A post-processing method, device, equipment and medium for target perception and tracking

Abstract

An embodiment of the present invention discloses a post - processing method, device, equipment and medium for target perception and tracking. The method includes: when the front - end sensor of the vehicle senses a target object, obtaining the measurement value of the target object system track at the current moment, and initializing the covariance matrix and score of the target object system track; determining the prior estimate value of the target object system track according to the measurement value, and updating the prior covariance matrix; matching the measurement value at the next moment with the prior estimate value, determining the first score and the posterior estimate value of the system track according to the matching result and the initialized score, and updating the posterior covariance matrix; if the first score is greater than the first preset threshold, determining the state information of the target object; if the first score is less than the second preset threshold, deleting the posterior estimate value. This technical solution can realize the perception and tracking of the target system track and maintain the life cycle of the system track, improving the accuracy and reliability of the target state information.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving, and particularly to a post-processing method, device, equipment and medium for target perception and tracking. Background Art

[0002] At present, the autonomous driving technology has developed rapidly, and mid- to high-end models at home and abroad are all loaded with autonomous driving functions, using single or multiple target detection sensors for detection and fusion to provide reliable environmental perception results for the system. However, due to the limitations of current sensor technologies, challenges are brought to target tracking, and the two most significant challenges are false alarms and missed detections. No matter which sensor is selected, almost these two challenges will be faced. For example, in the application of millimeter-wave radar, false alarms occur because objects in the background reflect enough electromagnetic wave energy to make the signal processor think that the object is a target of interest. Missed detections occur due to too low signal-to-noise ratio, weak signal intensity below the detection threshold, being blocked, etc.

[0003] Currently, a standard Kalman filter is usually used for the processing of target perception and tracking, but this method is only applicable to linear systems and cannot perform effective tracking when the vehicle is driving in a roundabout or on a curve. Summary of the Invention

[0004] The present invention provides a post-processing method, device, equipment and medium for target perception and tracking, which can support various complex application scenarios with multiple non-linear and multi-noises, has strong scalability and portability, and can improve the accuracy and reliability of the post-processing results of target perception and tracking.

[0005] According to one aspect of the present invention, a post-processing method for target perception and tracking is provided, and the method includes:

[0006] When the front-end sensor of the vehicle senses a target object, obtain the measurement value of the target object system track at the current moment, and initialize the covariance matrix and score of the target object system track;

[0007] Determine the prior estimate value of the target object system track according to the measurement value, and update the prior covariance matrix;

[0008] Match the measurement value at the next moment with the prior estimate value, determine the first score and the posterior estimate value of the system track according to the matching result and the initialized score, and update the posterior covariance matrix;

[0009] If the first score is greater than the first preset threshold, determine the state information of the target object; if the first score is less than the second preset threshold, delete the posterior estimate value.

[0010] According to another aspect of the present invention, there is provided a post-processing device for target perception tracking, including:

[0011] A current moment measurement value acquisition module, configured to acquire the current moment measurement value of the target object system track when the front-end sensor of the vehicle senses the target object, and initialize the covariance matrix and score of the target object system track;

[0012] A prior estimation module, configured to determine the prior estimation value of the target object system track according to the measurement value, and update the prior covariance matrix;

[0013] A posterior estimation module, configured to match the measurement value at the next moment with the prior estimation value, determine the first score and the posterior estimation value of the system track according to the matching result and the initialized score, and update the posterior covariance matrix;

[0014] A state information determination module, configured to determine the state information of the target object if the first score is greater than a first preset threshold; and delete the posterior estimation value if the first score is less than a second preset threshold.

[0015] According to another aspect of the present invention, there is provided an electronic device, including:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the post-processing method for target perception tracking according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement the post-processing method for target perception tracking according to any embodiment of the present invention when executed.

[0020] In the technical solution of the embodiment of the present invention, when the front-end sensor of the vehicle senses a target object, the measured value of the target object system track at the current moment is obtained, and the covariance matrix and score of the target object system track are initialized; the prior estimate value of the target object system track is determined according to the measured value, and the prior covariance matrix is updated; the measured value at the next moment and the prior estimate value are matched, and the first score and the posterior estimate value of the system track are determined according to the matching result and the initialized score, and the posterior covariance matrix is updated; if the first score is greater than the first preset threshold, the state information of the target object is determined; if the first score is less than the second preset threshold, the posterior estimate value is deleted. This technical solution can realize the perception and tracking of the target system track and maintain the life cycle of the system track, and improve the accuracy and reliability of the target state information.

