Figure 7 A flow chart of the present invention for calculating the estimated coordinates to meet the estimation time requirement;

[0059] Figure 8 It is a structural diagram of the vehicle-mounted system of the present invention. DETAILED DESCRIPTION

[0060] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0061] Embodiment 1:

[0062] like Figure 1 As shown, the present invention proposes a memory parking mapping method, comprising the following steps:

[0063] Step 101: Establish a local coordinate system and a global coordinate system, obtain local coordinate information of the local coordinate system in real time through vehicle-mounted sensors, obtain global coordinate information of the global coordinate system in real time through satellite positioning technology, and bind the local coordinate information with the global coordinate information;

[0064] Specifically, a local coordinate system and a global coordinate system are established. The global coordinate system can be the latitude and longitude coordinates obtained by the satellite GNSS, and the local coordinate system is a local coordinate system set for the vehicle posture. When establishing the local coordinate system, the ground projection point of the center of the rear axle of the vehicle is usually taken as the zero point, the front of the vehicle is the positive X-axis, and the vertical upward is the positive Z-axis. The right-hand coordinate system, because the local coordinate system mainly reflects the posture change and displacement of the vehicle during driving, is recorded in the local coordinate, and the translation vector and rotation component are recorded as the coordinate value. The translation component and the rotation component can be obtained through the vehicle's sensors such as steering gear, encoder, etc. Therefore, the local coordinate information, that is, the local coordinate can be expressed as:

[0065]

[0066] Where T is the local coordinate, P is the translation component, X1, Y1, Z1 are the translations of the translation component in each direction, q is the rotation component, and X2, Y2, Z2, Z2 are the components of the rotation component in each direction. The global coordinate system is usually a way to present the vehicle as a single moving point under the satellite, so the main parameters of the global coordinate in the path are latitude, longitude and altitude.

[0067] Furthermore, when recording, a certain time or a certain distance is used as the recording period. In this embodiment, recording is performed after a certain distance has been traveled, and the distance is usually set to 1 meter. When recording, the local coordinates and the global coordinates are recorded at the same time, and the local coordinates and the global coordinates are bound by the distance number. Therefore, the local coordinates can be expressed as (p i ,q i ), where i∈(0,1,2,3....), and p i+1 -p i <1.

[0068] The technical effect produced by this step is that by setting up the vehicle sensor to establish a local coordinate system and the satellite GNSS to establish a global coordinate system, during the mapping process, the vehicle local coordinates and global coordinates are obtained at the same time, and the global coordinates and local coordinates are bound so that the local coordinates can be queried when the global coordinates are queried.

[0069] Step 102: Acquire vehicle driving environment variables in real time, and obtain real-time position confidence according to the vehicle driving environment variables and global coordinate information;

[0070] Specifically, the surrounding environmental variables are acquired in real time. The environmental variables can be the characteristics of the surrounding environment. The vehicle-mounted sensors acquire images and capture features to determine the environmental variables, and further determine whether the current vehicle environment is an environment that may cause the global coordinates to fail. At the same time, combined with the global coordinate information, it can be substituted into the map to determine the reliability of the coordinates. The satellite can make its own judgment and output in the form of a flag.

[0071] The beneficial effect of this step is that it is used to obtain the confidence level of the actual position, and also judges by combining the local variables around the vehicle and the global variables monitored by the satellite, so that the judgment is more reasonable and effective.

[0072] Step 103: Determine whether the real-time position confidence is within a preset confidence threshold range. If so, record the global coordinate information at the current moment as the target coordinate information. Otherwise, calculate the target coordinate information based on the local coordinate information.

[0073] Specifically, the confidence of the real-time position in the above step is determined, and compared with the preset threshold to determine whether it is greater than the confidence threshold, or a confidence range is set. When the confidence falls within the confidence range, it means that the coordinate is a reliable coordinate. If the confidence value of the coordinate is in the non-confidence range, or is less than the confidence threshold, the global coordinate is calculated through the local coordinate. When calculating the coordinate, follow the rigid body transformation, and the transformation matrix is ​​as follows:

[0074]

[0075] When building a map in memory, two different coordinate systems, namely the global coordinate system and the local coordinate system, are set up. The coordinates of the global coordinate system and the local coordinate system are simultaneously acquired and bound to build the map. As a result, when there is a problem with the detection of the global coordinate system, the global coordinates can be reversed through the local coordinate system. This method can effectively achieve that when the global coordinates detected by the satellite have detection errors in some environments, they can be reversed through the local coordinates to complete the path, reduce the path discreteness in the mapping process, and also make the global coordinates corrected through the local coordinates.

