System and method for estimating ego-motion state of a vehicle

Abstract

The disclosure relates to method (200) and system (100) for estimating ego-motion state of vehicle. Method (200) comprises receiving sensor detection clusters comprising information about detected objects with doppler uncertainties, in vehicle's surroundings; creating doppler uncertainty model for each sensor detection cluster; initializing set of particles. Each particle in the set of particles represents probable ego-motion state of vehicle and each particle is assigned weight. Method (200) further comprises propagating each particle using prediction model to obtain updated set of particles with predicted state; updating weight of each updated particle based on sensor detection clusters and doppler uncertainty model; and estimating ego-motion state of vehicle from set of updated particles with updated state.

Description

[0001] 202403892

[0002] 1

[0003] SYSTEM AND METHOD FOR ESTIMATING EGO-MOTION STATE OF A VEHICLE TECHNICAL FIELD

[0004] The present disclosure in general relates to information system in a vehicle. More particularly, it relates to system and method for estimating ego-motion state of the vehicle.

[0005] BACKGROUND

[0006] For effective operation of driver assistance systems for vehicles, ego motion plays a critical role in enabling safe and accurate navigation. Ego motion refers to movement of a vehicle relative to its surroundings. Accurate estimation of the ego motion is essential for various application such as, collision avoidance, lane keeping, and obstacle detection.

[0007] Sensors, particularly Frequency Modulated Continuous Wave (FMCW) sensors, are widely used in automotive applications due to their robustness under various weather conditions and ability to detect objects at longer ranges. For sensor-based ego motion estimation, range rate, i.e. , the rate at which an object moves toward or away from the sensor and / or radial velocity of detection are key parameters. The range rate or radial velocity of the detections are derived from doppler shift in frequency of sensor signals. However, the FMCW sensors face a fundamental limitation known as ‘doppler ambiguity’ or “doppler uncertainty”, which causes uncertainty / ambiguity in measurement of the range rate. The doppler ambiguity poses challenges for accurate ego motion estimation.

[0008] Existing techniques to address the doppler ambiguity often rely on Inertial Measurement Units (IMUs) in conjunction with the sensors. However, the use of IMUs has several drawbacks related to IMU calibration, other and product constraints, such as in 2-wheelers where it is desirable to minimize sensor connections.

[0009] Thus, there is a need for an improved solution that is capable of solving the aforementioned problems of doppler ambiguity in sensor ego motion estimation.202403892

[0010] 2

[0011] SUMMARY

[0012] Though some recent studies have explored resolving doppler ambiguity without the use of Inertial Measurement Units (IMUs), relying solely on sensor data. While such methods reduce system complexity, the methods typically involve high computational loads, making them unsuitable for real-time implementation in automotive environments. Further, in the two-wheelers, capability to estimate ego-motion state without the IMUs may offer many advantages.

[0013] Therefore, there is a need for a solution that can effectively resolve Doppler ambiguity in sensor-based ego motion estimation without relying on IMUs, while also being suitable for real-time operation with low computational overhead.

[0014] It is therefore an object of the present disclosure to provide a system and method for estimating ego-motion state of a vehicle.

[0015] This and other objects are achieved by means of a system, a method, a computer program, and a computer-readable medium defined in the appended claims. The term exemplary is in the present context to be understood as serving as an instance, example or illustration.

[0016] According to an aspect of the present disclosure, a method for estimating ego-motion of a vehicle is disclosed. The method comprises receiving one or more sensor detection clusters comprising information about one or more detected objects with doppler uncertainties, in a vehicle’s surroundings. The method further comprises creating a doppler uncertainty model for each sensor detection cluster of the one or more sensor detection clusters. The method further comprises initializing a set of particles. Each particle in the set of particles represents a probable ego-motion state of the vehicle, and wherein each particle is assigned a weight. The method further comprises propagating each particle in the set of particles using a prediction model to obtain an updated set of particles with a predicted state. The method further comprises updating weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model. The method further comprises estimating the ego-motion state of the vehicle from the set of updated particles with updated state.202403892

[0017] 3

[0018] Optionally, the information comprises range rate, distance, velocity and azimuth angle of the detected objects.

[0019] Optionally, the ego-motion state is estimated using particle filter framework.