[0021] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 is a flowchart of a post-processing method for target perception and tracking according to Embodiment 1 of the present invention;

[0024] Figure 2 is a flowchart of a post-processing method for target perception and tracking according to Embodiment 2 of the present invention;

[0025] Figure 3 is a schematic diagram of the positional relationship between a vehicle and a target object in a post-processing method for target perception and tracking according to Embodiment 2 of the present invention;

[0026] Figure 4 is a schematic structural diagram of a post-processing device for target perception and tracking according to Embodiment 3 of the present invention;

[0027] Figure 5 is a schematic structural diagram of an electronic device for implementing the post-processing method for target perception and tracking in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] To enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0029] It should be noted that the terms "first", "second", "target", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0030] Embodiment 1

[0031] Figure 1 FIG. is a flowchart of a post-processing method for target perception and tracking provided in Embodiment 1 of the present invention. This embodiment is applicable to the situation of post-processing target perception and tracking in the surrounding environment. This method can be executed by a post-processing device for target perception and tracking. The post-processing device for target perception and tracking can be implemented in the form of hardware and / or software, and the post-processing device for target perception and tracking can be configured in an electronic device with data processing capabilities. As Figure 1 shown, the method includes:

[0032] S110. When the front-end sensor of the vehicle senses a target object, obtain the measurement value of the target object system track at the current moment, and initialize the covariance matrix and score of the target object system track.

[0033] Among them, the front-end sensor can be a detection device installed on the vehicle for sensing the surrounding environment of the vehicle and target objects, and can be an optical camera, lidar, millimeter-wave radar, ultrasonic radar, etc. The target object can be a pedestrian, a vehicle, an obstacle, etc. The system trajectory can be the movement trajectory of the target object detected by the front-end sensor of the vehicle. The measurement value can be the output value or the fusion value of the front-end sensor. For example, if the front-end sensor is a single sensor, the measurement value is the output value of the sensor; if the front-end sensor is a combined sensor, the measurement value is the fusion value after the output values of multiple sensors are fused. The measurement value can be the speed, distance, position, type, or acceleration of the target object, etc. The covariance matrix of the target object system trajectory can represent the strength of the correlation of each group of measurement values of the target object system trajectory. The score of the target object system trajectory can represent the quality of the target object system trajectory.

[0034] In the embodiment of the present invention, the front-end sensor of the vehicle senses the target object and obtains the measurement value of the target object system trajectory. If the front-end sensor is an optical camera, after the light reflected by the target object passes through the optical camera and is focused on the CCD / CMOS chip by the lens to generate a digital signal to determine that the target object is sensed, the measurement value of the target object system trajectory can be determined according to the generated digital signal; when the front-end sensor is lidar, laser pulses can be emitted into the surrounding environment of the vehicle to detect and range the target object, and the measurement value of the target object system trajectory can be determined according to the time difference and phase difference of the laser pulse signals; if the front-end sensor is a millimeter-wave radar, radio waves can be emitted and the reflected signals can be received, and the distance of the target relative to the vehicle can be calculated according to the flight time of the electromagnetic wave returning, and the speed of the target object relative to the vehicle can be calculated by detecting the frequency difference according to the Doppler principle; if the front-end sensor is an ultrasonic radar, the distance of the target object relative to the vehicle can be judged by the time of sound transmission in the air, and the distance of the target object relative to the vehicle can be determined according to the time difference between the emission and reception by using the propagation speed of ultrasonic waves in the air and measuring the time when the sound wave is reflected by the target object after emission; if the front-end sensor is a combined sensor of at least two of an optical camera, lidar, millimeter-wave radar, or ultrasonic radar, the measurement values of the local trajectories of the target object determined by each sensor can be fused by a data fusion algorithm to determine the measurement value of the system trajectory of the target object, where the data fusion algorithm can be a convex combination fusion algorithm or a Bar-Shalom-Campo fusion algorithm.

[0035] In addition, the covariance matrix and score of the target object system trajectory are initialized, and corresponding initial values can be set according to different types of front-end sensors.

[0036] It is understandable that the front-end sensor can be a single sensor or a combination of multiple sensors. The embodiments of the present invention do not make specific limitations on the number and types of front-end sensors.

[0037] S120. Determine a prior estimate of the target object system track based on the measurement value, and update the prior covariance matrix.

[0038] Among them, the prior estimate can be a value obtained by applying the system to the optimal estimate of the previous moment, which is a prior understanding of the motion law of the system itself. The prior covariance matrix can reflect the error and confidence of the determined prior estimate. In the embodiments of the present invention, determining the prior estimate of the target object system track based on the measurement value and updating the prior covariance matrix can be determined based on the extended Kalman filter or based on the particle filter.