[0076] Embodiment 2:

[0077] like Figure 2 to Figure 7 As shown, this embodiment further illustrates and optimizes the process proposed in Embodiment 1.

[0078] In some embodiments, a local coordinate system and a global coordinate system are established, local coordinate information of the local coordinate system is acquired in real time through vehicle-mounted sensors, global coordinate information of the global coordinate system is acquired in real time through satellite positioning technology, and the local coordinate information is bound to the global coordinate information; including:

[0079] Step 201: With the front of the vehicle as the positive X axis and the vertical upward as the positive Z axis, a local coordinate system is established in a right-handed coordinate system, and the real-time vehicle translation component and rotation component are recorded and combined into a local coordinate system;

[0080] Specifically, the ground projection point of the rear axle center of the vehicle is taken as the zero point, the front of the vehicle is the positive X axis, and the vertical upward is the positive Z axis, and the right-hand coordinate system is used. The rotation vector and translation vector are recorded, and the local coordinates can be expressed as:

[0081]

[0082] Where T is the local coordinate, P is the translation component, X1, Y1, Z1 are the translation amounts in each direction of the translation component, q is the rotation component, and X2, Y2, Z2, Z2 are the components of the rotation component in each direction.

[0083] Step 202: Establish a global coordinate system using the longitude, latitude and altitude detected by the satellite, and record the real-time longitude, latitude and altitude to form a global coordinate system;

[0084] Specifically, the longitude, latitude and altitude recorded by GNSS are aggregated into global coordinates, which are marked as LLA.

[0085] Step 203: after the vehicle is powered on, the initial second local coordinates of the vehicle local coordinate system are recorded, and the second local coordinates are transformed according to the vehicle driving state to obtain the first local coordinates;

[0086] Specifically, after establishing the coordinate system, wait for the vehicle to start, and after the vehicle is powered on, record the initial local coordinate system of the vehicle as the second local coordinate system:

[0087] The local coordinates can be expressed as:

[0088]

[0089] After each meter traveled, the local coordinates are recorded. Therefore, the local coordinates can be expressed as (p i ,q i ), where i∈(0,1,2,3....), and p i+1 -p i <1. It should be noted that when transforming coordinates, the rigid body transformation is followed, and the transformation matrix is ​​as follows:

[0090]

[0091] Step 204: After obtaining the initial mapping point, obtain the first global coordinate and confidence level at that moment, as well as the first local coordinate, and bind the first global coordinate and the first local coordinate.

[0092] Specifically, after obtaining the initial mapping point, the first global coordinate and confidence are recorded, and the real-time confidence is bound to the recorded global coordinate. The confidence is obtained by obtaining the surrounding environmental variables in real time. The environmental variables can be the characteristics of the surrounding environment. The image is obtained and the features are captured for the vehicle-mounted sensor to determine the environmental variables, and further determine whether the current vehicle environment is an environment that may cause the global coordinate to fail. At the same time, combined with the global coordinate information, it can be substituted into the map to determine the reliability under the coordinate, which can be determined by the satellite itself and output in the form of a flag.

[0093] The technical effect of this step is that by binding the obtained confidence to the obtained global coordinates, the reliability of the global coordinates can be judged by the confidence, and then it can be determined whether extrapolation is needed. At the same time, when doing path extrapolation, recommendations can be made based on the confidence to reduce coordinate avoidance in the path.

[0094] It should be further explained that there is no order of precedence for steps 201 to 204. When establishing the local coordinates and the global coordinates, they are established simultaneously, and the mapping point and the power-on point can be consistent, or can be set according to needs to reduce unnecessary road sections that are bound to the first local coordinates and the first global coordinates.

[0095] The local coordinate system is established through the vehicle's own moving posture, and the global coordinate system is established through the satellite. When building the map, the coordinate system recorded at a certain moment will record the local coordinate system and the global coordinate system and bind the coordinates. This allows the corresponding local coordinate system to be queried and the coordinates to be calculated when the global real-time position confidence is low.