[0020] Optionally, the method comprises computing an expected uncertain doppler reading for each of the sensor clusters using the doppler uncertainty model. Optionally, the doppler uncertainty model is generated using a non-linear model.

[0021] Optionally, the ego-motion state comprises ego velocity and yaw rate of the vehicle.

[0022] Optionally, the probable ego-motion state comprises at least one of: ego longitudinal velocities, lateral velocities and / or accelerations.

[0023] According to another aspect of the present disclosure, a system for estimating egomotion of a vehicle is disclosed. The system comprises processing circuitry. The processing circuitry is configured to receive one or more sensor detection clusters comprising information about one or more detected objects with doppler uncertainties, in a vehicle’s surroundings. The processing circuitry is further configured to generate a doppler uncertainty model for each sensor detection cluster of the one or more sensor detection clusters. The processing circuitry is further configured to initialize a set of particles. Each particle in the set of particles represents a probable motion state of the vehicle, and wherein each particle is assigned a weight. The processing circuitry is further configured to propagate each particle in the set of particles using a prediction model to obtain an updated set of particles with predicted state. The processing circuitry is further configured to update weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model. The processing circuitry is further configured to estimate the egomotion state of the vehicle from the set of updated particles with updated state.

[0024] According to another aspect of the present disclosure, there is provided a computer program when loaded and run on a system, causes a processing circuitry to perform corresponding steps of method for estimating ego-motion of a vehicle.

[0025] According to another aspect of the present disclosure, there is provided a computer-readable medium having stored thereon a computer program.202403892

[0026] 4

[0027] Some embodiments disclosed herein have one or more of the following advantages:

[0028] - The proposed method and system are not dependent on external sensors for estimating ego motions

[0029] - The proposed method and system have ability to recover from scenarios of ‘low confidence

[0030] - With the proposed method and system, orders of magnitude computationally cheaper than the alternative

[0031] - Independently estimated ego motion can be used for calibrating Inertial Measurement Unit (IMU)

[0032] - Rider assistance functionalities for two-wheelers independent of IMU is beneficial - By removing dependency of sensor with the IMU, the proposed method and system provides several benefits such as saving of printed circuit board (PCB) space, cost optimization, no complexity of mounting and calibration of IMU Other advantages may be readily apparent to one having skill in the art. Certain embodiments may have none, some, or all of the recited advantages.

[0033] BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The foregoing will be apparent from the following more particular description of the example embodiments, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the example embodiments.

[0035] FIG. 1 discloses a block diagram of a system for estimating ego-motion state of a vehicle, according to some embodiments;

[0036] FIG. 2 illustrates a flow chart for a method for estimating the ego-motion state of the vehicle, according to some embodiments;

[0037] FIG. 3 illustrates a flow chart disclosing additional details of the method for estimating the ego-motion state of the vehicle, according to some embodiments;

[0038] FIG. 4 illustrates another flow chart disclosing additional details of the method for estimating the ego-motion state of the vehicle, according to some embodiments; and202403892

[0039] 5

[0040] FIG. 5 discloses an example computing environment according to some embodiments.

[0041] DETAILED DESCRIPTION

[0042] Aspects of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings. The systems and methods disclosed herein can, however, be realized in many different forms and should not be construed as being limited to the aspects set forth herein. Like numbers in the drawings refer to like elements throughout.

[0043] The terminology used herein is for the purpose of describing particular aspects of the disclosure only and is not intended to limit the invention. It should be emphasized that the term “comprises / comprising” when used in this specification is taken to specify the presence of stated features, integers, steps, or components, but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0044] Embodiments of the present disclosure will be described and exemplified more fully hereinafter with reference to the accompanying drawings. The solutions disclosed herein can, however, be realized in many different forms and should not be construed as being limited to the embodiments set forth herein.

[0045] It will be appreciated that when the present disclosure is described in terms of a system and a method, it may also be embodied in one or more processors and one or more memories coupled to the one or more processors, wherein the one or more memories store one or more programs that perform the steps, services and functions disclosed herein when executed by the one or more processors.

