WO2021163846A1

WO2021163846A1 - Target tracking method and target tracking apparatus

Info

Publication number: WO2021163846A1
Application number: PCT/CN2020/075556
Authority: WO
Inventors: 周鹏; 冯源; 张欢; 李选富; 吴祖光; 李维
Original assignee: 华为技术有限公司
Priority date: 2020-02-17
Filing date: 2020-02-17
Publication date: 2021-08-26
Also published as: CN115039095A

Abstract

A target tracking method and a target tracking apparatus in the field of artificial intelligence. The method comprises: an actual observation quantity of a moving body captured by a sensor is obtained; if a capture time of the actual observation quantity is earlier than an update time of a current state quantity of the moving body, an estimated observation quantity is obtained according to a state transition model for the moving body; updating is performed on the current state quantity according to the actual observation quantity and the estimated observation quantity, and the updated state quantity serves as a state quantity corresponding to the update time. The present invention is able to implement target tracking when capture times of observation quantities are in the wrong order, preventing state quantity time stamps from jumping around in an incorrect order due to observation quantity time stamps becoming out of order.

Description

Target tracking method and target tracking device

Technical field

The embodiments of the present application relate to the field of artificial intelligence, and more specifically, to a target tracking method and a target tracking device.

Background technique

The target tracking method can be applied to many scenarios, such as target detection, lane line detection, positioning and so on. Common target tracking methods need to use sensor measurement data combined with the state transition model of the tracked object to perform Bayesian estimation. For example, Bayes filters can be used to achieve target tracking. Specifically, the measurement data of the sensor is input to the Bayesian filter to obtain an estimate of the state of the tracked object. However, in the case where there are multiple sensors, due to transmission paths and other reasons, there may be a time when the measurement data of one sensor reaches the Bayes filter later than the time when it should reach the Bayes filter. Fig. 1 is a schematic diagram of the measurement data input to the Bayesian filter appearing out of order in time. The abscissa in Figure 1 shows the real time. The data transmitted by the sensor includes a timestamp (header: Xms) and measurement data (data). The timestamp is used to indicate that the measurement or collection time of the measurement data is Xms. time. The measurement data measured by sensor 0 at 1ms is input into the Bayesian filter at 3ms, and the measurement data measured by sensor 1 at 0ms is input into the Bayesian filter at 4ms, resulting in the time to input the measurement data of the Bayesian filter. Out of order. In this case, the Bayesian filter first outputs the state quantity calculated based on the measurement data measured at 1ms with a time stamp of 1ms, and then outputs the state calculated based on the measurement data measured at 0ms with a time stamp of 0ms As a result, the time stamp of the output state quantity jumps with the change of the time stamp of the measurement data.

There are usually two ways to solve the above problems. One method is to determine whether the time stamp of the measurement data is later than the time stamp of the latest updated state quantity in the Bayesian filter, and only the measurement data whose collection time is later than the update time of the latest state quantity is input to the Bayesian filter. Yes filter. However, this method will directly cause the lack of observations, leading to a decrease in the confidence of the output results. Another method is to extrapolate the late observation data to the current moment. However, extrapolation in the observation space relies on strong assumptions and tends to enlarge errors.

Summary of the invention

The present application provides a target tracking method and a target tracking device, which can realize target tracking when the acquisition time of the observation is out of order, and avoid the out-of-order jump of the time stamp of the state quantity with the out-of-order change of the time stamp of the observation. .

In the first aspect, a target tracking method is provided, which includes: acquiring the actual observation of the moving object collected by the sensor; in the case that the acquisition time of the actual observation is earlier than the update time of the current state of the moving object, according to the moving object The corresponding state transition model obtains the estimated observation; the current state quantity is updated according to the actual observation and the estimated observation, and the updated state quantity is used as the state quantity corresponding to the update moment.

Observation can include the speed of the moving object, the acceleration of the moving object, or the position of the moving object. For example, the observation may include the position and/or speed of the moving object measured by the millimeter wave radar.

The state quantity may include the speed of the moving object, the acceleration of the moving object, or the position of the moving object.

The update time of the current state quantity of the moving object can be understood as the time when the state quantity of the moving object was updated last time.

Optionally, the state transition model of the moving object includes a kinematics model of the moving object. For example, the kinematics model may include a uniform linear motion model and the like.

According to the solution of the embodiment of the present application, the state transition is performed according to the state transition model corresponding to the moving object, and then the current state quantity is updated, so that a more accurate state quantity corresponding to the latest update time can be obtained. At the same time, the updated state quantity is the state quantity corresponding to the latest update time, rather than the state quantity corresponding to the collection time, so as to avoid the out-of-order jump of the output state quantity in time.

In combination with the first aspect, in some implementations of the first aspect, obtaining estimated observations according to the state transition model corresponding to the moving object includes: obtaining the likelihood of the observation corresponding to the acquisition time according to the state transition model corresponding to the moving object Function, the estimated observation is determined according to the likelihood function of the observation corresponding to the acquisition time, the state transition model is determined according to the first probability density function, the first probability density function is to transfer the current state quantity to the corresponding acquisition time The probability density function of the state quantity.

In combination with the first aspect, in some implementations of the first aspect, the current state quantity is updated based on actual observations and estimated observations, and the updated state quantity is used as the state quantity corresponding to the update time, including: The likelihood function of the observation corresponding to the time updates the probability density function of the current state quantity; the state quantity corresponding to the update time is determined according to the probability density function of the updated state quantity.

In combination with the first aspect, in some implementations of the first aspect, the likelihood function of the observation corresponding to the acquisition time satisfies:

g(z _k-1 |X _k ')=∫g(z _k-1 |X)f _k-1|k (X|X _k ')dX

Among them, g(z _k-1 |X _k ') represents the likelihood function of the observation z _k-1 _{corresponding to the time t k-1} , X represents the state quantity updated at the time _{t k-1} _{, and X k} 'represents the missing state quantity update time t _K in the case of Z _{_k-1, g (z k-1 |} X) Z represents Z _k-1 _k-1 is used on the time t where _k-1 is updated state quantity Likelihood function, f _k-1|k (X|X _k ') is the first probability density function, the first probability density function is to transfer the state quantity X _k ' _{updated at t k to} _{update at t k-1} The probability density function of the state quantity.

In combination with the first aspect, in some implementations of the first aspect, the probability density function of the updated state quantity satisfies:

Among them, f _k|k (X _k |z ^k ) represents the probability density function of the state quantity X _k _{updated at time t k} _{, and f'k|k} (X _k '|z _k ,z ^k-2 ) represents the absence of z In the case of _k-1 , the probability density function of the state quantity X _k _{'updated at time t k} _{, z k} represents the observations collected at time _{t k} ^{, and z k} represents the set of observations collected at k time {z ₁ , z ₂ ,...,z _k-2 ,z _k-1 ,z _k }, z ^k-2 represents the set of observations collected at k-2 moments {z ₁ ,z ₂ ,...,z _k-2 }.

In combination with the first aspect, in some implementations of the first aspect, the current state quantity is updated based on the actual observations and estimated observations, and the updated state quantity is used as the state quantity corresponding to the update time, including:

Determine the expectation of the state quantity corresponding to the update time according to the actual observations and estimated observations;

The expectation of the state quantity corresponding to the update time is regarded as the state quantity corresponding to the update time;

Among them, the expectation of the state quantity corresponding to the update time is related to the Kalman gain value. The Kalman gain value is related to the first covariance, the observation matrix at the acquisition time, the covariance of the observation matrix, and the variance of the observation at the acquisition time. A covariance refers to the covariance of the state quantity transferred from the update time to the collection time.