[0039] Taking the extended Kalman filter as an example for explanation, the prior estimate of the target object system track can be determined by the following formula:

[0040]

[0041] Among them, X k-1 is the state matrix of the target object system track at the previous moment, X k|k-1 is the prior estimate of the target object system track, f(X k-1 ) is the state transition matrix, is the Jacobian matrix obtained by performing a first-order Taylor expansion on the state transition matrix f(X k-1 );

[0042] The prior covariance matrix can be determined by the following formula:

[0043]

[0044] Among them, P k-1 is the covariance matrix of the target object system track at the previous moment, P k|k-1 is the prior covariance matrix of the target object system track, and Q is the prediction noise matrix.

[0045] S130. Match the measurement value of the next moment with the prior estimate, determine the first score and the posterior estimate of the system track according to the matching result and the initialization score, and update the posterior covariance matrix.

[0046] Among them, the first score can be an evaluation criterion for the prediction accuracy of the prior estimated value of the target object. The posterior estimated value can be a value obtained by correcting the prior estimated value based on the successfully matched prior estimated value and the measurement value at the next moment. The posterior covariance matrix can reflect the error and confidence level of the determined posterior estimated value. It can be understood that the prior estimated value determined based on the measurement value at the current moment is a predicted estimated value of the measurement value at the next moment.

[0047] In an embodiment of the present invention, in order to determine whether the prediction accuracy of the prior estimated value meets a preset standard, the measurement value at the next moment and the prior estimated value are matched to determine the first score and the posterior estimated value of the system track.

[0048] In an embodiment of the present invention, optionally, matching the measurement value and the prior estimated value includes performing threshold matching and ID matching on the measurement value at the next moment and the prior estimated value; correspondingly, matching the measurement value and the prior estimated value, and determining the first score of the system track according to the matching result and the initialization score, includes: performing ID matching on the measurement value at the next moment and the prior estimated value within the threshold range, if the ID of the measurement value at the next moment is the same as the ID of the prior estimated value, it is marked as a successful match, and points are added according to a preset rule based on the initialization score to determine the first score of the system track; otherwise, it is marked as a failed match, and points are deducted according to a preset rule based on the initialization score to determine the first score of the system track; correspondingly, matching the measurement value at the next moment and the prior estimated value, and determining the posterior estimated value of the system track according to the matching result and the initialization score, includes: when it is marked as a successful match, determining the Kalman gain value according to the measurement value at the next moment, the prior estimated value, and the measurement noise matrix; determining the posterior estimated value of the system track according to the prior estimated value, the measurement value at the next moment, and the Kalman gain value.

[0049] Among them, the threshold matching can be the error range between the measurement value at the next moment and the prior estimated value. The ID matching can be the type of the front-end sensor. ID matching is performed on the measurement value at the next moment and the prior estimated value whose error value between them satisfies the preset threshold range.

[0050] If the ID of the measurement value at the next moment is the same as the ID of the prior estimated value, it is marked as a successful match, and points are added according to a preset rule based on the initialization score to determine the first score of the system track, and posterior update is performed.

[0051] Among them, the preset rules can be adjusted according to the actual situation for the added or subtracted scores. The posterior update can be carried out by first determining the Kalman gain value to characterize whether the posterior estimate value believes more in the measurement value at the next moment or the prior estimate value, and then determining the posterior estimate value and the posterior covariance matrix based on the Kalman gain value. The Kalman gain value can be determined by the following formula:

[0052] K = P k|k-1 *(P k|k-1 + R) -1 ;

[0053] Among them, K is the Kalman gain value, and R is the measurement noise matrix. The measurement noise matrix R is a matrix that can characterize the measurement accuracy of the front-end sensor;

[0054] The posterior estimate value can be determined by the following formula:

[0055] X k = X k|k-1 + K*(M k - X k|k-1 );

[0056] Among them, X k is the posterior estimate value of the target object, and M k is the measurement value at the next moment;

[0057] The posterior covariance matrix can be determined by the following formula:

[0058] P k = (I - K*H)*P k|k-1 ;

[0059] Among them, P k is the posterior covariance matrix of the target object, I is the identity matrix, and H is the relationship matrix between the measurement value at the next moment and the prior estimate value.

[0060] If the ID of the measurement value at the next moment is different from the ID of the prior estimate value, it is marked as a matching failure, and points are deducted according to the preset rules based on the initialized score to determine the first score of the system track.