[0046] FIG. 1 discloses a block diagram of a system 100 for estimating ego-motion state of a vehicle (not shown in FIG.s), according to some embodiments. The system 100 refers to an electronic control system comprising an input / output unit 104, processing circuitry 106 and a communication unit 108. The processing circuitry 106 is communicatively coupled to a display / interface 110 of the vehicle and communicates via the communication unit 108.202403892

[0047] 6

[0048] In an example, the ego-motion may refer to movement of the vehicle relative to its surroundings. The system 100 may use sensor data (sensor detection clusters) to estimate the ego-motion state of the vehicle / vehicle’s motion, even when uncertainties in doppler measurements exist.

[0049] In an example, the vehicle may comprise a four-wheeler, a two-wheeler, a three-wheeler, a four-wheeler for example, a car of different type like hatchback, sedan etc. or a truck or any other four wheeler automobile.

[0050] As shown in FIG. 1, an aspect of this disclosure shows the system 100 that comprises of the processing circuitry 106. The processing circuitry 106 is configured to receive one or more sensor detection clusters comprising information about one or more detected objects with doppler uncertainties, in the vehicle’s surroundings. For each sensor detection cluster of the one or more sensor detection clusters, the processing circuitry 106 is configured to create a doppler uncertainty model (details are explained later in FIG. 3). The processing circuitry 106 is further configured to initialize a set of particles. Each particle in the set of particles represents a probable motion state of the vehicle, and each particle is assigned a weight.

[0051] The processing circuitry 106 is further configured to propagate each particle in the set of particles using a prediction model to obtain an updated set of particles with predicted state. In an example, each particle is propagated forward in time using the prediction model, which predicts a state of the particle at a next time step. In an example, the prediction model may include and not limited to constant velocity models or first order models.

[0052] The processing circuitry 106 is further configured to update the weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model. The processing circuitry 106 estimates the ego-motion state of the vehicle from the set of updated particles with updated state. In an example, the ego-motion state of the vehicle is estimated by calculating a weighted average of all the particles. Particles with higher weights contribute more to the final estimate.202403892

[0053] 7

[0054] In an embodiment, the information comprises range rate, distance, velocity and azimuth angle of the detected objects.

[0055] In an example, the information is received from one or more sensors 102. Each detection cluster comprises position and distance of the detected objects, velocity (doppler) measurements which may comprise uncertainties.

[0056] In an embodiment, the processing circuitry 106 is configured to receive the one or more sensor detection clusters that comprises information about one or more detected objects with the doppler uncertainties from the one or more sensors 102. In an example, the one or more sensors 102 comprise a depth sensor, for example, Radio Detection and Ranging sensor or Light Detection and Ranging sensor. Optionally, the system 100 may comprise at least one sensor of the one or more sensors 102 communicatively coupled with the system 100. In an example, the system 100 may be configured inside or outside of the vehicle.

[0057] In an example, the detected objects comprise traffic participants, stationary objects such as guard-rails, obstacles, buildings, barriers and so on.

[0058] In an embodiment, the processing circuitry 106 is configured to estimate the ego-motion state using particle filter framework. The particle filter framework is a well-known technique in the field of sensor fusion. The particle filter framework may be used for estimating state of the vehicle and tracking of the vehicle’s position, orientation, and motion in dynamic environment. In an example, each particle represents the probable ego-motion state of the vehicle. The probable ego-motion state represents a possible ego-motion state of the vehicle, such as position, velocity, and / or orientation. In an example, the set of particles are initialized randomly or based on prior knowledge, and each particle is assigned to a weight. For example, initial set of particles represent probable states like: position: (x=0, y=0), velocity: 60 km / h, orientation: facing north and each particle has a weight of 1.0 initially.

[0059] In an embodiment, the processing circuitry 106 is configured to compute an expected uncertain doppler reading for each of the sensor clusters using the doppler uncertainty model.202403892

[0060] 8

[0061] In an embodiment, the processing circuitry 106 is configured to generate doppler uncertainty model using a non-linear model.

[0062] In an embodiment, the ego-motion state comprises ego velocity and yaw rate of the vehicle.

[0063] In an embodiment, the probable ego-motion state comprises at least one of: ego longitudinal velocities, lateral velocities and / or accelerations.

[0064] In an example, the ego longitudinal velocities refer to velocities of the vehicle in a direction of its forward motion. In an example, the lateral velocities refer to velocities of an object moves sideways or perpendicular to its primary direction of motion. in other words, the lateral velocities are the speed of movement along the lateral (side to side) axis. In the vehicle, lateral velocities may represent the sideways motion of the vehicle.

