Method and device for multi-sensor data fusion for automated and autonomous vehicles
A computer-implemented method estimates a course of a roadway in a surrounding area of a vehicle on the basis of a state function describing the course of a roadway. The state function includes a clothoid spline. The method includes providing surrounding-area measurement data describing the course of a roadway at a current position of the vehicle, the surrounding-area measurement data having at least one polynomial function. The method also includes transforming the state function and the surrounding-area measurement data into a common coordinate system. The method further includes adapting the state function on the basis of the surrounding-area measurement data in the common coordinate system.
1. A computer-implemented method for estimating a course of a roadway in a surrounding area of a vehicle on the basis of a state function describing the course of a roadway, the state function comprising a clothoid spline, the method comprising:
providing surrounding-area measurement data describing the course of the roadway at a current position of the vehicle, the surrounding-area measurement data comprising at least one polynomial function;
transforming the state function into a different coordinate system such that the state function is transformed from a state space to a measurement space, wherein the state function and the surrounding-area measurement data are both in a common coordinate system as a result of said transforming; and
after said transforming, adapting the state function on the basis of the surrounding-area measurement data in the common coordinate system, wherein adapting the state function further comprises using a sampling method for selecting sample points from measurement models and from surrounding area measurement data such that there is a constant arc length between neighbouring sample points in each case when there are multiple sample points.
2. The computer-implemented method as claimed in claim 1 , wherein the surrounding-area measurement data is recorded by at least one camera.
3. The computer-implemented method as claimed in claim 1 , wherein the surrounding-area measurement data is provided by a map.
4. The computer-implemented method as claimed in claim 1 , wherein the surrounding-area measurement data comprises first surrounding-area measurement data recorded by at least one camera and second surrounding-area measurement data provided by a map.
5. The computer-implemented method as claimed in claim 1 , wherein the common coordinate system comprises location coordinates.
6. The computer-implemented method as claimed in claim 1 , wherein the common coordinate system comprises curvature values.
7. The computer-implemented method as claimed in claim 1 , wherein adapting the state function to the surrounding-area measurement data provides an adaptation of at least one curvature value of the state function to at least one further curvature value of the surrounding-area measurement data.
8. The computer-implemented method as claimed in claim 1 , wherein the state function identifies a course of a roadway marking of a roadway.
9. The computer-implemented method as claimed in claim 8 , further comprising ascertaining a course of a lane center by transforming the state function.
10. The computer-implemented method as claimed in claim 1 , further comprising providing validation measurement data and a validation of the estimation of the course of the roadway with the aid of the validation measurement data.
11. A device for estimating the course of a roadway in the surrounding area of the vehicle, comprising:
one or more interfaces configured to record surrounding-area measurement data;
a processing unit configured to perform the computer-implemented method as claimed in claim 1 .
12. A vehicle comprising the device as claimed in claim 11 .
13. The method of claim 1 wherein the state function is represented in the state space by arc lengths and curvature values, wherein a measurement model is used to transform the clothoid spline from the from the state space to the measurement space, and wherein the measurement space serves as the common coordinate system comprising two-dimensional location coordinates.
14. The method of claim 13 wherein the measurement model provides an approximate representation of the clothoid spline in the form of a polynomial, wherein sample points generated by the polynomial approximate the clothoid spline.
15. The method of claim 14 wherein the polynomial is a third degree polynomial.
16. The method of claim 13 wherein the measurement model provides a parameter representation of the clothoid spline, wherein sample points generated by the parameter representation approximate the clothoid spline.
17. The method of claim 16 wherein each of the sample points of the parameter representation is represented in a vector representation.
18. The method of claim 13 wherein a sampling method is used to generate sample points based on the measurement model, wherein the sampling method includes (i) determining sample points such that a distance between the sample points in an x-direction is constant, or (ii) determining sample points such that there is a constant arc length change between neighbouring sample points.
19. The method of claim 2 wherein images of roadway markings or roadway boundaries captured by said at least one camera are processed in order to approximate the course of the roadway by at least one polynomial function.