In combination with the first aspect, in some implementations of the first aspect, the Kalman gain value satisfies:

Var(z _k-1 ) satisfies:

P _k-1|k satisfies:

Wherein, P _{k-1 | k} represents t _K transferred from the update time to the acquisition time t covariance _k-1 state quantity, P _{'k | k} represents the lack Concept t _k-1 time captured measurement Z _k-1 In the case of t _k , the covariance of the state quantity updated at time _{t k, F k-1|k} represents the state transition matrix from t _k to t _k-1 _{, Q k} represents the covariance of the prediction matrix, H _k-1 Represents the observation matrix at time _{t k-1} _{, Var(z k-1} ) represents the variance of the observation z _k-1 _{collected at time t k-1} , and R _k-1 represents the covariance of the observation matrix.

In combination with the first aspect, in some implementations of the first aspect, the expectation of the state quantity corresponding to the update time meets:

Among them, x _k|k represents the expectation of the state quantity updated at time _{t k,}

Represents the estimated observation at time _{t k-1.}

In combination with the first aspect, in some implementations of the first aspect, the estimated value of the observation corresponding to the collection time satisfies:

in,

Represents t estimated observations _k-1 time point, H _k-1 represents a t the observation matrix _k-1 time, F _{k-1 | k} indicates a transition from t _K time to t _k-1 time of the state transition matrix, x ' _{k | k} represents the concept of a lack of time t _k-1 measurement state quantity acquired in the case where time t _k z _k-1 update is desired.

In combination with the first aspect, in some implementations of the first aspect, in the case where the acquisition time of the actual observation is earlier than the update time of the current state of the moving object, the estimated observation is obtained according to the state transition model corresponding to the moving object , Including: when the acquisition time of the observation is earlier than the update time of the current state quantity of the moving object, and the time difference between the acquisition time and the update time is less than or equal to the threshold value, the estimated observation is obtained according to the state transition model corresponding to the moving object quantity.

According to the solution of the embodiment of the present application, the state transition is performed only when the time difference between the collection time and the update time is less than or equal to the threshold, and then the state quantity is updated, which can avoid the use of This observation updates the state quantity, which reduces the confidence of the updated state quantity.

In a second aspect, a target tracking device is provided, which includes an acquisition module and a processing module, wherein the acquisition module is used to: acquire the actual observation of a moving object collected by the sensor; the processing module is used to: early in the acquisition time of the actual observation In the case of the update time of the current state quantity of the moving object, the estimated observation quantity is obtained according to the state transition model corresponding to the moving object; the current state quantity is updated according to the actual observation quantity and the estimated observation quantity, and the updated state quantity is used as the update The state quantity corresponding to the moment.

In conjunction with the second aspect, in some implementations of the second aspect, the processing module is used to obtain the likelihood function of the observation corresponding to the acquisition time according to the state transition model corresponding to the moving object, and the estimated observation is based on the acquisition time The likelihood function of the corresponding observation is determined, and the state transition model is determined according to the first probability density function. The first probability density function is the probability density function that transfers the current state quantity to the state quantity corresponding to the collection time.

With reference to the second aspect, in some implementations of the second aspect, the processing module is used to: update the probability density function of the current state quantity according to the likelihood function of the observation corresponding to the acquisition time; according to the updated state quantity The probability density function of determines the state quantity corresponding to the update moment.

In combination with the second aspect, in some implementations of the second aspect, the likelihood function of the observation corresponding to the acquisition time satisfies:

g(z _k-1 |X _k ')=∫g(z _k-1 |X)f _k-1|k (X|X _k ')dX

In combination with the second aspect, in some implementations of the second aspect, the probability density function of the updated state quantity satisfies:

In combination with the second aspect, in some implementations of the second aspect, the processing module is used to: determine the expectation of the state quantity corresponding to the update time according to the actual observation and the estimated observation; and use the expectation of the state quantity corresponding to the update time as the update The state quantity corresponding to the time; among them, the expectation of the state quantity corresponding to the update time is related to the Kalman gain value, the Kalman gain value and the first covariance, the observation matrix at the acquisition time, the covariance of the observation matrix and the observation at the acquisition time The variance of the quantity is related, and the first covariance refers to the covariance of the state quantity transferred from the update time to the collection time.

In combination with the second aspect, in some implementations of the second aspect, the Kalman gain value satisfies:

Var(z _k-1 ) satisfies:

P _k-1|k satisfies:

In combination with the second aspect, in some implementations of the second aspect, the expectation of the state quantity corresponding to the update time meets:

Represents the estimated observation at time _{t k-1.}

In combination with the second aspect, in some implementations of the second aspect, the estimated value of the observation corresponding to the collection time satisfies:

in,

With reference to the second aspect, in some implementations of the second aspect, the processing module is used to: the actual observation measurement acquisition time is earlier than the update time of the current state quantity of the moving object, and the time difference between the acquisition time and the update time When the value is less than or equal to the threshold, the estimated observation is obtained according to the state transition model corresponding to the moving object.

It should be understood that the expansion, limitation, explanation and description of the related content in the above-mentioned first aspect are also applicable to the same content in the second aspect.

In a third aspect, a target tracking device is provided. The device includes: a memory for storing a program; a processor for executing the program stored in the memory, and when the program stored in the memory is executed, the processor is configured to execute the first aspect In the method.

In a fourth aspect, a computer program product is provided. The computer program product includes: computer program code, which when the computer program product runs on a computer, causes the computer to execute the method in the first aspect.

In a fifth aspect, a computer-readable storage medium is provided, the computer-readable storage medium stores a computer program, and when the computer program runs on a computer, the computer executes the method in the first aspect.

It should be understood that, in this application, the method of the first aspect may specifically refer to the first aspect and a method in any one of the various implementation manners of the first aspect.

Description of the drawings

Figure 1 is a schematic diagram of measurement data input into Bayesian filtering;

Fig. 2 is a schematic structural diagram of a vehicle provided by an embodiment of the present application;

Fig. 3 is a schematic structural diagram of a computer system provided by an embodiment of the present application;

FIG. 4 is a schematic diagram of the application of a cloud-side command automatic driving vehicle provided by an embodiment of the present application;

FIG. 5 is a schematic structural diagram of a target tracking device provided by an embodiment of the present application;

FIG. 6 is a schematic flowchart of a target tracking method provided by an embodiment of the present application;

FIG. 7 is a schematic structural diagram of another target tracking device provided by an embodiment of the present application;

FIG. 8 is a schematic structural diagram of another target tracking device provided by an embodiment of the present application.

Detailed ways

The technical solution in this application will be described below in conjunction with the accompanying drawings.

Fig. 2 is a functional block diagram of a vehicle 100 provided by an embodiment of the present application.

Among them, the vehicle 100 may be a manually driven vehicle, or the vehicle 100 may be configured in a fully or partially automatic driving mode.

In one example, the vehicle 100 can control its own vehicle while in the automatic driving mode, and can determine the current state of the vehicle and its surrounding environment through human operations, determine the possible behavior of at least one other vehicle in the surrounding environment, and The confidence level corresponding to the possibility of other vehicles performing possible behaviors is determined, and the vehicle 100 is controlled based on the determined information. When the vehicle 100 is in the automatic driving mode, the vehicle 100 can be placed to operate without human interaction.

The vehicle 100 may include various subsystems, such as a traveling system 110, a sensing system 120, a control system 130, one or more peripheral devices 140 and a power supply 160, a computer system 150, and a user interface 170.

Alternatively, the vehicle 100 may include more or fewer subsystems, and each subsystem may include multiple elements. In addition, each of the subsystems and elements of the vehicle 100 may be wired or wirelessly interconnected.

