WO2024125062A1 - 用于评估行车信息来源的方法和装置 - Google Patents

用于评估行车信息来源的方法和装置 Download PDF

Info

Publication number
WO2024125062A1
WO2024125062A1 PCT/CN2023/124441 CN2023124441W WO2024125062A1 WO 2024125062 A1 WO2024125062 A1 WO 2024125062A1 CN 2023124441 W CN2023124441 W CN 2023124441W WO 2024125062 A1 WO2024125062 A1 WO 2024125062A1
Authority
WO
WIPO (PCT)
Prior art keywords
driving information
information source
evaluated
vehicle
source
Prior art date
Application number
PCT/CN2023/124441
Other languages
English (en)
French (fr)
Inventor
周翯男
王乃岩
Original Assignee
北京图森智途科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京图森智途科技有限公司 filed Critical 北京图森智途科技有限公司
Publication of WO2024125062A1 publication Critical patent/WO2024125062A1/zh

Links

Definitions

  • the present disclosure relates to an autonomous driving technology, and more particularly to a method, device, and computer-readable storage medium for evaluating a source of driving information.
  • Static calibration requires the installation of related equipment indoors in advance, which is costly and not suitable for large-scale deployment.
  • the pre-completed sensor calibration may fail during vehicle driving, thus affecting the consistency of information and causing misjudgment or even danger.
  • Calibration verification in natural scenes requires the assistance of high-precision maps, and the geometric consistency of the elements maintained in the high-precision map is used to determine whether the calibration is invalid.
  • high-precision maps are also expensive, and when the actual environment changes but the high-precision map has not been updated, the verification results will be inaccurate.
  • the present disclosure proposes a method, device and computer-readable storage medium for evaluating driving information sources, which can estimate the current status of individual driving information sources when there are multiple driving information sources (for example, sensors), thereby improving the efficiency of troubleshooting when the calibration of driving information sources fails, and at a low cost.
  • the first aspect of the present disclosure proposes a method for evaluating a driving information source, comprising obtaining information from a driving information source to be evaluated and multiple other driving information sources; based on this information, calculating multiple parameter scores between the driving information source to be evaluated and each of the multiple other driving information sources; based on the multiple parameter scores, obtaining multiple consistency information associated with multiple states of the driving information source to be evaluated; and based on the multiple consistency information, determining the current state of the driving information source to be evaluated.
  • the second aspect of the present disclosure provides a vehicle information source evaluation device, including one or more processors and a memory storing a program.
  • the program includes instructions, which, when executed by the processor, enable the vehicle information source evaluation device to execute the above-mentioned vehicle information source evaluation method.
  • the third aspect of the present disclosure proposes a computer-readable storage medium storing a program, wherein the program includes instructions, which, when executed by one or more processors of a computing device, cause the computing device to execute the above-mentioned driving information source evaluation method.
  • FIG1 is a schematic diagram showing a method for evaluating a source of driving information according to an embodiment of the present disclosure
  • FIG2 is a flow chart showing a method for evaluating a source of driving information according to an embodiment of the present disclosure
  • FIG3 is a schematic diagram showing a method for evaluating a source of driving information according to an embodiment of the present disclosure
  • FIG4 is a flow chart showing a method for calculating a posterior probability according to an embodiment of the present disclosure
  • FIG. 5 is a schematic diagram showing a vehicle in which the various techniques disclosed herein may be implemented
  • FIG. 6 is a schematic diagram showing a computing device according to an embodiment of the present disclosure.
  • the term “plurality” means two or more, unless otherwise specified.
  • the term “and/or” describes the association relationship of associated objects, covering any one of the listed objects and all possible combinations.
  • the character “/” generally indicates that the associated objects before and after are in an “or” relationship.
  • FIG. 1 is a schematic diagram showing a method for evaluating a source of driving information according to an embodiment of the present disclosure.
  • multiple driving information sources S1 to Si of the embodiment of the present disclosure are configured on a vehicle, for example.
  • the function of these driving information sources S1 to Si is that the computing device on the vehicle can receive information from at least one of these driving information sources S1 to Si, analyze it and determine the subsequent autonomous driving behavior, for example, determine the current speed according to the distance and speed of the preceding vehicle, or determine the driving path and trajectory according to the environment.
  • the driving information sources S1-Si include a plurality of sensors.
  • the driving information sources S1-Si include high-precision maps.
  • the driving information source evaluation device 100 can be used to execute the driving information source evaluation method to evaluate the current status C1 ⁇ Ci of the driving information sources S1 ⁇ Si, where the current status C1 ⁇ Ci, for example, includes each of the driving information sources S1 ⁇ Si being normal, failed, and/or a score associated with its health status.
  • the driving information source evaluation device 100 may include, for example, a front-end module 110 and a back-end module 120.
  • the front-end module 110 may calculate parameter scores between two driving information sources based on the information, and output the calculated parameter scores to the back-end module 120.
  • the back-end module 120 may evaluate the current state C1-Ci of any one of the driving information sources S1-Si based on the parameter scores between two driving information sources.
  • the front-end module 110 and the back-end module 120 may be, for example, two different functional modules in a software program, or may be, for example, two different hardware modules.
  • the parameter score between any two driving information sources is negatively correlated with the error (e.g., calibration error) between the two driving information sources.
  • the parameter score can also be called an external parameter score between sensors.
  • the front-end module 110 is designed to define the parameter scores between various types of driving information sources within the same preset range, such as but not limited to, 0 to 1.
  • the back-end module 120 allows the back-end module 120 to be maintained without considering the type or specification of the driving information source. More specifically, since the data formats and/or units of different driving information sources may be different, when adding and/or replacing the driving information source on the vehicle, as long as the parameter scores calculated by the front-end module 110 are defined within the same preset range, the back-end module 120 can operate normally without modification.
  • the driving information source evaluation device 100 is, for example, a computing system/device/equipment disposed on a vehicle.
  • the driving information source evaluation device 100 is, for example, a remote service system that can communicate with a communication system on a vehicle.
  • the driving information source evaluation device 100 is, for example, a combination of a computing system/device/equipment on a vehicle and a remote service system.
  • the computing system/device/equipment on the vehicle can be used to implement the function of the front-end module 110, and the remote service system can be used to implement the function of the back-end module 120.
  • FIG. 2 is a flow chart showing a method for evaluating a source of driving information according to an embodiment of the present disclosure
  • FIG. 3 is a schematic diagram showing a method for evaluating a source of driving information according to an embodiment of the present disclosure.
  • the vehicle information source evaluation device 100 obtains information from multiple vehicle information sources. Specifically, during vehicle driving, the vehicle information source evaluation device 100 (eg, the front end module 110 ) continuously obtains information from multiple vehicle information sources.
  • the multiple driving information sources include driving information source A, driving information source B, and driving information source C.
  • the driving information source evaluation device 100 continuously obtains information Ia from driving information source A, information Ib from driving information source B, and information Ic from driving information source C.
  • the driving information source A is selected as the driving information source to be evaluated.
  • the same method can also be applied to other driving information sources B and C.
  • the multiple driving information sources include multiple sensors on the vehicle.
  • the multiple sensors may include, for example, an inertial measurement unit (IMU), a global navigation satellite system (GNSS) transceiver (such as a global positioning system (GPS) transceiver), a radio detection and ranging device (RADAR), a laser detection and ranging system (LiDAR), an acoustic sensor, an ultrasonic sensor, and at least one of an image capture device (such as a camera).
  • IMU inertial measurement unit
  • GNSS global navigation satellite system
  • GPS global positioning system
  • RADAR radio detection and ranging device
  • LiDAR laser detection and ranging system
  • acoustic sensor such as a laser detection and ranging system
  • ultrasonic sensor such as a camera
  • the plurality of sensors may include, for example, a plurality of sensors for monitoring the vehicle (eg, an oxygen (O 2 ) monitor, a fuel gauge sensor, an engine oil pressure sensor, and temperature, humidity, pressure sensors, etc.).
  • a plurality of sensors for monitoring the vehicle eg, an oxygen (O 2 ) monitor, a fuel gauge sensor, an engine oil pressure sensor, and temperature, humidity, pressure sensors, etc.
  • the multiple driving information sources include at least two of an inertial measurement unit, a lidar, and a camera.
  • the multiple driving information sources only include multiple sensors on the vehicle. More specifically, the driving information source evaluation method introduced in the embodiments of the present disclosure can also evaluate the current state of each driving information source (sensor) when there is no standardized static calibration equipment or high-precision map.
  • multiple sources of driving information include high-precision maps.
  • the high-precision map used as the source of driving information may be, for example, stored in a memory on the vehicle, or may be, for example, stored in at least one node of the network.
  • the high-precision map may also be used as an evaluation object of the driving information source evaluation method proposed in the present disclosure, so as to detect errors in the high-precision map (for example, inconsistent with the actual scene) when the error occurs, without affecting the evaluation of other sources of driving information.
  • step S220 the driving information source evaluation device 100 calculates a plurality of parameter scores between the driving information source to be evaluated and each of the plurality of other driving information sources based on the information received in step S210 . Specifically, during the driving process of the vehicle, the driving information source evaluation device 100 (e.g., the front-end module 110 ) continuously calculates a plurality of parameter scores between the driving information source to be evaluated and each of the plurality of other driving information sources.
  • the driving information source evaluation device 100 calculates a parameter score X between the driving information source A to be evaluated and the driving information source B, and a parameter score Z between the driving information source A to be evaluated and the driving information source C.
  • the driving information source evaluation device 100 may calculate a parameter score between a camera and an inertial measurement unit, for example.
  • the driving information source evaluation device 100 obtains two frame images from the camera, such as a current frame image and a previous frame image, and the scenes in these two frame images may be displaced. If the calibration between the camera and the inertial measurement unit is consistent, the information of the inertial measurement unit should be able to accurately predict the position of a specific object in the previous frame image in the current frame image. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 selects a specific indicator in a previous frame image, for example.
  • the vehicle information source evaluation device 100 samples (eg, downsamples) specific indicators (eg, road markings) on the ground plane in the previous frame image to obtain a plurality of sampling points P l in the previous frame image.
  • specific indicators eg, road markings
  • the driving information source evaluation device 100 predicts the position of the specific indicator in the previous frame image in the current frame image, for example based on the information of the inertial measurement unit.
  • the vehicle information source evaluation device 100 calculates multiple intersection points Pik-1 of the rays from the optical center of the camera to the multiple sampling points P l and the ground plane in the coordinate system of the inertial measurement unit based on the camera intrinsic parameter K , the external parameter Ti c between the camera and the inertial measurement unit, the height of the inertial measurement unit from the ground, and other information. Based on the time difference between the two frame images, the vehicle information source evaluation device 100 can integrate the information of the inertial measurement unit to obtain the attitude change T d between the shooting time of the two frame images.
  • the driving information source evaluation device 100 may use the prediction accuracy as the basis for parameter scoring to calculate the parameter score.
  • the prediction accuracy may be obtained by comparing the predicted position and the actual position of the specific indicator in the current frame image.
  • the driving information source evaluation device 100 can calculate the proportion of multiple second prediction points I that actually fall within a specific indicator in the current frame image (for example, the number of second prediction points I that actually fall within the specific indicator in the current frame image/the total number of second prediction points I) as the parameter score.
  • the driving information source evaluation device 100 may calculate a parameter score between two cameras.
  • the driving information source evaluation device 100 obtains a frame of image from two cameras respectively, and the scenes in the two frames of image will be displaced. If the calibration between the two cameras is consistent, then using the external parameters between the inertial measurement unit and the two cameras, it should be possible to accurately predict the position of a specific object in one frame of image in the other frame of image. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 selects a specific indicator from a first frame image from a first camera, for example.
  • the driving information source evaluation device 100 samples (eg, downsamples) specific indicators (eg, road markings) on the ground plane in the first frame image to obtain a plurality of sampling points P l1 in the first frame image.
  • specific indicators eg, road markings
  • the driving information source evaluation device 100 predicts the position of the specific indicator in the first frame image in the second frame image based on the external parameters between the inertial measurement unit and the two cameras, for example.
  • the vehicle information source evaluation device 100 calculates a plurality of intersection points Pic1 of rays from the optical center of the first camera to a plurality of sampling points P l1 and the ground plane in the coordinate system of the inertial measurement unit based on the first camera intrinsic parameter K 1 , the extrinsic parameter Ti c1 between the first camera and the inertial measurement unit, the height of the inertial measurement unit from the ground, etc.
  • the driving information source evaluation device 100 may use the prediction accuracy as the basis for parameter scoring to calculate the parameter score.
  • the prediction accuracy may be obtained by comparing the predicted position of the specific indicator in the second frame image with the actual position.
  • the driving information source evaluation device 100 can calculate the proportion of multiple prediction points I that actually fall within a specific indicator in the second frame image (for example, the number of prediction points I that actually fall within the specific indicator in the second frame image/the total number of prediction points I) as the parameter score.
  • the driving information source evaluation device 100 may, for example, calculate a parameter score between a camera and a high-precision map.
  • the driving information source evaluation device 100 obtains the current frame image from the camera. If there is no error or the calibration between the camera and the high-precision map is consistent, the position of a specific object in the high-precision map in the current frame image should be accurately predicted by using the external parameters between the inertial measurement unit and the camera and the posture of the inertial measurement unit in the high-precision map coordinate system. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 selects, for example, specific indicators in a high-precision map.
  • the driving information source evaluation device 100 samples (eg, upsamples) specific indicators (eg, road markings) on the high-precision map to obtain a plurality of sampling points P m of the high-precision map.
  • specific indicators eg, road markings
  • the driving information source evaluation device 100 predicts the position of a specific indicator in the high-precision map in the current frame image based on the external parameters between the inertial measurement unit and the camera and the posture of the inertial measurement unit in the high-precision map coordinate system.
  • the vehicle information source evaluation device 100 may be based on the camera intrinsic parameter K, the relationship between the camera and the inertial measurement unit,
  • the driving information source evaluation device 100 may use the prediction accuracy as the basis for parameter scoring to calculate the parameter score.
  • the prediction accuracy may be obtained by comparing the predicted position and the actual position of the specific indicator in the current frame image.
  • the driving information source evaluation device 100 can calculate the proportion of multiple prediction points I that actually fall within a specific indicator in the current frame image (for example, the number of prediction points I that actually fall within the specific indicator in the current frame image/the total number of prediction points I) as the parameter score.
  • the driving information source evaluation device 100 may calculate another parameter score between the camera and the high-precision map.
  • the driving information source evaluation device 100 obtains the current frame image from the camera. If there is no error or the calibration between the camera and the high-precision map is consistent, the position of the specific object in the current frame image in the high-precision map can be accurately predicted by using the external parameters between the camera and the inertial measurement unit and the posture of the inertial measurement unit in the high-precision map coordinate system. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 selects, for example, a specific indicator in the current frame image and the same specific indicator in the high-precision map.
  • the vehicle information source evaluation device 100 samples (e.g., downsamples) a specific indicator (e.g., road markings) on the ground plane in the current frame image to obtain a plurality of image sampling points P l in the current frame image.
  • the vehicle information source evaluation device 100 samples (e.g., upsamples) the same specific indicator (e.g., road markings) on the high-precision map to obtain a plurality of map sampling points P m of the high-precision map.
  • the driving information source evaluation device 100 predicts the position of a specific indicator in the current frame image in the high-precision map based on the external parameters between the camera and the inertial measurement unit and the posture of the inertial measurement unit in the high-precision map coordinate system.
  • the driving information source evaluation device 100 may use the predicted recall rate as the basis for parameter scoring to calculate the parameter score.
  • the predicted recall rate may be obtained by comparing the predicted position of a specific indicator in a high-precision map with the actual position.
  • the driving information source evaluation device 100 can traverse multiple prediction points P im , search for map sampling points P m with a specific radius (e.g., 0.5 meters), and calculate the search success ratio (e.g., the number of prediction points P im that successfully searched for map sampling points P m within the specific radius/the total number of prediction points P im ) as the parameter score.
  • a specific radius e.g., 0.5 meters
  • the search success ratio e.g., the number of prediction points P im that successfully searched for map sampling points P m within the specific radius/the total number of prediction points P im
  • the driving information source evaluation device 100 may calculate a parameter score between a laser radar and an inertial measurement unit, for example.
  • the driving information source evaluation device 100 uses a LiDAR odometer to calculate the linear velocity of the LiDAR in the coordinate system of the inertial measurement unit. If the calibration between the LiDAR and the inertial measurement unit is consistent, the calculated linear velocity should be the same as the linear velocity observed by the inertial measurement unit itself. Therefore, the consistency of the two linear velocities can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 calculates the posture change between two time points using, for example, a laser radar odometer, and converts the posture change into the coordinate system of the inertial measurement unit based on the external parameters between the laser radar and the inertial measurement unit.
  • the driving information source evaluation device 100 calculates the linear velocity of the laser radar in the coordinate system of the inertial measurement unit, for example.
  • the driving information source evaluation device 100 may calculate the parameter score by, for example, using the consistency between the calculated linear velocity and the linear velocity observed by the inertial measurement unit itself as the basis for the parameter score. This consistency may be obtained, for example, by comparing the two linear velocities.
  • the linear velocity observed by the inertial measurement unit itself is V i .