[0096] In some embodiments, the vehicle driving environment variables are acquired in real time, and the real-time position confidence is obtained according to the vehicle driving environment variables and global coordinate information; including:

[0097] Step 301: Acquire real-time global coordinate information through the GNSS status flag;

[0098] Specifically, the longitude and latitude information of GNSS is collected, and according to the working status of GNSS and the scene perception results, the flag bit can be obtained by autonomously judging the accuracy of the coordinates according to the map or satellite, and the flag bit can be set to output the judgment result.

[0099] The beneficial effect of this step is that the confidence of the global coordinates is judged from the perspective of the global coordinates through the judgment of the satellite.

[0100] Step 302: Acquire vehicle driving environment variables through SLAM technology;

[0101] Specifically, the surrounding environment is acquired through SLAM technology, and the surrounding environment information can be acquired through a vehicle-mounted camera module, a sensor module, etc.

[0102] The beneficial effect of this step is that, through the SLAM technology, the environment around the vehicle is judged from a specific local angle, and then the confidence of the global coordinates is reversed, providing a second angle for confidence judgment.

[0103] Step 303: According to the global coordinate information and the vehicle driving environment variables, a preset confidence list is queried to obtain the real-time position confidence.

[0104] Specifically, the global coordinate information and environmental coordinate information collected through the above steps are used to query the preset confidence list. In this step, the two types of information can be trained and analyzed through big data, and the trained environmental judgment model can be used to combine the global coordinate information and the environmental coordinate information to perform confidence inference. Specifically, the confidence can be comprehensively output through the surrounding environmental characteristics and the characteristics of the global coordinate surroundings in longitude and latitude.

[0105] The beneficial effect of this step is that the confidence is judged from two aspects, namely, a joint judgment is made through global coordinate information and surrounding environment information. By judging from two angles, the accuracy of the judgment is improved, and the confidence judgment has reference value.

[0106] When judging the confidence, a marker will be set on the GNSS satellite for confidence judgment. At the same time, in order to further improve the accuracy of the judgment, the surrounding environment is judged through SLAM technology to obtain the surrounding environmental variables and further analyze the confidence of the real-time position.

[0107] In some embodiments, determining whether the real-time position confidence is within a preset confidence threshold range, if so, recording the global coordinate information at the current moment as the target coordinate information, otherwise calculating the target coordinate information based on the local coordinate information includes:

[0108] Step 401: obtaining the real-time location confidence, and determining whether the real-time location confidence is greater than a preset real-time location confidence threshold;

[0109] Specifically, the confidence is obtained according to the above method, and a real-time position confidence threshold is set. The threshold can be a specific value or a threshold range can be specified. If the data falls within the threshold range, it means that the condition is met. The threshold range is relatively more accurate, but the relative calculation complexity will also increase. Therefore, this embodiment pre-sets the confidence threshold. When it is greater than the threshold, it means that the confidence is high, which further indicates that the global coordinates are reliable. Otherwise, it means that the global coordinates are unreliable.

[0110] The beneficial effect of this step is that, through the confidence threshold, the global real-time position confidence can be judged by the threshold whether it is credible.

[0111] Step 402: If it is greater than, record the current first global coordinate and set the first confidence state.

[0112] Otherwise, the second global coordinates are calculated according to the first local coordinates, the second global coordinates are recorded and the second confidence state is set to replace the first global coordinates, and the second global coordinates and the second confidence state are recorded in the map construction.

[0113] Specifically, in this step, when the confidence is bound to the global coordinates, another parameter, namely the confidence state, is also bound, so the global coordinates finally include:

[0114]

[0115] Among them, LLA is the latitude, longitude and altitude, P is the confidence, FLAG is the confidence state, and the confidence state is usually a flag bit such as 0, 1, 2, etc., which is used to indicate whether the confidence is greater than the threshold range. The reason for setting the confidence state is that the confidence is usually a fractional value, which is usually cumbersome to express and prone to confusion. When applied, it will not be referenced according to the specific value of the confidence, so it is blurred into low confidence, high confidence, etc., represented by flag bits 0 or 1 or other values. Setting the first confidence state is a state with high confidence, and setting the second confidence state is a state with low confidence, that is, the coordinates that need to be calculated. When the confidence state is the second confidence state, it is usually recorded in the coordinate system, but when drawing the path, calculating the path, and recommending the path, the coordinates of the second confidence state usually have a lower priority. Therefore, it can be further concluded that the confidence state also has the function of defining the priority of the coordinates, especially when recommending the path later, the path can be made more accurate based on the coordinates with high confidence.

