CN117031512A - Target object tracking method and device and electronic equipment - Google Patents
Target object tracking method and device and electronic equipment Download PDFInfo
- Publication number
- CN117031512A CN117031512A CN202311098186.4A CN202311098186A CN117031512A CN 117031512 A CN117031512 A CN 117031512A CN 202311098186 A CN202311098186 A CN 202311098186A CN 117031512 A CN117031512 A CN 117031512A
- Authority
- CN
- China
- Prior art keywords
- state quantity
- target object
- determining
- target
- kalman
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 101
- 238000001914 filtration Methods 0.000 claims abstract description 65
- 238000001514 detection method Methods 0.000 claims abstract description 63
- 230000008569 process Effects 0.000 claims abstract description 55
- 239000011159 matrix material Substances 0.000 claims abstract description 46
- 230000007704 transition Effects 0.000 claims abstract description 23
- 238000004590 computer program Methods 0.000 claims description 16
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 10
- 238000012937 correction Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 description 30
- 238000012544 monitoring process Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 8
- 230000003044 adaptive effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/38—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system
- G01S19/39—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system the satellite radio beacon positioning system transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/42—Determining position
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/13—Receivers
- G01S19/35—Constructional details or hardware or software details of the signal processing chain
- G01S19/37—Hardware or software details of the signal processing chain
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a target object tracking method, a target object tracking device and electronic equipment, wherein the method comprises the following steps: acquiring a historical state quantity of a target object at a historical moment, and predicting a predicted state quantity of the target object at the current moment according to the historical state quantity, a state transition matrix and a Kalman filtering system; determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system; the Kalman gain is determined based on the uncertainty and the observation noise of the Kalman system at the current moment, and the predicted state quantity is updated based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated predicted state quantity. The scheme of the embodiment of the invention solves the problem of inaccurate target tracking caused by the fact that the process noise and the observation noise are fixed values in the current Kalman filtering, realizes the self-adaptive adjustment of the noise, and improves the target tracking precision.
Description
Technical Field
The present invention relates to the field of target tracking technologies, and in particular, to a method and an apparatus for tracking a target object, and an electronic device.
Background
Target tracking is one of the hot spots of current autopilot domain research, and kalman filters are typically used in target tracking to fuse detection information with historical frame information. However, the performance of the traditional kalman filter is greatly affected by hardware noise and environmental noise, which can lead to the degradation and even divergence of the filter performance and seriously affect the accuracy of target tracking. Two key parameters describing noise in Kalman filtering, process noise and observation noise are often determined according to industry experience, and no change occurs in the iterative calculation process. Once unreasonable and determined, this can lead to a substantial degradation of overall tracking performance.
In view of this problem, there are some methods of attempting to adjust adaptive parameters during calculation, in which sliding windows are used for adjustment, but in this method, the size of the sliding window is an empirical value, and the size of the sliding window has a large influence on the final result. Alternatively, the measurement data of a plurality of sensors is used to perform noise statistics, thereby obtaining observation noise. However, for the automatic driving field, the observation noise can be basically obtained from the sensor firmware parameters, and the observation noise is less required to be obtained through a data statistics mode, so that the noise in the adjustment process is a more critical problem.
Therefore, how to adjust the process noise and the observation noise to improve the accuracy of the target tracking is a technical problem to be solved.
Disclosure of Invention
The invention provides a target object tracking method, a target object tracking device and electronic equipment, which are used for realizing self-adaptive adjustment of process noise and observation noise and improving the target tracking precision.
According to an aspect of the present invention, there is provided a tracking method of a target object, including:
acquiring a historical state quantity of a target object at a historical moment, and predicting a predicted state quantity of the target object at a current moment according to the historical state quantity, a state transition matrix and a Kalman filtering system;
determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system;
and determining a Kalman gain based on the uncertainty and the observation noise of the Kalman filtering system at the current moment, and updating the prediction state quantity based on the Kalman gain and the detection state quantity to determine tracking information of the target object based on the updated prediction state quantity.