  • the driving information source evaluation device 100 can use a function whose value range falls between 0 and 1 and whose value is negatively correlated with the difference between the two linear velocities (for example, exp(-1.0*normalize(V i -V li )), where (V i -V li ) is, for example, a three-dimensional velocity vector, and normalize(V i -V li ) is to take the two-norm of the velocity vector) to represent the parameter score.
  • the driving information source evaluation device 100 may, for example, calculate a parameter score between two laser radars.
  • the driving information source evaluation device 100 uses the laser radar odometer to calculate the two linear velocities of the two laser radars in the coordinate system of the inertial measurement unit. If the calibration between the two laser radars is consistent, the two calculated linear velocities should be the same. Therefore, the consistency of the two linear velocities can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 uses a first laser radar odometer to calculate a first posture change between two time points, and converts the first posture change to the coordinate system of the inertial measurement unit based on the external parameters between the first laser radar and the inertial measurement unit; and uses a second laser radar odometer to calculate a second posture change between two time points, and converts the second posture change to the coordinate system of the inertial measurement unit based on the external parameters between the second laser radar and the inertial measurement unit.
  • the driving information source evaluation device 100 calculates, for example, a first linear velocity of the first laser radar in the coordinate system of the inertial measurement unit; and calculates a second linear velocity of the second laser radar in the coordinate system of the inertial measurement unit.
  • the vehicle information source evaluation device 100 may calculate the parameter score by taking the calculated consistency between the first linear speed and the second linear speed as the basis for the parameter score.
  • the consistency may be obtained by comparing the two linear speeds, for example.
  • the driving information source evaluation device 100 can use a function whose value range falls between 0 and 1 and whose value is negatively correlated with the difference between the first linear velocity V l1i and the second linear velocity V l2i (for example, exp(-1.0*normalize(V l1i -V l2i )), where (V l1i -V l2i ) is, for example, a three-dimensional velocity vector, and normalize(V l1i -V l2i ) is taking the second norm of the velocity vector) to represent the parameter score.
  • a function whose value range falls between 0 and 1 and whose value is negatively correlated with the difference between the first linear velocity V l1i and the second linear velocity V l2i (for example, exp(-1.0*normalize(V l1i -V l2i )), where (V l1i -V l2i ) is, for example, a three-dimensional velocity vector, and normalize(V l1i
  • the driving information source evaluation device 100 may calculate a parameter score between a laser radar and a camera, for example.
  • the driving information source evaluation device 100 obtains the current frame image from the camera and uses the laser radar to detect the position of the specific object. If the laser radar and the camera are calibrated in the same way, the position of the specific object in the current frame image can be accurately predicted using the external parameters between the laser radar and the camera. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring. Therefore, the accuracy of this prediction can be used as the basis for parameter scoring.
  • the driving information source evaluation device 100 detects specific indicators using, for example, a laser radar.
  • the vehicle information source evaluation device 100 may utilize a laser radar sensor to extract points belonging to a specific indicator (eg, a vehicle) to obtain a plurality of sampling points P l .
  • a specific indicator eg, a vehicle
  • the driving information source evaluation device 100 predicts the position of the specific indicator detected by the laser radar in the current frame image based on the external parameters between the radar and the camera.
  • the driving information source evaluation device 100 may use the prediction accuracy as the basis for parameter scoring to calculate the The prediction accuracy can be obtained by, for example, comparing the predicted position of a specific indicator in the current frame image with the actual position.
  • the driving information source evaluation device 100 can calculate the proportion of multiple prediction points I that actually fall within a specific indicator in the current frame image (for example, the number of prediction points I that actually fall within the specific indicator in the current frame image/the total number of prediction points I) as the parameter score.
  • step S230 the driving information source evaluation device 100 (eg, the backend module 120 ) obtains a plurality of consistency information associated with a plurality of states of the driving information source to be evaluated based on the plurality of parameter scores calculated in step S220 .
  • the multiple parameter scores represent whether the calibration between the vehicle information source to be evaluated and each of the other vehicle information sources is consistent or harmonious.
  • the parameter score is the result jointly generated by the two vehicle information sources. Based on whether the calibration between the vehicle information source to be evaluated and each of the other vehicle information sources is consistent or harmonious, it is possible to infer the current state of the calibration of the vehicle information source to be evaluated.
  • the consistency information is used to indicate the degree of consistency between the current state of the driving information source to be evaluated and each possible state. If the degree of consistency between the current state displayed by the consistency information and the specific state is higher, it means that the current state of the driving information source to be evaluated is closer to this specific state.
  • the parameter score is defined within a preset range (e.g., 0 to 1), it can be used to determine the conditional probability function of the parameter score of the to-be-evaluated driving information source and each of the other driving information sources under different state combinations (the specific determination method will be described in detail later). In this way, the consistency information of the to-be-evaluated driving information source under each state can be represented by the posterior probability.
  • a preset range e.g., 0 to 1
  • the multiple states of the driving information source to be evaluated include, for example, two states (eg, normal and failure).
  • the present disclosure does not limit the number of possible states of the driving information source to be evaluated.
  • the normal state of the driving information source A is recorded as a, and the failure state is recorded as a'; the normal state of the driving information source B is recorded as b, and the failure state is recorded as b'; the normal state of the driving information source C is recorded as c, and the failure state is recorded as c'.
  • the prior probability that driving information source A is in a normal state is recorded as P(a)
  • the prior probability that driving information source A is in a failed state is recorded as P(a’)
  • the prior probability that driving information source B is normal is recorded as P(b)
  • the prior probability that driving information source B is failed is recorded as P(b’
  • the prior probability that driving information source C is normal is recorded as P(c)
  • the prior probability that driving information source C is failed is recorded as P(c’).
  • the various state combinations associated with the driving information source A to be evaluated include (a,b), (a’,b), (a,b’), (a’,b’), (a,c), (a’,c), (a,c’), (a’,c’).
  • the conditional probability functions of multiple parameter scores include P(X
  • the posterior probability of the driving information source A to be evaluated being in a normal state can be calculated as: P(a
  • x,y,z) kP(a)*[P(b)*P(x
  • x,y,z) kP(a')*[P(b)*P(x
  • x,y,z) of the driving information source A to be evaluated in the normal state is high, it means that the current state of the driving information source A to be evaluated is closer to the normal state; if the posterior probability P(a’
  • FIG. 4 is a flow chart showing a method for calculating a posterior probability according to an embodiment of the present disclosure.
  • step S230 is, for example, based on the multiple parameter scores calculated in step S220, obtaining multiple posterior probabilities associated with multiple states of the driving information source to be evaluated.
  • step S230 further includes step S231 and step S232.
  • step S231 the vehicle information source evaluation device 100 determines a plurality of conditional probability functions associated with each parameter score of the vehicle information source to be evaluated based on a plurality of state combinations.
  • the parameter score associated with the driving information source A to be evaluated includes parameter score X and parameter score Z.
  • the multiple conditional probability functions of parameter score X based on multiple state combinations include P(X
  • the multiple conditional probability functions of parameter score Z based on multiple state combinations include P(Z
  • step S231 includes step S2311 and step S2312 .
  • step S2311 the driving information source evaluation device 100 determines a first distribution function of an expected result that satisfies one of the plurality of state combinations based on a plurality of first score values corresponding to the first parameter score in the plurality of parameter scores.
  • step S2312 the driving information source evaluation device 100 determines a plurality of first conditional probability functions in the plurality of conditional probability functions corresponding to the first parameter score based on the first distribution function.
  • each state combination can have an expected result.
  • the conditional probability function of the parameter score in each state combination can be described using a specific distribution function.
  • conditional probability distribution function of the parameter score when all the driving information sources corresponding to the state combination are in normal state, it can be expected that the conditional probability distribution function of the parameter score will be a strictly increasing function, and the parameter score has a high probability of being close to the upper limit of the preset range; when there is at least one driving information source that is in an invalid state, it can be expected that the conditional probability distribution function of the parameter score will be a strictly decreasing function, and the parameter score has a high probability of being close to the lower limit of the preset range. It is worth mentioning that the domain of this conditional probability distribution function is defined within the same preset range as the parameter score.
  • the first parameter score is, for example, the parameter score X.
  • the first distribution function that meets the expected result of the state combination (a, b) can be, for example, the cumulative distribution function (CDF) of the truncated normal distribution:
  • ⁇ and ⁇ are, for example, respectively the mean and variance of a plurality of first score values xi of the first parameter score X within a preset period of time (eg, 500 frames).
  • the first distribution function may also be other types of functions, and the present disclosure is not limited thereto.
  • a',b') 1-F(x; ⁇ , ⁇ ,m,n),
  • a,b) may also be determined as [F(x+0.5 ⁇ ; ⁇ , ⁇ ,m,n)-F(x-2 ⁇ ; ⁇ , ⁇ ,m,n)] described by the first distribution function.
  • the second parameter score is, for example, a parameter score Z.
  • a',c') corresponding to the second parameter score Z can also be determined, so the specific process is not repeated here.
  • step S232 the driving information source evaluation device 100 calculates multiple posterior probabilities based on the multiple conditional probability functions and multiple current parameter scores among the multiple parameter scores.
  • the prior probability of the source of the driving information to be evaluated in each state is equal by default. For example, if the source of the driving information to be evaluated has three possible states, the prior probability of each state is 1/3 by default; if the source of the driving information to be evaluated has two possible states, the prior probability of each state is 1/2 by default.
  • the prior probability of each driving information source in the same state is equal by default. For example, if each driving information source has two possible states, a normal state and a failure state, the prior probability of each driving information source in the normal state is equal (for example, all p), and the prior probability of each driving information source in the failure state is also equal (for example, all 1-p).
  • the prior probability of each driving information source in each state can be preset through historical data, factory data of the driving information source, or other information.
  • the posterior probability of the driving information source A to be evaluated in a normal state is: P(a
  • x,y,z) kP(a)*[P(b)*P(x
  • x,y,z) kP(a')*[P(b)*P(x
  • the prior probabilities of each driving information source A, B, and C in the normal state and the failure state are equal (for example, all are 0.5).
  • the scores x, y, and z of the multiple current parameter scores are substituted into the multiple conditional probability functions determined in step S231, and each posterior probability of the driving information source A to be evaluated in the normal state and the failure state can be obtained.
  • the driving information source evaluation device 100 since the multiple posterior probabilities are only used for comparison in the future and the actual values do not need to be taken into consideration, the driving information source evaluation device 100 does not calculate the normalization constant k (eg, setting k to 1).
  • step S240 the driving information source evaluation device 100 determines the current state of the driving information source to be evaluated based on the plurality of consistency information obtained in step S230 .
  • the consistency information is used to indicate the consistency between the current state of the driving information source to be evaluated and each state.
  • the consistency information is, for example, a posterior probability
  • the driving information source evaluation device 100 may determine, for example, that the current state is the state associated with the largest one of the plurality of posterior probabilities calculated in step S232 .
  • the driving information source evaluation device 100 calculates, for example, in step S232 the posterior probability of the driving information source A to be evaluated being in a normal state and the posterior probability of the driving information source A to be evaluated being in a failed state. If the posterior probability of the driving information source A to be evaluated being in a normal state is greater than the posterior probability of the driving information source A to be evaluated being in a failed state, the normal state will be determined as the current state of the driving information source A to be evaluated; if the posterior probability of the driving information source A to be evaluated being in a normal state is less than the posterior probability of the driving information source A to be evaluated being in a failed state, the failed state will be determined as the current state of the driving information source A to be evaluated.
  • the current state of each driving information source can be determined, and based on this, the subsequent troubleshooting of the driving information source failure/calibration failure/expiration is also more convenient.
  • the driving information source evaluation device 100 may display the evaluation score (e.g., health score) of the driving information source to be evaluated, so that the operator/technician can instantly grasp the status of the driving information source to be evaluated.
  • the evaluation score of the driving information source A to be evaluated may be its posterior probability in a specific state (e.g., normal state).
  • the driving information source evaluation device 100 (e.g., the backend module 120) normalizes multiple posterior probabilities of the driving information source to be evaluated in multiple states to obtain multiple normalized posterior probabilities, and then outputs the normalized posterior probability corresponding to one of the specific states.
  • the vehicle information source evaluation device 100 does not set k to 1, but calculates a normalization constant k to calculate a plurality of normalized posterior probabilities, and then outputs the normalized posterior probability corresponding to the vehicle information source A to be evaluated as its evaluation score.
  • the driving information source evaluation device 100 can display the evaluation score of each driving information source, so that the operator/technician can understand the status of all driving information sources at the same time.
  • the driving information source evaluation device 100 may issue a warning according to the current state or evaluation score of each driving information source. For example, the driving information source evaluation device 100 may determine whether the current state of each driving information source is faulty/calibration invalid/expired, or whether the evaluation score is lower than a preset threshold, and issue a warning for the driving information source whose current state is faulty/calibration invalid/expired or whose evaluation score is lower than the preset threshold. For example, the warning may be issued in different display colors, warning sounds, etc.
  • the driving information source evaluation device 100 may, for example, determine whether to execute a risk reduction strategy during the driving process of the vehicle based on the current state or evaluation score of the driving information source to be evaluated. For example, the driving information source evaluation device 100 may determine to execute a risk reduction strategy during the driving process of the vehicle when a specific driving information source to be evaluated is faulty/decalibrated invalid/expired, or the evaluation score is lower than a preset threshold.
  • the driving information source evaluation device 100 may, for example, determine whether to execute a risk reduction strategy during the driving process of the vehicle based on multiple current states or evaluation scores of multiple driving information sources. For example, the driving information source evaluation device 100 may determine to execute a risk reduction strategy during the driving process of the vehicle when multiple current states or evaluation scores of multiple driving information sources meet a preset condition (e.g., more than a preset number of driving information sources are faulty/de-calibrated/expired, or the sum of the evaluation scores is lower than a preset threshold).
  • a preset condition e.g., more than a preset number of driving information sources are faulty/de-calibrated/expired, or the sum of the evaluation scores is lower than a preset threshold.
  • the risk reduction strategy includes, for example, downgrading the autonomous driving level or executing a minimum risk strategy (Minimal Risk Maneuver, MRM).
  • the risk reduction strategy includes, for example, shutting down one of the multiple sensors.
  • the driving information source evaluation device 100 can shut down the sensor.
  • the risk reduction strategy can be to shut down only the driving information source that is currently in a faulty/calibrated invalid/expired state (for example, a driving information source with an evaluation score lower than a preset threshold), while keeping the vehicle moving and other driving information sources running.
  • the driving information source evaluation device 100 may, for example, calibrate multiple external parameters of the multiple sensors according to the current state of the driving information source to be evaluated. For example, when the calibration of the sensor serving as the driving information source to be evaluated fails or the evaluation score is lower than a preset threshold, the driving information source evaluation device 100 may calibrate the external parameters between it and other sensors.
  • the driving information source evaluation device 100 can, for example, perform online recalibration to correct external parameters between sensors. For example, the driving information source evaluation device 100 can automatically and iteratively adjust the calibration of a sensor with a failed calibration to restore the failed calibration sensor to a normal state, or optimize the sum of the evaluation scores of all driving information sources (for example, to a maximum value). In some embodiments, the sensor can also be returned to the factory for recalibration. The sensor is recalibrated to correct the external parameters between sensors.
  • FIG. 5 is a schematic diagram showing a vehicle in which the various techniques disclosed herein may be implemented.
  • vehicle 500 may be a car, truck, motorcycle, bus, boat, airplane, helicopter, lawnmower, excavator, snowmobile, aircraft, recreational vehicle, amusement park vehicle, farm equipment, construction equipment, tram, golf cart, train, trolley bus, or other vehicle.
  • Vehicle 500 may be operated in full or in part in an autonomous driving mode.
  • Vehicle 500 may control itself in the autonomous driving mode, for example, vehicle 500 may determine the current state of the vehicle and the current state of the environment in which the vehicle is located, determine the predicted behavior of at least one other vehicle in the environment, determine the trust level corresponding to the possibility of the at least one other vehicle performing the predicted behavior, and control vehicle 500 itself based on the determined information.
  • vehicle 500 may operate without human interaction.
  • the vehicle 500 may include various vehicle systems, such as a drive system 542, a sensor system 544, a control system 546, a user interface system 548, a computing system 550, and a communication system 552.
  • the vehicle 500 may include more or fewer systems, each of which may include multiple units. Further, each system and unit of the vehicle 500 may be interconnected.
  • the computing system 550 may communicate data with one or more of the drive system 542, the sensor system 544, the control system 546, the user interface system 548, and the communication system 552.
  • the functions described in the vehicle 500 may be divided into additional functional components or physical components, or combined into a smaller number of functional components or physical components.
  • additional functional components or physical components may be added to the example shown in Figure 5.
  • the drive system 542 may include a plurality of operable components (or units) that provide kinetic energy to the vehicle 500.
  • the drive system 542 may include an engine or motor, wheels, a transmission, an electronic system, and a power source (or power source).
  • the engine or motor may be any combination of the following devices: an internal combustion engine, an electric motor, a steam engine, a fuel cell engine, a propane engine, or other forms of engines or motors.
  • the engine may convert a power source into mechanical energy.
  • the drive system 542 may include a plurality of engines or motors.
