阅读说明：本技术 用于物体跟踪的航向角估算 (Course angle estimation for object tracking ) 是由 W·党 J·K·希夫曼 K·威什瓦吉特 S·陈 于 2020-04-29 设计创作，主要内容包括：一种跟踪物体的说明性示例方法包括：随时间检测物体上点以获得多个检测；确定检测中的每一个的位置；确定所确定的位置之间的关系；以及基于该关系确定该物体的估算的航向角。(An illustrative example method of tracking an object includes: detecting points on the object over time to obtain a plurality of detections; determining a location of each of the detections; determining a relationship between the determined locations; and determining an estimated heading angle of the object based on the relationship.)

1. A method of tracking an object, the method comprising:

detecting one or more points on the object over time to obtain a plurality of detections;

determining a location of each of the detections;

determining a spatial relationship between the determined locations; and

an estimated heading angle of the object is determined based on the spatial relationship.

2. The method of claim 1, wherein the step of removing the metal oxide layer comprises removing the metal oxide layer from the metal oxide layer

Determining the spatial relationship comprises defining a shape that encompasses the determined location; and is

Determining the estimated heading angle of the object includes determining an orientation of a defined shape.

3. The method of claim 2, wherein the defined shape is at least one of an arc, a line, and a rectangle.

4. The method of claim 2, wherein determining the estimated heading angle is based on a determined orientation and range-rate information related to at least some of the detections.

5. The method of claim 1, comprising: detecting the one or more points at least until a distance spanned by the determined location exceeds a preselected threshold distance.

6. The method of claim 1, comprising: correcting a previously estimated heading angle of the object using the determined estimated heading angle.

7. The method of claim 6, wherein the previously estimated heading angle of the object is determined by a Kalman filter.

8. The method of claim 1, wherein determining the estimated heading angle comprises using:

the estimation of the likelihood is carried out,

the least-squares estimation is performed,

principal component analysis, or

And (4) Hough transform.

9. An apparatus for tracking an object, the apparatus comprising:

a detector that detects one or more points on the object over time; and

a processor configured to: determining a position of each of the detected one or more points, determining a spatial relationship between the determined positions, and determining an estimated heading angle of the object based on the relationship.

10. The device of claim 9, wherein the processor is configured to:

determining the spatial relationship by defining a shape that encompasses the determined location; and

determining the estimated heading angle of the object by determining an orientation of a defined shape.

11. The apparatus of claim 10, wherein the defined shape is at least one of an arc, a line, and a rectangle.

12. The device of claim 10, wherein the processor is configured to: determining the estimated heading angle based on the determined orientation and range-rate information related to at least some of the detected one or more points.

13. The apparatus of claim 9, wherein the distance spanned by the determined locations exceeds a preselected threshold distance.

14. The apparatus of claim 9, comprising a kalman filter that determines an initial estimated heading angle of the object, and wherein the processor is configured to provide the determined estimated heading angle to the kalman filter to correct the initial estimated heading angle of the object.

15. The device of claim 9, wherein the processor configured for determining the estimated heading angle comprises using:

the estimation of the likelihood is carried out,

the least-squares estimation is performed,

principal component analysis, or

And (4) Hough transform.

16. An apparatus for tracking an object, the apparatus comprising:

detecting means for detecting one or more points on the object over time; and

Determining means for determining: determining a position of each of the detected one or more points, determining a spatial relationship between the determined positions; and determining an estimated heading angle of the object based on the relationship.

17. The apparatus of claim 16,

the spatial relationship defines a shape that encompasses the determined location; and is

The determining means determines the estimated heading angle of the object by determining an orientation of a defined shape.

18. The apparatus of claim 17, wherein the defined shape is at least one of an arc, a line, and a rectangle.

19. The apparatus of claim 16, wherein the determining means determines the estimated heading angle based on the determined orientation and range-rate information related to at least some of the detected one or more points.

20. The apparatus of claim 16, wherein the distance spanned by the determined locations exceeds a preselected threshold distance.

Background

Various sensor types have proven useful for detecting objects near the vehicle or in the path of the vehicle. Example sensor types include ultrasound, radio detection and ranging (RADAR), and light detection and ranging (LIDAR). The manner in which such sensors are used in passenger vehicles has increased.

One challenge associated with tracking objects using such sensors is that objects may have varying shapes and sizes that impair the ability of the sensors to determine certain characteristics of the object, such as the direction in which the object is heading. Known Kalman filters are designed to quickly provide heading angle or direction information about the tracked object. Although kalman filter estimation is often very useful, the estimation may not be accurate.