[0116] The effect of this step is to set the confidence state, and further clearly understand the confidence level of the global coordinates through the confidence state.

[0117] The reliability of the current global coordinates is determined by the relationship between the real-time position confidence and the confidence threshold. When the coordinates are unreliable, the global coordinates are calculated and corrected through the local coordinates, and then the entire path is corrected to further prevent the global coordinates from becoming discrete.

[0118] In some embodiments, calculating a second global coordinate according to the first local coordinate, recording the second global coordinate and replacing the first global coordinate to set a second confidence state, and recording the second global coordinate and the second confidence state in the map construction includes:

[0119] Step 501: Obtain the first local coordinate in the map construction, the third global coordinate in the first confidence state N moments before, and the third local coordinate in the first N moments before;

[0120] Step 502: convert the third global coordinate into a fourth local coordinate in the northeastern sky station center coordinate system, and calculate a fifth local coordinate according to the first local coordinate, the fourth local coordinate and the third local coordinate;

[0121] Step 503: convert the fifth local coordinate into a sixth global coordinate in the satellite global coordinate system, assign the sixth global coordinate to the second global coordinate, and set the second global coordinate to a third confidence state.

[0122] Specifically, the above steps can be summarized into the following formula:

[0123]

[0124] Among them, the Northeastern Station center coordinate system is also the local coordinate system; the longitude and latitude are also the global coordinate system.

[0125] F ENU2LLA {} indicates the conversion of the Northeastern Station center coordinate system to the longitude and latitude, that is, the conversion of the local coordinate system to the global coordinate system. LLA2ENU () indicates the conversion of the Northeastern Station center coordinate system to longitude, latitude and altitude, that is, the conversion from the global coordinate system to the local coordinate system. i-n Indicates the latitude and longitude when the state is 1 n frames ago, that is, the third global coordinate, T(pq) i-n Represents the third local coordinate of the mapping coordinate system of the first n frames, T(pq) i Indicates the current first local coordinate, F LLA2ENU (LLA i-n ) is the local coordinate transformation of the third global coordinate, i.e., the fourth local coordinate, is the fifth local coordinate, LLA odo , which is the sixth global coordinate.

[0126] The coordinate system of the Northeastern Station Center is based on the starting point of the map as the zero point, the east and north directions are the positive X-axis and the positive Y-axis, and the vertical direction is the positive Z-axis. When the LLA.P low threshold is reached, the current longitude and latitude are judged to be unreliable, and the local coordinates of the map trajectory will be used to calculate the longitude and latitude to obtain the estimated longitude and latitude high value LLA odo , and record the cumulative estimation time of GNSS signal as t odo2LLA .

[0127] When extrapolating, according to the local coordinates corresponding to the current global coordinates, query the global coordinates with high confidence closest to the current time and their corresponding local coordinates in the previous N recorded moments, and convert the global coordinates into coordinates in the Northeast Sky Station Center coordinate system that can be operated with the local coordinates for operation. Through the rigid body change principle, the extrapolated coordinates in the Northeast Sky Station Center coordinate system are obtained, and then converted into global coordinates for recording. In this way, the extrapolation of local coordinates and global coordinates is realized. The extrapolated coordinates are set to the third confidence state to mark the coordinates as extrapolated coordinates.

[0128] In some embodiments, calculating the second global coordinates according to the first local coordinates, recording the second global coordinates and setting the second confidence state, and recording the second global coordinates and the second confidence state in the map construction further includes:

[0129] Step 601: After the second global coordinates are calculated from the first local coordinates, the first calculation time is recorded;

[0130] Specifically, the accumulated first estimated time of the GNSS signal is recorded as t odo2LLA .

[0131] Step 602: Determine whether the first estimated time is greater than the estimated time threshold; if so, set the second confidence state for the second global coordinate; otherwise, set the third confidence state.

[0132] Specifically, set the time threshold th_time odo2LLA , when the first estimated time is t odo2LLA When the time threshold th_time odo2LLA In contrast, when t odo2LLA Less than a certain time threshold th_time odo2LLA When LLA odo The estimated value is assigned to LLA.LLA, and the state is set to the third confidence state. The third confidence state has a lower priority than the first confidence state and higher than the second confidence state. This state is a coordinate state that is estimated and meets the estimated time. Since this state is estimated, when recommending a map, its priority is lower than the coordinates of the first confidence state with higher confidence.