  • a hybrid vehicle may include a gasoline engine and an electric motor, and may also include other situations.
  • the wheels of vehicle 500 can be standard wheels.
  • the wheels of vehicle 500 can be various types of wheels, including single-wheel, double-wheel, three-wheel, or four-wheel forms, such as four wheels on a car or truck. Other numbers of wheels are also possible, such as six or more wheels.
  • One or more wheels of vehicle 500 can be operated to rotate in a different direction than the other wheels.
  • the wheel can be at least one wheel fixedly connected to a transmission.
  • the wheel can include a combination of metal and rubber, or a combination of other substances.
  • the transmission can include a unit that can be operated to transmit the mechanical power of the engine to the wheel.
  • the transmission can include a gearbox, a clutch, a differential gear, and a drive shaft.
  • the transmission can also include other units.
  • the drive shaft can include one or more axles that match the wheels.
  • the electronic system can include a unit for transmitting or controlling electronic signals of vehicle 500. These electronic signals can be used to start multiple lights, multiple servo mechanisms, multiple motors, and other electronic drive or control devices in vehicle 500.
  • the power source can be an energy source that provides power to the engine or motor in whole or in part. That is, the engine or motor can The power source is converted into mechanical energy.
  • the power source may include gasoline, oil, petroleum-based fuels, propane, other compressed gas fuels, ethanol, fuel cells, solar panels, batteries, and other electrical energy sources.
  • the power source may additionally or alternatively include any combination of a fuel tank, a battery, a capacitor, or a flywheel.
  • the power source may also provide energy for other systems of the vehicle 500.
  • the sensor system 544 may include a plurality of sensors for sensing information about the environment and conditions of the vehicle 500.
  • the sensor system 544 may include an inertial measurement unit (IMU), a global navigation satellite system (GNSS) transceiver (e.g., a global positioning system (GPS) transceiver), a radio detection and ranging device (RADAR, referred to as radar), a laser detection and ranging system (LiDAR, referred to as laser radar), an acoustic sensor, an ultrasonic sensor, and an image capture device (e.g., a camera).
  • IMU inertial measurement unit
  • GNSS global navigation satellite system
  • GPS global positioning system
  • RADAR radio detection and ranging device
  • LiDAR laser detection and ranging system
  • acoustic sensor e.g., a laser radar
  • ultrasonic sensor e.g., a camera
  • image capture device e.g., a camera
  • the sensor system 544 may include a plurality of sensors for monitoring the vehicle 500 (e.g., an oxygen (O 2 ) monitor, a fuel gauge sensor, an engine oil pressure sensor, and temperature, humidity, pressure sensors, etc.). Other sensors may also be configured. One or more sensors included in the sensor system 544 may be driven individually or collectively to update the position, orientation, or both of the one or more sensors.
  • O 2 oxygen
  • Other sensors may also be configured.
  • One or more sensors included in the sensor system 544 may be driven individually or collectively to update the position, orientation, or both of the one or more sensors.
  • the IMU may include a combination of sensors (e.g., accelerometers and gyroscopes) for sensing attitude changes (e.g., position changes and direction changes) of the vehicle 500 based on inertial acceleration.
  • the GPS transceiver may be any sensor for estimating the geographic location of the vehicle 500.
  • the GPS transceiver may include a receiver/transmitter to provide location information of the vehicle 500 relative to the earth.
  • GPS is an example of a global navigation satellite system, so in some embodiments, the GPS transceiver may be replaced by a Beidou satellite navigation system transceiver or a Galileo satellite navigation system transceiver.
  • the radar unit may use radio signals to sense objects in the environment where the vehicle 500 is located.
  • the radar unit may also be used to sense the speed and direction of objects approaching the vehicle 500.
  • the LiDAR unit may be any sensor that uses lasers to sense objects in the environment where the vehicle 500 is located.
  • the LiDAR unit may include a laser source, a laser scanner, and a detector.
  • the LiDAR unit is used to operate in a continuous (e.g., using heterodyne detection) or discontinuous detection mode.
  • the image capture device may include a device for capturing multiple images of the environment where the vehicle 500 is located.
  • An example of an image capture device is a camera, which may be a still image camera or a motion video camera.
  • the control system 546 is used to control the operation of the vehicle 500 and its components (or units). Accordingly, the control system 546 may include various units, such as a steering unit, a power control unit, a braking unit, and a navigation unit.
  • the steering unit can be a combination of machines that adjust the direction of the vehicle 500.
  • the power control unit for example, a throttle
  • the braking unit can include a combination of machines for decelerating the vehicle 500.
  • the braking unit can use friction to decelerate the vehicle in a standard manner. In other embodiments, the braking unit can convert the kinetic energy of the wheels into electric current.
  • the braking unit can also take other forms.
  • the navigation unit can be any system that determines the driving path or route for the vehicle 500.
  • the navigation unit can also dynamically update the driving path while the vehicle 500 is moving.
  • the control system 546 can also additionally or optionally include other components (or units) that are not shown or described.
  • the user interface system 548 may be used to allow the vehicle 500 to communicate with external sensors, other vehicles, other computer systems, and other
  • the user interface system 548 may include, for example, a standard visual display device (e.g., a plasma display, a liquid crystal display (LCD), a touch screen display, a head mounted display, or other similar display), a speaker or other audio output device, and a microphone or other audio input device.
  • a standard visual display device e.g., a plasma display, a liquid crystal display (LCD), a touch screen display, a head mounted display, or other similar display
  • a speaker or other audio output device e.g., a speaker or other audio output device
  • the user interface system 548 may also include a navigation interface and an interface for controlling the interior environment of the vehicle 500 (e.g., temperature, fans, etc.).
  • the communication system 552 can provide a way for the vehicle 500 to communicate with one or more devices or other vehicles in the vicinity.
  • the communication system 552 can communicate with one or more devices directly or through a communication network.
  • the communication system 552 can be, for example, a wireless communication system.
  • the communication system can use 3G cellular communication (e.g., CDMA, EVDO, GSM/GPRS) or 4G cellular communication (e.g., WiMAX or LTE), and can also use 5G cellular communication.
  • 3G cellular communication e.g., CDMA, EVDO, GSM/GPRS
  • 4G cellular communication e.g., WiMAX or LTE
  • the communication system can communicate with a wireless local area network (WLAN) (e.g., using
  • WLAN wireless local area network
  • the communication system 552 can communicate directly with one or more devices or other vehicles around it, for example, using infrared, Bluetooth, or other communication methods. , or ZIGBEE.
  • Other wireless protocols such as various vehicle-mounted communication systems, are also within the scope of the present disclosure.
  • the communication system may include one or more dedicated short-range communication (DSRC) devices, V2V devices, or V2X devices that communicate data with vehicles and/or roadside stations in public or private.
  • DSRC dedicated short-range communication
  • the computing system 550 can control some or all functions of the vehicle 500.
  • the autonomous driving control unit in the computing system 550 can be used to identify, evaluate, and avoid or cross potential obstacles in the environment where the vehicle 500 is located.
  • the autonomous driving control unit can be used to control the vehicle 500 without a driver, or to assist the driver in controlling the vehicle.
  • the autonomous driving control unit is used to combine data from sensors, such as data from a GPS transceiver, radar data, LiDAR data, camera data, and data from other vehicle systems to determine the driving path or trajectory of the vehicle 500.
  • the autonomous driving control unit can be activated to enable the vehicle 500 to be driven in an autonomous driving mode.
  • the computing system 550 may include at least one processor (which may include at least one microprocessor) that executes processing instructions (i.e., machine executable instructions) stored in a non-volatile computer-readable medium (e.g., a data storage device or memory).
  • the computing system 550 may also be a plurality of computing devices that control components or systems of the vehicle 500 in a distributed manner.
  • the memory may contain processing instructions (e.g., program logic) that are executed by the processor to implement various functions of the vehicle 500.
  • the computing system 550 is capable of data communication with the drive system 542, the sensor system 544, the control system 546, the user interface system 548, and/or the communication system 552.
  • the interface in the computing system is used to facilitate data communication between the computing system 550 and the drive system 542, the sensor system 544, the control system 546, the user interface system 548, and the communication system 552.
  • the memory may also include other instructions, including instructions for data transmission, instructions for data reception, instructions for interaction, or instructions for controlling the drive system 542 , the sensor system 544 , or the control system 546 or the user interface system 548 .
  • the memory may store a variety of information or data, such as image processing parameters, road maps, and path information. This memory may be used during operation of the vehicle 500 in an automated, semi-automated, and/or manual mode. This information may be used by vehicle 500 and computing system 550 .
  • the autonomous driving control unit is shown as being separate from the processor and memory, it should be understood that in some embodiments, some or all of the functions of the autonomous driving control unit may be implemented using program code instructions residing in one or more memories (or data storage devices) and executed by one or more processors, and the autonomous driving control unit may in some cases be implemented using the same processor and/or memory (or data storage device). In some embodiments, the autonomous driving control unit may be implemented at least in part using various dedicated circuit logic, various processors, various field programmable gate arrays (“FPGA”), various application specific integrated circuits (“ASIC”), various real-time controllers, and hardware.
  • FPGA field programmable gate arrays
  • ASIC application specific integrated circuits
  • Computing system 550 can control functions of vehicle 500 based on input received from various vehicle systems (e.g., drive system 542, sensor system 544, and control system 546), or input received from user interface system 548. For example, computing system 550 can use input from control system 546 to control a steering unit to avoid an obstacle detected by sensor system 544. In some embodiments, computing system 550 can be used to control multiple aspects of vehicle 500 and its systems.
  • vehicle systems e.g., drive system 542, sensor system 544, and control system 546
  • computing system 550 can use input from control system 546 to control a steering unit to avoid an obstacle detected by sensor system 544.
  • computing system 550 can be used to control multiple aspects of vehicle 500 and its systems.
  • FIG. 5 shows various components (or units) integrated into the vehicle 500
  • these components may be mounted on the vehicle 500 or separately associated with the vehicle 500.
  • the computing system 550 may exist partially or completely independently of the vehicle 500.
  • the vehicle 500 can exist in the form of separate or integrated equipment units.
  • the equipment units that make up the vehicle 500 can communicate with each other in the form of wired or wireless communications.
  • additional components or units may be added to each system or one or more components or units (e.g., the LiDAR or radar shown in FIG. 5) may be removed from the system.
  • FIG. 6 is a schematic diagram showing a computing device according to an embodiment of the present disclosure.
  • the computing system 550 may be implemented in the form of a computing device 600 .
  • the instruction set when executed and/or the processing logic when started can make the machine perform any one or more of the methods described and/or required herein.
  • the machine operates as a standalone device, or can be connected (e.g., networked) to other machines. In a networked deployment, the machine can operate as a server or client machine in a server-client network environment, or as a peer machine in a peer (or distributed) network environment.
  • the machine can be a personal computer (PC), a laptop computer, a tablet computing system, a personal digital assistant (PDA), a cellular phone, a smart phone, a network application, a set-top box (STB), a network router, a switch or a bridge or any machine capable of executing an instruction set (in succession or otherwise) specifying the action to be taken by the machine or starting processing logic.
  • PC personal computer
  • PDA personal digital assistant
  • STB set-top box
  • STB set-top box
  • switch or a bridge any machine capable of executing an instruction set (in succession or otherwise) specifying the action to be taken by the machine or starting processing logic.
  • the computing device 600 may include a data processor 602 (e.g., a system on a chip (SoC), a general processing core, a graphics core, and optional other processing logic) and a storage 604 (e.g., memory) that may communicate with each other via a bus 606 or other data transfer system.
  • the computing device 600 may also include various input/output (I/O) devices and/or interfaces 610, such as a touch screen display, an audio jack, a voice interface, and an optional network interface 612.
  • I/O input/output
  • the network interface 612 may include one or more radio transceivers configured to communicate with any one or more standard wireless and/or cellular protocols or access technologies (e.g., second generation (2G), 2.5 generation, third generation (3G), fourth generation (4G) and next generation radio access of cellular systems, Global System for Mobile Communications (GSM), General Packet Radio Service (GPRS), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), LTE, CDMA2000, WLAN, Wireless Router (WR) Mesh, etc.).
  • GSM Global System for Mobile Communications
  • GPRS General Packet Radio Service
  • EDGE Enhanced Data GSM Environment
  • WCDMA Wideband Code Division Multiple Access
  • LTE Long Term Evolution
  • CDMA2000 Code Division Multiple Access 2000
  • WLAN Wireless Router
  • the network interface 612 may also be configured to communicate with various other wired and/or wireless communication protocols (including TCP/IP, UDP, SIP, SMS, RTP, WAP, CDMA, TDMA, UMTS, UWB, WiFi, WiMax, In essence, network interface 612 may include or support virtually any wired and/or wireless communication and data processing mechanism by which information/data may be transmitted between computing device 600 and another computing or communication system via network 614.
  • wired and/or wireless communication protocols including TCP/IP, UDP, SIP, SMS, RTP, WAP, CDMA, TDMA, UMTS, UWB, WiFi, WiMax.
  • the memory 604 may represent a machine-readable medium (or computer-readable storage medium) on which one or more instruction sets, software, firmware, or other processing logic (e.g., logic 608) are stored that implement any one or more of the methods or functions described and/or required herein.
  • the logic 608 or a portion thereof may also reside completely or at least partially within the processor 602.
  • the memory 604 and the processor 602 may also constitute a machine-readable medium (or computer-readable storage medium).
  • the logic 608 or a portion thereof may also be configured as a processing logic or logic, at least a portion of which is partially implemented in hardware.
  • the logic 608 or a portion thereof may also be transmitted or received over a network 614 via a network interface 612.
  • machine-readable medium (or computer-readable storage medium) of an example embodiment may be a single medium
  • machine-readable medium should be understood to include a single non-transitory medium or multiple non-transitory media (e.g., a centralized or distributed database and/or associated cache and computing system) storing one or more instruction sets.
  • the term “machine-readable medium” (or computer-readable storage medium) may also be understood to include any non-transitory medium that can store, encode, or carry an instruction set for execution by a machine and cause the machine to perform any one or more of the methods of the various embodiments or that can store, encode, or carry a data structure utilized by or associated with such an instruction set.
  • the term “machine-readable medium” (or computer-readable storage medium) may therefore be understood to include, but is not limited to, solid-state memories, optical media, and magnetic media.
  • the disclosed and other embodiments, modules, and functional operations described in this document may be implemented in digital electronic circuit systems, or in computer software, firmware or hardware (including the structures disclosed in this document and their structural equivalents), or a combination of one or more of them.
  • the disclosed and other embodiments may be implemented as one or more computer program products, that is, one or more modules of computer program instructions encoded on a computer-readable medium to be executed by a data processing device or to control the operation of the data processing device.
  • the computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, a material composition that affects a machine-readable propagation signal, or a combination of one or more of them.
  • data processing device covers all devices, equipment, and machines for processing data, including, for example, a programmable processor, a computer or multiple processors or computers.
  • the device may also include code that creates an execution environment for the computer program in question, such as code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
  • a propagation signal is an artificially generated signal, such as an electrical signal, an optical signal, or an electromagnetic signal generated by a machine, which is generated to encode information for transmission to a suitable receiver device.
  • a computer program (also referred to as a program, software, software application, script, or code) can be written in any form of programming language (including compiled or interpreted languages), and the computer program can be deployed in any form, including being deployed as an independent program or as a module, component, subroutine, or another unit suitable for use in a computing environment.
  • a computer program does not necessarily correspond to a file in a file system.
  • a program may be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), or in a single file dedicated to the program under discussion, or in multiple collaborative files (e.g., files storing one or more modules, subroutines, or partial codes).
  • a computer program may be deployed to be executed on one computer or on multiple computers located at one site or distributed in multiple sites and interconnected by a communication network.
  • the processes and logic flows described in this document may be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output.
  • the processes and logic flows may also be performed by a dedicated logic circuit system (e.g., FPGA (field programmable gate array) or ASIC (application-specific integrated circuit)), and the apparatus may also be implemented as a dedicated logic circuit (e.g., FPGA (field programmable gate array) or ASIC (application-specific integrated circuit)).
  • processors suitable for executing computer programs include, for example, both general-purpose microprocessors and special-purpose microprocessors and any one or more processors of any kind of digital computer.
  • the processor will receive instructions and data from a read-only memory or a random access memory or both.
  • the necessary elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data.
  • a computer will also include one or more mass storage devices (such as magnetic disks, magneto-optical disks, or optical disks) for storing data, or the computer will also be operatively connected to receive data from the one or more mass storage devices or to transfer data to the one or more mass storage devices or to perform both.
  • the computer does not need to have such a device.
  • Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, including, for example, semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM disks and DVD-ROM disks.
  • semiconductor memory devices such as EPROM, EEPROM, and flash memory devices
  • magnetic disks such as internal hard disks or removable disks
  • magneto-optical disks and CD-ROM disks and DVD-ROM disks.
  • the processor and memory may be supplemented by a dedicated logic circuit system or may be incorporated into the dedicated logic circuit system.
  • Computer-readable media may include removable and non-removable storage devices, including but not limited to read-only memory (ROM), random access memory (RAM), compact disks (CD), digital versatile disks (DVD), etc. Therefore, computer-readable media may include non-transitory storage media.
  • program modules may include routines, programs, objects, components, data structures, etc., which perform specific tasks or implement specific abstract data types.
  • Computer or processor executable instructions, associated data structures, and program modules represent examples of program code for executing the steps of the methods disclosed herein.
  • a specific sequence of such executable instructions or associated data structures represents an example of a program code for implementing these methods.
  • a hardware circuit implementation may include discrete analog and/or digital components, which may be integrated as part of a printed circuit board, for example.
  • the disclosed components or modules may be implemented as application specific integrated circuits (ASICs) and/or field programmable gate arrays (FPGAs) devices.
  • ASICs application specific integrated circuits
  • FPGAs field programmable gate arrays
  • some implementations may include a digital signal processor (DSP), which is a dedicated microprocessor with an architecture optimized for the operational needs of digital signal processing associated with the disclosed functionality of the present application.
  • DSP digital signal processor
  • various components or subassemblies within each module may be implemented with software, hardware or firmware. Any connection method and medium known in the art may be used to provide a connection between a module and/or a component within a module, including but not limited to communication via the Internet, a wired network, or a wireless network using an appropriate protocol.