The kalman filter is designed to operate based on tracking the movement of a single point. Since the three-dimensional object has multiple points that can be detected by the sensor, the kalman filter may interpret the detector information about the multiple points as if it indicates movement of a single point. In other words, the kalman filter is unable to distinguish between the detection of multiple different points on an object and the movement of a single point on the object. Given the possibility of interpreting different points on an object as if they were the same point, sensor information about these points may be mistaken for movement of a single point, resulting in inaccurate tracking information.

Disclosure of Invention

An illustrative example method of tracking an object includes: detecting one or more points on the object over time to obtain a plurality of detections; determining a location of each of the detections; determining a relationship between the determined locations; and determining an estimated heading angle (heading angle) of the object based on the relationship.

In an example embodiment having one or more features of the method of the previous paragraph, determining the spatial relationship includes: defining a shape that encompasses the determined location; and determining the estimated heading angle of the object includes determining an orientation of the defined shape.

In an example embodiment having one or more features of the method of the previous paragraph, the defined shape is at least one of an arc, a line, and a rectangle.

In an example embodiment having one or more features of the method of the preceding paragraph, determining the estimated heading angle is based on the determined orientation and range rate information related to at least some of the detections.

Example embodiments having one or more features of the method of the previous paragraph include: one or more points are detected at least until the distance spanned by the determined location exceeds a preselected threshold distance.

Example embodiments having one or more features of the method of the previous paragraph include: the determined estimated heading angle is used to correct a previously estimated heading angle of the object.

In an example embodiment having one or more features of the method of the preceding paragraph, the previously estimated heading angle of the object is determined by a kalman filter.

In an example embodiment having one or more features of the method of any of the preceding paragraphs, determining the estimated heading angle includes using: likelihood estimation, least squares estimation, principal component analysis, or hough transform.

An illustrative example apparatus for tracking an object includes: a detector that detects one or more points on the object over time; and a processor configured to: the method further includes determining a position of each of the detected one or more points, determining a spatial relationship between the determined positions, and determining an estimated heading angle of the object based on the relationship.

In an example embodiment of one or more features of the apparatus of the previous paragraph, the processor is configured to: determining the spatial relationship by defining a shape that encompasses the determined location; and determining an estimated heading angle of the object by determining an orientation of the defined shape.

In an example embodiment of one or more features of the apparatus of the previous paragraph, the defined shape is at least one of an arc, a line, and a rectangle.

In an example embodiment of one or more features of the apparatus of the previous paragraph, the processor is configured to: an estimated heading angle is determined based on the determined orientation and range-rate information associated with at least some of the detected one or more points.

In an example embodiment having one or more features of the apparatus of the preceding paragraph, the distance spanned by the determined locations exceeds a preselected threshold distance.

An example embodiment having one or more features of the apparatus of the preceding paragraph includes a kalman filter that determines an initial estimated heading angle of the object, and the processor is configured to provide the determined estimated heading angle to the kalman filter to correct the initial estimated heading angle of the object.

In an example embodiment having one or more features of the device of any of the preceding paragraphs, the processor is configured to: determining an estimated heading angle using: likelihood estimation, least squares estimation, principal component analysis, or hough transform.

An illustrative example apparatus for tracking an object includes: detecting means for detecting one or more points on the object over time; and determining means for determining: determining a position of each of the detected one or more points, determining a spatial relationship between the determined positions; and determining an estimated heading angle of the object based on the relationship.

In an example embodiment having one or more features of the apparatus of the preceding paragraph, the spatial relationship defines a shape that encompasses the determined location; and the determining means determines an estimated heading angle of the object by determining an orientation of the defined shape.

In an example embodiment of one or more features of the apparatus of the previous paragraph, the defined shape is at least one of an arc, a line, and a rectangle.

In an example embodiment having one or more features of the method of the apparatus of the preceding paragraph, the determining means determines the estimated heading angle based on the determined orientation and range-rate information relating to at least some of the detected one or more points.

In an example embodiment having one or more features of the apparatus of the preceding paragraph, the distance spanned by the determined locations exceeds a preselected threshold distance.

The various features and advantages of at least one disclosed example embodiment will become apparent to those skilled in the art from the following detailed description. The drawings that accompany the detailed description can be briefly described as follows.

Drawings

Fig. 1 diagrammatically shows an example use of the apparatus for tracking an object.

FIG. 2 schematically illustrates selected portions of an example object tracking device.

FIG. 3 is a flow chart summarizing an example method of tracking an object.

Fig. 4 graphically illustrates log-likelihood functions used in example embodiments.

FIG. 5 graphically illustrates an angle grid used in an example embodiment to determine an object heading angle.

FIG. 6 shows another example angular grid having a finer resolution than the grid shown in FIG. 5.