[0065] In another aspect, FIG. 2 illustrates a flow chart for a method 200 for estimating the ego-motion state of the vehicle. The method 200 may be performed by the system 100 as discussed above.

[0066] At step 202, the method 200 comprises receiving the one or more sensor detection clusters comprising information about the one or more detected objects with doppler uncertainties, in the vehicle’s surroundings.

[0067] At step 204, the method 200 comprises creating the doppler uncertainty model for each sensor detection cluster of the one or more sensor detection clusters.

[0068] At step 206, the method 200 comprises initializing the set of particles. Each particle in the set of particles represents the probable ego-motion state of the vehicle. Each particle is assigned the weight.

[0069] At step 208, the method 200 comprises propagating each particle in the set of particles using the prediction model to obtain the updated set of particles with the predicted state.

[0070] The method 200 at step 210 comprises updating the weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model.202403892

[0071] 9

[0072] The method 200 at step 212 comprises estimating the ego-motion state of the vehicle from the set of updated particles with the updated state.

[0073] Details of the method 200 are similar to the details of the system 100 as discussed above and hence are not repeated for the sake of brevity.

[0074] In an example, FIG. 3 illustrates a flow chart disclosing additional details 300 of the method 200 and the system 100. At step 302, the system 100 initiates initialization of the particles. In an example, a set of particles is created, each representing the probable state / possible ego-motion state of the vehicle. The initial state of each particle may include position (x, y, z), velocity (vx, vy, vz), orientation (yaw, pitch, roll) and / or other relevant parameters such as acceleration, angular velocity, etc.

[0075] At step 304, the processing circuitry 106 propagates each particle in the set of particles using the prediction model to obtain an updated set of particles with a predicted state. The propagated particles represent potential future states of the vehicle.

[0076] At step 304, the weight of each updated particle is updated based on the one or more sensor detection clusters / sensor detections data and the doppler uncertainty model. The doppler uncertainty model is used to calculate likelihood of state of each particle given in the sensor detection clusters / sensor measurements. Particles with higher likelihoods are assigned higher weights, indicating a higher probability of being the correct ego-motion state of the vehicle.

[0077] Optionally, at step 306, particle resampling of the set of particles is performed. Particle resampling is a technique used to reduce the number of particles with low weights with respect to weight of other particles in the set of particles and increase the number of particles with high weights with respect to weight of other particles in the set of particles. This helps in filtering the particles on most likely regions of state space of the set of particles. The resampling may be performed using various methods such as symmetric resampling or multinomial resampling or stratified resampling.

[0078] In an example embodiment, doppler uncertainty model (also referred to as doppler ambiguity model) is described now. The doppler uncertainty model refers to a mathematical model that helps to account uncertainty or ambiguity that arises in202403892

[0079] 10

[0080] measurement of doppler frequency shifts in Radio Detection and Ranging systems. Doppler uncertainty occurs when a measured doppler frequency shift exceeds sampling or pulse repetition frequency (PRF) of a system, causing the true frequency shift to appear as an alias or incorrect value. The doppler uncertainty model aims to address the uncertainty by computing the expected uncertain doppler readings. In an example, the doppler uncertainty model considers a given frame or scan containing ND detections / sensor detections from the one or more sensor detection clusters (sensor clusters). For each ithsensor cluster / detections / sensor detections, an azimuth angle value (0f) is determined, which represents the direction of the detected objects. The doppler uncertainty model simulates the expected doppler readings from the stationary objects / detected object at 0;, when the ego motion is given by the set of particles (also referred to as hypothesis). A hypothesis is a guess of ego-motion states of the vehicle and may be represented by x and y components velocity (vx, vy). Suppose (vx, vy) are ego speed values for a hypothesis and vaf is an ambiguity free limit. Then the corresponding doppler uncertainty model / doppler ambiguity model is represented as shown in below equation.

[0081] < > > >

[0082]

[0083] The ambiguity free limit is a threshold (vaf) beyond which the doppler uncertainty becomes significant.

[0084] In an example, the ego motion state comprises the ego longitudinal and lateral velocities and accelerations X = (vx, vy, ax, ay). Then the prediction model is the random-walk model as shown in the equation below.