According to another aspect of the present invention, there is provided a tracking apparatus of a target object, including:
the prediction state quantity determining module is used for obtaining the historical state quantity of the target object at the historical moment and predicting the prediction state quantity of the target object at the current moment according to the historical state quantity, the state transition matrix and the Kalman filtering system;
the uncertainty determining module is used for determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system;
the prediction state quantity updating module determines a Kalman gain based on uncertainty and observation noise of the Kalman filtering system at the current moment, and updates the prediction state quantity based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated prediction state quantity.
According to another aspect of the present invention, there is provided an electronic apparatus including:
at least one processor;
and a memory communicatively coupled to the at least one processor;
Wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of tracking a target object according to any one of the embodiments of the present invention.
According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to execute a tracking method of a target object according to any embodiment of the present invention.
According to the technical scheme, the historical state quantity of the target object at the historical moment is obtained, and the predicted state quantity of the target object at the current moment is predicted according to the historical state quantity, the state transition matrix and the Kalman filtering system; determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system; the Kalman gain is determined based on the uncertainty and the observation noise of the Kalman system at the current moment, and the predicted state quantity is updated based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated predicted state quantity. The problem of inaccurate target tracking caused by the fact that process noise and observation noise are fixed values in the current Kalman filtering is solved, self-adaptive adjustment of noise is achieved, and the target tracking accuracy is improved.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a tracking method for a target object according to a first embodiment of the present invention;
fig. 2 is a schematic structural diagram of a tracking device for a target object according to a third embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device implementing a tracking method of a target object according to a fourth embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
Fig. 1 is a flowchart of a method for tracking a target object according to an embodiment of the present invention, where the method may be performed by a tracking device for a target object, and the device may be implemented in hardware and/or software, and the device may be configured in a computer device. As shown in fig. 1, the method includes:
S110, acquiring a historical state quantity of a target object at a historical moment, and predicting a predicted state quantity of the target object at a current moment according to the historical state quantity, a state transition matrix and a Kalman filtering system.
In the embodiment of the present invention, the target object may be any object to be tracked, for example, in the field of automatic driving, the target object may be a target vehicle to be tracked, a pedestrian in front of a driving road, other movable obstacles, etc., and in the embodiment, the target object is merely illustrated, and the target object is not particularly limited. The historical moment refers to the moment above the current moment, for example, the current moment is T, the corresponding historical moment is T-1, and the historical state quantity refers to the physical state of the target object at the historical moment, wherein the physical state includes, but is not limited to, the speed, the coordinates, the size, the orientation and the like of the target object. The state transition matrix is a matrix for describing the state change of the target object, the Kalman filtering system refers to a system for predicting or estimating the state of the target object through a Kalman filtering algorithm, and the corresponding predicted state quantity is the physical state of the target object predicted by the Kalman filtering system at the current moment.
In the application scenario of the embodiment, if the target object is to be tracked, firstly, the historical state quantity of the target object at the historical moment is obtained, and the state quantity of the target object at the current moment is predicted by combining the state transition matrix through the kalman filter system, so as to obtain the predicted state quantity of the target object.
Based on the scheme, predicting the predicted state quantity of the target object at the current moment according to the historical state quantity, the state transition matrix and the Kalman filtering system comprises the following steps: and determining the product of the historical state quantity and the state transition matrix, and correcting the predicted state quantity of the target object at the current moment based on the product and the system correction. Specifically, the state transition matrix is multiplied by the historical state quantity of the target object at the historical moment, and the system correction quantity is added to obtain the predicted state quantity of the target object at the current moment.
And S120, determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system.
In this embodiment, the detection state quantity refers to a state quantity obtained by detecting a state of a target object at a current moment by a sensing device, the target matching distance refers to a matching distance between a predicted state quantity of the target object and the detection state quantity, the process noise of the kalman filter system refers to noise existing in the prediction process when the kalman system predicts the current state of the target object, and the uncertainty of the kalman filter system is used for representing the uncertainty of the kalman filter system.
On the basis of the above embodiment, the determining the target matching distance according to the predicted state quantity and the detected state quantity of the target object includes: determining a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used at the current moment, and constructing a matching matrix based on the matching distances to be used between the predicted state quantities to be used and the detection state quantities to be used; processing the matching matrix through a Hungary matching algorithm, and determining a predicted state quantity and a detection state quantity corresponding to the target object from a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used; and taking the matching distance to be used between the predicted state quantity and the detection state quantity as the target matching distance.