Illustratively, the travel system 110 may include components for providing power movement to the vehicle 100. In one embodiment, the travel system 110 may include an engine 111, a transmission 112, an energy source 113, and wheels 114/tires. The engine 111 may be an internal combustion engine, an electric motor, an air compression engine, or other types of engine combinations; for example, a hybrid engine composed of a gasoline engine and an electric motor, or a hybrid engine composed of an internal combustion engine and an air compression engine. The engine 111 can convert the energy source 113 into mechanical energy.

Exemplarily, the energy source 113 may include gasoline, diesel, other petroleum-based fuels, propane, other compressed gas-based fuels, ethanol, solar panels, batteries, and other power sources. The energy source 113 may also provide energy for other systems of the vehicle 100.

Exemplarily, the transmission device 112 may include a gearbox, a differential, and a drive shaft; wherein, the transmission device 112 may transmit mechanical power from the engine 111 to the wheels 114.

In an embodiment, the transmission device 112 may also include other devices, such as a clutch. Among them, the drive shaft may include one or more shafts that can be coupled to one or more wheels 114.

Exemplarily, the sensing system 120 may include several sensors that sense information about the environment around the vehicle 100.

For example, the sensing system 120 may include a positioning system 121 (for example, a GPS system, a Beidou system or other positioning systems), an inertial measurement unit 122 (IMU), a radar 123, a laser rangefinder 124, and a camera 125. The sensing system 120 may also include sensors of the internal system of the monitored vehicle 100 (for example, an in-vehicle air quality monitor, a fuel gauge, an oil temperature gauge, etc.). Sensor data from one or more of these sensors can be used to detect objects and their corresponding characteristics (position, shape, direction, speed, etc.). Such detection and identification are key functions for the safe operation of the autonomous vehicle 100.

Among them, the positioning system 121 can be used to estimate the geographic location of the vehicle 100. The IMU 122 may be used to sense changes in the position and orientation of the vehicle 100 based on inertial acceleration. In an embodiment, the IMU 122 may be a combination of an accelerometer and a gyroscope.

Exemplarily, the radar 123 may use radio signals to sense objects in the surrounding environment of the vehicle 100. In some embodiments, in addition to sensing the object, the radar 123 may also be used to sense the speed and/or direction of the object.

Exemplarily, the laser rangefinder 124 may use laser light to sense objects in the environment where the vehicle 100 is located. In some embodiments, the laser rangefinder 124 may include one or more laser sources, laser scanners, and one or more detectors, as well as other system components.

Exemplarily, the camera 125 may be used to capture multiple images of the surrounding environment of the vehicle 100. For example, the camera 125 may be a still camera or a video camera.

As shown in FIG. 2, the control system 130 controls the operation of the vehicle 100 and its components. The control system 130 may include various elements, such as a steering system 131, a throttle 132, a braking unit 133, a computer vision system 134, a route control system 135, and an obstacle avoidance system 136.

Illustratively, the steering system 131 may be operated to adjust the forward direction of the vehicle 100. For example, it may be a steering wheel system in one embodiment. The throttle 132 may be used to control the operating speed of the engine 111 and thereby control the speed of the vehicle 100.

Illustratively, the braking unit 133 may be used to control the deceleration of the vehicle 100; the braking unit 133 may use friction to slow down the wheels 114. In other embodiments, the braking unit 133 may convert the kinetic energy of the wheels 114 into electric current. The braking unit 133 may also take other forms to slow down the rotation speed of the wheels 114 to control the speed of the vehicle 100.

As shown in FIG. 2, the computer vision system 134 may be operable to process and analyze the images captured by the camera 125 in order to identify objects and/or features in the surrounding environment of the vehicle 100. The aforementioned objects and/or features may include traffic signals, road boundaries and obstacles. The computer vision system 134 may use object recognition algorithms, structure from motion (SFM) algorithms, video tracking, and other computer vision technologies. In some embodiments, the computer vision system 134 may be used to map the environment, track objects, estimate the speed of objects, and so on.

Illustratively, the route control system 135 may be used to determine the travel route of the vehicle 100. In some embodiments, the route control system 135 may combine data from sensors, GPS, and one or more predetermined maps to determine a travel route for the vehicle 100.

As shown in FIG. 2, the obstacle avoidance system 136 may be used to identify, evaluate, and avoid or otherwise cross potential obstacles in the environment of the vehicle 100.

In one example, the control system 130 may additionally or alternatively include components other than those shown and described. Alternatively, a part of the components shown above may be reduced.

As shown in FIG. 2, the vehicle 100 can interact with external sensors, other vehicles, other computer systems, or users through a peripheral device 140; wherein, the peripheral device 140 can include a wireless communication system 141, an onboard computer 142, a microphone 143 and/ Or speaker 144.

In some embodiments, the peripheral device 140 may provide a means for the vehicle 100 to interact with the user interface 170. For example, the onboard computer 142 may provide information to the user of the vehicle 100. The user interface 116 can also operate the onboard computer 142 to receive user input; the onboard computer 142 can be operated through a touch screen. In other cases, the peripheral device 140 may provide a means for the vehicle 100 to communicate with other devices located in the vehicle. For example, the microphone 143 may receive audio (eg, voice commands or other audio input) from the user of the vehicle 100. Similarly, the speaker 144 may output audio to the user of the vehicle 100.

As shown in FIG. 2, the wireless communication system 141 may wirelessly communicate with one or more devices directly or via a communication network. For example, the wireless communication system 141 can use 3G cellular communication; for example, code division multiple access (CDMA), EVD0, global system for mobile communications (GSM)/general packet radio service (general packet radio service) packet radio service, GPRS), or 4G cellular communication, such as long term evolution (LTE); or, 5G cellular communication. The wireless communication system 141 can communicate with a wireless local area network (WLAN) by using wireless Internet access (WiFi).

In some embodiments, the wireless communication system 141 may directly communicate with the device using an infrared link, Bluetooth, or ZigBee; other wireless protocols, such as various vehicle communication systems, for example, the wireless communication system 141 may include one or Multiple dedicated short range communications (DSRC) devices, these devices may include public and/or private data communications between vehicles and/or roadside stations.

As shown in FIG. 2, the power supply 160 may provide power to various components of the vehicle 100. In one embodiment, the power source 160 may be a rechargeable lithium ion or lead-acid battery. One or more battery packs of such batteries may be configured as a power source to provide power to various components of the vehicle 100. In some embodiments, the power source 160 and the energy source 113 may be implemented together, such as in some all-electric vehicles.

Exemplarily, part or all of the functions of the vehicle 100 may be controlled by the computer system 150, where the computer system 150 may include at least one processor 151, and the processor 151 is executed in a non-transitory computer readable medium stored in the memory 152, for example. The instruction 153. The computer system 150 may also be multiple computing devices that control individual components or subsystems of the vehicle 100 in a distributed manner.

For example, the processor 151 may be any conventional processor, such as a commercially available CPU.

Alternatively, the processor may be a dedicated device such as an ASIC or other hardware-based processor. Although FIG. 2 functionally illustrates the processor, the memory, and other elements of the computer in the same block, those of ordinary skill in the art should understand that the processor, computer, or memory may or may not actually include Multiple processors, computers or memories in the same physical enclosure. For example, the memory may be a hard disk drive or other storage medium located in a housing other than the computer. Therefore, a reference to a processor or computer will be understood to include a reference to a collection of processors or computers or memories that may or may not operate in parallel. Rather than using a single processor to perform the steps described here, some components such as steering components and deceleration components may each have its own processor that only performs calculations related to component-specific functions .

In the various aspects described herein, the processor may be located away from the vehicle and wirelessly communicate with the vehicle. In other aspects, some of the processes described herein are executed on a processor disposed in the vehicle and others are executed by a remote processor, including taking the necessary steps to perform a single manipulation.