[0061] S140. If the first score is greater than the first preset threshold, the state information of the target object is determined; if the first score is less than the second preset threshold, the posterior estimate value is deleted.

[0062] Among them, both the first preset threshold and the second preset threshold can be the score range of the system track quality. The state information can be the output data information determined according to the measurement value and the estimate value.

[0063] It can be understood that the first score characterizes the quality of the target object system trajectory. The larger the first score, the better and more reliable the system trajectory quality. Conversely, the smaller the first score, the worse and less reliable the system trajectory quality. When the first score exceeds a certain level, that is, is greater than the first preset threshold, it indicates that the accuracy or reliability of the target object state information determined based on the measurement value and the estimated value meets the preset standard range. When the first score is lower than a certain level, that is, is less than the second preset threshold, it indicates that the accuracy or reliability of the target object state information determined based on the measurement value and the estimated value is lower than the preset standard and cannot be output and adopted, and its data information needs to be cleared.

[0064] The post-processing method for target perception and tracking provided by the embodiments of the present invention includes: when the front-end sensor of the vehicle senses a target object, obtaining the measurement value of the target object system trajectory at the current moment, and initializing the covariance matrix and score of the target object system trajectory; determining the prior estimated value of the target object system trajectory according to the measurement value, and updating the prior covariance matrix; matching the measurement value at the next moment with the prior estimated value, determining the first score and the posterior estimated value of the system trajectory according to the matching result and the initialized score, and updating the posterior covariance matrix; if the first score is greater than the first preset threshold, determining the state information of the target object; if the first score is less than the second preset threshold, deleting the posterior estimated value. By using the measurement values at the current moment and the next moment to determine the prior estimated value and the posterior estimated value of the target object, the present technical solution realizes the perception and tracking of the target object, and manages the score of the target object system trajectory to maintain the life cycle of the system trajectory, thereby improving the accuracy and reliability of the post-processing result of target perception and tracking.

[0065] Embodiment 2

[0066] Figure 2 FIG. is a flowchart of a post-processing method for target perception and tracking provided by Embodiment 2 of the present invention. This embodiment is optimized based on the above embodiment. Specifically, the optimization is that the measurement value at the current moment and the measurement value at the next moment are obtained in a Cartesian coordinate system established with the center of the rear axle of the vehicle as the origin.

[0067] It can be understood that, for the convenience of the front-end sensor of the vehicle to sense and obtain the data information of the target objects around the vehicle, the measurement value at the current moment and the measurement value at the next moment are obtained in a Cartesian coordinate system established with the center of the rear axle of the vehicle as the origin.

[0068] As Figure 2 shown, the method of this embodiment specifically includes the following steps:

[0069] S210. When the front - end sensor of the vehicle senses a target object, obtain the positioning information of the vehicle in the geodetic coordinate system and the measured value of the target object system track at the current moment.

[0070] Among them, the geodetic coordinate system is a coordinate system established with the reference ellipsoid as the reference surface in geodetic surveying. The position of a ground point is represented by geodetic longitude, geodetic latitude, and geodetic height. The positioning information can represent the precise position of the vehicle in the geodetic coordinate system. For example, it can include the position coordinate value, speed value, and heading angle of the vehicle.

[0071] By obtaining the positioning information of the vehicle in the geodetic coordinate system, it can be matched to a high - precision map, which is beneficial for the vehicle to obtain real - time and accurate road information, determine the precise vehicle navigation route information, and determine the position information of the target object during driving.

[0072] S220. Obtain the first timestamp corresponding to the positioning information and the second timestamp corresponding to the measured value at the current moment.

[0073] Among them, the first timestamp and the second timestamp can use GPS time service. The acquisition method of the first timestamp and the second timestamp can obtain the standard time signal from GPS satellites and transmit this information through various interface types, so as to achieve the time synchronization of the entire system.

[0074] S230. Determine the prior estimate value of the target object system track according to the measured value and update the prior covariance matrix.

[0075] S240. Match the measured value at the next moment with the prior estimate value, determine the first score and the posterior estimate value of the system track according to the matching result and the initialization score, and update the posterior covariance matrix.

[0076] S250. If the first score is greater than the first preset threshold, determine the state information of the target object according to the positioning information and the posterior estimate value; if the first score is less than the second preset threshold, delete the posterior estimate value.