It will be appreciated that there may be a plurality of predicted state amounts and a plurality of detected state amounts at the present moment, so that a matching pair corresponding to the target object may be determined from the plurality of predicted state amounts and the detected state amounts, that is, the predicted state amounts and the detected state amounts of the target object may be determined by a hungarian matching method.
Specifically, calculating the matching distance between each predicted state quantity to be used and each detected state quantity to be used to obtain a plurality of matching distances to be used, filling the matching distances to be used into a matching matrix, processing the matching matrix through a Hungary matching algorithm, determining the predicted state quantity to be used and the detected state quantity to be used with the highest matching degree from the matching matrix, taking the predicted state quantity and the detected state quantity to be used corresponding to the target object as the predicted state quantity and the detected state quantity, and further taking the matching distance to be used between the predicted state quantity and the detected state quantity as the target matching distance.
On the basis of the above embodiment, the constructing a matching matrix based on the to-be-used matching distance between the to-be-used prediction state quantity and the to-be-used detection state quantity includes: calculating a plurality of matching distances to be used based on the size information of each object prediction frame and each object detection frame at the current moment and the coincidence information of the object prediction frames and the object detection frames; and constructing the matching matrix based on a plurality of the matching distances to be used.
In the implementation scenario of the present embodiment, the object detection frame and the object prediction frame may be circumscribed rectangular frames of the target object in the target tracking, the size information of the object prediction frame and the object detection frame refers to the length and width of the rectangular frames, and the overlapping information of the object prediction frame and the object detection frame refers to the overlapping area or overlapping volume between the rectangular frames.
In the process of calculating the matching distance to be used, the distance between the center points of the object prediction frame and the object detection frame, the sizes of the object detection frame and the object prediction frame, the two-dimensional overlapping area and the three-dimensional overlapping volume of the object prediction frame and the object detection frame, the orientation of the target object and the like are calculated, and then the calculated numerical values of the center point distance, the size of the frame, the overlapping area and the orientation are weighted and summed to obtain a result as the matching distance to be used. For another example, the matching distance between the detection frame and the prediction frame is calculated by the following formula (1):
Dis match ＝w 1 *Dis center +w 2 *Dis size +w 3 *Dis 2diou +w 4 *Dis 3diou +w 5 *Dis dir (1)
Wherein, distance is the center point distance, distance is the target frame size, 2diou is the overlapping area, 3diou is the overlapping volume, dis dio is the target orientation, dis match is the matching distance, and w is the calculation weight of each factor.
On the basis of the foregoing embodiment, the determining, based on the target matching distance and the process noise of the kalman filter system, the uncertainty of the kalman filter system at the current moment includes: determining a matching score based on the target matching distance, the matching score being used as a first weight of the process noise; and determining the uncertainty of the Kalman filtering system at the current moment according to the first weight, the process noise and the uncertainty of the Kalman filtering system at the historical moment.
In the embodiment of the invention, the matching score is a score value obtained by converting the target matching distance through a score conversion formula, the first weight is a weight value of process noise, and the first weight is changed along with the target matching distance; in practical application, after determining the target matching distance, converting the target matching distance into a matching score, further taking the matching score as a first weight of process noise, and further calculating the uncertainty of the Kalman system at the current moment through a formula by the aid of the first weight, the process noise and the uncertainty of the Kalman system at the historical moment.
In practical application, the calculation formula (2) defining the matching score is:
wherein Score match Is a matching score.
A time-varying adaptive weighting is performed on the process noise Q by alpha t And (3) representing.
α t ＝Score match (3)
On the basis, the calculation of the system uncertainty of the time Kalman filtering system is described, and the calculation is carried out through the following formula (4):
in the above formula, F is a state transition matrix, P' is a system uncertainty estimated at the current time, pt-1 is a system uncertainty at a historical time, μ is a system correction amount, and using the state transition matrix, the above formula (4) can obtain a prediction of the state quantity at the current time by using the state quantity Xt-1 at the previous time, and calculate the currently predicted system uncertainty at the same time.