[0085] Xk = FXk-i + v

[0086] Here, F is the state-transition matrix for random-walk model and v is the ‘process noise’. The process noise refers to a mathematical framework for accounting for approximation202403892

[0087] 11

[0088] of the prediction model. In an example the process noise may include and not limited to gaussian noise.

[0089]

[0090] For every particle (or velocity hypothesis), the doppler uncertainty model (vamb) is used assuming all stationary sensor detection clusters. The weight w of the particle is proportional to the agreement of the doppler uncertainty model with the detections. That is for ND detections,

[0091]

[0092] >

[0093] In an example, referring to FIG. 4, additional details 400 of the method 200 and the system 100 are shown. At step 402, the processing circuitry 106 initiates the initialization of the particles. In an example, the set of particles is created, each representing the probable state / possible ego-motion state of the vehicle. At step 404, the processing circuitry 106 propagates each particle in the set of particles using the prediction model to obtain an updated set of particles with the predicted state.

[0094] At step 406, the weight of each updated particle is updated based on the one or more sensor detection clusters / sensor detections and the doppler uncertainty model. In an example, at step 406, each particle represents a hypothesis on the state which comprises of ego-motion and acceleration. Thus, each state is a 4x1 vector corresponding to (vx, vy, ax, ay). For each particle, the weight w is initialized as 0. The ambiguity free limit (vaf) is read from the sensor data. For each sensor detection in the sensor data, the uncertain doppler reading (Vambi) is computed using the doppler uncertainty model, as shown in below equation.

[0095] >

[0096]

[0097] 202403892

[0098] 12

[0099] The weight w of the particle is proportional to the agreement of the doppler uncertainty model with the detections. That is for ND detections,

[0100]

[0101] >

[0102] At step 408, the processing circuitry 106 estimates the ego motion states from the set of updated particles. The ego-motion state is calculated as a weighted average of all the particles. The particles with higher weights contribute more to the final estimate.

[0103] At step 410, confidence estimation is performed. The processing circuitry 106 compares the received sensor detections with the expected uncertain doppler readings calculated using the estimated ego-motion states and the doppler uncertainty model. A confidence metric is calculated based on the actual sensor detections and the probable ego-motion state. A higher confidence metric indicates a more accurate ego-motion estimate.

[0104] FIG. 5 illustrates an example-computing environment 500 implementing the system 100 and method 200 as shown in FIG.s 1 and 2 for estimating ego-motion state of the vehicle. As depicted in FIG. 5, the computing environment 500 comprises at least one data processing unit 506 that is equipped with a control unit 502 and an Arithmetic Logic Unit (ALU) 504, a plurality of networking devices 508 and a plurality Input output, I / O devices 510, a memory 512, a storage 514. The data processing unit 506 may be responsible for implementing the system 100 and method 200 described in FIG.s 1 and 2, respectively. For example, the data processing unit 506 in some embodiments be equivalent to the processing circuitry of the platform described above in conjunction with FIG. 1. For another example, the data processing unit 506 in some embodiments be equivalent to the processing circuitry of the platform described above in conjunction with FIG. 1. The data processing unit 506 is capable of executing software instructions stored in memory 512. The data processing unit 506 receives commands from the control unit 502 in order to perform its processing. Further, any logical and arithmetic operations involved in the execution of the instructions are computed with the help of the ALU 504.

[0105] The computer program is loadable into the data processing unit 506, which may, for example, be comprised in an electronic apparatus (such as the platform). When loaded202403892

[0106] 13

[0107] into the data processing unit 506, the computer program may be stored in the memory 512 associated with or comprised in the data processing unit 506. According to some embodiments, the computer program may, when loaded into and run by the data processing unit 506, cause execution of method steps according to, for example, any of the methods illustrated in FIG. 2, FIG. 3 and FIG. 4 as described herein.

[0108] The overall computing environment 500 may be composed of multiple homogeneous and / or heterogeneous cores, multiple CPUs of different kinds, special media and other accelerators. Further, the plurality of data processing unit 506 may be located on a single chip or over multiple chips.

[0109] The algorithm comprising of instructions and codes required for the implementation are stored in either the memory 512 or the storage 514 or both. At the time of execution, the instructions may be fetched from the corresponding memory 512 and / or storage 514, and executed by the data processing unit 506.