[0077] It can be understood that the measured value of the target object at the current moment and the measured value at the next moment are obtained in the Cartesian coordinate system established with the center of the rear axle of the vehicle as the origin, while the positioning information of the vehicle is determined in the geodetic coordinate system. It is necessary to perform coordinate transformation on the posterior estimate value further calculated based on the measured value at the current moment and the measured value at the next moment to facilitate the vehicle to obtain the state information of the target object during driving.

[0078] Figure 3It is a schematic diagram of the positional relationship between a vehicle and a target object in a post - processing method for target perception and tracking provided by Embodiment 2 of the present invention. As Figure 3 shown, the position coordinate information of the vehicle is established based on the geodetic coordinate system. Its horizontal coordinate system takes o as the coordinate origin and establishes a coordinate system with the x utm and y utm axes. Then the position coordinates of vehicle A in this coordinate system are (x u0 , y u0 ), and the speeds are respectively and The acceleration is The position coordinate information of the target object B takes the center of the rear axle of the vehicle as the origin and the direction x local perpendicular to the rear axle of the vehicle and the direction y local of the rear axle of the vehicle to establish a coordinate system. Then the position coordinates of the target object in this coordinate system are (x a , y a ), and the speeds are respectively and The acceleration is The heading angle is θ.

[0079] Among them, the position coordinates of the target object B in the geodetic coordinate system can be determined through conversion by the following formula:

[0080] x ua =x a ·sinθ - y a ·cosθ + x u0 ;

[0081] y ua =x a ·cosθ + y a ·sinθ + y u0 ;

[0082] Among them, x ua is the coordinate information of the target object on the x utm axis of the geodetic coordinate system, and y ua is the coordinate information of the target object on the y utm axis of the geodetic coordinate system.

[0083] S260. Perform motion compensation on the state information of the target object according to the first timestamp, the second timestamp, and the state information of the target object.

[0084] Among them, the motion compensation can be to eliminate the error of the target object state information according to the time error between the first timestamp corresponding to the vehicle positioning information and the second timestamp corresponding to the current moment measurement value, and improve the accuracy of the target object state information.

[0085] In an embodiment of the present invention, optionally, performing motion compensation on the state information of the target object according to the first timestamp, the second timestamp, and the state information of the target object includes: determining the absolute speed of the target object in the current time period according to the first timestamp, the third timestamp of the measurement value at the next moment, and the position information corresponding to the first timestamp and the third timestamp respectively; based on a standard Kalman filter, and the absolute speed of the target object in the current time period and the speed estimation value of the target object in the previous time period, determining the speed estimation value and the Kalman gain value of the target object at the current moment, and updating the covariance matrix; determining a first heading angle of the target object according to the speed estimation value at the current moment, and determining a second heading angle of the target object through a pose detection sensor of the vehicle; determining a target heading angle according to the first heading angle and the second heading angle.

[0086] It can be understood that performing motion compensation on the target object may include speed compensation and heading angle compensation.

[0087] When performing speed compensation on the target object, where the current time period may be the time period between the current moment and the next moment, and the previous time period may be the time period between the current moment and the previous moment. Determining the absolute speed of the target object in the current time period according to the first timestamp, the third timestamp corresponding to the measurement value at the next moment, the positioning information of the vehicle, and the position information in the state information of the target object may include steps A1 - A3:

[0088] A1: Determining the duration of the current time period according to the first timestamp and the third timestamp corresponding to the measurement value at the next moment;

[0089] A2: Determining the distance information between the vehicle and the target object according to the positioning information of the vehicle and the position information of the target object in the geodetic coordinate system;

[0090] A3: Determining the absolute speed of the target object in the current time period according to the duration of the current time period and the distance information.

[0091] Based on a standard Kalman filter, and the absolute speed of the target object in the current time period and the speed estimation value of the target object in the previous time period, determining the speed estimation value and the second Kalman gain value of the target object at the current moment, and updating the covariance matrix. Among them, the speed estimation value of the target object at the current moment can be determined by the following formula:

[0092] V k|k-1 = A * V k-1 ;

[0093] Where V k|k-1is the estimated value of the target object's speed at the current moment, A is the state transition matrix, and V k-1 is the estimated value of the target object's speed in the previous time period.

[0094] The second Kalman gain value can be determined by the following formula:

[0095] K' = V k|k-1 *(V k|k-1 + R') -1 ;

[0096] where K' is the second Kalman gain value and R' is the measurement noise matrix. The measurement noise matrix R is a matrix that can characterize the measurement accuracy of the front-end sensor;

[0097] The posterior covariance matrix can be determined by the following formula:

[0098] V k = (I - K'*H')*V k|k-1 ;

[0099] where V k is the posterior covariance matrix of the target object's speed, I is the identity matrix, and H' is the relationship matrix between the absolute speed of the target object in the current time period and the estimated value of the target object's speed in the previous time period.