[0133] During the calculation, the calculation time is recorded to determine the validity of the calculation. When the calculation time is too long, even if the global coordinates are calculated, the coordinates themselves are still discrete. At this time, the calculated coordinates cannot be recorded, so the second confidence state is still set.

[0134] In some embodiments, further comprising

[0135] Step 701: when the second global coordinate is in the third confidence state, the first local coordinate in the map construction, the seventh global coordinate in the first confidence state N-1 moments before, and the fourth local coordinate in the first N-1 moments before are reacquired;

[0136] Step 702: recalculate the second global coordinates according to the above method, record the second calculation time, and determine whether the second calculation time is greater than the calculation time threshold;

[0137] Step 703: If yes, recalculate and advance one moment until the calculated second estimated time is less than the estimated time threshold, set the second global coordinate at this moment to the second confidence state, and record it in the map.

[0138] Specifically, if the calculated coordinates are in the second confidence state, it means that the coordinates are still divergent after the calculation, and it is necessary to find new global coordinates including the first confidence state from the previously recorded coordinates and bring them into the above-mentioned method such as steps 501 to 503 for calculation, and record the calculation time. If the calculation time is still greater than the calculation time threshold, continue to repeat step 701, otherwise, set the calculated coordinates to the third confidence state and record them.

[0139] When the estimated time is too long, recalculation will be performed twice or more to make the estimated time less than the estimated time threshold, thereby making the data valid and preventing the data from being discrete.

[0140] In some embodiments, after the engine is turned off and the power is turned off, the coordinates with a confidence level higher than a confidence level threshold recorded most recently at the time of the engine being turned off are recorded.

[0141] Specifically, if the user starts the car in a scene with no GNSS signal at all, and at th_time odo2LLA There is no reliable latitude and longitude data of the first confidence level within the threshold to calculate the latitude and longitude forward. This will result in a part of the path or even the entire path without valid latitude and longitude values.

[0142] To avoid this situation, the latitude and longitude at the last moment need to be saved when the vehicle is turned off, so that the valid latitude and longitude information can be read in the next time the vehicle is started and powered on.

[0143] The purpose is to prevent the initial coordinates from being discrete during the second startup due to changes in the environment or being placed in an environment with low confidence.

[0144] Embodiment 3:

[0145] Based on Example 1 and Example 2, this example describes the application scenario of a memory parking mapping method proposed in the above embodiments, including path recommendation and repositioning in autonomous driving.

[0146] In some embodiments, after mapping by Example 1 and / or Example 2, it is necessary to push the mapped path close to the current vehicle position in real time in the subsequent memory parking function. This requires the use of global coordinate information, i.e., longitude, latitude and altitude, to calculate the path distance, and push the path information that meets the distance requirements to the interface display for user selection.

[0147] The following classification processes are performed on map tracks bound with different states of latitude and longitude information:

[0148] 1. Determine the current vehicle status: Use the method in Section 2 to obtain the latitude, longitude and altitude information of the current vehicle, recorded as LLA cur .

[0149] 2. Recommendation in a confirmed state: The confirmed state means that the current vehicle state and the LLA state of the path mapping trajectory point are both in the first confidence state or the third confidence state. At this time, when making a path recommendation judgment, the recommended distance threshold will be appropriately tightened, recorded as th_dis low .

[0150] Traverse the longitude and latitude bound to the path trajectory and calculate the value closest to the current vehicle on the path, recorded as Dis trajIf Dis traj Less than the threshold th-dis low , then this path is taken as a candidate recommended path.

[0151] After traversing all paths, follow the Dis traj The numerical value recommends candidate recommendation paths from near to far.

[0152] 3. Recommendations in an uncertain state

[0153] The uncertain state means that the current vehicle state and the LLA state of the path mapping trajectory point have the first confidence state or both are the second confidence state. In this case, the distance threshold is appropriately relaxed and recorded as th_dis high The path recommendation method is the same as Article 2 above.

[0154] In some embodiments, after the user selects a recommended route, the current vehicle's position relative to the route needs to be located. At this time, a visual relocalization algorithm is used for scene recognition and position recovery.

[0155] The success rate of relocalization can be improved by using the longitude and latitude bound to the map track and its status. According to the path recommendation in Section 3, the point closest to the current vehicle on the path can be found in real time, and key frames can be searched nearby. This not only effectively speeds up the relocalization speed, but also reduces the probability of mismatching during relocalization.

