Method and apparatus for representing a map element and method and apparatus for locating a vehicle/robot

• US 10,643,103 B2
• Filed: 11/20/2017
• Issued: 05/05/2020
• Est. Priority Date: 06/18/2015
• Status: Active Grant

• Current Assignee: Bayerische Motoren Werke AG
• Original Assignee: Bayerische Motoren Werke AG
• Inventors: Isert, Carsten, Jiang, Wanli, Li, Qianshan
• Primary Examiner(s): Garcia, Santiago
• Application Number:
  US15/817,392
• Publication Number:
  US 20180165544A1
• Time in Patent Office: 897 Days
• CPC Class Codes: G01C 21/3635 Guidance using 3D or perspective road maps
• CPC Class Codes: G05D 1/0231 using optical position detecting means position-fixing by using electromagnetic waves other than radio waves, e.g. optical position detecting means G01S5/16
• CPC Class Codes: G06F 18/2137 based on criteria of topology preservation, e.g. multidimensional scaling or self-organising maps
• CPC Class Codes: G06T 17/10 Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes
• CPC Class Codes: G06V 20/10 Terrestrial scenes scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59

Abstract

A method and an apparatus are provided for representing a map element and for locating a vehicle/robot based thereupon. The method for representing a map element includes: generating a Gaussian Mixture Model for the map element; generating a signature for identifying the map element, wherein the signature includes properties of the map element; and generating a Signatured Gaussian Mixture Model for representing the map element, wherein the Signatured Gaussian Mixture Model includes the Gaussian Mixture Model and the signature.

Claims

1. A computer-implemented method for representing a map element, comprising the acts of:

generating a Gaussian Mixture Model for the map element;

generating a signature for identifying the map element, wherein the signature comprises properties of the map element; and

generating a Signatured Gaussian Mixture Model for representing the map element, wherein the Signatured Gaussian Mixture Model comprises the Gaussian Mixture Model and the signature, and the Signatured Gaussian Mixture Model further comprises a dynamic existence probability of the map element.

2. The method according to claim 1, wherein the map element is a geometry or point cluster in a digital map or an image segment in an image.

3. The method according to claim 1, wherein generating the Gaussian Mixture Model for the map element comprises:

using Gaussian Mixture Models of geometric primitives to approximate a shape of the map element.

4. The method according to claim 3, wherein the geometric primitives comprise at least one of:

a line segment, an arc segment, a spiral segment, a planar rectangle, a planar triangle, a planar circle/sphere, a planar ellipse, a sphere surface patch, a cylinder surface patch, a 3D cube, a 3D sphere and a 3D cylinder.

5. The method according to claim 1, wherein the properties of the map element

comprise at least of one of type, reflectivity and flatness of the map element.

6. A computer implemented method for locating a vehicle/robot, comprising the acts of:

reading a Signatured Gaussian Mixture Map for a map section which the vehicle/robot is currently located in, wherein the Signatured Gaussian Mixture Map comprises Signatured Gaussian Mixture Models for map elements within the map section, and wherein the Signatured Gaussian Mixture Models are generated by generating a Gaussian Mixture Model for each of the map elements;

generating a signature for identifying each of the map elements, wherein each signature comprises properties of a respective map element and generating a Signatured Gaussian Mixture Model for representing each of the map elements, wherein each of the Signatured Gaussian Mixture Models comprises the Gaussian Mixture Model and the signature;

generating the Signatured Gaussian Mixture Models for the map elements within a real-time point cloud or an image acquired by the vehicle/robot;

establishing one or more correspondences between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for the map elements within the real-time point cloud or image based on signatures of the Signatured Gaussian Mixture Models for the map elements within the real-time point cloud or image; and

matching the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for the map elements within the real-time point cloud or image based on the one or more correspondences established.

7. The method according to claim 6, further comprising the act of:

reading information on the map section which the vehicle/robot is currently located in.

8. The method according to claim 6, further comprising the act of:

reading information on prior pose of the vehicle/robot and data reflecting the uncertainty area of the prior pose.

9. The method according to claim 6, wherein matching the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for map elements within the realtime point cloud or image comprises:

optimizing a Euclidean distance between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image.

10. The method according to claim 9, wherein optimizing the Euclidean distance between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image comprises:

optimizing the Euclidean distance with pose of the vehicle/robot constrained on predefined space and within the uncertainty area.

11. An apparatus for representing a map element, comprising:

a memory, having computer executable instructions stored therein; and

a processor, coupled to the memory and configured to;

generate a Gaussian Mixture Model for the map element;

generate a signature for identifying the map element, wherein the signature comprises properties of the map element; and

generate a Signatured Gaussian Mixture Model for representing the map element, wherein the Signatured Gaussian Mixture Model comprises the Gaussian Mixture Model and the signature, and the Signatured Gaussian Mixture Model further comprises a dynamic existence probability of the map element.

12. The apparatus according to claim 11, wherein generating a Gaussian Mixture Model for the map element comprises:

using Gaussian Mixture Models of geometric primitives to approximate a shape of the map element.

13. An apparatus for locating a vehicle/robot, comprising:

a memory, having computer executable instructions stored therein; and

a processor, coupled to the memory and configured to;

read a Signatured Gaussian Mixture Map for a map section which the vehicle/robot is currently located in, wherein the Signatured Gaussian Mixture Map comprises Signatured Gaussian Mixture Models for map elements within the map section, and wherein the Signatured Gaussian Mixture Models are generated by generating a Gaussian Mixture Model for each of the map elements;

generating a signature for identifying each of the map elements, wherein each signature comprises properties of a respective map element and generating a Signatured Gaussian Mixture Model for representing each of the map elements, wherein each of the Signatured Gaussian Mixture Models comprises the Gaussian Mixture Model and the signature;

generate the Signatured Gaussian Mixture Models for the map elements within a realtime point cloud or an image acquired by the vehicle/robot;

establish one or more correspondences between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the realtime point cloud or image based on signatures of the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image; and

match the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image based on the correspondence established.

14. The apparatus according to claim 13, wherein the processor is further configured to:

read information on prior pose of the vehicle/robot and data reflecting the uncertainty area of the prior pose.

15. The apparatus according to claim 13, wherein matching the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for map elements within the realtime point cloud or image comprises:

optimizing the Euclidean distance between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image.

16. The apparatus according to claim 15, wherein optimizing the Euclidean distance between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image comprises:

optimizing the Euclidean distance with pose of the vehicle/robot constrained on predefined space and within the uncertainty area.

17. A non-transient storage medium having instructions

stored thereon that when executed cause a processor to implement computer-implemented method for representing a map element, comprising:

instructions for causing the processor to generate a Gaussian Mixture Model for the map element;

instructions for causing the processor to generate a signature for identifying the map element, wherein the signature comprises properties of the map element; and

instructions for causing the processor to generate a Signatured Gaussian Mixture Model for representing the map element, wherein the Signatured Gaussian Mixture Model comprises the Gaussian Mixture Model and the signature, and the Signatured Gaussian Mixture Model further comprises a dynamic existence probability of the map element.

18. A non-transient storage medium having instructions stored thereon that when executed cause a processor to implement computer-implemented method for locating a vehicle/robot, comprising:

instructions for causing the processor to read a Signatured Gaussian Mixture Map for a map section which the vehicle/robot is currently located in, wherein the Signatured Gaussian Mixture Map comprises Signatured Gaussian Mixture Models for map elements within the map section, and wherein the Signatured Gaussian Mixture Models are generated by generating a Gaussian Mixture Model for each of the map elements;

generating a signature for identifying each of the map elements, wherein each signature comprises properties of a respective map element and generating a Signatured Gaussian Mixture Model for representing each of the map elements, wherein each of the Signatured Gaussian Mixture Models comprises the Gaussian Mixture Model and the signature;

instructions for causing the processor to generate the Signatured Gaussian Mixture Models for the map elements within a real-time point cloud or an image acquired by the vehicle/robot;

instructions for causing the processor to establish one or more correspondences between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image based on signatures of the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image; and

instructions for causing the processor to match the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image based on the one or more correspondences established.

19. An apparatus for representing a map element,

comprising:

a Gaussian Mixture Model generation unit configured to generate a Gaussian Mixture Model for the map element;

a signature generation unit configured to generate a signature for identifying the map element, wherein the signature comprises properties of the map element; and

a Signatured Gaussian Mixture Model generation unit configured to generate a Signatured Gaussian Mixture Model for representing the map element, wherein the Signatured Gaussian Mixture Model comprises the Gaussian Mixture Model and the signature, and the Signatured Gaussian Mixture Model further comprises a dynamic existence probability of the map element.

20. An apparatus for locating a first vehicle/robot, comprising:

a Signature Gaussian Mixture Map read unit configured to read Signature Gaussian Mixture Map for a map section which the first vehicle/robot is currently located in, wherein the Signatured Gaussian Mixture Map comprises Signatured Gaussian Mixture Models for map elements within the map section, and wherein the Signatured Gaussian Mixture Models are generated by generating a Gaussian Mixture Model for each of the map elements;

generating a signature for identifying each of the map elements, wherein each signature comprises properties of a respective map element and generating a Signatured Gaussian Mixture Model for representing each of the map elements, wherein each of the Signatured Gaussian Mixture Models comprises the Gaussian Mixture Model and the signature;

a Signatured Gaussian Mixture Model generation unit configured to generate the Signatured Gaussian Mixture Models for the map elements within a real-time point cloud or an image acquired by the first vehicle/robot;

a correspondence establishing unit configured to establish one or more correspondences between the Signatured Gaussian Mixture Map and the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image based on signatures of the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image; and

a matching unit configured to match the Signatured Gaussian Mixture Map with the Signatured Gaussian Mixture Models for map elements within the real-time point cloud or image based on the one or more correspondences established.

21. A second vehicle comprising the apparatus according to claim 20.