[0100] When compensating the heading angle of the target object, the first heading angle of the target object is determined according to the estimated value of the speed at the current moment, and the second heading angle of the target object is determined by the pose detection sensor of the vehicle; the target heading angle is determined according to the first heading angle and the second heading angle.

[0101] Among them, since the first heading angle is determined according to the estimated value of the speed at the current moment and the second heading angle is determined by the pose detection sensor of the vehicle, the direction of the first heading angle is more accurate than that of the second heading angle, and the angular value of the second heading angle is more accurate than that of the first heading angle. If the direction of the second heading angle detected by the pose detection sensor of the vehicle is opposite to the direction of the first heading angle, the direction of the second heading angle is reversely adjusted. The advantage of such a setting is that it can improve the determination accuracy and precision of the target object's heading angle.

[0102] On the basis of the above embodiments, optionally, the state information includes at least one of the speed information, acceleration information, position information, heading angle information, size, or category of the target object.

[0103] The post - processing method for target perception and tracking provided by the embodiment of the present invention coordinates the state information of the target object relative to the vehicle into the geodetic coordinate system, which is beneficial for the vehicle to obtain real - time and accurate road information, determine accurate vehicle navigation route information, and determine the position information of the target object during driving; by compensating the movement of the target object through the timestamp error determined when obtaining the measurement values by vehicle positioning and the front - end sensor, the accuracy of target perception and tracking is further improved.

[0104] Embodiment III

[0105] Figure 4 It is a schematic structural diagram of a post - processing device for target perception and tracking provided by Embodiment III of the present invention. This device can execute the post - processing method for target perception and tracking provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. As Figure 4 shown, the device includes:

[0106] The current - moment measurement value acquisition module 410 is used to, when the front - end sensor of the vehicle senses a target object, acquire the current - moment measurement value of the target object's system track, and initialize the covariance matrix and score of the target object's system track;

[0107] The prior - estimation module 420 is used to determine the prior - estimation value of the target object's system track according to the measurement value, and update the prior covariance matrix;

[0108] The posterior - estimation module 430 is used to match the next - moment measurement value with the prior - estimation value, determine the first score and the posterior - estimation value of the system track according to the matching result and the initialized score, and update the posterior covariance matrix;

[0109] The state - information determination module 440 is used to, if the first score is greater than the first preset threshold, determine the state information of the target object; if the first score is less than the second preset threshold, delete the posterior - estimation value.

[0110] The post - processing device for target perception and tracking provided by the embodiment of the present invention, when the front - end sensor of the vehicle senses a target object, obtains the measured value of the target object's system track at the current moment, and initializes the covariance matrix and score of the target object's system track; determines the prior estimate value of the target object's system track according to the measured value, and updates the prior covariance matrix; matches the measured value at the next moment with the prior estimate value, determines the first score and the posterior estimate value of the system track according to the matching result and the initialized score, and updates the posterior covariance matrix; if the first score is greater than the first preset threshold, determines the state information of the target object; if the first score is less than the second preset threshold, deletes the posterior estimate value. In this technical solution, the prior estimate value and the posterior estimate value of the target object are determined through the measured values at the current moment and the next moment to realize the perception and tracking of the target object, and the life cycle of the system track is maintained by managing the score of the target object's system track, thereby improving the accuracy and reliability of the post - processing result of target perception and tracking.

[0111] Further, the state information includes at least one of the speed information, acceleration information, position information, heading angle information, size or category of the target object.

[0112] Further, matching the measured value at the next moment with the prior estimate value includes performing threshold matching and ID matching on the measured value and the prior estimate value;

[0113] Correspondingly, the posterior estimation module 430 includes:

[0114] The first score determination unit is used to perform ID matching on the measured value at the next moment and the prior estimate value within the threshold range. If the ID of the measured value is the same as the ID of the prior estimate value, it is marked as a successful match, and points are added according to a preset rule on the basis of the initialized score to determine the first score of the system track; otherwise, it is marked as a failed match, and points are deducted according to a preset rule on the basis of the initialized score to determine the first score of the system track;

[0115] Correspondingly, the posterior estimation module 430 includes:

[0116] The Kalman gain value determination unit is used to determine the Kalman gain value according to the measured value at the next moment, the prior estimate value, and the measurement noise matrix when it is marked as a successful match;

[0117] The posterior estimation unit is used to determine the posterior estimate value of the system track according to the prior estimate value, the measured value at the next moment, and the Kalman gain value.