Landscapes

  • Engineering & Computer Science (AREA)
  • Traffic Control Systems (AREA)
  • Manufacturing & Machinery (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)

Abstract

本公开提供一种行车信息来源评估方法，包括取得来自待评估行车信息来源与多个其他行车信息来源的信息；基于上述信息，计算待评估行车信息来源关联于多个其他行车信息来源中的每一者之间的多个参数评分；基于多个参数评分，得到关联于待评估行车信息来源的多种状态的多个一致性信息；及基于多个一致性信息，决定待评估行车信息来源的当前状态。除此之外，本公开也提供了可执行上述方法的行车信息来源评估装置和计算机可读存储介质。

Description

用于评估行车信息来源的方法和装置
本申请要求在2022年12月13日提交中国专利局、申请号为202211603851.6、发明名称为“用于评估行车信息来源的方法和装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本公开涉及一种自动驾驶技术，特别涉及一种用于对行车信息来源进行评估的方法、装置和计算机可读存储介质。
背景技术
随着汽车科技的进步，于车辆上设置多个传感器相当常见，为了能够准确的同时使用来自不同传感器的信息，这些传感器的良好标定是极为重要的。传统上，通常通过静态标定或是利用自然场景下的语意要素来验证传感器的标定。
静态标定需要事先在室内搭建相关设备而使得成本高昂，并不适合大规模的部署。此外，由于路况、天气、障碍物等各种因素，事先完成的传感器标定可能会在车辆行驶期间失效，从而影响到信息的一致性而造成错误判断，甚至发生危险。
自然场景下的标定验证则需要借助高精地图的辅助，通过与高精地图中保持的要素验证几何一致性来判断标定是否失效。然而，高精地图同样的成本高昂，且当实际的环境发生变化但高精地图尚未更新时会导致验证结果失准。
除此之外，在使用传统方法通过两个传感器的外参来验证这两个传感器的标定时，只能判断出两个传感器中至少有一个标定失效，但无法推理出具体是哪个传感器的标定失效，例如，发生了非预期的位置和/或姿态变换。因此，对于当代设置有越来越多传感器的车辆来说，传感器的维护十分困难。
发明内容
有鉴于此，本公开提出一种用于对行车信息来源进行评估的方法、装置和计算机可读储存介质，可以在具有多个行车信息来源(例如，传感器)时，推估出个别行车信息来源的当前状态，提升行车信息来源的标定失效时的排解效率，且成本低廉。
本公开的第一个方面提出了一种行车信息来源评估方法，包括取得来自待评估行车信息来源与多个其他行车信息来源的信息；基于此信息，计算待评估行车信息来源关联于多个其他行车信息来源中的每一者之间的多个参数评分；基于多个参数评分，得到关联于待评估行车信息来源的多种状态的多个一致性信息；及基于多个一致性信息，决定待评估行车信息来源的当前状态。
本公开的第二个方面提出了一种行车信息来源评估装置，包括一个或多个处理器，以及存储程序的存储器。程序包括指令，此指令在由处理器执行时使行车信息来源评估装置执行上述的行车信息来源评估方法。
本公开的第三个方面提出了一种存储有程序的计算机可读存储介质，此程序包括指令，此指令在由计算装置的一个或者多个处理器执行时，致使计算装置执行上述的行车信息来源评估方法。
附图说明
附图示例性地示出了实施例并且构成说明书的一部分，与说明书的文字描述一起用于讲解实施例的示例性实施方式。显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。在所有附图中，相同的附图标记指代类似但不一定相同的要素。
图1是示出根据本公开一实施例的行车信息来源评估方法的示意图；
图2是示出根据本公开一实施例的行车信息来源评估方法的流程图；
图3是示出根据本公开一实施例的行车信息来源评估方法的示意图；
图4是示出根据本公开一实施例的后验概率计算方法的流程图；
图5是示出可以在其中实现本文公开的各种技术的车辆的示意图；
图6是示出根据本公开一实施例的计算设备的示意图。
具体实施方式
为了使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明作进一步地详细描述，显然，所描述的实施例仅仅是本发明的一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
在本公开中，术语“多个”是指两个或两个以上，除非另有说明。在本公开中，术语“和/或”，描述关联对象的关联关系，涵盖所列出的对象中的任何一个以及全部可能的组合方式。字符“/”一般表示前后关联对象是一种“或”的关系。
在本公开中，除非另有说明，否则使用术语“第一”、“第二”等是用于区别类似的对象，而不意图限定其位置关系、时序关系或重要性关系。应该理解这样使用的术语在适当情况下可以互换，以便这里描述的本发明的实施例能够以除了在这里图示或描述的那些以外的方式实施。
此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、系统、产品或 设备固有的其它步骤或单元。
图1是示出根据本公开一实施例的行车信息来源评估方法的示意图。
请参考图1，本公开实施例的多个行车信息来源S1～Si例如配置于车辆。这些行车信息来源S1～Si的作用在于，车辆上的计算装置能够接收来自这些行车信息来源S1～Si中至少一个的信息，加以分析并决定后续的自动驾驶行为，例如，根据前车距离与前车速度决定当前速度、或根据环境决定行驶的路径和轨迹等。
在一些实施例中，行车信息来源S1～Si包括多个传感器。
在一些实施例中，行车信息来源S1～Si包括高精地图。
请回到图1，在本公开实施例中，行车信息来源评估装置100可用于执行行车信息来源评估方法，以对行车信息来源S1～Si的当前状态C1～Ci进行评估，其中当前状态C1～Ci例如包括行车信息来源S1～Si中的每一者正常、失效、和/或与其健康度关联的评分等。
在一些实施例中，行车信息来源评估装置100例如可包括前端模块110与后端模块120。前端模块110取得来自行车信息来源S1～Si的信息后，可基于这些信息计算出两两行车信息来源之间的参数评分，并且将所计算出来的参数评分输出至后端模块120。后端模块120基于两两行车信息来源之间的参数评分，可評估出行车信息来源S1～Si中任一者的当前状态C1～Ci。必须说明的是，前端模块110与后端模块120可以例如是软件程序中的两个不同功能模块，也可以例如是两个不同硬件模块。
在一些实施例中，任意两个行车信息来源之间的参数评分与这两个行车信息来源之间的误差(例如，标定误差)成负相关。换句话说，两个行车信息来源之间的标定误差愈大，表示这两个行车信息来源之间的标定愈不一致或愈不和谐，则参数评分愈低；两个行车信息来源之间的标定误差愈小，表示这两个行车信息来源之间的标定愈一致或愈和谐，则参数评分愈高。
在一些实施例中，当对应的行车信息来源都是传感器时，参数评分也可以称为传感器之间的外参评分。
在一些实施例中，前端模块110经设计以将各种类型的行车信息来源之间的参数评分都定义于相同的预设范围内，例如但不限于，0到1。有利地，这样的设计使得后端模块120可以无需考量行车信息来源的类型或规格，便于维护。更详细来说，由于不同的行车信息来源的资料格式和/或单位可能不同，在新增和/或更换车辆上的行车信息来源时，只要考量前端模块110所计算的参数评分都能被定义于相同的预设范围内，后端模块120可以不必改动而照常运行。
在一些实施例中，行车信息来源评估装置100例如是设置于车辆上的计算系统/装置/设备。
在一些实施例中，行车信息来源评估装置100例如是可与车辆上的通信系统互相通信的远程服务系统。
在一些实施例中，行车信息来源评估装置100例如是车辆上的计算系统/装置/设备与远程服务系统的组合。举例来说，车辆上的计算系统/装置/设备可用于实现前端模块110的功能，而远程服务系统可用于实现后端模块120的功能。
以下通过实施例，针对本公开所提出的行车信息来源评估方法作进一步的细节说明。
图2是示出根据本公开一实施例的行车信息来源评估方法的流程图；图3是示出根据本公开一实施例的行车信息来源评估方法的示意图。
必须说明的是，在与图2、图3实施例关联的描述中将会搭配图1的行车信息来源评估装置100进行说明，但本公开并不限于这些描述。例如，为了方便说明，在图3实施例中以三个行车信息来源为例子，但本领域普通技术人员应理解，本公开所提出的行车信息来源评估方法也可套用于三个以上的行车信息来源的情况。
请参考图2，在步骤S210中，行车信息来源评估装置100会取得来自多个行车信息来源的信息。具体来说，在车辆行驶过程中，行车信息来源评估装置100(例如，前端模块110)会持续从多个行车信息来源取得信息。
请参考图3，多个行车信息来源包括行车信息来源A、行车信息来源B以及行车信息来源C。在车辆行驶过程中，行车信息来源评估装置100会持续从行车信息来源A取得信息Ia、从行车信息来源B取得信息Ib以及从行车信息来源C取得信息Ic。
在以下的说明当中，将选定行车信息来源A作为待评估行车信息来源。然而，相同的方法也可以类推至其他行车信息来源B、C。
在一些实施例中，多个行车信息来源包括车辆上的多个传感器。
多个传感器例如可以包括惯性测量单元(IMU)、全球导航卫星系统(GNSS)收发器(例如全球定位系统(GPS)收发器)、无线电探测和测距装置(RADAR，简称为雷达)、激光探测及测距系统(LiDAR，简称为激光雷达)、声学传感器、超声波传感器以及图像捕捉装置(例如相机)的至少其中之一。
多个传感器例如可以包括用于监控车辆的多个感应器(例如，氧气(O2)监控器、油量表传感器、发动机油压传感器，以及温度、湿度、压力传感器等等)。
在一些实施例中，多个行车信息来源包括惯性测量单元、激光雷达和相机中的至少其中两种。
在一些实施例中，多个行车信息来源仅包括车辆上的多个传感器。更明确的说，本公开实施例所介绍的行车信息来源评估方法在没有标准化的静态标定设备或高精地图时，也可以对每一个行车信息来源(传感器)的当前状态作出评估。
在一些实施例中，多个行车信息来源包括高精地图。作为行车信息来源的高精地图可以例如是存储在车辆上的存储器中，也可以例如是存储在网路的至少一个节点。有利地，高精地图也可以作为本公开所提出的行车信息来源评估方法的评估对象，以在高精地图出现误差(例如，与实际场景不一致)时将其检测出来，且不影响到其他行车信息来源的评估。
请回到图2，在步骤S220中，行车信息来源评估装置100会基于步骤S210中所接收到的信息，计算待评估行车信息来源关联于多个其他行车信息来源中的每一者之间的多个参数评分。具体来说，在车辆行驶过程中，行车信息来源评估装置100(例如，前端模块110)会持续计算待评估行车信息来源关联于多个其他行车信息来源中的每一者之间的多个参数评分。
请参考图3，行车信息来源评估装置100会计算待评估行车信息来源A关联于行车信息来源B之间的参数评分X，以及待评估行车信息来源A关联于行车信息来源C之间的参数评分Z。
在一些实施例中，行车信息来源评估装置100例如会计算相机与惯性测量单元之间的参数评分。
在一些实施例中，行车信息来源评估装置100会从相机取得两张帧图像，例如一张当前帧图像和一张先前帧图像，而这两张帧图像中的场景会存在位移。倘若相机与惯性测量单元之间的标定一致，则利用惯性测量单元的信息应当可以准确预测先前帧图像中的特定物在当前帧图像中的位置。因此，这个预测的准确度可作为参数评分的基础。
首先，行车信息来源评估装置100例如会选取先前帧图像中的特定指标。
例如，行车信息来源评估装置100会对先前帧图像中地平面上的特定指标(例如，道路标线)进行采样(例如，降采样)，以得到先前帧图像中的多个采样点Pl。
接着，行车信息来源评估装置100例如会基于惯性测量单元的信息，预测先前帧图像中的特定指标在当前帧图像中的位置。
例如，行车信息来源评估装置100会基于相机内参K、相机与惯性测量单元之间的外参Ti c、惯性测量单元的离地高度等信息，计算相机的光心到多个采样点Pl的射线在惯性测量单元的座标系下与地平面的多个交点Pik-1。基于两张帧图像的时间差，行车信息来源评估装置100可以积分惯性测量单元的信息，以得到两张帧图像的拍摄时间之间的姿态变化Td。因此，基于姿态变化Td，行车信息来源评估装置100可以预测多个交点Pik-1在当前帧图像的拍摄时间时，在惯性测量单元的座标系下的多个第一预测点Pik＝Td*Pik-1。接着，行车信息来源评估装置100例如会基于相机内参K、相机与惯性测量单元之间的外参Ti c，将这些预测点投影到当前帧图像的图像平面上，以得到多个第二预测点I＝K*(Ti c)-1*Pik。
最后，行车信息来源评估装置100例如会将预测准确度作为参数评分的基础，来计算参数评分。预测准确度例如可以通过比较特定指标在当前帧图像中的预测位置与实际位置来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以计算多个第二预测点I实际落入当前帧影像中的特定指标内的比例(例如，实际落入当前帧影像中的特定指标内的第二预测点I的数量/第二预测点I的总数量)，作为参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算两个相机之间的参数评分。
在一些实施例中，行车信息来源评估装置100会分别从两个相机取得一帧图像，而这两张帧图像中的场景会存在位移。倘若两个相机之间的标定一致，则利用惯性测量单元分别与两个相机之间的外参，应当可以准确预测其中一帧图像中的特定物在另一帧图像中的位置。因此，这个预测的准确度可作为参数评分的基础。
首先，行车信息来源评估装置100例如会选取来自第一相机的第一帧图像中的特定指标。
例如，行车信息来源评估装置100会对第一帧图像中地平面上的特定指标(例如，道路标线)进行采样(例如，降采样)，以得到第一帧图像中的多个采样点Pl1。
接着，行车信息来源评估装置100例如会基于惯性测量单元分别与两个相机之间的外参，预测第一帧图像中的特定指标在第二帧图像中的位置。
例如，行车信息来源评估装置100会基于第一相机内参K1、第一相机与惯性测量单元之间的外参Ti c1、惯性测量单元的离地高度等信息，计算第一相机的光心到多个采样点Pl1的射线在惯性测量单元的座标系下与地平面的多个交点Pic1。接著，行车信息来源评估装置100会基于第二相机内参K2以及第二相机与惯性测量单元之间的外参Ti c2，将这些交点Pic1投影到第二帧图像的图像平面上，以得到多个预测点I＝K2*(Ti c2)-1*Pic1。
最后，行车信息来源评估装置100例如会将预测准确度作为参数评分的基础，来计算参数评分。预测准确度例如可以通过比较特定指标在第二帧图像中的预测位置与实际位置来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以计算多个预测点I实际落入第二帧影像中的特定指标内的比例(例如，实际落入第二帧影像中的特定指标内的预测点I的数量/预测点I的总数量)，作为参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算相机与高精地图之间的参数评分。