S130, determining a Kalman gain based on uncertainty and observation noise of the Kalman filtering system at the current moment, and updating the prediction state quantity based on the Kalman gain and the detection state quantity to determine tracking information of the target object based on the updated prediction state quantity.
On the basis of the scheme, the method for determining the Kalman gain based on the uncertainty and the observation noise of the Kalman filtering system at the current moment comprises the following steps: and determining a second weight of the observed noise according to the state information of the sensor at the current moment, and calculating the Kalman gain based on the second weight, the observed noise and the uncertainty of the Kalman filtering system at the current moment.
The state information of the sensor at the current moment refers to the hardware state of the sensor, the second weight refers to the weight of the observation noise, and the Kalman gain can be calculated through the second weight, the observation noise and the uncertainty of the Kalman filtering system at the current moment. Illustratively, the Kalman gain K may be calculated by the following equation (5):
in the above formula, H is an observation matrix, and P' t Is of the system at the current momentUncertainty, R is observed noise, beta t Is to self-adaptively weight the time variation of the observation noise, because the observation noise R is determined according to the hardware performance of the sensor, in practical application, the hardware state of the sensor is monitored, if the monitoring value signal is normal, beta t The value of (2) is always kept to be 1, if the monitoring value signal is abnormal, the sensor hardware or the data transmission process is indicated to have problems, and beta is increased t The final state result would be more inclined to believe the predicted result if the value of (c) is 10.
Wherein Normal represents that the sensor monitoring value is Normal, and abNormal represents that the sensor monitoring value is abNormal.
In an embodiment of the present invention, the updating the predicted state quantity based on the kalman gain and the detected state quantity to determine tracking information of the target object based on the updated predicted state quantity includes: determining a difference value between a target detection state quantity and the prediction state quantity, and determining a state quantity to be updated based on a product of the difference value and the Kalman gain; and updating the predicted state quantity based on the state quantity to be updated, and taking the updated predicted state quantity as target tracking information of the target object at the current moment.
The target detection state quantity refers to a state quantity of a target object detected by the sensor, the state quantity to be updated refers to a value obtained by multiplying a difference value between the target detection state quantity and the predicted state quantity by Kalman gain, and the predicted state quantity is added to the state quantity to be updated to update the predicted state quantity.
In a preferred embodiment, after said determining a kalman gain based on uncertainty and observed noise of said kalman filter system at the current moment, further comprising: and updating the uncertainty of the Kalman filtering system at the current moment based on the Kalman gain to obtain the actual uncertainty of the Kalman filtering system at the current moment.
Illustratively, after the Kalman gain is determined, the predicted state quantity is updated and the actual uncertainty of the system is determined, as in equation (7).
Wherein Zt is a target detection state quantity, and Pt is the actual uncertainty of the system.
The key of the whole calculation flow is that the weight of the process noise is defined by utilizing the matching distance according to the matching characteristic, a large amount of extra calculation is not needed to be added, and the whole calculation flow is matched with the whole algorithm flow. The observation noise is adjusted by monitoring the sensor state, so that the process noise and the observation noise can be adaptively adjusted at the same time, and the final state estimation comprises the influence of the adaptive process noise and the observation noise.
According to the technical scheme, the historical state quantity of the target object at the historical moment is obtained, and the predicted state quantity of the target object at the current moment is predicted according to the historical state quantity, the state transition matrix and the Kalman filtering system; determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system; the Kalman gain is determined based on the uncertainty and the observation noise of the Kalman system at the current moment, and the predicted state quantity is updated based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated predicted state quantity. The problem of inaccurate target tracking caused by the fact that process noise and observation noise are fixed values in the current Kalman filtering is solved, self-adaptive adjustment of noise is achieved, and the target tracking accuracy is improved.