[0110] In case of any hardware implementations various networking devices 508 or external I / O devices 510 may be connected to the computing environment to support the implementation through the networking devices 508 and the I / O devices 510.

[0111] The embodiments disclosed herein can be implemented through at least one software program running on at least one hardware device and performing network management functions to control the elements. The elements shown in FIG. 5 include blocks which can be at least one of a hardware device, or a combination of hardware device and software module.202403892

[0112] 14

[0113] Reference signs

[0114] • System 100

[0115] • One or more sensors 102

[0116] • Input / output unit 104

[0117] • Processing circuitry 106

[0118] • Communication unit 108

[0119] • Display / interface 110

[0120] • Computing environment 500 • Control unit 502

[0121] • Arithmetic Logic Unit (ALU) 504 • Processing Unit (PU) 506

[0122] • Networking devices 508

[0123] • Input / output (I / O) devices 510 • Memory 512

[0124] • Storage 514

Claims

20240389215Claims:

1. A method (200) for estimating ego-motion state of a vehicle, c h a r a ct e r i z e d i n t h a t, the method (200) comprises:receiving (202) one or more sensor detection clusters comprising information about one or more detected objects with doppler uncertainties, in a vehicle’s surroundings;creating (204) a doppler uncertainty model for each sensor detection cluster of the one or more sensor detection clusters;initializing (206) a set of particles, wherein each particle in the set of particles represents a probable ego-motion state of the vehicle, and wherein each particle is assigned a weight;propagating (208) each particle in the set of particles using a prediction model to obtain an updated set of particles with a predicted state;updating (210) weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model; andestimating (212) the ego-motion state of the vehicle from the set of updated particles with updated state.

2. The method (200) according to claim 1 , wherein the information comprises at least one of range rate, distance, velocity and azimuth angle of the detected objects.

3. The method (200) according to any of the claims 1 or 2, wherein the ego-motion state is estimated using particle filter framework.

4. The method (200) according to any of the preceding claims, wherein the method (200) comprises:computing an expected uncertain doppler reading for each of the sensor clusters using the doppler uncertainty model.202403892165. The method (200) according to any of the preceding claims, wherein the doppler uncertainty model is generated using a non-linear model.

6. The method (200) according to any of the preceding claims, wherein the ego-motion state comprises ego velocity and yaw rate of the vehicle.

7. The method (200) according to any of the preceding claims, wherein the probable ego-motion state comprises at least one of: ego longitudinal velocities, lateral velocities and / or accelerations.

8. A system (100) for estimating ego-motion of a vehicle, c h a r a c t e r i z e d i n t h a t, the system comprises:processing circuitry (106) configured to:receive one or more sensor detection clusters comprising information about one or more detected objects with doppler uncertainties, in a vehicle’s surroundings;create a doppler uncertainty model for each sensor detection cluster of the one or more sensor detection clusters;initialize a set of particles, wherein each particle in the set of particles represents a probable motion state of the vehicle, and wherein each particle is assigned a weight;propagate each particle in the set of particles using a prediction model to obtain an updated set of particles with predicted state;update weight of each updated particle in the set of updated particles based on the one or more sensor detection clusters and the doppler uncertainty model; andestimate the ego-motion state of the vehicle from the set of updated particles with updated state.202403892179. The system (100) according to claim 8, wherein the information comprises range rate, distance, velocity and azimuth angle of the detected objects.

10. The system (100) according to any of the claims 8 or 9, wherein the processing circuitry (106) is configured to estimate the ego-motion state using particle filter framework.

11. The system (100) according to any of the claims 8-10, wherein the processing circuitry (106) is configured to compute an expected uncertain doppler reading for each of the sensor clusters using the doppler uncertainty model.

12. The system (100) according to any of the claims 8-11 , wherein the processing circuitry (106) is configured to generate doppler uncertainty model using a non-linear model.

13. The system (100) according to any of the claims 8-12, wherein the ego-motion state comprises ego velocity and yaw rate of the vehicle and wherein the probable egomotion state comprises at least one of: ego longitudinal velocities, lateral velocities and / or accelerations.

14. A computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out a method of any one of the claims 1 to 7.

15. A computer-readable medium having stored thereon the computer program of the claim 14.

Images