在一些实施例中，行车信息来源评估装置100会从相机取得当前帧图像。倘若相机与高精地图之间无误差或标定一致，则利用惯性测量单元与相机之间的外参以及惯性测量单元在高精地图座标系下的姿态，应当可以准确预测高精地图中的特定物在当前帧图像中的位置。因此，这个预测的准确度可作为参数评分的基础。
首先，行车信息来源评估装置100例如会选取高精地图中的特定指标。
例如，行车信息来源评估装置100会对高精地图中上的特定指标(例如，道路标线)进行采样(例如，升采样)，以得到高精地图的多个采样点Pm。
接着，行车信息来源评估装置100例如会基于惯性测量单元与相机之间的外参以及惯性测量单元在高精地图座标系下的姿态，预测高精地图中的特定指标在当前帧图像中的位置。
例如，行车信息来源评估装置100会基于相机内参K、相机与惯性测量单元之间的 外参Ti c、以及惯性测量单元在高精地图座标系下的姿态Ti等信息，将多个采样点Pm投影到当前帧图像的图像平面上，以得到多个预测点I＝K*(Ti c)-1*(Ti)-1*Pm。
最后，行车信息来源评估装置100例如会将预测准确度作为参数评分的基础，来计算参数评分。预测准确度例如可以通过比较特定指标在当前帧图像中的预测位置与实际位置来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以计算多个预测点I实际落入当前帧影像中的特定指标内的比例(例如，实际落入当前帧影像中的特定指标内的预测点I的数量/预测点I的总数量)，作为参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算相机与高精地图之间的另一种参数评分。
在一些实施例中，行车信息来源评估装置100会从相机取得当前帧图像。倘若相机与高精地图之间无误差或标定一致，则利用相机与惯性测量单元之间的外参以及惯性测量单元在高精地图座标系下的姿态，应当可以准确预测当前帧图像中的特定物在高精地图中的位置。因此，这个预测的准确度可作为参数评分的基础。
首先，行车信息来源评估装置100例如会选取当前帧图像中的特定指标，以及高精地图中相同的特定指标。
例如，行车信息来源评估装置100会对当前帧图像中地平面上的特定指标(例如，道路标线)进行采样(例如，降采样)，以得到当前帧图像中的多个图像采样点Pl。此外，行车信息来源评估装置100会对高精地图中上的相同特定指标(例如，道路标线)进行采样(例如，升采样)，以得到高精地图的多个地图采样点Pm。
接着，行车信息来源评估装置100例如会基于相机与惯性测量单元之间的外参以及惯性测量单元在高精地图座标系下的姿态，预测当前帧图像中的特定指标在高精地图中的位置。
例如，行车信息来源评估装置100会基于相机内参K、相机与惯性测量单元之间的外参Ti c、惯性测量单元的离地高度等信息，计算相机的光心到多个图像采样点Pl的射线在惯性测量单元的座标系下与地平面的多个交点Pi。接著，行车信息来源评估装置100会基于惯性测量单元在高精地图座标系下的姿态Ti，将这些交点Pi转换到高精地图的座标系上，以得到多个预测点Pim＝Ti*Pi。
最后，行车信息来源评估装置100例如会将预测召回率作为参数评分的基础，来计算参数评分。预测召回率例如可以通过比较特定指标在高精地图中的预测位置与实际位置来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以遍历多个预测点Pim，以特定半径(例如，0.5米)来搜索地图采样点Pm，并且计算搜索成功的比例(例如，在特定半径内成功搜索到地图采样点Pm的预测点Pim数量/预测点Pim的总数量)，作为参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算激光雷达与惯性测量单元之间的参数评分。
在一些实施例中，行车信息来源评估装置100会利用激光雷达里程计(LiDAR Odometry)来计算激光雷达在惯性测量单元的座标系下的线速度。倘若激光雷达与惯性测量单元之间的标定一致，则计算出来的线速度应当与惯性测量单元自身观测到的线速度相同。因此，这两个线速度的一致性可作为参数评分的基础。
首先，行车信息来源评估装置100例如会利用激光雷达里程计来计算两个时间点间的姿态变化，并且基于激光雷达与惯性测量单元之间的外参将这个姿态变化转换到惯性测量单元的座标系。
例如，行车信息来源评估装置100会利用激光雷达里程计来计算两个时间点间的姿态变化Td，然后基于激光雷达与惯性测量单元之间的外参Ti l将这个姿态变化Td转换为惯性测量单元的座标系下的姿态变化Tdi＝Ti l*Td*(Ti l)-1。
接着，行车信息来源评估装置100例如会计算激光雷达在惯性测量单元的座标系下的线速度。
例如，行车信息来源评估装置100会基于激光雷达在惯性测量单元的座标系下的姿态变化Tdi以及两个时间点的时间差td，计算激光雷达在惯性测量单元的座标系下的线速度Vli＝Tdi/td。
最后，行车信息来源评估装置100例如会将计算出来的线速度以及惯性测量单元自身观测到的线速度之间的一致性作为参数评分的基础，来计算参数评分。这个一致性例如可以通过比较两个线速度来得到。
例如，惯性测量单元自身观测到的线速度为Vi。倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以利用值域落在0到1之间，且值与两个线速度的差值成负相关的函数(例如，exp(-1.0*normalize(Vi-Vli))，其中(Vi-Vli)例如是三维的速度矢量，而normalize(Vi-Vli)为对此速度矢量取二范数)來表示参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算两个激光雷达之间的参数评分。
在一些实施例中，行车信息来源评估装置100会利用激光雷达里程计来分别计算两个激光雷达在惯性测量单元的座标系下的两个线速度。倘若两个激光雷达之间的标定一致，则计算出来的两个线速度应当相同。因此，这两个线速度的一致性可作为参数评分的基础。
首先，行车信息来源评估装置100例如会利用第一激光雷达里程计来计算两个时间点间的第一姿态变化，并且基于第一激光雷达与惯性测量单元之间的外参将这个第一姿态变化转换到惯性测量单元的座标系；以及利用第二激光雷达里程计来计算两个时间点间的第二姿态变化，并且基于第二激光雷达与惯性测量单元之间的外参将这个第二姿态变化转换到惯性测量单元的座标系。
例如，行车信息来源评估装置100会利用第一激光雷达里程计来计算两个时间点间的第一姿态变化Td1，然后基于第一激光雷达与惯性测量单元之间的外参Ti l1将这个第一姿态变化Td1转换为惯性测量单元的座标系下的第一姿态变化Td1i＝Ti l1*Td1*(Ti l1)-1；以及利用第二激光雷达里程计来计算两个时间点间的第二姿态变化Td2，然后基于第二激光雷达与惯性测量单元之间的外参Ti l2将这个第二姿态变化Td2转换为惯性测量单元的座标系下的第二姿态变化Td2i＝Ti l2*Td2*(Ti l2)-1。
接着，行车信息来源评估装置100例如会计算第一激光雷达在惯性测量单元的座标系下的第一线速度；以及计算第二激光雷达在惯性测量单元的座标系下的第二线速度。
例如，行车信息来源评估装置100会基于第一激光雷达在惯性测量单元的座标系下的第一姿态变化Td1i以及两个时间点的时间差td，计算第一激光雷达在惯性测量单元的座标系下的第一线速度Vl1i＝Td1i/td；以及基于第二激光雷达在惯性测量单元的座标系下的第二姿态变化Td2i以及两个时间点的时间差td，计算第二激光雷达在惯性测量单元的座标系下的第二线速度Vl2i＝Td2i/td。
最后，行车信息来源评估装置100例如会将计算出来的第一线速度以及第二线速度之间的一致性作为参数评分的基础，来计算参数评分。这个一致性例如可以通过比较两个线速度来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以利用值域落在0到1之間，且值与第一线速度Vl1i和第二线速度Vl2i的差值成负相关的函数(例如，exp(-1.0*normalize(Vl1i-Vl2i))，其中(Vl1i-Vl2i)例如是三维的速度矢量，而normalize(Vl1i-Vl2i)为对此速度矢量取二范数)來表示参数评分。
在一些实施例中，行车信息来源评估装置100例如会计算激光雷达与相机之间的参数评分。
在一些实施例中，行车信息来源评估装置100会从相机取得当前帧图像，并且利用激光雷达检测特定物的位置。倘若激光雷达与相机与之间的标定一致，则利用激光雷达与相机之间的外参，应当可以准确预测特定物在当前帧图像中的位置。因此，这个预测的准确度可作为参数评分的基础。。因此，这个预测的准确度可作为参数评分的基础。
首先，行车信息来源评估装置100例如会利用激光雷达检测特定指标。
例如，行车信息来源评估装置100会利用激光雷达感测器提取属于特定指标(例如，车辆)的点，以得到多个采样点Pl。
接着，行车信息来源评估装置100例如会基于雷達与相机之间的外参，预测激光雷达检测到的特定指标在当前帧图像中的位置。
例如，行车信息来源评估装置100会基于相机内参K、以及激光雷达与相机之间的外参Tc l，将多个采样点Pl投影到当前帧图像的图像平面上，以得到多个预测点I＝K*Tc l*Pl。
最后，行车信息来源评估装置100例如会将预测准确度作为参数评分的基础，来计 算参数评分。预测准确度例如可以通过比较特定指标在当前帧图像中的预测位置与实际位置来得到。
例如，倘若参数评分是定义在预设范围0到1之间，则行车信息来源评估装置100可以计算多个预测点I实际落入当前帧影像中的特定指标内的比例(例如，实际落入当前帧影像中的特定指标内的预测点I的数量/预测点I的总数量)，作为参数评分。
请回到图2，在步骤S230中，行车信息来源评估装置100(例如，后端模块120)会基于步骤S220中所计算出来的多个参数评分，得到关联于待评估行车信息来源的多种状态的多个一致性信息。
具体来说，多个参数评分代表的是待评估行车信息来源与其他行车信息来源中的每一者之间是否标定一致或是否和谐。也就是说，参数评分是两个行车信息来源所共同产生的结果。基于待评估行车信息来源与其他行车信息来源中的每一者之间是否标定一致或是否和谐，能够推估出待评估行车信息来源的标定处于何种当前状态。
进一步来说，一致性信息用以指示待评估行车信息来源的当前状态与每一种可能状态之间的一致性程度，若一致性信息显示的当前状态与特定状态之间一致性程度愈高，表示待评估行车信息来源的当前状态愈接近于这个特定状态。
在一些实施例中，由于参数评分定义于预设范围(例如，0到1)内，因此可以用于决定待评估行车信息来源与其他行车信息来源中的每一者在不同状态组合下，参数评分的条件概率函数(具体的决定方式容后详述)。如此一来，待评估行车信息来源在每一种状态下的一致性信息则可以通过后验概率来表示。
在一些实施例中，待评估行车信息来源的多种状态例如包括2种(例如，正常与失效)。然而，本公开并不限制待评估行车信息来源可能的状态数量。
请参考图3，在本公开实施例中，将行车信息来源A的正常状态记为a，失效状态记为a’；行车信息来源B的正常状态记为b，失效状态记为b’；行车信息来源C的正常状态记为c，失效状态记为c’。
基此，行车信息来源A为正常状态的先验概率记为P(a)，行车信息来源A为失效状态的先验概率记为P(a’),行车信息来源B正常的先验概率记为P(b),行车信息来源B失效的先验概率记为P(b’),行车信息来源C正常的先验概率记为P(c),行车信息来源C失效的先验概率记为P(c’)。
因此，关联于待评估行车信息来源A的多种状态组合包括(a,b)、(a’,b)、(a,b’)、(a’,b’)、(a,c)、(a’,c)、(a,c’)、(a’,c’)。在这些状态组合下，多个参数评分(例如，X；Y；Z)的条件概率函数包括P(X|a,b)、P(X|a’,b)、P(X|a,b’)、P(X|a’,b’)、P(Z|a,c)、P(Z|a’,c)、P(Z|a,c’)、P(Z|a’,c’)。
基于上述描述以及后验概率的数学概念，在当前检测出多个参数评分(例如，X＝x；Y＝y；Z＝z)的基础上，可以计算待评估行车信息来源A在正常状态的后验概率为：
P(a|x,y,z)＝kP(a)*[P(b)*P(x|a,b)+P(b’)*P(x|a,b’)+P(c)*P(z|a,c)+P(c’)*P(z|a,c’)]；
以及待评估行车信息来源A在失效状态的后验概率为：
P(a’|x,y,z)＝kP(a’)*[P(b)*P(x|a’,b)+P(b’)*P(x|a’,b’)+P(c)*P(z|a’,c)+P(c’)*P(z|a’,c’)]，
其中k为归一化常数。
如果待评估行车信息来源A在正常状态的后验概率P(a|x,y,z)较高，表示待评估行车信息来源A的当前状态较接近于正常状态；如果待评估行车信息来源A在失效状态的后验概率P(a’|x,y,z)较高，表示待评估行车信息来源A的当前状态较接近于失效状态。
图4是示出根据本公开一实施例的后验概率计算方法的流程图。
在一些实施例中，步骤S230例如是基于步骤S220中所计算出来的多个参数评分，得到关联于待评估行车信息来源的多种状态的多个后验概率。在这个情况下，如图4所示，步骤S230又包括步骤S231与步骤S232。
请参考图4，在步骤S231中，行车信息来源评估装置100会决定关联于待评估行车信息来源的每一个参数评分在多种状态组合的基础上的多个条件概率函数。
请参考图3，关联于待评估行车信息来源A的参数评分包括参数评分X与参数评分Z。参数评分X在多种状态组合的基础上的多个条件概率函数包括P(X|a,b)、P(X|a’,b)、P(X|a,b’)、P(X|a’,b’)；参数评分Z在多种状态组合的基础上的多个条件概率函数包括P(Z|a,c)、P(Z|a’,c)、P(Z|a,c’)、P(Z|a’,c’)。
请回到图4，步骤S231例如包括步骤S2311与步骤S2312。
在步骤S2311中，行车信息来源评估装置100会基于多个参数评分中对应于第一参数评分的多个第一评分值，决定符合多种状态组合的其中之一的预期结果的第一分布函数。接着，在步骤S2312中，行车信息来源评估装置100基于第一分布函数，决定对应于第一参数评分的多个条件概率函数中的多个第一条件概率函数。
由于参数评分是关联于两个行车信息来源之间的标定一致性或和谐性，因此在参数评分被定义于预设范围内时，每一种状态组合都可以存在预期结果。换句话说，参数评分在每一种状态组合的条件概率函数可以使用某一种特定的分布函数来描述。
举例来说，当状态组合对应的所有的行车信息来源都是正常状态时，可以预期参数评分的条件概率分布函数会是一种严格递增函数，且参数评分有高机率是接近预设范围的上限；当存在至少一个行车信息来源都是失效状态时，可以预期参数评分的条件概率分布函数会是一种严格递减函数，且参数评分有高机率是接近预设范围的下限。值得一提的是，这个条件概率分布函数的定义域被定义于与参数评分相同的预设范围内。
请参考图3，第一参数评分例如是参数评分X。符合状态组合(a,b)的预期结果的第一分布函数例如可以是截断正态分布的累计分布函数(Cumulative Distribution Function，CDF)：
其中：