Example two
The second embodiment of the present invention is a method for tracking a target object, and the present embodiment is a preferred embodiment between the foregoing embodiments. The method comprises the following steps:
based on a mass-production automatic driving algorithm, the realization of multi-target tracking mainly comprises two parts of matching and filtering. According to the actual condition of the mass production scene, the matching part calculates the matching distance according to multiple factors, and the matching part comprises the following steps: center point distance, target frame size, 2diou,3diou, target orientation. Namely:
Dis match ＝w 1 *Dis center +w 2 *Dis size +w 3 *Dis 2diou +w 4 *Dis 3diou +w 5 *Dis dir
Where w is the calculated weight for each factor. And respectively calculating the results of the 5 indexes according to the prediction state and the detection result, and finally obtaining the matching distance between each detected obstacle in the matching matrix and the predicted obstacle in the track. Dis (Dis) match And the matching matrix is used for filling in a matching matrix, and a matching pair of a detection result and a prediction result is obtained by a Hungary matching method, so that the matching of the current frame detection result and the historical track is finally realized. Based on the calculation result, dis match The smaller the matching degree is, the higher the matching degree is, and the correction of the prediction noise can be fed back by using the distance result of the matching calculation.
After the matching is completed, the adaptive Kalman filtering of each frame is divided into two stages, namely a prediction stage and an update stage.
Prediction stage:
and using a state transition matrix, obtaining the prediction of the state quantity at the current moment by using the state quantity at the last moment, and simultaneously calculating the uncertainty of the currently predicted system.
Where F is the state transition matrix, P' is the system uncertainty calculated at the current time, and this calculation involves a time-varying adaptive weighting of the process noise Q, with alpha, in addition to the conventional t-1 time system uncertainty and the process noise Q t And (3) representing.
α t ＝Score match
The calculation formula defining the matching score is:
Score match a value closer to 0 indicates a higher degree of matching between the predicted result and the detected result, and thus can be determined from Score match And the noise of the prediction state is adjusted. Since the predicted result includes calculation of speed, acceleration, etc., these time domain information detection results cannot be outputted, and the specific gravity thereof can be reduced at the time of noise in the subsequent adaptive adjustment process. The final result matching distance of the matching calculation is directly processed in numerical value, no extra calculation is added, and other influences on the original algorithm are avoided. The regulation of the process noise is also met in a physical sense.
Updating:
the method mainly comprises the calculation of Kalman gain, the final calculation of state quantity at the current moment and the final calculation of system uncertainty at the current moment
The formula for the Kalman gain is as follows:
wherein R is observation noise, beta t The time-varying self-adaptive weighting of the observation noise is carried out, because the observation noise is determined according to the hardware performance of the sensor, the hardware state of the sensor is monitored in the automatic driving code of the red flag mass production, if the monitoring value signal is normal, the signal is beta t The value of (2) is always kept to be 1, if the monitoring value signal is abnormal, the sensor hardware or the data transmission process is indicated to have problems, and beta is increased t The final state result would be more inclined to believe the predicted result if the value of (c) is 10.
After the Kalman gain is determined, the final calculation result of the current frame state quantity and the system uncertainty is updated.
The key of the whole calculation flow is that the weight of the process noise is defined by utilizing the matching distance according to the matching characteristic, a large amount of extra calculation is not needed to be added, and the whole calculation flow is matched with the whole algorithm flow. The observation noise is adjusted by monitoring the sensor state, so that the process noise and the observation noise can be adaptively adjusted at the same time, and the final state estimation comprises the influence of the adaptive process noise and the observation noise.
The scheme of the embodiment of the invention improves the problem that the influence of process noise and observation noise cannot be adjusted according to the change of the situation in the traditional Kalman filter, and adjusts the weight parameters in a self-adaptive way. Based on a mass-production automatic driving framework, the weight calculation of the process noise is bound with the matching logic depth, a concept of a matching score is provided according to a matching calculation result, and then the weight of the process noise is determined by the height of the matching score without adding extra calculation. The self-adaptive adjustment process of decoupling process noise and observation noise changes the problem that only one noise can be adjusted when new information is introduced, and the self-adaptive weight of the observation noise is changed through monitoring signals of sensor hardware. Weighting parameters of a process noise matrix are measured by combining the matched calculation results; the weight of the observed noise is defined by the detection of the state of the sensor. Therefore, the problem that noise and observation noise influence cannot be adaptively adjusted in the filtering stage at the same time is solved, the performance of the filter is improved, and the accuracy and the effect of target tracking in a mass-produced automatic driving algorithm are further improved. And based on the flow and the signal quantity of the mass production algorithm, the practical application of the vehicle-mounted algorithm is easier to realize.