[0118] Further, the current moment measurement value and the next moment are obtained in the Cartesian coordinate system established with the center of the vehicle's rear axle as the origin.

[0119] Further, the device further includes:

[0120] A vehicle positioning information acquisition module, configured to acquire the positioning information of the vehicle in the geodetic coordinate system when the front-end sensor of the vehicle senses a target object before acquiring the current moment measurement value of the target object system track;

[0121] Correspondingly, the state information determination module 440 includes:

[0122] A state information determination unit, configured to determine the state information of the target object according to the positioning information and the posteriori estimation value.

[0123] Further, the device further includes:

[0124] A timestamp acquisition module, configured to acquire a first timestamp corresponding to the positioning information and a second timestamp corresponding to the current moment measurement value after acquiring the positioning information of the vehicle in the geodetic coordinate system;

[0125] Correspondingly, the device further includes:

[0126] A motion compensation module, configured to perform motion compensation on the state information of the target object according to the first timestamp, the second timestamp, and the state information of the target object after outputting the state information of the target object if the first score is greater than a first preset threshold.

[0127] Further, the motion compensation module includes:

[0128] A target absolute velocity determination unit, configured to determine the absolute velocity of the target object in the current time period according to the first timestamp, the previous timestamp of the measurement value, the third timestamp of the next moment measurement value, and the position information corresponding to the first timestamp and the third timestamp respectively;

[0129] A target velocity estimation unit, configured to determine the velocity estimation value and the Kalman gain value of the target object at the current moment and update the covariance matrix based on a standard Kalman filter, the absolute velocity of the target object in the current time period, and the velocity estimation value of the target object at the previous moment;

[0130] A heading angle determination unit, configured to determine a first heading angle of the target object according to the velocity estimation value at the current moment and determine a second heading angle of the target object through the pose detection sensor of the vehicle;

[0131] A target heading angle unit for determining a target heading angle according to the first heading angle and the second heading angle.

[0132] The post-processing device for target perception and tracking provided by an embodiment of the present invention can execute the post-processing method for target perception and tracking provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution of the method.

[0133] Embodiment 4

[0134] Figure 5 FIG. shows a schematic structural diagram of an electronic device 10 that can be used to implement the embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device (such as a helmet, glasses, a watch, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present invention described herein and / or claimed.

[0135] As Figure 5 shown, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by the at least one processor. The processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0136] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0137] The processor 11 may be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the post-processing method for target perception tracking.

[0138] In some embodiments, the post-processing method for target perception tracking may be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the post-processing method for target perception tracking described above may be executed. Alternatively, in other embodiments, the processor 11 may be configured to execute the post-processing method for target perception tracking in any other suitable manner (e.g., by means of firmware).

[0139] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0140] The computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer programs are executed by the processor, the functions / operations specified in the flowchart and / or block diagram are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0141] Embodiment Five

[0142] In the context of the present invention, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0143] To provide for interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0144] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0145] A computing system may include a client and a server. The client and the server are generally far from each other and usually interact via a communication network. The client-server relationship is created by computer programs running on respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0146] It should be understood that various forms of the processes shown above can be used, steps can be reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0147] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A post - processing method for target perception and tracking, characterized in that, the method includes: When the front - end sensor of the vehicle senses a target object, obtain the measurement value of the target object's system track at the current moment, and initialize the covariance matrix and score of the target object's system track; Determine the prior estimate value of the target object's system track according to the measurement value, and update the prior covariance matrix; Match the measurement value at the next moment with the prior estimate value, determine the first score and the posterior estimate value of the system track according to the matching result and the initialized score, and update the posterior covariance matrix; If the first score is greater than the first preset threshold, determine the state information of the target object; if the first score is less than the second preset threshold, delete the posterior estimate value; Wherein, the current - moment measurement value and the next moment are obtained in the Cartesian coordinate system established with the center of the vehicle's rear axle as the origin; Before obtaining the measurement value of the target object's system track at the current moment, it includes: When the front - end sensor of the vehicle senses a target object, obtain the positioning information of the vehicle in the geodetic coordinate system; obtain the first timestamp corresponding to the positioning information, and the second timestamp corresponding to the current - moment measurement value; Correspondingly, if the first score is greater than the first preset threshold, determining the state information of the target object includes: Determine the state information of the target object according to the positioning information and the posterior estimate value; After determining the state information of the target object according to the positioning information and the posterior estimate value, it further includes: Determine the absolute speed of the target object in the current time period according to the first timestamp, the previous timestamp of the measurement value, the third timestamp of the measurement value at the next moment, and the position information corresponding to the first timestamp and the third timestamp respectively; based on the standard Kalman filter, and the absolute speed of the target object in the current time period and the speed estimate value of the target object at the previous moment, determine the speed estimate value and the Kalman gain value of the target object at the current moment, and update the covariance matrix; determine the first heading angle of the target object according to the speed estimate value at the current moment, and determine the second heading angle of the target object through the pose detection sensor of the vehicle; determine the target heading angle according to the first heading angle and the second heading angle.