Z＝Φ(β)-Φ(α)，
其中，μ和σ例如分别是第一参数评分X的在一段预设时间内(例如，500帧)的多个第一评分值xi的均值和方差。
必须说明的是，只要能够符合预期结果，第一分布函数也可以选用其他类型的函数，本公开并不在此限制。
接着，对应第一参数评分X的多个第一条件概率函数P(X|a,b)、P(X|a’,b)、P(X|a,b’)、P(X|a’,b’)可以用第一分布函数来描述，例如决定為：
P(X＝x|a,b)＝F(x；μ,σ,m,n)；
P(X＝x|a’,b)＝1-F(x；μ,σ,m,n)；
P(X＝x|a,b’)＝1-F(x；μ,σ,m,n)；
P(X＝x|a’,b’)＝1-F(x；μ,σ,m,n)，
以使这些第一条件概率函数都符合对应的预期结果。
必须说明的是，只要能够符合预期结果，本公开并不限制第一条件概率函数的具体决定方式。
在一些例子中，第一条件概率函数P(X|a,b)也可例如是决定为用第一分布函数来描述的[F(x+0.5σ；μ,σ,m,n)-F(x-2σ；μ,σ,m,n)]。
此外，第二参数评分例如是参数评分Z。通过类似的方式，对应第二参数评分Z的多个第二条件概率函数P(Z|a,c)、P(Z|a’,c)、P(Z|a,c’)、P(Z|a’,c’)也能够被决定出来，故具体的过程不在此重复。
请回到图4，决定了多个条件概率函数后，在步骤S232中，行车信息来源评估装置100会基于多个条件概率函数及多个参数评分中的多个当前参数评分，计算多个后验概率。
从上述分析可知，为了计算多个后验概率，除了将当前参数评分的评分值代入多个条件概率函数之外，还需要每一个行车信息来源在每一个状态的先验概率。
在一些实施例中，待评估行车信息来源在每一个状态的先验概率默认为相等。举例来说，倘若待评估行车信息来源共有三种可能的状态，则每一个状态的先验概率默认为1/3；倘若待评估行车信息来源共有两种可能的状态，则每一个状态的先验概率默认为1/2。
在一些实施例中，每一个行车信息来源在相同状态的先验概率默认为相等。举例来说，倘若每一个行车信息来源都有正常状态与失效状态两种可能的状态，则每一个行车信息来源在正常状态的先验概率都相等(例如，皆为p)，并且每一个行车信息来源在失效状态的先验概率也都相等(例如，皆为1-p)。
在一些实施例中，每一个行车信息来源在每一个状态的先验概率可以通过历史资料、行车信息来源的出厂数据或其他信息来进行预设。
请参考图3，待评估行车信息来源A在正常状态的后验概率为：
P(a|x,y,z)＝kP(a)*[P(b)*P(x|a,b)+P(b’)*P(x|a,b’)+P(c)*P(z|a,c)+P(c’)*P(z|a,c’)]；
以及待评估行车信息来源A在失效状态的后验概率为：
P(a’|x,y,z)＝kP(a’)*[P(b)*P(x|a’,b)+P(b’)*P(x|a’,b’)+P(c)*P(z|a’,c)+P(c’)*P(z|a’,c’)]。
默认每一个行车信息来源A、B、C在正常状态与失效状态的先验概率皆相等(例如，皆为0.5)，则基于默认的多个行车信息来源的多个先验概率，再将多个当前参数评分的评分值x、y、z代入步骤S231所决定的多个条件概率函数中，就可以得到待评估行车信息来源A在正常状态与失效状态下的每一个后验概率。
在一些实施例中，由于多个后验概率在后续仅用作大小的比较，并无需在意实际数值。因此，行车信息来源评估装置100并不计算归一化常数k(例如，令k为1)。
请回到图2，在步骤S240中，行车信息来源评估装置100会基于步骤S230中得到的多个一致性信息，决定待评估行车信息来源的当前状态。
如前所述，一致性信息用以指示待评估行车信息来源的当前状态与每一种状态之间的一致性程度，当前状态与特定状态之间一致性程度愈高，表示待评估行车信息来源的当前状态愈接近于这个特定状态。因此，行车信息来源评估装置100(例如，后端模块120)例如可基于一致性信息，决定一致性程度最高的特定状态为当前状态。
在一些实施例中，一致性信息例如是后验概率，而行车信息来源评估装置100例如会决定当前状态为步骤S232中所计算出来的多个后验概率中的最大者所关联的状态。
请参考图3，行车信息来源评估装置100例如在步骤S232中计算出待评估行车信息来源A在正常状态的后验概率以及待评估行车信息来源A在失效状态的后验概率。倘若待评估行车信息来源A在正常状态的后验概率大于待评估行车信息来源A在失效状态的后验概率，则正常状态会被决定为待评估行车信息来源A的当前状态；倘若待评估行车信息来源A在正常状态的后验概率小于待评估行车信息来源A在失效状态的后验概率，则失效状态会被决定为待评估行车信息来源A的当前状态。
利用本公开实施例所介绍的方法，每一个行车信息来源的当前状态都可以被决定出来。基于此，关于行车信息来源的故障/标定失效/过期等的后续排解也更加便利。
在一些实施例中，行车信息来源评估装置100例如会将待评估行车信息来源的评估分数(例如，健康度分数)进行显示，让操作员/技术员能够即时掌握待评估行车信息来源的状况。以图3为例，待评估行车信息来源A的评估分数可以是其在特定状态(例如，正常状态)下的后验概率。
在一些实施例中，为了让所显示的评估分数具有意义，行车信息来源评估装置100(例如，后端模块120)例如会归一化待评估行车信息来源的在多个状态下的多个后验概率，以得到多个归一化后验概率，然后再输出其中一个特定状态所对应的归一化后验概 率，以作为待评估行车信息来源的评估分数。以图3为例，行车信息来源评估装置100例如不令k为1，而是会计算出归一化常数k，以计算出多个归一化后验概率，随后再输出待评估行车信息来源A所对应的归一化后验概率作为其评估分数。
在一些实施例中，重复上面介绍的方法，行车信息来源评估装置100可以显示出每一个行车信息来源的评估分数，让操作员/技术员能够同时掌握所有行车信息来源的状况。
在一些实施例中，行车信息来源评估装置100例如会根据每一个行车信息来源的当前状态或评估分数来发出警示。举例来说，行车信息来源评估装置100可以判断每一个行车信息来源的当前状态是否为故障/标定失效/过期，或评估分数是否低于预设阀值，并且针对当前状态为故障/标定失效/过期或评估分数低于预设阀值的行车信息来源发出警示。举例来说，警示可以是通过不同显示颜色、警示音等方式来发出。
在一些实施例中，行车信息来源评估装置100例如会根据待评估行车信息来源的当前状态或评估分数，决定是否要在车辆的行进过程中执行降低风险策略。举例来说，行车信息来源评估装置100可以在特定的待评估行车信息来源故障/标定失效/过期，或评估分数低于预设阀值时，决定在车辆的行进过程中执行降低风险策略。
在一些实施例中，行车信息来源评估装置100例如会根据多个行车信息来源的多个当前状态或评估分数，决定是否要在车辆的行进过程中执行降低风险策略。举例来说，行车信息来源评估装置100可以在多个行车信息来源的多个当前状态或评估分数满足预设条件(例如，超过预设数量的行车信息来源都故障/标定失效/过期，或评估分数加总低于预设阀值)时，决定在车辆的行进过程中执行降低风险策略。
在一些实施例中，降低风险策略例如包括降级自动驾驶等级或执行最小风险策略(Minimal Risk Maneuver，MRM)等。
在一些实施例中，当多个行车信息来源包括车辆上的多个传感器时，降低风险策略例如包括关闭多个传感器的其中之一。举例来说，当作为待评估行车信息来源的传感器故障/标定失效/过期或评估分数低于预设阀值时，行车信息来源评估装置100可以关闭这个传感器。举例而言，降低风险策略可以是仅关闭当前状态为故障/标定失效/过期的行车信息来源(例如是评估分数低于预设阀值的行车信息来源)，同时保持车辆的行进以及其它行车信息来源的运行。
在一些实施例中，当多个行车信息来源包括车辆上的多个传感器时，行车信息来源评估装置100例如会根据待评估行车信息来源的当前状态，校正多个传感器的多个外参。举例来说，当作为待评估行车信息来源的传感器标定失效或评估分数低于预设阀值时，行车信息来源评估装置100可以校正其与其他传感器之间的外参。
在一些实施例中，行车信息来源评估装置100例如可以进行在线的重新标定，以校正传感器与传感器之间的外参。举例来说，行车信息来源评估装置100可以自动迭代调整标定失效的传感器的标定，以使标定失效的传感器恢复到正常状态，或优化所有行车信息来源的评估分数加总(例如，达到极大值)。在一些实施例中，也可以回厂对传感 器进行重新标定，以校正传感器与传感器之间的外参。
图5是示出可以在其中实现本文公开的各种技术的车辆的示意图。
请参考图5，车辆500可以是轿车、卡车、摩托车、公共汽车、船只、飞机、直升机、割草机、挖土机、摩托雪橇、航空器、旅游休闲车、游乐园车辆、农场装置、建筑装置、有轨电车、高尔夫车、火车、无轨电车，或其它车辆。车辆500可以完全地或部分地以自动驾驶模式进行运行。车辆500在自动驾驶模式下可以控制其自身，例如车辆500可以确定车辆的当前状态以及车辆所处环境的当前状态，确定在该环境中的至少一个其它车辆的预测行为，确定该至少一个其它车辆执行所预测行为的可能性所对应信任等级，并且基于所确定的信息来控制车辆500自身。在处于自动驾驶模式时，车辆500可以在无人交互的情况下运行。
车辆500可以包括各种车辆系统，例如驱动系统542、传感器系统544、控制系统546、用户接口系统548、计算系统550以及通信系统552。车辆500可以包括更多或更少的系统，每个系统可以包括多个单元。进一步地，车辆500的每个系统和单元之间可以是互联的。例如，计算系统550能够与驱动系统542、传感器系统544、控制系统546、用户接口系统548和通信系统552中的一个或多个进行数据通信。从而，车辆500的一个或多个所描述的功能可以被划分为附加的功能性部件或者实体部件，或者结合为数量更少的功能性部件或者实体部件。在更进一步的例子中，附加的功能性部件或者实体部件可以增加到如图5所示的示例中。
驱动系统542可以包括为车辆500提供动能的多个可操作部件(或单元)。在一些实施例中，驱动系统542可以包括发动机或电动机、车轮、变速器、电子系统、以及动力(或动力源)。发动机或者电动机可以是如下装置的任意组合：内燃机、电机、蒸汽机、燃料电池发动机、丙烷发动机、或者其它形式的发动机或电动机。在一些实施例中，发动机可以将一种动力源转换为机械能。在一些实施例中，驱动系统542可以包括多种发动机或电动机。例如，油电混合车辆可以包括汽油发动机和电动机，也可以包括其它的情况。
车辆500的车轮可以是标准车轮。车辆500的车轮可以是多种形式的车轮，包括独轮、双轮、三轮、或者四轮形式，例如轿车或卡车上的四轮。其它数量的车轮也是可以的，例如六轮或者更多的车轮。车辆500的一个或多个车轮可被操作为与其他车轮的旋转方向不同。车轮可以是至少一个与变速器固定连接的车轮。车轮可以包括金属与橡胶的结合，或者是其他物质的结合。变速器可以包括可操作来将发动机的机械动力传送到车轮的单元。出于这个目的，变速器可以包括齿轮箱、离合器、差动齿轮和传动轴。变速器也可以包括其它单元。传动轴可以包括与车轮相匹配的一个或多个轮轴。电子系统可以包括用于传送或控制车辆500的电子信号的单元。这些电子信号可用于启动车辆500中的多个灯、多个伺服机构、多个电动机，以及其它电子驱动或者控制装置。动力源可以是全部或部分地为发动机或电动机提供动力的能源。也即，发动机或电动机能够 将动力源转换为机械能。示例性地，动力源可以包括汽油、石油、石油类燃料、丙烷、其它压缩气体燃料、乙醇、燃料电池、太阳能板、电池以及其它电能源。动力源可以附加的或者可选地包括燃料箱、电池、电容、或者飞轮的任意组合。动力源也可以为车辆500的其它系统提供能量。
传感器系统544可以包括多个传感器，这些传感器用于感测车辆500的环境和条件的信息。例如，传感器系统544可以包括惯性测量单元(IMU)、全球导航卫星系统(GNSS)收发器(例如全球定位系统(GPS)收发器)、无线电探测和测距装置(RADAR，简称为雷达)、激光探测及测距系统(LiDAR，简称为激光雷达)、声学传感器、超声波传感器以及图像捕捉装置(例如相机)。传感器系统544可以包括用于监控车辆500的多个感应器(例如，氧气(O2)监控器、油量表传感器、发动机油压传感器，以及温度、湿度、压力传感器等等)。还可以配置其它传感器。包括在传感器系统544中的一个或多个传感器可以被单独驱动或者被集体驱动，以更新一个或多个传感器的位置、方向，或者这二者。
IMU可以包括传感器的结合(例如加速器和陀螺仪)，用于基于惯性加速来感应车辆500的姿态变化(例如，位置变化和方向变化)。GPS收发器可以是任何用于估计车辆500的地理位置的传感器。出于该目的，GPS收发器可以包括接收器/发送器以提供车辆500相对于地球的位置信息。需要说明的是，GPS是全球导航卫星系统的一个示例，因此，在一些实施例中，GPS收发器可以替换为北斗卫星导航系统收发器或者伽利略卫星导航系统收发器。雷达单元可以使用无线电信号来感应车辆500所在环境中的对象。在一些实施例中，除感应对象之外，雷达单元还可以用于感应接近车辆500的物体的速度和前进方向。LiDAR单元可以是任何使用激光来感应车辆500所在环境中的物体的传感器。在一些实施例中，LiDAR单元可以包括激光源、激光扫描仪、以及探测器。LiDAR单元用于以连续(例如使用外差检测)或者不连续的检测模式进行工作。图像捕捉装置可以包括用于捕捉车辆500所在环境的多个图像的装置。图像捕捉装置的一个例子是相机，相机可以是静态图像相机或者动态视频相机。
控制系统546用于控制对车辆500及其部件(或单元)的操作。相应地，控制系统546可以包括各种单元，例如转向单元、动力控制单元、制动单元和导航单元。
转向单元可以是调整车辆500前进方向的机械的组合。动力控制单元(例如可以为油门)，例如可以被用于控制发动机的运转速度，进而控制车辆500的速度。制动单元可以包括用于对车辆500进行减速的机械的组合。制动单元可以以标准方式利用摩擦力来使车辆减速。在其他实施例中，制动单元可以将车轮的动能转化为电流。制动单元也可以采用其它形式。导航单元可以是任何为车辆500确定驾驶路径或路线的系统。导航单元还可以在车辆500行进的过程中动态的更新驾驶路径。控制系统546还可以附加地或者可选地包括其它未示出或未描述的部件(或单元)。