[0156] Embodiment 4:

[0157] like Figure 8 As shown, this embodiment is based on a memory parking mapping method proposed in Embodiment 1 and Embodiment 2, and a specific application in Embodiment 3. This embodiment proposes a vehicle-mounted system, which at least includes:

[0158] A mapping module is used to establish a local coordinate system and a global coordinate system, obtain the local coordinate information of the local coordinate system in real time through the vehicle-mounted sensor, obtain the global coordinate information of the global coordinate system in real time through the satellite positioning technology, and bind the local coordinate information with the global coordinate information;

[0159] A coordinate calculation module is used to determine whether the real-time position confidence is within a preset confidence threshold range. If so, the global coordinate information at the current moment is recorded as the target coordinate information; otherwise, the target coordinate information is obtained by calculating according to the local coordinate information;

[0160] The confidence judgment module is used to obtain vehicle driving environment variables in real time, and obtain real-time position confidence based on the vehicle driving environment variables and global coordinate information.

[0161] In some embodiments, it also includes:

[0162] The confidence calculation module is used to query a preset confidence list according to the global coordinate information and the vehicle surrounding environment variables to obtain real-time position confidence.

[0163] The route recommendation module is used to make further path recommendations and analysis based on the confidence status in the map.

[0164] The estimated time comparison module is used to record the estimated time and compare the estimated time.

[0165] The relocalization module is used to locate the current vehicle's position relative to the recommended path after the user selects the recommended path. At this time, the visual relocalization algorithm will be used for scene recognition and position recovery.

[0166] Embodiment 5:

[0167] Based on Embodiments 1 to 4, this embodiment proposes a computer-readable storage medium, including at least one execution instruction stored in the storage medium. When the executable instruction is executed on the computer as in Embodiment 4, the device executes the memory parking mapping method as in any one of Embodiments 1 to 2 and the path application method in Embodiment 3.

[0168] In some embodiments, through a computer-readable storage medium, it is possible to establish a local coordinate system and a global coordinate system, obtain local coordinate information of the local coordinate system in real time through a vehicle-mounted sensor, obtain global coordinate information of the global coordinate system in real time through a satellite positioning technology, and bind the local coordinate information with the global coordinate information;

[0169] Obtain vehicle driving environment variables in real time, and obtain real-time position confidence based on vehicle driving environment variables and global coordinate information;

[0170] Determine whether the real-time position confidence is within the preset confidence threshold range. If so, record the global coordinate information at the current moment as the target coordinate information. Otherwise, calculate the target coordinate information based on the local coordinate information.

[0171] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system or other device. In addition, the embodiments of the present invention are not directed to any particular programming language.

[0172] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. Similarly, in order to simplify the present invention and help understand one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. Wherein, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself is a separate embodiment of the present invention.

[0173] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive.

[0174] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be understood as limitations on the order of execution.

Claims

1. A memory parking mapping method, characterized in that: include: Establishing a local coordinate system and a global coordinate system, obtaining local coordinate information of the local coordinate system in real time through vehicle-mounted sensors, obtaining global coordinate information of the global coordinate system in real time through satellite positioning technology, and binding the local coordinate information with the global coordinate information; Acquire vehicle driving environment variables in real time, and obtain real-time position confidence according to the vehicle driving environment variables and global coordinate information; It is determined whether the real-time position confidence is within a preset confidence threshold range. If so, the global coordinate information at the current moment is recorded as the target coordinate information. Otherwise, the target coordinate information is obtained by calculating based on the local coordinate information.

2. The memory parking mapping method according to claim 1, characterized in that: The method of establishing a local coordinate system and a global coordinate system, acquiring local coordinate information of the local coordinate system in real time through a vehicle-mounted sensor, acquiring global coordinate information of the global coordinate system in real time through a satellite positioning technology, and binding the local coordinate information with the global coordinate information includes: The local coordinate system is established with the front of the vehicle as the positive X axis and the vertical upward as the positive Z axis in a right-handed coordinate system, and the real-time vehicle translation component and rotation component are recorded and combined into the local coordinate system; Establishing the global coordinate system with the longitude, latitude and altitude detected by the satellite, and recording the real-time longitude, latitude and altitude to form the global coordinate system; After the vehicle is powered on, the initial second local coordinates of the local coordinate system of the vehicle are recorded, and the second local coordinates are transformed according to the driving state of the vehicle to obtain the real-time first local coordinates; After the initial mapping point is obtained, the first global coordinate and confidence level at that moment, as well as the first local coordinate are obtained, and the first global coordinate and the first local coordinate are bound.