According to the technical scheme, the historical state quantity of the target object at the historical moment is obtained, and the predicted state quantity of the target object at the current moment is predicted according to the historical state quantity, the state transition matrix and the Kalman filtering system; determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system; the Kalman gain is determined based on the uncertainty and the observation noise of the Kalman system at the current moment, and the predicted state quantity is updated based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated predicted state quantity. The problem of inaccurate target tracking caused by the fact that process noise and observation noise are fixed values in the current Kalman filtering is solved, self-adaptive adjustment of noise is achieved, and the target tracking accuracy is improved.
Example III
Fig. 2 is a schematic structural diagram of a tracking device for a target object according to a third embodiment of the present invention. As shown in fig. 2, the apparatus includes:
a predicted state quantity determining module 310, configured to obtain a historical state quantity of a target object at a historical moment, and predict a predicted state quantity of the target object at a current moment according to the historical state quantity, a state transition matrix and a kalman filter system;
An uncertainty determining module 320, configured to determine a target matching distance according to the predicted state quantity and the detected state quantity of the target object, and determine an uncertainty of the kalman filter system at a current time based on the target matching distance and a process noise of the kalman filter system;
a predicted state quantity updating module 330, configured to determine a kalman gain based on an uncertainty of the kalman filter system and an observed noise at a current time, update the predicted state quantity based on the kalman gain and the detected state quantity, and determine tracking information of the target object based on the updated predicted state quantity.
According to the technical scheme, the historical state quantity of the target object at the historical moment is obtained, and the predicted state quantity of the target object at the current moment is predicted according to the historical state quantity, the state transition matrix and the Kalman filtering system; determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system; the Kalman gain is determined based on the uncertainty and the observation noise of the Kalman system at the current moment, and the predicted state quantity is updated based on the Kalman gain and the detection state quantity so as to determine tracking information of the target object based on the updated predicted state quantity. The problem of inaccurate target tracking caused by the fact that process noise and observation noise are fixed values in the current Kalman filtering is solved, self-adaptive adjustment of noise is achieved, and the target tracking accuracy is improved.
Optionally, the uncertainty determination module 320 includes:
the matrix construction module is used for determining a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used at the current moment, and constructing a matching matrix based on the matching distances to be used between the predicted state quantities to be used and the detection state quantities to be used;
the state quantity determining module is used for processing the matching matrix through a Hungary matching algorithm and determining a predicted state quantity and a detection state quantity corresponding to the target object from a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used;
and the matching distance determining module is used for taking the matching distance to be used between the predicted state quantity and the detection state quantity as the target matching distance.
Optionally, the matrix construction module includes:
the to-be-used matching distance calculating sub-module is used for calculating a plurality of to-be-used matching distances based on the size information of each object prediction frame and each object detection frame at the current moment and the coincidence information of the object prediction frames and the object detection frames;
and the matrix construction sub-module is used for constructing the matching matrix based on a plurality of the matching distances to be used.
Optionally, the uncertainty determination module 320 includes:
a first weight calculation sub-module for determining a matching score based on the target matching distance, the matching score being used as a first weight of the process noise;
and the uncertainty calculation submodule is used for determining the uncertainty of the Kalman filtering system at the current moment according to the first weight, the process noise and the uncertainty of the Kalman filtering system at the historical moment.
Optionally, the prediction state quantity updating module 330 includes:
and the Kalman gain calculation module is used for determining a second weight of the observed noise according to the state information of the sensor at the current moment, and calculating the Kalman gain based on the second weight, the observed noise and the uncertainty of the Kalman filtering system at the current moment.
Optionally, the prediction state quantity updating module 330 includes:
the state quantity to be updated determining module is used for determining a difference value between a target detection state quantity and the target prediction state quantity, and determining the state quantity to be updated based on a product of the difference value and the Kalman gain;
and the target tracking information determining module is used for updating the predicted state quantity based on the state quantity to be updated, and taking the updated predicted state quantity as target tracking information of the target object at the current moment.