In some embodiments, the memory 152 may contain instructions 153 (eg, program logic), which may be executed by the processor 151 to perform various functions of the vehicle 100, including those functions described above. The memory 152 may also contain additional instructions, for example, including sending data to, receiving data from, interacting with, and/or performing data to one or more of the traveling system 110, the sensing system 120, the control system 130, and the peripheral device 140. Control instructions.

Exemplarily, in addition to the instructions 153, the memory 152 may also store data, such as road maps, route information, the position, direction, and speed of the vehicle, and other such vehicle data, as well as other information. Such information may be used by the vehicle 100 and the computer system 150 during the operation of the vehicle 100 in autonomous, semi-autonomous, and/or manual modes.

As shown in FIG. 2, the user interface 170 may be used to provide information to or receive information from a user of the vehicle 100. Optionally, the user interface 170 may include one or more input/output devices in the set of peripheral devices 140, for example, a wireless communication system 141, a car computer 142, a microphone 143, and a speaker 144.

In an embodiment of the present application, the computer system 150 may control the functions of the vehicle 100 based on inputs received from various subsystems (for example, the traveling system 110, the sensing system 120, and the control system 130) and from the user interface 170. For example, the computer system 150 may use input from the control system 130 in order to control the braking unit 133 to avoid obstacles detected by the sensing system 120 and the obstacle avoidance system 136. In some embodiments, the computer system 150 is operable to provide control of many aspects of the vehicle 100 and its subsystems.

Optionally, one or more of these components described above may be installed or associated with the vehicle 100 separately. For example, the storage 152 may exist partially or completely separately from the vehicle 100. The above-mentioned components may be communicatively coupled together in a wired and/or wireless manner.

Optionally, the above-mentioned components are only an example. In actual applications, the components in the above-mentioned modules may be added or deleted according to actual needs. FIG. 2 should not be construed as a limitation to the embodiment of the present application.

Optionally, the vehicle 100 may be an autonomous vehicle traveling on a road, and may recognize objects in its surrounding environment to determine the adjustment to the current speed. The object may be other vehicles, traffic control equipment, or other types of objects. In some examples, each recognized object can be considered independently, and based on the respective characteristics of the object, such as its current speed, acceleration, distance from the vehicle, etc., can be used to determine the speed to be adjusted by the self-driving car.

Optionally, the vehicle 100 or a computing device associated with the vehicle 100 (such as the computer system 150, the computer vision system 134, and the memory 152 as shown in FIG. 2) may be based on the characteristics of the identified object and the state of the surrounding environment (for example, traffic, Rain, ice on the road, etc.) to predict the behavior of the identified object.

Optionally, each recognized object depends on each other's behavior. Therefore, all recognized objects can also be considered together to predict the behavior of a single recognized object. The vehicle 100 can adjust its speed based on the predicted behavior of the identified object. In other words, the self-driving car can determine based on the predicted behavior of the object that the vehicle will need to be adjusted (e.g., accelerate, decelerate, or stop) to a stable state. In this process, other factors may also be considered to determine the speed of the vehicle 100, such as the lateral position of the vehicle 100 on the road on which it is traveling, the curvature of the road, the proximity of static and dynamic objects, and so on.

In addition to providing instructions to adjust the speed of the self-driving car, the computing device can also provide instructions to modify the steering angle of the vehicle 100 so that the self-driving car follows a given trajectory and/or maintains an object near the self-driving car (for example, , The safe horizontal and vertical distances of cars in adjacent lanes on the road.

The above-mentioned vehicle 100 may be a car, truck, motorcycle, bus, boat, airplane, helicopter, lawn mower, recreational vehicle, playground vehicle, construction equipment, tram, golf cart, train, and trolley, etc. The application examples are not particularly limited.

In a possible implementation manner, the vehicle 100 shown in FIG. 2 may be an automatic driving vehicle, and the automatic driving system will be described in detail below.

Fig. 3 is a schematic diagram of an automatic driving system provided by an embodiment of the present application.

The automatic driving system shown in FIG. 3 includes a computer system 201, where the computer system 201 includes a processor 203, and the processor 203 is coupled to a system bus 205. The processor 203 may be one or more processors, where each processor may include one or more processor cores. The display adapter 207 (video adapter) can drive the display 209, and the display 209 is coupled to the system bus 205. The system bus 205 may be coupled to an input/output (I/O) bus 213 through a bus bridge 211, and an I/O interface 215 is coupled to an I/O bus. The I/O interface 215 communicates with a variety of I/O devices, such as input devices 217 (such as keyboard, mouse, touch screen, etc.), media tray 221, (such as CD-ROM, multimedia interface, etc.) . The transceiver 223 can send and/or receive radio communication signals, and the camera 255 can capture landscape and dynamic digital video images. Wherein, the interface connected to the I/O interface 215 may be the USB port 225.

The processor 203 may be any traditional processor, such as a reduced instruction set computer (RISC) processor, a complex instruction set computer (CISC) processor, or a combination of the foregoing.

Optionally, the processor 203 may be a dedicated device such as an application specific integrated circuit (ASIC); the processor 203 may be a neural network processor or a combination of a neural network processor and the foregoing traditional processors.

Optionally, in various embodiments described herein, the computer system 201 may be located far away from the autonomous driving vehicle, and may wirelessly communicate with the autonomous driving vehicle. In other respects, some of the processes described herein are executed on a processor provided in an autonomous vehicle, and others are executed by a remote processor, including taking actions required to perform a single manipulation.

The computer system 201 can communicate with the software deployment server 249 through the network interface 229. The network interface 229 may be a hardware network interface, such as a network card. The network 227 may be an external network, such as the Internet, or an internal network, such as an Ethernet or a virtual private network (VPN). Optionally, the network 227 may also be a wireless network, such as a wifi network, a cellular network, and so on.

As shown in FIG. 3, the hard disk drive interface is coupled with the system bus 205, the hardware drive interface 231 can be connected with the hard drive 233, and the system memory 235 is coupled with the system bus 205. The data running in the system memory 235 may include an operating system 237 and application programs 243. The operating system 237 may include a parser 239 (shell) and a kernel 241 (kernel). The shell 239 is an interface between the user and the kernel of the operating system. The shell can be the outermost layer of the operating system; the shell can manage the interaction between the user and the operating system, for example, waiting for the user's input, interpreting the user's input to the operating system, and processing various operating systems The output result. The kernel 241 may be composed of those parts of the operating system that are used to manage memory, files, peripherals, and system resources. Directly interact with the hardware. The operating system kernel usually runs processes and provides inter-process communication, providing CPU time slice management, interrupts, memory management, IO management, and so on. Application programs 243 include programs that control auto-driving cars, such as programs that manage the interaction between autonomous vehicles and obstacles on the road, programs that control the route or speed of autonomous vehicles, and programs that control interaction between autonomous vehicles and other autonomous vehicles on the road. . The application program 243 also exists on the system of the software deployment server 249. In one embodiment, the computer system 201 may download the application program from the software deployment server 249 when the automatic driving-related program 247 needs to be executed.

Exemplarily, the sensor 253 may be associated with the computer system 201, and the sensor 253 may be used to detect the environment around the computer 201.

For example, the sensor 253 can detect animals, cars, obstacles, and crosswalks. Further, the sensor can also detect the surrounding environment of the above-mentioned animals, cars, obstacles, and crosswalks, such as: the environment around the animals, for example, when the animals appear around them. Other animals, weather conditions, the brightness of the surrounding environment, etc.

Optionally, if the computer system 201 is located in an auto-driving car, the sensor may be a camera, an infrared sensor, a chemical detector, a microphone, etc.

Exemplarily, there may be multiple sensors 253. Multiple sensors can be used to detect the location of obstacles around the vehicle, and the location of the obstacles can be obtained based on the data obtained by the multiple sensors. Specifically, obtaining the location of the obstacle based on the data acquired by multiple sensors can be implemented by the target tracking method in the embodiment of the present application.