Detailed Description

Fig. 1 diagrammatically shows an apparatus 20 for tracking a moving object 22. In this example, the device 20 is located on a vehicle 24. For purposes of discussion, the moving object 22 is another vehicle that is located forward of the vehicle 24 and at least partially in the path of the vehicle 24. The device 20 tracks the vehicle 22 using RADAR signaling as schematically shown at 26.

As schematically shown in fig. 1, the vehicle 22 includes a center of mass 32. The vehicle 22 travels along a curved trajectory, such as a turn (to the right according to the drawing). In such a case, the velocity vector of the center of mass 32 of the vehicle 22 is located at the heading angle 34 in the case shown in FIG. 1. The coordinate system in which the heading angle 34 is determined may be based on a world coordinate system. Alternatively, the coordinate system may be fixed relative to the vehicle 24 or the apparatus 20.

In this document, the pointing angle refers to the body orientation angle of a moving object such as the vehicle 22 that the apparatus 20 is tracking. The body orientation angle or pointing angle is the azimuth direction in which the center line or longitudinal axis of the moving body points.

In this document, the heading angle is the direction of movement of a particular reference point on a moving object (such as the vehicle 22). It is noted that in certain contexts, such as aviation, the term "heading angle" is used to refer to an angle referred to in this document as a "heading angle". Also, in the airborne context, the term "tracking" is used to refer to tracking referred to in this document as "heading angle".

Fig. 2 schematically illustrates selected portions of the apparatus 20 including an emitter 42 and a detector 44. Emitter 42 emits radiation in an outward direction, and when such radiation reflects from an object, the reflected radiation is received and detected by detector 44. In some exemplary embodiments, the emitter 42 and the detector 44 operate according to known RADAR principles and techniques. Other embodiments include emitter and detector configurations that may be used in LIDAR or ultrasonic detection technologies.

The apparatus 20 includes a filter 46, the filter 46 being configured to estimate dynamic quantities of the tracked object, such as the position, velocity, acceleration and trajectory curvature of the object. In some example embodiments, the filter 46 operates according to known principles of a Kalman filter. In this example, the filter 46 provides information indicative of a heading angle of at least one reference point on the moving object 22. For example, the filter 46 provides information indicative of the heading angle 34 of the center of mass 32 of the vehicle 22.

The filter 46 is able to provide information about the heading angle of the center of mass of the moving object, however, the filter 46 is unable to distinguish between multiple detection points on the object or vehicle 22. For example, in FIG. 1, the detector 44 may receive reflected radiation from a plurality of points A, B, C, D at various locations on the vehicle 22. The filter 46 cannot distinguish between information about those detected points a-D and information about a single point detected multiple times. In other words, the filter 46 will interpret information about the points A-D (such as the rate of change of distance) as if the information indicates movement of a single point rather than movement of multiple different points. When the filter 46 operates as a kalman filter, it treats all detected point information as if it were related to a single point.

The device 20 includes a processor 50, and the processor 50 may be part of a dedicated microprocessor or another computing device supported on the vehicle 24. In this embodiment, the processor 50 is configured for determining an estimated heading angle based on position information over time with respect to one or more detection points A-D on the object 22.

A memory 52 is associated with the processor 50. In some example embodiments, the memory 52 includes computer-executable instructions that cause the processor 50 to operate for the purpose of tracking a moving object and determining a pointing angle or a body orientation angle of the object. In some example embodiments, memory 52 at least temporarily contains information about various characteristics or features of detected points on tracked object 22 to assist processor 50 in making a desired determination as to the position or movement of such object.

Example embodiments of the present invention allow for more accurate determination of the heading angle of a moving object, such as vehicle 22. This feature is particularly useful at the beginning of a tracking session when the filter 46 has relatively limited information for determining the heading angle. Object tracking is improved by the processor 50 providing heading angle estimates to the filter 46. Accordingly, embodiments of the present invention provide improvements in tracking technology and improvements in vehicle control based on information about moving objects in the vicinity of the vehicle or in the path of the vehicle.

In some embodiments, the filter 46 operates to provide an indication of the presence of an object even though the heading angle has not been defined. The method allows for early object detection and subsequent determination or refinement of the heading angle.

FIG. 3 is a flowchart 60 that outlines an example method of tracking an object 22, including estimating a heading angle using position information about one or more points A-D on the tracked object. At 62, detector 44 detects radiation reflected from one or more points on object 22 over time. Over time, multiple detections provide information about the movement of object 22 as object 22 moves. At 64, the processor 50 determines a location for each of the detections, which corresponds to a respective location of the points A-D detected over time. In this example, the location information is determined in a world coordinate system.

At 66, the processor 50 determines the relationship between the locations determined at 64. The relationship in this example includes the detected relative position over the area defining the shape. For example, the determined relationship defines an arc, line, or rectangle that encompasses or includes the determined location. The processor determines the orientation of the shape within the world coordinate system and uses the orientation to determine an estimated heading angle of the object at 68. Thus, the determination of the heading angle is based on the relationship determined at 66.