Optionally, the prediction state quantity determining module 310 includes:
and the prediction module is used for determining the product of the historical state quantity and the state transition matrix, and correcting the predicted state quantity of the target object at the current moment based on the product and the system correction.
Optionally, the tracking device of the target object further includes:
and the actual uncertainty determining module is used for updating the uncertainty of the Kalman filtering system at the current moment based on the Kalman gain to obtain the actual uncertainty of the Kalman filtering system at the current moment.
The target object tracking device provided by the embodiment of the invention can execute the target object tracking method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Example IV
Fig. 3 is a schematic structural diagram of an electronic device implementing a tracking method of a target object according to a fourth embodiment of the present invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 3, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform 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 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.
Various 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, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the respective methods and processes described above, for example, a tracking method of the target object.
In some embodiments, the method of tracking a target object may be implemented as a computer program, which is tangibly embodied on 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 tracking method of the target object described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the tracking method of the target object by any other suitable means (e.g. by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present invention, a computer-readable storage medium may 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 may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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.
To provide for interaction with a user, the systems and techniques described here 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 a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end 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: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method of tracking a target object, comprising:
acquiring a historical state quantity of a target object at a historical moment, and predicting a predicted state quantity of the target object at a current moment according to the historical state quantity, a state transition matrix and a Kalman filtering system;
determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system;
And determining a Kalman gain based on the uncertainty and the observation noise of the Kalman system at the current moment, and updating the predicted state quantity based on the Kalman gain and the detection state quantity to determine tracking information of the target object based on the updated predicted state quantity.
2. The method of claim 1, wherein said determining a target matching distance from said predicted state quantity and said detected state quantity of said target object comprises:
determining a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used at the current moment, and constructing a matching matrix based on the matching distances to be used between the predicted state quantities to be used and the detection state quantities to be used;
processing the matching matrix through a Hungary matching algorithm, and determining a predicted state quantity and a detection state quantity corresponding to the target object from a plurality of predicted state quantities to be used and a plurality of detection state quantities to be used;
and taking the matching distance to be used between the predicted state quantity and the detection state quantity as the target matching distance.
3. The method of claim 2, wherein the constructing a matching matrix based on the to-be-used matching distance between the to-be-used predicted state quantity and the to-be-used detected state quantity comprises:
Calculating a plurality of matching distances to be used based on the size information of each object prediction frame and each object detection frame at the current moment and the coincidence information of the object prediction frames and the object detection frames;
and constructing the matching matrix based on a plurality of the matching distances to be used.
4. The method of claim 1, wherein the determining uncertainty of the kalman filter system at a current time based on the target matching distance and a process noise of the kalman filter system comprises:
determining a matching score based on the target matching distance, the matching score being used as a first weight of the process noise;
and determining the uncertainty of the Kalman filtering system at the current moment according to the first weight, the process noise and the uncertainty of the Kalman filtering system at the historical moment.
5. The method of claim 1, wherein the determining a kalman gain based on uncertainty and observed noise of the kalman filter system at a current time comprises:
and determining a second weight of the observed noise according to the state information of the sensor at the current moment, and calculating the Kalman gain based on the second weight, the observed noise and the uncertainty of the Kalman filtering system at the current moment.
6. The method according to claim 1, wherein the updating the predicted state quantity based on the kalman gain and the detected state quantity to determine tracking information of the target object based on the updated predicted state quantity includes:
determining a difference value between a target detection state quantity and the target prediction state quantity, and determining a state quantity to be updated based on a product of the difference value and the Kalman gain;
and updating the predicted state quantity based on the state quantity to be updated, and taking the updated predicted state quantity as target tracking information of the target object at the current moment.
7. The method according to claim 1, wherein predicting the predicted state quantity of the target object at the current time based on the historical state quantity, the state transition matrix, and the kalman filter system comprises:
and determining the product of the historical state quantity and the state transition matrix, and correcting the predicted state quantity of the target object at the current moment based on the product and the system correction.