In an example, the computer system 150 shown in FIG. 2 may also receive information from other computer systems or transfer information to other computer systems. Alternatively, the sensor data collected from the sensor system 120 of the vehicle 100 may be transferred to another computer to process the data.

For example, as shown in FIG. 4, data from the computer system 312 may be transmitted to the server 320 on the cloud side via the network for further processing. The network and intermediate nodes can include various configurations and protocols, including the Internet, World Wide Web, Intranet, virtual private network, wide area network, local area network, private network using one or more company’s proprietary communication protocols, Ethernet, WiFi and HTTP, And various combinations of the foregoing; this communication can be by any device capable of transferring data to and from other computers, such as modems and wireless interfaces.

In one example, the server 320 may include a server with multiple computers, such as a load balancing server group, which exchanges information with different nodes of the network for the purpose of receiving, processing, and transmitting data from the computer system 312. The server may be configured similarly to the computer system 312, with a processor 330, a memory 340, instructions 350, and data 360.

Exemplarily, the data 360 of the server 320 may include information related to road conditions around the vehicle. For example, the server 320 may receive, detect, store, update, and transmit information related to the road conditions of the vehicle.

The following is a detailed introduction to the relevant content of the Bayesian filter.

State space refers to a collection of state variables that describe all possible states of a system. The car can be regarded as a system, and the operation of the car by the user can be regarded as input variables. When there is an input of an operation signal, it will have a clear impact on the speed, acceleration, angular velocity and other variables of the car. These affected variables can all be regarded as the system. The component of the state variable.

Observation refers to the process of obtaining state variable estimates directly or indirectly through a certain measurement method.

Bayesian estimation means that starting from any time k-1, the probability distribution of the state a priori estimate at the next time k is calculated, which is called prediction; then after obtaining the observation value at time k, the prediction link is obtained The a priori estimate of is revised, and the posterior estimate of the state at time k is obtained, which is called update.

Bayesian recursive estimation is called Bayesian filtering when used in actual engineering.

The Bayesian filter is divided into two parts: prediction and update.

The predict process satisfies formula (1):

f _k+1|k (X|z ^k )=∫f _k+1|k (X|X')f _k|k (X'|z ^k )dX' (1)

The update process satisfies formula (2) and formula (3):

f _k+1 (z _k+1 |z ^k )=∫f _k+1 (z _k+1 |X)f _k+1|k (X|z ^k )dX (3)

Among them, X is the state quantity, that is, the output quantity of the Bayesian filter, and the space where the state quantity is located is the state space. z is the observation, that is, the input of the Bayesian filter, and the space where the observation is located is the observation space. z _k+1 is the k+1 observation, z ^k = {z ₀ , z ₁ ,..., z _k } is the set of the previous k observations; f(*) is the probability density function.

Fig. 5 is a schematic diagram of a target tracking device according to an embodiment of the present application. The target tracking device 402 can be applied to the computer system 401. The target tracking device includes a prediction unit 410 and an update unit 420. The prediction unit 410 includes a judgment module 411, and the update unit 420 includes an observation transfer module 422.

Further, the prediction unit 410 may also include a state transition module 412. Further, the update unit 420 may also include an update module 421.

In order to better understand the execution process of the target tracking method in the embodiment of the present application, the function of each module in FIG. 5 is briefly described below.

The judging module 411 is used to judge whether the collection time of the observation is earlier than the update time of the current state quantity. That is to judge whether the observation is late. The current state quantity refers to the state quantity of the latest update. Specifically, when the collection time of the observation is earlier than the latest update time of the state quantity, the observation is a late observation.

It should be understood that, in the embodiments of the present application, the observation collected by the sensor is the actual observation.

The state transition module 412 is configured to obtain the predicted value of the state amount of the current update according to the state amount of the previous update when the collection time of the observation is not earlier than the update time of the current state amount. The specific process satisfies formula (1).

The update module 421 is configured to update the state quantity according to the predicted value of the current state quantity obtained by the state transition module 412. The specific process satisfies formula (2) and formula (3).

The observation transfer unit 422 is configured to perform state transition on the observation when the acquisition time of the observation is earlier than the update time of the current state quantity to obtain the estimated observation. The observation transfer unit 422 is further configured to update the current state quantity according to the actual observation and the estimated observation, and use the updated state quantity as the state quantity corresponding to the update time.

The target tracking method 500 of the embodiment of the present application will be described in detail below in conjunction with FIG. 6. FIG. 6 is a schematic flowchart of a target tracking method 500 according to an embodiment of the present application. The method shown in FIG. 6 may be executed by the target tracking device in the embodiment of the present application. The method 500 includes steps S510 to S560. Steps S510 to S560 will be described in detail below.

S510: Obtain actual observations of the moving object collected by the sensor. It should be understood that, in the embodiments of the present application, the observations collected by the sensors are the actual observations, and the actual observations may also be referred to as "observations" in the embodiments of the present application. Observation can include the speed of the moving object, the acceleration of the moving object, or the position of the moving object. For example, the observation may include the position and/or speed of the moving object measured by the millimeter wave radar.

S520: Determine whether the collection time of the observation is earlier than the update time of the current state quantity of the moving object.

If the collection time of the observation is not earlier than the update time of the current state quantity of the moving object, step S530 is executed. If the collection time of the observation is earlier than the update time of the current state quantity of the moving object, step S540 is executed.

The acquisition time of the observation is earlier than the update time of the current state quantity of the moving object. It can be understood that the observation can be used to update the current state quantity of the moving object, but it cannot actually be used for the current state quantity of the moving object. Update.

The collection time of the observation can be indicated by the time stamp of the observation.

The state quantity may include the speed of the moving object, the acceleration of the moving object, or the position of the moving object. For example, when the method 500 is used to track the position of a target, the state quantity may be the result of the tracking, and the result of the tracking may be the position of the moving object.

Among them, the state quantity of the moving object and the update time of the state quantity can be stored in the tracking list in the Bayesian filter.

Specifically, judging whether the collection time of the observation is earlier than the update time of the current state quantity of the moving object may be judging whether the time stamp of the observation is earlier than the latest update time in the tracking list.

S530: Update the current state quantity of the moving object, and use the updated state quantity as the state quantity corresponding to the collection moment of the observation.

Specifically, the current state quantity of the moving object can be updated through the Bayesian filter. For example, the current state quantity of the moving object can be updated according to the above formula (1), formula (2) and formula (3).

S540: Determine whether the time difference between the collection time of the observation and the update time of the current state quantity of the moving object is greater than a first threshold.

If the time difference is greater than the first threshold, the observation is discarded. That is, the input observation is not used to update the state quantity. In this way, when the time difference is too large, using the observation to update the state quantity may reduce the confidence of the updated state quantity.

If the time difference is less than or equal to the first threshold, step S550 is executed.

Optionally, step S540 may also be to determine whether the time difference between the collection time of the observation and the update time of the current state quantity of the moving object is greater than or equal to the first threshold.

If the time difference is greater than or equal to the first threshold, then the observation is discarded. That is, the input observation is not used to update the state quantity. If the time difference is less than the first threshold, step S550 is executed.

It should be noted that step S520, step S530, and step S540 are optional steps, and the method 500 of the embodiment of the present application may execute step S550 after step S510.

S550: In a case where the collection time of the observation is earlier than the update time of the current state quantity of the moving object, obtain the estimated observation according to the state transition model corresponding to the moving object.

When the method 500 is used to track the position of a target, the state transition model may be a kinematics model of a moving object. For example, the kinematics model may include a uniform linear motion model and the like.

Specifically, the likelihood function of the observation corresponding to the acquisition time is obtained according to the state transition model corresponding to the moving object. The estimated observation is determined based on the likelihood function of the observation corresponding to the acquisition time. The state transition model is based on the first The probability density function is determined, and the first probability density function is a probability density function that transfers the current state quantity to the state quantity corresponding to the collection time.

