Incremental updates of pose graphs for generating high definition maps for navigating autonomous vehicles

Patent metadata

Field	Value
Publication number	US-10267635-B2
Application number	US-201715857611-A
Country	US
Kind code	B2
Filing date	Dec 28, 2017
Priority date	Dec 30, 2016
Publication date	Apr 23, 2019
Grant date	Apr 23, 2019

Abstract

A high-definition map system receives sensor data from vehicles travelling along routes and combines the data to generate a high definition map for use in driving vehicles, for example, for guiding autonomous vehicles. A pose graph is built from the collected data, each pose representing location and orientation of a vehicle. The pose graph is optimized to minimize constraints between poses. Points associated with surface are assigned a confidence measure determined using a measure of hardness/softness of the surface. A machine-learning-based result filter detects bad alignment results and prevents them from being entered in the subsequent global pose optimization. The alignment framework is parallelizable for execution using a parallel/distributed architecture. Alignment hot spots are detected for further verification and improvement. The system supports incremental updates, thereby allowing refinements of subgraphs for incrementally improving the high-definition map for keeping it up to date.

First claim

What is claimed is:

1. A method for incremental updates of pose graphs for generating high-definition maps, the method comprising:
receiving track data collected by sensors of autonomous vehicles, each track data collected by sensors of an autonomous vehicle driving through a path in a geographical region;
generating a pose graph, wherein each node of the pose graph represents a pose of a vehicle, the pose comprising a location and orientation of the vehicle, and wherein each edge between a pair of nodes represents a transformation between nodes of the pair of nodes;
receiving a set of new track data for updating the pose graph;
performing alignment across poses within the set of new track data;
performing alignment across pairs of poses, wherein each of the pairs of poses comprises a pose from the generated pose graph and a pose from the set of new tracks;
updating the pose graph by adding poses from the set of new tracks to the pose graph;
performing pose graph optimization of the updated pose graph;
generating a high-definition map based on the optimized updated pose graph; and
sending the high-definition map to one or more autonomous vehicles for navigating the autonomous vehicles in the geographical region.

2. The method of claim 1 , wherein the sensor data comprises data collected by LIDAR, data collected by global positioning system (GPS), or data collected by an inertial measurement unit (IMU).

3. The method of claim 1 , wherein performing alignment across poses within the set of new track data comprises:
performing single track alignment across a pair of poses, wherein each pose of the pair is within the same track data.

4. The method of claim 1 , wherein performing alignment across poses within the set of new track data comprises:
performing cross track alignment across a pair of poses, wherein a first pose from the pair is from a first track data and a second pose from the pair is from a second track data.

5. The method of claim 1 , wherein performing pose graph optimization of the updated pose graph comprises:
keeping data of poses from the pose graph fixed while changing poses from the set of new track data.

6. The method of claim 1 , wherein performing pose graph optimization of the updated poser graph comprises:
optimizing poses while changing poses from pose graph and poses from the set of new track data.

7. The method of claim 1 , further comprising:
sending information describing a portion of the pose graph for presentation via a user interface, wherein the portion of the pose graph is for a portion of a geographical region; and
receiving a request to edit the pose graph, the editing comprising one of: removing an edge of the pose graph or adding an edge to the pose graph.

8. The method of claim 7 , wherein the portion of the geographical region is specified using a bounding box specified using latitude and longitude values.

9. The method of claim 8 , further comprising:
responsive to editing the pose graph, reoptimizing the pose graph, the reoptimizing keeping poses associated with locations outside the bounding box fixed and modifying poses associated with locations within the bounding box.

10. The method of claim 1 , further comprising:
responsive to determining that a track data is older than a threshold value, deprecating the track data, the deprecating comprising, deleting all edges based on the track data.

11. A non-transitory computer readable storage medium storing instructions for:
receiving track data collected by sensors of autonomous vehicles, each track data collected by sensors of an autonomous vehicle driving through a path in a geographical region;
generating a pose graph, wherein each node of the pose graph represents a pose of a vehicle, the pose comprising a location and orientation of the vehicle, and wherein each edge between a pair of nodes represents a transformation between nodes of the pair of nodes;
receiving a set of new track data for updating the pose graph;
performing alignment across poses within the set of new track data;
performing alignment across pairs of poses, wherein each of the pairs of poses comprises a pose from the generated pose graph and a pose from the set of new tracks;
updating the pose graph by adding poses from the set of new tracks to the pose graph;
performing pose graph optimization of the updated pose graph;
generating a high-definition map based on the optimized updated pose graph; and
sending the high-definition map to one or more autonomous vehicles for navigating the autonomous vehicles in the geographical region.

12. The non-transitory computer readable storage medium of claim 11 , wherein the sensor data comprises data collected by LIDAR, data collected by global positioning system (GPS), and data collected by an inertial measurement unit (IMU).

13. The non-transitory computer readable storage medium of claim 11 , wherein instructions for performing alignment across poses within the set of new track data comprise instructions for:
performing single track alignment across a pair of poses, wherein each pose of the pair is within the same track data.

14. The non-transitory computer readable storage medium of claim 11 , wherein instructions for performing alignment across poses within the set of new track data comprise instructions for:
performing cross track alignment across a pair of poses, wherein a first pose from the pair is from a first track data and a second pose from the pair is from a second track data.

15. The non-transitory computer readable storage medium of claim 11 , wherein instructions for performing pose graph optimization of the updated pose graph comprise instructions for:
keeping data of poses from the pose graph fixed while changing data of poses from the set of new track data.

16. The non-transitory computer readable storage medium of claim 11 , wherein instructions for performing pose graph optimization of the updated pose graph comprise instructions for:
optimizing poses while changing poses from pose graph and poses from the set of new track data.

17. The non-transitory computer readable storage medium of claim 11 , wherein the stored instructions are further for:
sending information describing a portion of the pose graph for presentation via a user interface, wherein the portion of the pose graph is for a portion of a geographical region; and
receiving a request to edit the pose graph, the editing comprising one of: removing an edge of the pose graph or adding an edge to the pose graph.

18. A computer system comprising:
an electronic processor; and
a non-transitory computer readable storage medium storing instructions executable by the electronic processor, the instructions for:
receiving track data collected by sensors of autonomous vehicles, each track data collected by sensors of an autonomous vehicle driving through a path in a geographical region;
generating a pose graph, wherein each node of the pose graph represents a pose of a vehicle, the pose comprising a location and orientation of the vehicle, and wherein each edge between a pair of nodes represents a transformation between nodes of the pair of nodes;
receiving a set of new track data for updating the pose graph;
performing alignment across poses within the set of new track data;
performing alignment across pairs of poses, wherein each of the pairs of poses comprises a pose from the generated pose graph and a pose from the set of new tracks;
updating the pose graph by adding poses from the set of new tracks to the pose graph;
perform

Assignees

• Deepmap Inc

Inventors

• Chen Chen

Classifications

• G01C11/12 Primary

  the pictures being supported in the same relative position as when they were taken · CPC title

• G06T17/05 Primary

  Geographic models · CPC title

• G01S19/47

  the supplementary measurement being an inertial measurement, e.g. tightly coupled inertial · CPC title

• G01S19/46

  the supplementary measurement being of a radio-wave signal type · CPC title

• B60W2552/00

  Input parameters relating to infrastructure · CPC title