2. The method according to claim 1, characterized in that, the state information includes at least one of the speed information, acceleration information, position information, heading - angle information, size or category of the target object.

3. The method according to claim 1, characterized in that, Matching the measurement value at the next moment with the prior estimate value includes performing threshold matching and ID matching on the measurement value and the prior estimate value; Correspondingly, matching the measurement value at the next moment with the prior estimate value, and determining the first score of the system track according to the matching result and the initialized score includes: Match the next moment measurement value with the prior estimate value within the threshold range. If the ID of the measurement value is the same as the ID of the prior estimate value, mark it as a successful match, and add points according to the preset rules based on the initial score to determine the first score of the system track; otherwise, mark it as a failed match, and subtract points according to the preset rules based on the initial score to determine the first score of the system track; Correspondingly, match the next moment measurement value with the prior estimate value, and determine the posterior estimate value of the system track according to the matching result and the initial score, including: When marked as a successful match, determine the Kalman gain value according to the next moment measurement value, the prior estimate value, and the measurement noise matrix; Determine the posterior estimate value of the system track according to the prior estimate value, the next moment measurement value, and the Kalman gain value.

4. A post-processing device for target perception and tracking, characterized in that, the device includes: A current moment measurement value acquisition module, configured to acquire the current moment measurement value of the target object system track when the front-end sensor of the vehicle senses the target object, and initialize the covariance matrix and score of the target object system track; A prior estimate module, configured to determine the prior estimate value of the target object system track according to the measurement value, and update the prior covariance matrix; A posterior estimate module, configured to match the next moment measurement value with the prior estimate value, determine the first score and the posterior estimate value of the system track according to the matching result and the initial score, and update the posterior covariance matrix; A state information determination module, configured to determine the state information of the target object if the first score is greater than the first preset threshold; if the first score is less than the second preset threshold, delete the posterior estimate value; wherein, the current moment measurement value and the next moment are acquired in the Cartesian coordinate system established with the center of the rear axle of the vehicle as the origin; wherein, the device further includes: A vehicle positioning information acquisition module, configured to acquire the positioning information of the vehicle in the geodetic coordinate system when the front-end sensor of the vehicle senses the target object before acquiring the current moment measurement value of the target object system track; Correspondingly, the state information determination module includes: A state information determination unit, configured to determine the state information of the target object according to the positioning information and the posterior estimate value; wherein, the device further includes: A timestamp acquisition module, configured to acquire a first timestamp corresponding to the positioning information and a second timestamp corresponding to the current moment measurement value after acquiring the positioning information of the vehicle in the geodetic coordinate system; Correspondingly, the device further includes: A motion compensation module, configured to perform motion compensation on the state information of the target object according to the first timestamp, the second timestamp, and the state information of the target object after outputting the state information of the target object if the first score is greater than the first preset threshold; wherein, the motion compensation module includes: A target absolute velocity determination unit, configured to determine the absolute velocity of the target object in the current time period according to the first timestamp, the previous timestamp of the measurement value, the third timestamp of the measurement value at the next moment, and the position information corresponding to the first timestamp and the third timestamp respectively; A target velocity estimation unit, configured to determine the velocity estimation value and the Kalman gain value of the target object at the current moment based on a standard Kalman filter, the absolute velocity of the target object in the current time period, and the velocity estimation value of the target object at the previous moment, and update the covariance matrix; A heading angle determination unit, configured to determine a first heading angle of the target object according to the velocity estimation value at the current moment, and determine a second heading angle of the target object through a pose detection sensor of the vehicle; A target heading angle unit, configured to determine a target heading angle according to the first heading angle and the second heading angle.

5. An electronic device, characterized in that, the electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the post-processing method for target perception and tracking according to any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, the computer-readable storage medium stores computer instructions, and the computer instructions are used to implement the post-processing method for target perception and tracking according to any one of claims 1-3 when executed by a processor.

Images