The step S550 will be described below in conjunction with a formula.

_{The observation z k} collected at the time t _{k is} input to the Bayesian filter, and the Bayesian filter _{updates the state quantity according to the observation z k} , and uses the obtained state quantity X _k 'as the current state quantity X _k '. Correspondingly, the latest update of the state quantity is time t _k , that is, the update time of the current state quantity X _k _{′ is time t k} . After the time t _k, t _k-1 time View captured measurement z _k-1 input Bayesian filter. The acquisition time of the above observation is earlier than the update time of the current state quantity. It can be understood that the Bayesian filter receives the observation z _k and updates the state quantity and then receives the observation z _k-1 .

If the concept of time measurements z _k-1 input to the Bayesian-filter concept of earlier measurement time Bayesian-filter input z _k, t is the likelihood function g _k-1 time (z _{_k-1} | X _k- ₁ ).

If the time when the observation z _{k-1 is} input to the Bayesian filter is not earlier than the time when the observation z _{k is} input to the Bayes filter, then the state transition model from time _{t k to} time t _k-1 _{is f k-1 |k} (X|X _k '), this state transition is Markov transition. f _k-1|k (X|X _k ') is the above-mentioned first probability density function, and the first probability density function is to transfer _{the state quantity X k} ' _{updated at t k to} the state quantity updated at _{t k-1} The probability density function.

The likelihood function of the observation corresponding to the acquisition time satisfies:

g(z _k-1 |X _k ')=∫g(z _k-1 |X)f _k-1|k (X|X _k ')dX

Among them, g(z _k-1 |X _k ') represents the likelihood function of the observation z _k-1 _{corresponding to the time t k-1} , X represents the state quantity updated at the time _{t k-1} _{, and X k} 'represents the missing state quantity update time t _K in the case of Z _{_k-1, g (z k-1 |} X) Z represents Z _k-1 _k-1 is used on the time t where _k-1 is updated state quantity Likelihood function.

S560: Update the current state quantity according to the actual observation and the estimated observation, and use the updated state quantity as the state quantity corresponding to the update time.

Specifically, the probability density function of the current state quantity may be updated according to the likelihood function of the observation corresponding to the acquisition time. The state quantity corresponding to the update time is determined according to the probability density function of the updated state quantity.

The step S560 will be described below in conjunction with a formula.

The probability density function of the updated state quantity satisfies:

Among them, f _k|k (X _k |z ^k ) represents the probability density function of the state quantity X _k _{updated at time t k} _{, and f'k|k} (X _k '|z _k ,z ^k-2 ) represents the absence of z In the case of _k-1 , the probability density function of the state quantity X _k _{'updated at time t k.} That is, the posterior probability corresponding to the current state quantity X _k'. z _k represents the observations collected at time _{t k} ^{, z k} represents the set of observations collected at k time {z ₁ ,z ₂ ,...,z _k-2 ,z _k-1 ,z _k }, z ^{k -2} represents the set of observations collected at k-2 moments {z ₁ , z ₂ ,..., z _k-2 }.

Further, the Kalman filter is an implementation form of the Bayes filter. Taking the Kalman filter as an example, step S560 will be described.

The above step S560 may include determining the expectation of the state quantity corresponding to the update time according to the actual observation and the estimated observation; and taking the expectation of the state quantity corresponding to the update time as the state quantity corresponding to the update time.

Among them, the expectation of the state quantity corresponding to the update time is related to the Kalman gain value. The Kalman gain value is related to the first covariance, the observation matrix at the time of collection, the covariance of the observation matrix, and the variance of the observation at the time of collection. The first covariance refers to the covariance of the state quantity transferred from the update time to the collection time.

Optionally, the Kalman gain value satisfies:

Var(z _k-1 ) satisfies:

P _k-1|k satisfies:

The expectation of the state quantity corresponding to the update time meets:

Represents the estimated observation at time _{t k-1.} The expected state quantity x _k|k corresponding to the update time is taken as the state quantity corresponding to the update time.

Optionally, the estimated value of the observation corresponding to the collection time satisfies:

in,

Represents t estimated observations _k-1 time point, H _k-1 represents a t the observation matrix _k-1 time, F _{k-1 | k} indicates a transition from t _K time to t _k-1 time of the state transition matrix, x ' _{k | k} t _k represents a time measurement in the case where z _k-1 is updated in a desired state quantity of the absence of the concept of time t _k-1 collected.

The estimated observation at time t _k-1 can be understood as the state transition matrix F _{k-1|k to} transfer the expected state quantity _x'k|k to time t _k-1 , and then based on the observation matrix H _k-1 Obtain the expectation of the observation at _{t k-1}

As the estimated observation at the time _{t k-1.}

Further, the covariance of the updated state quantity at time _{t k is calculated.} The covariance of the updated state at t _{k satisfies:}

P _k|k = (IK' _k H _k-1 )P _k-1|k

Among them, P _k|k represents the covariance of the state quantity updated at time _{t k.}

According to the solution of the embodiment of the present application, the state transition is performed according to the state transition model corresponding to the moving object, and then the current state quantity is updated, and the more accurate state quantity corresponding to the latest update time can be obtained, instead of the state quantity corresponding to the collection time. , To avoid out-of-order jumps of the output state quantity in time.

Hereinafter, taking the method 500 applied to tracking a target as an example, step S550 and step S560 will be described.

R represents the distance between the sensor and the target. x=(R) represents a one-dimensional vector.

In the absence of the time t _k-1 Concept captured measurement time t _K in the case where Z _k-1 is updated to meet the state quantity x _k

x _k ～N(x; x′ _k|k ,P′ _k|k ) indicates that the distance between the sensor and the target is a random number _{with an expectation of x′ k|k} and a variance of P′ _k|k. _P'k|k can be 1.

x _{'k | k} t _k represents a time measurement in the case where z _k-1 is updated in a desired state quantity of the absence of the concept of time t _k-1 collected. P _{'k | k} represents the missing covariance Concept time t _k-1 measured state quantity acquisition case where z _k-1 t _k is the time updates.

For example, x _{'k | k} can be 3, P' _{k | k} may be _1, t _k-1 time View captured measurement z _k-1 may be four. In this case, P _k-1|k satisfies:

The variance Var(z _k-1 ) of the observation z _k-1 collected at the time t _k-1 satisfies:

The Kalman gain value satisfies:

The estimated observation at time t _{k-1 satisfies:}

_k t state the amount of time to update the expectations x _{k | k} satisfy:

Among them, x _k|k can be used as the output result, that is, the state quantity corresponding to the time _{t k.}

t _k renewed state quantity covariance P _{k | k} satisfying:

P _k|k =(IK' _k H _k-1 )P _k-1|k =(1-1*1)*1=0

It should be understood that the foregoing examples are intended to help those skilled in the art understand the embodiments of the present application, and are not intended to limit the embodiments of the present application to the specific numerical values or specific scenarios illustrated. Those skilled in the art can obviously make various equivalent modifications or changes based on the examples given, and such modifications or changes also fall within the scope of the embodiments of the present application.

The target tracking method of the embodiment of the present application is described in detail above with reference to FIG. 6, and the device embodiment of the present application will be described in detail below with reference to FIG. 7 to FIG. 8. It should be understood that the target tracking device in the embodiment of the present application can execute the target tracking method of the foregoing embodiment of the present application, that is, the specific working process of the following various products can refer to the corresponding process in the foregoing method embodiment.

Fig. 7 is a schematic block diagram of a target tracking device provided by an embodiment of the present application. It should be understood that the target tracking device 1000 can execute the target tracking method shown in FIG. 6. The target tracking device 1000 includes: an acquiring unit 1010 and a processing unit 1020.