用户接口系统548可以用于允许车辆500与外部传感器、其它车辆、其它计算机系 统和/或车辆500的用户之间的互动。例如，用户接口系统548可以包括标准视觉显示装置(例如，等离子显示器、液晶显示器(LCD)、触屏显示器、头戴显示器，或其它类似的显示器)，扬声器或其它音频输出装置，麦克风或者其它音频输入装置。例如，用户接口系统548还可以包括导航接口以及控制车辆500的内部环境(例如温度、风扇，等等)的接口。
通信系统552可以为车辆500提供与一个或多个设备或者周围其它车辆进行通信的方式。在一个示例性的实施例中，通信系统552可以直接或者通过通信网络与一个或多个设备进行通信。通信系统552例如可以是无线通信系统。例如，通信系统可以使用3G蜂窝通信(例如CDMA、EVDO、GSM/GPRS)或者4G蜂窝通信(例如WiMAX或LTE)，还可以使用5G蜂窝通信。可选地，通信系统可以与无线本地局域网(WLAN)进行通信(例如，使用)。在一些实施例中，通信系统552可以直接与一个或多个设备或者周围其它车辆进行通信，例如，使用红外线，蓝牙，或者ZIGBEE。其它无线协议，例如各种车载通信系统，也在本申请公开的范围之内。例如，通信系统可以包括一个或多个专用短程通信(DSRC)装置、V2V装置或者V2X装置，这些装置会与车辆和/或路边站进行公开或私密的数据通信。
计算系统550能控制车辆500的部分或者全部功能。计算系统550中的自动驾驶控制单元可以用于识别、评估、以及避免或越过车辆500所在环境中的潜在障碍。通常，自动驾驶控制单元可以用于在没有驾驶员的情况下控制车辆500，或者为驾驶员控制车辆提供辅助。在一些实施例中，自动驾驶控制单元用于将来自传感器的数据，例如GPS收发器的数据、雷达数据、LiDAR数据、相机数据、以及来自其它车辆系统的数据结合起来，来确定车辆500的行驶路径或轨迹。自动驾驶控制单元可以被激活以使车辆500能够以自动驾驶模式被驾驶。
计算系统550可以包括至少一个处理器(其可以包括至少一个微处理器)，处理器执行存储在非易失性计算机可读介质(例如数据存储装置或存储器)中的处理指令(即机器可执行指令)。计算系统550也可以是多个计算装置，这些计算装置分布式地控制车辆500的部件或者系统。在一些实施例中，存储器中可以包含被处理器执行来实现车辆500的各种功能的处理指令(例如，程序逻辑)。在一些实施例中，计算系统550能够与驱动系统542、传感器系统544、控制系统546、用户接口系统548、和/或通信系统552进行数据通信。计算系统中的接口用于促进计算系统550和驱动系统542、传感器系统544、控制系统546、用户接口系统548、以及通信系统552之间的数据通信。
存储器还可以包括其它指令，包括用于数据发送的指令、用于数据接收的指令、用于互动的指令、或者用于控制驱动系统542、传感器系统544、或控制系统546或用户接口系统548的指令。
除存储处理指令之外，存储器可以存储多种信息或数据，例如图像处理参数、道路地图、和路径信息。在车辆500以自动方式、半自动方式和/或手动模式运行的期间，这 些信息可以被车辆500和计算系统550所使用。
尽管自动驾驶控制单元被示为与处理器和存储器分离，但是应当理解，在一些实施方式中，自动驾驶控制单元的某些或全部功能可以利用驻留在一个或多个存储器(或数据存储装置)中的程序代码指令来实现并由一个或多个处理器执行，并且自动驾驶控制单元在某些情况下可以使用相同的处理器和/或存储器(或数据存储装置)来实现。在一些实施方式中，自动驾驶控制单元可以至少部分地使用各种专用电路逻辑，各种处理器，各种现场可编程门阵列(“FPGA”)，各种专用集成电路(“ASIC”)，各种实时控制器和硬件来实现。
计算系统550可以根据从各种车辆系统(例如，驱动系统542，传感器系统544，以及控制系统546)接收到的输入，或者从用户接口系统548接收到的输入，来控制车辆500的功能。例如，计算系统550可以使用来自控制系统546的输入来控制转向单元，来避开由传感器系统544检测到的障碍物。在一些实施例中，计算系统550可以用来控制车辆500及其系统的多个方面。
虽然图5中显示了集成到车辆500中的各种部件(或单元)，这些部件(或单元)中的一个或多个可以搭载到车辆500上或单独关联到车辆500上。例如，计算系统550可以部分或者全部地独立于车辆500存在。从而，车辆500能够以分离的或者集成的设备单元的形式而存在。构成车辆500的设备单元之间可以以有线通信或者无线通信的方式实现相互通信。在一些实施例中，可以将附加部件或单元添加到各个系统或从系统中移除一个或多个以上的部件或单元(例如，图5所示的LiDAR或雷达)。
图6是示出根据本公开一实施例的计算设备的示意图。
请参考图6，在一些实施例中，计算系统550可以通过计算设备600的形式来实现。
在一些实施例中，在计算设备600内，指令集在被执行时和/或处理逻辑在被启动时可以使该机器执行本文中所描述和/或要求的方法中的任何一种或多种。在备选实施例中，机器作为独立设备操作，或可以被连接(例如联网)到其他机器。在联网部署中，机器可以在服务器-客户端网络环境下以服务器或客户端机器的身份操作，或在对等(或分布式)网络环境中作为对等机操作。机器可以是个人计算机(PC)、膝上型计算机、平板计算系统、个人数字助理(PDA)、蜂窝电话、智能电话、网络应用、机顶盒(STB)、网络路由器、交换机或桥接器或能够执行指定将由该机器采取的动作的指令集(相继或以其他方式)或启动处理逻辑的任何机器。进一步地，虽然只图示了单个机器，但是术语“机器”也可以被理解为包括单独地或联合地执行用以执行本文中所描述和/或要求的方法中的任何一种或多种的指令集(或多个指令集)的机器的任何集合。
计算设备600可以包括可以经由总线606或其他数据传送系统彼此通信的数据处理器602(例如系统芯片(SoC)、通用处理核心、图形核心和可选其他处理逻辑)和存储器604(例如，内存)。计算设备600还可以包括各种输入/输出(I/O)设备和/或接口610，诸如触摸屏显示器、音频插孔、语音接口和可选网络接口612。在示例实施例 中，网络接口612可以包括一个或多个无线电收发器，其被配置成与任何一个或多个标准无线和/或蜂窝协议或接入技术(例如蜂窝系统的第二代(2G)、2.5代、第三代(3G)、第四代(4G)和下一代无线电接入、全球移动通信系统(GSM)、通用分组无线电服务(GPRS)、增强型数据GSM环境(EDGE)、宽带码分多址(WCDMA)、LTE、CDMA2000、WLAN、无线路由器(WR)网格等)。网络接口612还可以被配置成与各种其他有线和/或无线通信协议(包括TCP/IP、UDP、SIP、SMS、RTP、WAP、CDMA、TDMA、UMTS、UWB、WiFi、WiMax、IEEE802.11x等)一起使用。本质上，网络接口612可以实际上包括或支持任何有线和/或无线通信和数据处理机构，通过该机构，信息/数据可以经由网络614在计算设备600与另一计算或通信系统之间传播。
存储器604可以表示机器可读介质(或计算机可读存储介质)，在机器可读介质(或计算机可读存储介质)上存储实施本文中所描述和/或要求的方法或功能中的任何一个或多个的一个或多个指令集、软件、固件或其他处理逻辑(例如逻辑608)。在由计算设备600执行期间，逻辑608或其一部分也可以完全或至少部分地驻留在处理器602内。如此，存储器604和处理器602也可以构成机器可读介质(或计算机可读存储介质)。逻辑608或其一部分也可以被配置为处理逻辑或逻辑，该处理逻辑或逻辑的至少一部分被部分地实现于硬件中。逻辑608或其一部分还可以经由网络接口612来通过网络614被传输或接收。虽然示例实施例的机器可读介质(或计算机可读存储介质)可以是单种介质，但是术语“机器可读介质”(或计算机可读存储介质)应被理解为包括存储一个或多个指令集的单种非暂时性介质或多种非暂时性介质(例如集中式或分布式数据库和/或相关联的高速缓存和计算系统)。术语“机器可读介质”(或计算机可读存储介质)也可以被理解为包括能够存储、编码或携带指令集以供机器执行并且使机器执行各种实施例的方法中的任何一种或多种或能够存储、编码或携带被这种指令集利用或与之相关联的数据结构的任何非暂时性介质。术语“机器可读介质”(或计算机可读存储介质)可以因此被理解为包括但不限于固态存储器、光学介质和磁性介质。
所公开的和其他实施例、模块以及本文档中所描述的功能操作可以在数字电子电路系统中、或在计算机软件、固件或硬件中(包括本文档中所公开的结构和其结构等效物)或它们中的一个或多个的组合中被实现。所公开的和其他实施例可以被实现为一个或多个计算机程序产品，也就是说，被编码在计算机可读介质上以由数据处理装置执行或以控制该数据处理装置的操作的计算机程序指令的一个或多个模块。计算机可读介质可以是机器可读存储设备、机器可读存储衬底、存储器设备、影响机器可读传播信号的物质合成物或它们中的一个或多个的组合。术语“数据处理装置”涵盖了用于处理数据的所有装置、设备和机器，包括例如可编程处理器、计算机或多个处理器或计算机。除了硬件之外，该装置还可以包括为探讨中的计算机程序创建执行环境的代码，例如构成处理器固件、协议栈、数据库管理系统、操作系统或它们中的一个或多个的组合的代码。传播信号是人工生成的信号，例如由机器生成的电信号、光信号或电磁信号，该信号被生成 以对要传输给适合的接收器装置的信息进行编码。
计算机程序(也被称为程序、软件、软件应用、脚本或代码)可以以任何形式的编程语言(包括编译语言或解译语言)被写入，并且该计算机程序可以以任何形式被部署，包括被部署为独立的程序或部署为模块、部件、子例程或适合在计算环境中使用的另一单元。计算机程序并非必须与文件系统中的文件对应。程序可以被存储在保持其他程序或数据(例如被存储在标记语言文档中的一个或多个脚本)的文件的一部分中，或被存储在专用于探讨中的程序的单个文件中，或被存储在多个协作文件(例如存储一个或多个模块、子程序或部分代码的文件)中。计算机程序可以被部署成在一个计算机上执行或在被定位于一个站点处或被分布在多个站点中并且通过通信网络被互连的多个计算机上被执行。
本文档中所描述的过程和逻辑流可以被执行一个或多个计算机程序的一个或多个可编程处理器执行以通过对输入数据进行操作并且生成输出来执行功能。过程和逻辑流还可以被专用逻辑电路系统(例如FPGA(现场可编程门阵列)或者ASIC(专用集成电路))执行，并且装置还可以被实现为专用逻辑电路(例如FPGA(现场可编程门阵列)或者ASIC(专用集成电路))。
适合执行计算机程序的处理器包括例如通用微处理器和专用微处理器两者以及任何种类的数字计算机的任何一个或多个处理器。通常，处理器将接收来自只读存储器或随机存取存储器或两者的指令和数据。计算机的必要元件是用于执行指令的处理器和用于存储指令和数据的一个或多个存储器设备。通常，计算机还会包括用于存储数据的一个或多个海量存储设备(例如磁盘、磁光盘或光盘)，或计算机还会被操作地联接以接收来自该一个或多个海量存储设备的数据或将数据传送给该一个或多个海量存储设备或进行两者。然而，计算机不需要具有这种设备。适合于存储计算机程序指令和数据的计算机可读介质包括所有形式的非易失性存储器、介质和存储器设备，包括例如半导体存储器设备，例如EPROM、EEPROM和闪速存储器设备；磁盘，例如内部硬盘或可移除盘；磁光盘；以及CD-ROM盘和DVD-ROM盘。处理器和存储器可以被专用逻辑电路系统补充或可以被并入该专用逻辑电路系统中。
本文描述的一些实施例是在方法或过程的一般上下文中所描述的，其在一个实施例中可以由包括在计算机可读介质中的计算机程序产品来实施，该计算机程序产品可以包括计算机可执行指令(如程序代码)，计算机可执行指令例如可以由联网环境中的计算机执行。计算机可读介质可以包括可移动和不可移动存储设备，包括但不限于只读存储器(ROM)、随机存取存储器(RAM)、光盘(CD)、数字多功能盘(DVD)等。因此，计算机可读介质可以包括非暂态存储介质。通常，程序模块可以包括例程、程序、对象、部件、数据结构等，其执行特定任务或实施特定抽象数据类型。计算机或处理器可执行指令、相关联的数据结构和程序模块表示用于执行本文所公开的方法的步骤的程序代码的示例。这样的可执行指令或相关联的数据结构的特定序列表示用于实现在这些 步骤或过程中描述的功能的对应动作的示例。
所公开的实施例中的一些实施例可以使用硬件电路、软件或其组合来实现为设备或模块。举例来说，硬件电路实施可以包括离散模拟和/或数字部件，其例如可以集成为印刷电路板的部分。备选地或附加地，所公开的部件或模块可以被实施为专用集成电路(ASIC)和/或现场可编程门阵列(FPGA)设备。附加地或备选地，一些实施可以包括数字信号处理器(DSP)，其是具有针对与本申请的所公开功能性相关联的数字信号处理的操作需要而优化的架构的专用微处理器。类似地，每个模块内的各种部件或子组件可以用软件、硬件或固件来实施。可以使用本领域已知的任何一种连接方法和介质来提供模块和/或模块内部件之间的连接，包括但不限于通过因特网、有线网络或使用适当协议的无线网络的通信。
虽然本文包括许多细节，但是这些细节不应被解释为对所要求保护的发明的范围的限制，而是被解释为对特定实施例特有的特征的描述。本文中在不同的实施例的上下文中描述的某些特征也可以组合在单个实施例中。相反，在单个实施例的上下文中描述的各种特征也可以分开或以任何合适的子组合在多个实施例中实施。此外，在上文中虽然特征可以被描述为以某些组合作用并且甚至最初被要求保护，但是来自要求保护的组合的一个或多个特征在一些情况下可以从组合中被去除，并且所要求保护的组合可以针对于子组合或子组合的变型。类似地，虽然在图式中以特定次序描绘操作，但此不应被理解为要求以所示的特定次序或以循序次序执行此类操作、或执行所有所说明的操作以实现期望的结果。
仅描述了几个实施方式和示例，并且可以基于在本公开中描述和示出的内容来进行其他实施、增强和变化。