3. The memory parking mapping method according to claim 1, characterized in that: The real-time acquisition of vehicle driving environment variables and obtaining real-time position confidence according to the vehicle driving environment variables and global coordinate information includes: Acquire the real-time global coordinate information through the GNSS status flag; Acquire the vehicle driving environment variables through SLAM technology; According to the global coordinate information and the vehicle driving environment variables, a preset confidence list is queried to obtain the real-time position confidence.

4. The memory parking mapping method according to claim 1, characterized in that: The determining whether the real-time position confidence is within a preset confidence threshold range, if so, recording the global coordinate information at the current moment as the target coordinate information, otherwise calculating and obtaining the target coordinate information according to the local coordinate information, includes: Obtaining the real-time location confidence, and determining whether the real-time location confidence is greater than a preset real-time location confidence threshold; If it is greater than, record the current first global coordinate and set the first confidence state. Otherwise, the second global coordinate is calculated according to the first local coordinate, the second global coordinate is recorded and a second confidence state is set instead of the first global coordinate, and the second global coordinate and the second confidence state are recorded in the map.

5. The memory parking mapping method according to claim 4, characterized in that: The step of calculating the second global coordinate according to the first local coordinate, recording the second global coordinate and replacing the first global coordinate to set a second confidence state, and recording the second global coordinate and the second confidence state in the map construction includes: Obtain the first local coordinates in the map construction, the third global coordinates in the first confidence state at the previous N moments, and the third local coordinates at the previous N moments; The third global coordinate is converted into a fourth local coordinate in the coordinate system of the northeastern sky station center, and the fifth local coordinate is calculated according to the first local coordinate, the fourth local coordinate and the third local coordinate; The fifth local coordinate is converted into a sixth global coordinate in the satellite global coordinate system, the sixth global coordinate is assigned to the second global coordinate, and the second global coordinate is set to a third confidence state.

6. The memory parking mapping method according to claim 5, characterized in that: The method further includes calculating the second global coordinates according to the first local coordinates, recording the second global coordinates and setting a second confidence state, and recording the second global coordinates and the second confidence state in the map construction, and further including: After the first local coordinates are used to calculate the second global coordinates, the first calculation time is recorded; It is determined whether the first estimated time is greater than the estimated time threshold. If so, a second confidence state is set for the second global coordinate; otherwise, a third confidence state is set.

7. The memory parking mapping method according to claim 6, characterized in that: If the calculated second global coordinate is in the second confidence state, then re-acquire the first local coordinate in the map, the seventh global coordinate in the first confidence state N-1 moments before, and the fourth local coordinate in the first N-1 moments before; and recalculating the second global coordinates according to the above method, and recording the second calculation time, and determining whether the second calculation time is greater than the calculation time threshold; If yes, recalculate and move forward one moment until the calculated second estimated time is less than the estimated time threshold, set the second global coordinate at this moment to the third confidence state, and record it in the map.

8. The memory parking mapping method according to claim 6, characterized in that: After the engine is turned off and the power is turned off, the eighth global coordinate of the engine turning off time is recorded.

9. The vehicle-mounted system is characterized in that: include: A mapping module is used to establish a local coordinate system and a global coordinate system, obtain the local coordinate information of the local coordinate system in real time through the vehicle-mounted sensor, obtain the global coordinate information of the global coordinate system in real time through the satellite positioning technology, and bind the local coordinate information with the global coordinate information; A coordinate calculation module is used to determine whether the real-time position confidence is within a preset confidence threshold range. If so, the global coordinate information at the current moment is recorded as the target coordinate information; otherwise, the target coordinate information is obtained by calculating according to the local coordinate information; The confidence judgment module is used to obtain vehicle driving environment variables in real time, and obtain real-time position confidence based on the vehicle driving environment variables and global coordinate information.

10. A computer-readable storage medium, characterized in that: The storage medium includes at least one execution instruction stored therein, and when the executable instruction is executed on the vehicle-mounted system as claimed in claim 9, the device executes the memory parking mapping method as claimed in any one of claims 1 to 8.

Images