Wherein, the acquiring unit 1010 is used to acquire the actual observation of the moving object collected by the sensor. The processing unit 1020 is used to obtain the estimated observation according to the state transition model corresponding to the moving object when the acquisition time of the actual observation is earlier than the update time of the current state quantity of the moving object; The state quantity is updated, and the updated state quantity is used as the state quantity corresponding to the update time.

Optionally, the processing unit 1020 is configured to: obtain the likelihood function of the observation corresponding to the acquisition time according to the state transition model corresponding to the moving object, and the estimated observation is determined according to the likelihood function of the observation corresponding to the acquisition time, The state transition model is determined according to the first probability density function, and the first probability density function is the probability density function that transfers the current state quantity to the state quantity corresponding to the collection time.

Optionally, the processing unit 1020 is configured to: update the probability density function of the current state quantity according to the likelihood function of the observation corresponding to the acquisition time; determine the state quantity corresponding to the update time according to the probability density function of the updated state quantity .

Optionally, the likelihood function of the observation corresponding to the acquisition moment satisfies:

g(z _k-1 |X _k ')=∫g(z _k-1 |X)f _k-1|k (X|X _k ')dX

Optionally, the probability density function of the updated state quantity satisfies:

Optionally, the processing unit 1020 is configured to: determine the expectation of the state quantity corresponding to the update time according to actual observations and estimated observations; use the expectation of the state quantity corresponding to the update time as the state quantity corresponding to the update time; wherein, the update time corresponds to The expectation of the state quantity is related to the Kalman gain value. The Kalman gain value is related to the first covariance, the observation matrix at the acquisition time, the covariance of the observation matrix, and the variance of the observation at the acquisition time. The first covariance refers to It is the covariance of the state amount from the update time to the collection time.

Optionally, the Kalman gain value satisfies:

Var(z _k-1 ) satisfies:

P _k-1|k satisfies:

Optionally, the expectation of the state quantity corresponding to the update time meets:

Represents the estimated observation at time _{t k-1.}

in,

Optionally, the processing unit 1020 is configured to: in the case where the acquisition time of the observation is earlier than the update time of the current state quantity of the moving object, and the time difference between the acquisition time and the update time is less than or equal to a threshold, corresponding to the moving object The state transition model is estimated to be observed.

It should be noted that the above-mentioned target tracking device 1000 is embodied in the form of a functional unit. The term "unit" herein can be implemented in the form of software and/or hardware, which is not specifically limited.

For example, a "unit" may be a software program, a hardware circuit, or a combination of the two that realizes the above-mentioned functions. The hardware circuit may include an application specific integrated circuit (ASIC), an electronic circuit, and a processor for executing one or more software or firmware programs (such as a shared processor, a dedicated processor, or a group processor). Etc.) and memory, merged logic circuits and/or other suitable components that support the described functions.

Therefore, the units of the examples described in the embodiments of the present application can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.

FIG. 8 is a schematic diagram of the hardware structure of the target tracking device provided by an embodiment of the present application.

As shown in FIG. 8, the target tracking apparatus 1200 (the target tracking 1200 may specifically be a computer device) includes a memory 1201, a processor 1202, a communication interface 1203, and a bus 1204. Among them, the memory 1201, the processor 1202, and the communication interface 1203 implement communication connections between each other through the bus 1204.

The memory 1201 may be a read only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1201 may store a program. When the program stored in the memory 1201 is executed by the processor 1202, the processor 1202 is configured to execute each step of the target tracking method of the embodiment of the present application, for example, execute each step shown in FIG. 6.

It should be understood that the target tracking device shown in the embodiment of the present application may be a server, for example, it may be a server in the cloud, or may also be a chip configured in a server in the cloud.

The processor 1202 may adopt a general central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits for executing related programs to realize this The target tracking method of the application method embodiment.

The processor 1202 may also be an integrated circuit chip with signal processing capability. In the implementation process, each step of the target tracking method of the present application can be completed by an integrated logic circuit of hardware in the processor 1202 or instructions in the form of software.

The above-mentioned processor 1202 may also be a general-purpose processor, a digital signal processing (digital signal processing, DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (field programmable gate array, FPGA) or other programmable logic devices, Discrete gates or transistor logic devices, discrete hardware components. The methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like. The steps of the method disclosed in the embodiments of the present application may be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers. The storage medium is located in the memory 1201, and the processor 1202 reads the information in the memory 1201, and combines its hardware to complete the functions required by the units included in the target tracking device shown in FIG. 7 in the implementation of this application, or execute the method of this application The target tracking method shown in FIG. 6 of the embodiment.

The communication interface 1203 uses a transceiving device such as but not limited to a transceiver to implement communication between the target tracking device 1200 and other devices or a communication network.

The bus 1204 may include a path for transferring information between various components of the target tracking device 1200 (for example, the memory 1201, the processor 1202, and the communication interface 1203).

It should be noted that although the above-mentioned target tracking device 1200 only shows a memory, a processor, and a communication interface, in the specific implementation process, those skilled in the art should understand that the target tracking device 1200 may also include other necessary for normal operation. Device. At the same time, according to specific needs, those skilled in the art should understand that the above-mentioned target tracking device 1200 may also include hardware devices that implement other additional functions.

In addition, those skilled in the art should understand that the above-mentioned target tracking device 1200 may also only include the necessary components to implement the embodiments of the present application, and not necessarily include all the components shown in FIG. 8.

A person of ordinary skill in the art may realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed system, device, and method can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or It can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

In addition, the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.

If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solution of the present application essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or an access network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (read-only Memory, ROM), random access memory (random access memory, RAM), magnetic disk or optical disk and other media that can store program code .

The above are only specific implementations of this application, but the protection scope of this application is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application. Should be covered within the scope of protection of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

A target tracking method, characterized in that it comprises:

Obtain the actual measurement of the moving object collected by the sensor;

In the case that the acquisition time of the actual observation is earlier than the update time of the current state quantity of the moving object, obtaining the estimated observation according to the state transition model corresponding to the moving object;

The current state quantity is updated according to the actual observation and the estimated observation, and the updated state quantity is used as the state quantity corresponding to the update moment.
The method according to claim 1, wherein the obtaining the estimated observations according to the state transition model corresponding to the moving object comprises:

Obtaining the likelihood function of the observation corresponding to the acquisition time according to the state transition model corresponding to the moving object, and the estimated observation is determined according to the likelihood function of the observation corresponding to the acquisition time, The state transition model is determined according to a first probability density function, and the first probability density function is a probability density function that transfers the current state quantity to the state quantity corresponding to the collection time.
The method according to claim 2, wherein the current state quantity is updated according to the actual observation and the estimated observation, and the updated state quantity is used as the state corresponding to the update time Quantity, including:

Updating the probability density function of the current state quantity according to the likelihood function of the observation corresponding to the acquisition time;

The state quantity corresponding to the update time is determined according to the probability density function of the updated state quantity.
The method according to claim 2 or 3, wherein the likelihood function of the observation corresponding to the acquisition time satisfies:

g(z k-1 |X k ')=∫g(z k-1 |X)f k-1|k (X|X k ')dX;

Among them, g(z k-1 |X k ') represents the likelihood function of the observation z k-1 corresponding to the time t k-1 , X represents the state quantity updated at the time t k-1 , and X k 'represents the missing state quantity update time t K in the case of Z k-1, g (z k-1 | X) Z represents Z k-1 k-1 is used on the time t where k-1 is updated state quantity likelihood function, f k-1 | k ( X | X k ') of the first probability density function, the first probability density function is a t k renewed state quantity X k' is transferred to to t k -1 The probability density function of the updated state quantity.
The method according to claim 4, wherein the probability density function of the updated state quantity satisfies:

Among them, f k|k (X k |z k ) represents the probability density function of the state quantity X k updated at time t k , and f'k|k (X k '|z k ,z k-2 ) represents the absence of z In the case of k-1 , the probability density function of the state quantity X k 'updated at time t k , z k represents the observations collected at time t k , and z k represents the set of observations collected at k time {z 1 , z 2 ,...,z k-2 ,z k-1 ,z k }, z k-2 represents the set of observations collected at k-2 moments {z 1 ,z 2 ,...,z k-2 }.
The method of claim 1, wherein the current state quantity is updated according to the actual observation and the estimated observation, and the updated state quantity is used as the state corresponding to the update time Quantity, including:

Determine the expectation of the state quantity corresponding to the update time according to the actual observation and the estimated observation;

Taking the expectation of the state quantity corresponding to the update time as the state quantity corresponding to the update time;

Wherein, the expectation of the state quantity corresponding to the update time is related to the Kalman gain value, the Kalman gain value and the first covariance, the observation matrix at the acquisition time, the covariance of the observation matrix and the observation at the acquisition time The variance of the quantity is related, and the first covariance refers to the covariance of the state quantity transferred from the update time to the collection time.
The method according to claim 6, wherein the Kalman gain value satisfies:

Var(z k-1 ) satisfies:

P k-1|k satisfies:

Wherein, P k-1 | k represents t K transferred from the update time to the acquisition time t covariance k-1 state quantity, P 'k | k represents the lack Concept t k-1 time captured measurement Z k-1 In the case of t k , the covariance of the state quantity updated at time t k, F k-1|k represents the state transition matrix from t k to t k-1 , Q k represents the covariance of the prediction matrix, H k-1 Represents the observation matrix at time t k-1 , Var(z k-1 ) represents the variance of the observation z k-1 collected at time t k-1 , and R k-1 represents the covariance of the observation matrix.
The method according to claim 7, wherein the expectation of the state quantity corresponding to the update time meets:

Among them, x k|k represents the expectation of the state quantity updated at time t k,
Represents the estimated observation at time t k-1.
The method according to any one of claims 6 to 8, wherein the estimated observation quantity satisfies:

in,
Represents t estimated observations k-1 time point, H k-1 represents a t the observation matrix k-1 time, F k-1 | k indicates a transition from t K time to t k-1 time of the state transition matrix, x ' k | k t k represents a time measurement in the case where z k-1 is updated in a desired state quantity of the absence of the concept of time t k-1 collected.
The method according to any one of claims 1 to 9, characterized in that, in the case that the acquisition time of the observation is earlier than the update time of the current state quantity of the moving object, the The state transition model corresponding to the object gets estimated observations, including:

In the case that the acquisition time of the actual observation is earlier than the update time of the current state quantity of the moving object, and the time difference between the acquisition time and the update time is less than or equal to a threshold value, according to the moving object The corresponding state transition model gets estimated observations.
A target tracking device, characterized in that it comprises an acquisition module and a processing module, wherein,

The acquisition module is used to: acquire the actual observation of the moving object collected by the sensor;

The processing module is used for:

In the case that the acquisition time of the actual observation is earlier than the update time of the current state quantity of the moving object, obtaining the estimated observation according to the state transition model corresponding to the moving object;

The current state quantity is updated according to the actual observation and the estimated observation, and the updated state quantity is used as the state quantity corresponding to the update moment.
The device of claim 11, wherein the processing module is configured to:

Obtaining the likelihood function of the observation corresponding to the acquisition time according to the state transition model corresponding to the moving object, and the estimated observation is determined according to the likelihood function of the observation corresponding to the acquisition time, The state transition model is determined according to a first probability density function, and the first probability density function is a probability density function that transfers the current state quantity to the state quantity corresponding to the collection time.
The device of claim 12, wherein the processing module is configured to:

Updating the probability density function of the current state quantity according to the likelihood function of the observation corresponding to the acquisition time;

The state quantity corresponding to the update time is determined according to the probability density function of the updated state quantity.
The device according to claim 12 or 13, wherein the likelihood function of the observation corresponding to the acquisition time satisfies:

g(z k-1 |X k ')=∫g(z k-1 |X)f k-1|k (X|X k ')dX;

Wherein, g (z k-1 | X k ') denotes Views on T k-1 corresponding to the time measurement z likelihood function k-1, X is t k 1-time updating of the state quantity, X k' represents missing state quantity update time t K in the case of Z k-1, g (z k-1 | X) Z represents Z k-1 k-1 is used on the time t where k-1 is updated state quantity likelihood function, f k-1 | k ( X | X k ') of the first probability density function, the first probability density function is a t k renewed state quantity X k' is transferred to to t k -1 The probability density function of the updated state quantity.
The device according to claim 14, wherein the probability density function of the updated state quantity satisfies:

Among them, f k|k (X k |z k ) represents the probability density function of the state quantity X k updated at time t k , and f'k|k (X k '|z k ,z k-2 ) represents the absence of z In the case of k-1 , the probability density function of the state quantity X k 'updated at time t k , z k represents the observations collected at time t k , and z k represents the set of observations collected at k time {z 1 , z 2 ,...,z k-2 ,z k-1 ,z k }, z k-2 represents the set of observations collected at k-2 moments {z 1 ,z 2 ,...,z k-2 }.
The device of claim 11, wherein the processing module is configured to:

Determine the expectation of the state quantity corresponding to the update time according to the actual observation and the estimated observation;

Taking the expectation of the state quantity corresponding to the update time as the state quantity corresponding to the update time;

Wherein, the expectation of the state quantity corresponding to the update time is related to the Kalman gain value, the Kalman gain value and the first covariance, the observation matrix at the acquisition time, the covariance of the observation matrix and the observation at the acquisition time The variance of the quantity is related, and the first covariance refers to the covariance of the state quantity transferred from the update time to the collection time.
The apparatus according to claim 16, wherein the Kalman gain value satisfies:

Var(z k-1 ) satisfies:

P k-1|k satisfies:

Wherein, P k-1 | k represents t K transferred from the update time to the acquisition time t covariance k-1 state quantity, P 'k | k represents the lack Concept t k-1 time captured measurement Z k-1 In the case of t k , the covariance of the state quantity updated at time t k, F k-1|k represents the state transition matrix from t k to t k-1 , Q k represents the covariance of the prediction matrix, H k-1 Represents the observation matrix at time t k-1 , Var(z k-1 ) represents the variance of the observation z k-1 collected at time t k-1 , and R k-1 represents the covariance of the observation matrix.
The device according to claim 17, wherein the expectation of the state quantity corresponding to the update time meets:

Among them, x k|k represents the expectation of the state quantity updated at time t k,
Represents the estimated observation at time t k-1.
The device according to any one of claims 16 to 18, wherein the estimated observation quantity satisfies:

in,
Represents t estimated observations k-1 time point, H k-1 represents a t the observation matrix k-1 time, F k-1 | k indicates a transition from t K time to t k-1 time of the state transition matrix, x ' k | k t k represents a time measurement in the case where z k-1 is updated in a desired state quantity of the absence of the concept of time t k-1 collected.
The device according to any one of claims 11 to 19, wherein the processing module is configured to:

In the case that the acquisition time of the actual observation is earlier than the update time of the current state quantity of the moving object, and the time difference between the acquisition time and the update time is less than or equal to a threshold value, according to the moving object The corresponding state transition model gets estimated observations.
A target tracking device, characterized by comprising at least one processor and a memory, the at least one processor is coupled to the memory, and is configured to read and execute instructions in the memory to execute as claimed in claims 1 to 10. The method of any one of 10.
A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program runs on a computer, the computer executes any one of claims 1 to 10 The method described in the item.