Claims (21)

  1. 一种行车信息来源评估方法，其特征在于，包括：
    取得来自待评估行车信息来源与多个其他行车信息来源的信息；
    基于所述信息，计算所述待评估行车信息来源关联于所述多个其他行车信息来源中的每一者之间的多个参数评分；
    基于所述多个参数评分，得到关联于所述待评估行车信息来源的多种状态的多个一致性信息；及
    基于所述多个一致性信息，决定所述待评估行车信息来源的当前状态。
  2. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源及所述多个其他行车信息来源关联于处于行进程中的车辆，且所述行车信息来源评估方法更包括：
    基于所述当前状态，使所述车辆于所述行进过程中执行降低风险策略。
  3. 如权利要求2所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源与所述多个其他行车信息来源包括安装于所述车辆上的多个传感器，其中所述降低风险策略包括关闭所述多个传感器的至少其中之一。
  4. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源及所述多个其他行车信息来源包括安装于车辆上的多个传感器，且所述行车信息来源评估方法更包括：
    基于所述当前状态，校正所述多个传感器的多个外参。
  5. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述多个一致性信息包括多个后验概率，其中基于所述多个参数评分，得到关联于所述待评估行车信息来源的多种状态的所述多个一致性信息包括：
    基于所述多个参数评分，计算关联于所述待评估行车信息来源的多种状态的所述多个后验概率。
  6. 如权利要求5所述的行车信息来源评估方法，其特征在于，所述行车来源评估方法更包括：
    归一化所述多个后验概率，以得到多个归一化后验概率；
    基于所述多种状态中的特定状态所对应的所述多个归一化后验概率的其中之一，输出所述待评估行车信息来源的评估分数。
  7. 如权利要求5所述的行车信息来源评估方法，其特征在于，基于所述多个参数评分，计算关联于所述待评估行车信息来源的所述多种状态的所述多个后验概率包括：
    决定关联于所述待评估行车信息来源的多个参数评分在多种状态组合的基础上的多个条件概率函数；及
    基于所述多个条件概率函数及所述多个参数评分中的多个当前参数评分，计算所述多个后验概率。
  8. 如权利要求7所述的行车信息来源评估方法，其特征在于，决定关联于所述待评估行车信息来源、在所述多种状态组合的基础上的所述多个参数的所述多个条件概率函数包括：
    基于所述多个参数评分中的第一参数评分的多个第一评分值，决定符合所述多种状态组合的其中之一的预期结果的第一分布函数；及
    基于所述第一分布函数，决定对应于所述第一参数评分的所述多个条件概率函数中的多个第一条件概率函数。
  9. 如权利要求8所述的行车信息来源评估方法，其特征在于，所述第一分布函数为截断正态分布的累计分布函数，并且基于所述多个参数评分中的所述第一参数评分的所述多个第一评分值，决定符合所述多种状态组合的所述其中之一的所述预期结果的所述第一分布函数包括：
    基于所述多个第一评分值，决定所述截断正态分布的均值与方差。
  10. 如权利要求7所述的行车信息来源评估方法，其特征在于，所述参数评分定义于预设范围内，并且所述多个条件概率函数中的每一者的定义域定义于所述预设范围内。
  11. 如权利要求10所述的行车信息来源评估方法，其特征在于，所述参数评分与其所对应的两个行车信息来源之间的标定误差成负相关。
  12. 如权利要求10所述的行车信息来源评估方法，其特征在于，所述预设范围为0到1。
  13. 如权利要求7所述的行车信息来源评估方法，其特征在于，所述多个条件概率函数中的每一者通过截断正态分布的累计分布函数来描述。
  14. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述多个一致性信息包括多个后验概率，其中基于所述多个一致性信息，决定所述待评估行车信息来源的所述当前状态包括：
    决定所述当前状态为所述多个后验概率中的最大者所关联的所述多种状态的其中之一。
  15. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源与所述多个其他行车信息来源包括高精地图。
  16. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源及所述多个其他行车信息来源关联于处于行进过程中的车辆，且所述待评估行车信息来源与所述多个其他行车信息来源包括安装于所述车辆上的多个传感器。
  17. 如权利要求16所述的行车信息来源评估方法，其特征在于，所述多个传感器包括以下至少其中之一：相机、激光雷达、惯性测量单元。
  18. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述多个参数评分包括多个外参评分，其中所述多个外参评分的每一者来自于计算安装于车辆上的两个传感器之间的关联。
  19. 如权利要求1所述的行车信息来源评估方法，其特征在于，所述待评估行车信息来源的所述多个状态的多个先验概率默认为相等。
  20. 一种行车信息来源评估装置，包括：
    一个或者多个处理器，和
    存储程序的存储器，所述程序包括指令，所述指令在由所述处理器执行时使所述行车信息来源评估装置执行根据权利要求1至19中任一项所述的方法。
  21. 一种存储有程序的计算机可读存储介质，所述程序包括指令，所述指令在由计算装置的一个或者多个处理器执行时，致使所述计算装置执行根据权利要求1至19中任一项所述的方法。
PCT/CN2023/124441 2022-12-13 2023-10-13 用于评估行车信息来源的方法和装置 WO2024125062A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211603851.6A CN118190020A (zh) 2022-12-13 2022-12-13 用于评估行车信息来源的方法和装置
CN202211603851.6 2022-12-13

Publications (1)

Publication Number Publication Date
WO2024125062A1 true WO2024125062A1 (zh) 2024-06-20

Family

ID=91395459

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/124441 WO2024125062A1 (zh) 2022-12-13 2023-10-13 用于评估行车信息来源的方法和装置

Country Status (2)

Country Link
CN (1) CN118190020A (zh)
WO (1) WO2024125062A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107430591A (zh) * 2015-01-26 2017-12-01 Trw汽车美国有限责任公司 车辆驾驶员辅助系统
CN109345596A (zh) * 2018-09-19 2019-02-15 百度在线网络技术（北京）有限公司 多传感器标定方法、装置、计算机设备、介质和车辆
CN109523129A (zh) * 2018-10-22 2019-03-26 吉林大学 一种无人车多传感器信息实时融合的方法
CN109781163A (zh) * 2018-12-18 2019-05-21 北京百度网讯科技有限公司 标定参数有效性检验方法、装置、设备及存储介质
US20200331484A1 (en) * 2019-04-18 2020-10-22 International Business Machines Corporation Addressing vehicle sensor abnormalities
CN112508317A (zh) * 2020-09-01 2021-03-16 中国汽车技术研究中心有限公司 一种基于多源动力总成车型驾驶性主客观关联评分方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107430591A (zh) * 2015-01-26 2017-12-01 Trw汽车美国有限责任公司 车辆驾驶员辅助系统
CN109345596A (zh) * 2018-09-19 2019-02-15 百度在线网络技术（北京）有限公司 多传感器标定方法、装置、计算机设备、介质和车辆
CN109523129A (zh) * 2018-10-22 2019-03-26 吉林大学 一种无人车多传感器信息实时融合的方法
CN109781163A (zh) * 2018-12-18 2019-05-21 北京百度网讯科技有限公司 标定参数有效性检验方法、装置、设备及存储介质
US20200331484A1 (en) * 2019-04-18 2020-10-22 International Business Machines Corporation Addressing vehicle sensor abnormalities
CN112508317A (zh) * 2020-09-01 2021-03-16 中国汽车技术研究中心有限公司 一种基于多源动力总成车型驾驶性主客观关联评分方法

Also Published As

Publication number Publication date
CN118190020A (zh) 2024-06-14

Similar Documents

Publication Publication Date Title
US10867188B2 (en) System and method for image localization based on semantic segmentation
CN110914707B (zh) 基于相机和lidar数据的车辆位置和速度估计的系统和方法
US11729520B2 (en) Sensor layout for autonomous vehicles
US10752246B2 (en) System and method for adaptive cruise control with proximate vehicle detection
WO2021184218A1 (zh) 一种相对位姿标定方法及相关装置
EP4083943A1 (en) Method and apparatus for detecting lane lines, electronic device and storage medium
WO2021189210A1 (zh) 一种车辆换道方法及相关设备
CN112527003B (zh) 数据传输装置和系统
EP4180295A1 (en) Method and device for recognizing vehicle motion state
US20240017719A1 (en) Mapping method and apparatus, vehicle, readable storage medium, and chip
EP4087218A1 (en) Method and apparatus for processing sensor data, computing device and storage medium
US20220270377A1 (en) Method and apparatus for detecting corner points of lane lines, electronic device and storage medium
WO2024125062A1 (zh) 用于评估行车信息来源的方法和装置
CN115056784B (zh) 车辆控制方法、装置、车辆、存储介质及芯片
CN115221151A (zh) 车辆数据的传输方法、装置、车辆、存储介质及芯片
CN115035357A (zh) 目标检测模型构建方法、目标检测方法、装置和计算设备
CN112527002B (zh) 监控装置和监控方法
CN118279845A (zh) 车道线检测的方法、装置和计算机可读存储介质