8. The method of claim 1, further comprising, after said determining a kalman gain based on uncertainty and observed noise of said kalman filter system at a current time,:
And updating the uncertainty of the Kalman filtering system at the current moment based on the Kalman gain to obtain the actual uncertainty of the Kalman filtering system at the current moment.
9. A tracking device for a target object, comprising:
the prediction state quantity determining module is used for obtaining the historical state quantity of the target object at the historical moment and predicting the prediction state quantity of the target object at the current moment according to the historical state quantity, the state transition matrix and the Kalman filtering system;
the uncertainty determining module is used for determining a target matching distance according to the predicted state quantity and the detection state quantity of the target object, and determining uncertainty of the Kalman filtering system at the current moment based on the target matching distance and process noise of the Kalman filtering system;
and the prediction state quantity updating module is used for determining Kalman gain based on uncertainty and observation noise of the Kalman filtering system at the current moment, updating the prediction state quantity based on the Kalman gain and the detection state quantity, and determining tracking information of the target object based on the updated prediction state quantity.
10. An electronic device, the electronic device comprising:
At least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the tracking method of the target object of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311098186.4A CN117031512A (en) | 2023-08-29 | 2023-08-29 | Target object tracking method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311098186.4A CN117031512A (en) | 2023-08-29 | 2023-08-29 | Target object tracking method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117031512A true CN117031512A (en) | 2023-11-10 |
Family
ID=88626473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311098186.4A Pending CN117031512A (en) | 2023-08-29 | 2023-08-29 | Target object tracking method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117031512A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117765026A (en) * | 2023-12-21 | 2024-03-26 | 北京斯年智驾科技有限公司 | Target tracking method, system, device and storage medium |
-
2023
- 2023-08-29 CN CN202311098186.4A patent/CN117031512A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117765026A (en) * | 2023-12-21 | 2024-03-26 | 北京斯年智驾科技有限公司 | Target tracking method, system, device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117031512A (en) | Target object tracking method and device and electronic equipment | |
| CN112528927B (en) | Confidence determining method based on track analysis, road side equipment and cloud control platform | |
| CN113469025A (en) | Target detection method and device applied to vehicle-road cooperation, road side equipment and vehicle | |
| CN114677655A (en) | Multi-sensor target detection method and device, electronic equipment and storage medium | |
| CN116740355A (en) | Automatic driving image segmentation method, device, equipment and storage medium | |
| CN114771283A (en) | Crawling control method and device, electric vehicle and storage medium | |
| CN114677653A (en) | Model training method, vehicle key point detection method and corresponding devices | |
| CN113688920A (en) | Model training and target detection method and device, electronic equipment and road side equipment | |
| CN116794619A (en) | Radar debugging processing method and device, electronic equipment and storage medium | |
| CN116311893A (en) | Traffic jam time prediction method and device, electronic equipment and storage medium | |
| CN115235527B (en) | Sensor external parameter calibration method and device and electronic equipment | |
| CN115546597A (en) | Sensor fusion method, device, equipment and storage medium | |
| CN114842305A (en) | Depth prediction model training method, depth prediction method and related device | |
| CN117346782B (en) | Positioning optimization method, device, electronic equipment and storage medium | |
| CN117289686B (en) | Parameter calibration method and device, electronic equipment and storage medium | |
| CN117872346A (en) | Object tracking method, device, equipment and storage medium | |
| CN115755019A (en) | Target tracking method, device, equipment and medium | |
| CN117864172B (en) | Automatic driving control method, device and equipment | |
| CN114694138B (en) | Road surface detection method, device and equipment applied to intelligent driving | |
| CN114049615B (en) | Traffic object fusion association method and device in driving environment and edge computing equipment | |
| CN116520255A (en) | Filtering method | |
| CN115096328B (en) | Positioning method and device of vehicle, electronic equipment and storage medium | |
| CN117725834A (en) | Gradient estimation parameter optimization method and device, electronic equipment and storage medium | |
| CN117808848B (en) | Identification tracking method and device, electronic equipment and storage medium | |
| CN116338727A (en) | Point cloud information processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |