Method and apparatus for optimizing scan data and method and apparatus for correcting trajectory

Patent metadata

Field	Value
Publication number	US-11567496-B2
Application number	US-201716610127-A
Country	US
Kind code	B2
Filing date	Sep 20, 2017
Priority date	Dec 26, 2016
Publication date	Jan 31, 2023
Grant date	Jan 31, 2023

Abstract

A method and an apparatus optimizes scan data obtained by sensors on vehicle, and corrects trajectory for a vehicle/robot based on the optimized scan data. The method for optimizing the scan data obtained by scanning environment elements, includes: step of obtaining the scan data, including obtaining at least two frames of scan data respectively corresponding to different timings; step of cluster processing, based on the characteristic of the data points, including classifying the plurality of data points in each frame of the scan data into one or more clusters; step of establishing correspondence, among the at least two frames of scan data, including searching and obtaining at least one set of clusters having correspondence; step of optimizing clusters, among the at least two frames of scan data, including conducting calculation to each set of the at least one set of clusters having correspondence, to obtain optimized clusters respectively corresponding to each set of the at least one set of clusters having correspondence; and step of optimizing the scan data, including accumulating all optimized clusters to obtain an optimized scan date for the at least two frames of scan data.

First claim

What is claimed is:

1. A method for optimizing scan data obtained by scanning environment elements, the method comprising the steps of:
obtaining the scan data by obtaining at least two frames of scan data respectively corresponding to different timings, wherein
each frame of the scan data includes a plurality of data points;

cluster processing, based on a characteristic of the data points, by classifying the plurality of data points in each frame of the scan data into one or more clusters, wherein
the one or more clusters represent map elements corresponding to the environment elements;

establishing correspondence among the at least two frames of scan data by searching and obtaining at least one set of clusters having correspondence;
optimizing clusters among the at least two frames of scan data by calculating each set of the at least one set of clusters having correspondence to obtain optimized clusters respectively corresponding to each set of the at least one set of clusters having correspondence; and
optimizing the scan data by accumulating all optimized clusters to obtain an optimized scan data for the at least two frames of scan data.

2. The method for optimizing scan data according to claim 1 , wherein
in the step of cluster processing, attaching a descriptor for each cluster, wherein
the descriptor describes properties of the cluster.

3. The method for optimizing scan data according to claim 2 , wherein, in the step of establishing correspondence,
among the at least two frames of scan data, searching and obtaining a set of clusters having correspondence, by comparing content of the descriptors of the clusters.

4. The method for optimizing scan data according to claim 3 , wherein
the difference of the descriptors of the set of clusters having correspondence is lower than a predetermined threshold.

5. The method for optimizing scan data according to claim 1 , wherein, in the step of cluster processing,
based on a Euclidean distance between the data points and a consistency qualification as satisfied by data points, classifying the plurality of data points in each frame of the scan data into one or more clusters, wherein
the consistency qualification is based on the relationship between the characteristic of different data points.

6. The method for optimizing scan data according to claim 5 , wherein, in the step of cluster processing,
classifying the plurality of data points in each frame of the scan data into one or more clusters, by:
selecting a data point which does not belong to any cluster as a seed point;
searching adjacent data points around the seed point, wherein
the Euclidean distance between the adjacent data point and the seed point are lower than a predetermined length; and

searching points which have a relationship with the seed point satisfying the consistency qualification among the adjacent data points.

7. The method for optimizing scan data according to claim 1 , wherein
in the step of optimizing clusters, obtaining optimized clusters respectively corresponding to each set of the at least one set of clusters having correspondence by calculating relative poses between the different clusters for each set of clusters having correspondence.

8. The method for optimizing scan data according to claim 7 , wherein, for each set of the at least one set of clusters having correspondence,
the step of optimizing clusters includes:
obtaining a distance between two adjacent frames among the at least two frames of scan data by calculating a sum between both clusters belonging to the same set of clusters having correspondence in the two adjacent frames;
obtaining the relative poses between the said two frames by minimizing the distance, and obtaining all of the relative poses for each pair of adjacent frames among the at least two frames of scan data; and
obtaining optimized clusters respectively corresponding to the set of clusters having correspondence, by transferring all of the relative poses into a same coordinate system.

9. The method for optimizing scan data according to claim 1 , wherein, in the step of obtaining the scan data,
obtaining the at least two frames of scan data periodically.

10. The method for optimizing scan data according to claim 1 , wherein, in the step of cluster processing,
obtaining Gaussian Mixture Models for geometric primitives to simulate shape of the map element by using a Gaussian distribution calculation based on the characteristic of the data points.

11. A method for correcting trajectory of a vehicle/robot, the method comprising the steps of:
obtaining pose data by obtaining a plurality of first pose data, wherein
the pose data represents the position and orientation of the vehicle/robot;

obtaining a first trajectory, among the plurality of first pose data, by choosing at least two of first pose data and obtaining the first trajectory based on the pose data as chosen;
obtaining scan data by obtaining at least two frames of scan data respectively corresponding to different timings, wherein
each frame of the scan data includes a plurality of data points, wherein the scan data is obtained by scanning environment elements;

cluster processing, based on a characteristic of the data points, by classifying the plurality of data points in each frame of the scan data into one or more clusters, wherein
the one or more clusters represent map elements corresponding to the environment elements;

establishing correspondence among the at least two frames of scan data by searching and obtaining at least one set of clusters having correspondence;
optimizing clusters among the at least two frames of scan data by calculating each set of the at least one set of clusters having correspondence to obtain optimized clusters respectively corresponding to each set of the at least one set of clusters having correspondence;
optimizing the scan data by accumulating all optimized clusters to obtain an optimized scan data for the at least two frames of scan data; and
correcting the first trajectory to obtain a corrected trajectory of a vehicle/robot based on the difference between each frame of scan data and the optimized scan data.

12. The method for correcting trajectory of a vehicle/robot according to claim 11 , wherein, in the step of correcting trajectory comprises:
re-calculating pose data by synchronizing timestamp of the at least two of frames of scan data and the at least two of the first pose data, and according to the respective timestamp of the at least two of frames of scan data, obtaining second pose data on the first trajectory respectively corresponding to the at least two of frames of scan data;
correcting the pose data based on the difference between each frame of scan data and the optimized scan data by correcting the second pose data to obtain at least two of optimized pose data corresponding to the at least two of frames of scan data; and
obtaining a second trajectory based on the at least two of optimized pose data by obtaining the second trajectory as the corrected trajectory of a vehicle/robot.

13. The method for correcting trajectory of a vehicle/robot according to claim 12 , further comprising:
choosing at least two of first pose data which are continuous and satisfy predetermined smooth criteria, and
obtaining the first trajectory based on the pose data as chosen.

14. The method for optimizing scan data according to claim 13 , wherein, in the step of cluster processing,
attaching a descriptor for each cluster, wherein
the descriptor describes properties of the cluster.

15. The method for optimizing sca

Assignees

• Bayerische Motoren Werke Ag

Inventors

• Doemling Maximilian
• Granzow Sebastian
• Jiang Wanli
• Li Qianshan
• Li Jianpeng
• Lv Shuhan
• Xu Tao
• Xu Hongshan

Classifications

• G05D2201/0213

  Physics · mapped topic

• G05D1/0212 Primary

  with means for defining a desired trajectory (involving a plurality of land vehicles G05D1/0287) · CPC title

• G06F9/30192

  according to data descriptor, e.g. dynamic data typing · CPC title

• G06F9/3891

  organised in groups of units sharing resources, e.g. clusters · CPC title

• G06V20/56

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title