In some cases, the orientation of the shape defined by the relationship determined at 66 will indicate the path that the tracked object is moving, and not the direction of movement along the path. In this example, the processor 50 uses the range rate information about the plurality of detections to determine the direction of movement.

At 70, the processor 50 provides the determined estimated heading angle to the filter 46. The filter 46 will update the previous heading angle determined by the filter 46 based on the estimated heading angle from the processor 50. In some examples, the filter 46 will replace the previously determined heading angle with the estimated heading angle from the processor 50.

Having sufficient detection over time allows the processor 50 to make a more accurate estimate of the heading angle. In an example embodiment, the detected locations span a distance that exceeds a preselected threshold. Having a sufficiently large distance between the farthest detected positions increases the likelihood that the object 22 has moved sufficiently for the position information determined by the processor 50 to accurately estimate the heading angle. In view of this description, those skilled in the art will be able to select an appropriate threshold distance to meet the needs of their particular situation.

The processor determines the estimated heading angle using one of several techniques. In one embodiment, the likelihood estimation provides the most likely heading angle based on the detected position.

Suppose that there is a world coordinate system with a position x ═ x₁,...,x_N]And y ═ y₁,...,y_N]N tests. The log-likelihood of detection of a heading angle θ, l (x, y | θ), roughly describes the degree to which the detection trace can fit the trajectory at the heading angle θ. This embodiment includes the following assumptions: each detection has a uniform position likelihood inside the object 22. Due to detectionThe possibility that the detection position is outside the object 22 gradually decreases to zero as the detection position is further away from the object 22.

This example includes defining a body coordinate system of the object such that its longitudinal axis is parallel to the object heading angle 34 and its origin is the same as the origin of the world coordinates. If the object tracking has a centroid (x) in world coordinates_t,y_t) Then, the position of the centroid in the body coordinate system of the object is:

detection in the body coordinate System of the object (x)_i，y_i) The positions of (A) are:

log-likelihood of single detection l (x)_t,y_t| θ) can be adequately represented by detecting the orthogonal position relative to the centroid of the object, which can be expressed as:

FIG. 4 shows as O'_iAn example of a log-likelihood of detection of the function of (a). The width W of the cone on the left and right sides of the graph 74 may depend on the nominal standard deviation of the object pose, object range, and angular error in detection. It follows that

W＝W_min+|cos β|(max(r，r_min)-r_min)e

Where β is the object heading angle relative to the host vehicle 24, r is the object range, and e is the nominal standard deviation of the angular error in detection. Parameter W_minAnd r_minIs predetermined.

Assuming independence between the multiple detections, the overall log-likelihood can be written as:

in this example, the total likelihood is determined by the processor 50 as the sum of the individual likelihoods normalized by the number of detections. In one example embodiment, if the longitudinal span of the shape determined from the relationship determined at 66 is less than the threshold distance, the total likelihood will take a low value. Detection of a near current object position is preferred over detection of more distant to account for any manipulation of object 22 at longer distances. Additionally, when there are too few detections available over time, the processor 50 assigns a low value to the total likelihood.

The maximum likelihood estimate of the heading angle may be obtained by solving the following equation:

the optimization equation cannot be solved analytically. To numerically address this issue, an exemplary embodiment includes two angular searches to find the optimal heading angle π. Note that since the detection likelihoods at the heading angles θ and θ + pi are the same, it is sufficient to search inside θ ∈ [0, pi ].

Fig. 5 and 6 show two steps of angle search. The corresponding detection likelihood is first evaluated on a coarse angular grid, as shown in fig. 5. If none of the likelihoods exceeds the preselected threshold at 80, the processor 50 will stop the heading angle determination due to insufficient information in the detection. When at least one of the detections meets the threshold 80, the processor selects angle points for which the likelihood is above the threshold 80 and builds a fine angle grid covering a range of these angle points, as shown in fig. 6. Since the heading angles 0, 0.8 π and 0.9 π in FIG. 5 each exceed the threshold 80, the processor uses these angles as the angles of interest in the fine grid of FIG. 6. The resolution in fig. 5 is pi/10 and the resolution in fig. 6 is pi/60.

As can be appreciated from this illustration, processor 50 evaluates a plurality of candidate angles around angles 0, 0.8 π and 0.9 π. In this example, the best heading angle that maximizes the likelihood is selected among those candidates as the estimated heading angle pi.

The preceding description is exemplary rather than limiting in nature. Variations and modifications to the disclosed examples may become apparent to those skilled in the art that do not necessarily depart from the essence of this invention. The scope of legal protection given to this invention can only be determined by studying the following claims.