Navigation map updating method and apparatus and robot using the same

Patent metadata

Field	Value
Publication number	US-11629964-B2
Application number	US-202016843923-A
Country	US
Kind code	B2
Filing date	Apr 9, 2020
Priority date	Dec 20, 2019
Publication date	Apr 18, 2023
Grant date	Apr 18, 2023

Abstract

The present disclosure discloses a navigation map updating method as well as an apparatus, and a robot using the same. The method includes: controlling a robot to move along a designated path after a successful relocalization of the robot, and recording key frame data of each frame on the designated path and a corresponding pose; creating a new navigation map, and copying information in an original navigation map into the new navigation map; and covering the key frame data of each frame on the designated path onto the new navigation map to obtain an updated navigation map. In this manner, there is no need for the user to manipulate the robot to recreate the map at the environment where the robot is operated, which saves a lot of time and manpower.

First claim

What is claimed is:

1. A computer-implemented navigation map updating method, comprising:
providing a robot comprising a processor, a storage storing an existing navigation map and a preset probability grid map, and a sensor, wherein the storage and the sensor are electrically coupled to the processor;
performing, by the processor, a relocalization of the robot;
controlling, by the processor, the robot to move along a designated path, and recording key frame data of each frame on the designated path and a corresponding pose, wherein the key frame data of each frame on the designated path is obtained by the sensor;
creating, by the processor, a navigation map in the storage, and copying information in the existing navigation map into the created navigation map to obtain a first navigation map;
covering, by the processor, the key frame data of each frame on the designated path onto the first navigation map to obtain a second navigation map;
creating, by the processor, a probability grid map in the storage, and copying information in the preset probability grid map into the created probability grid map to obtain a first probability grid map;
covering, by the processor, the key frame data of each frame on the designated path into the first probability grid map to obtain a second probability grid map; and
denoising, by the processor, the second navigation map based on the second probability grid map to obtain a third navigation map, and sending the third navigation map to a remote server.

2. The method of claim 1 , wherein the sensor comprises a lidar, and recording the key frame data of each frame on the designated path and the corresponding pose comprises:
collecting, by the lidar, laser data of a current frame at a preset data collection frequency, and determining, by the processor, an odometer value of the laser data of the current frame;
calculating, by the processor, an estimated pose of the laser data of the current frame based on the odometer value of the laser data of the current frame and a pose and an odometer value of laser data of a previous frame of the current frame;
determining, by the processor, whether the laser data of the current frame is different from reference key frame data based on the estimated pose of the laser data of the current frame and a pose of the reference key frame data, wherein the reference key frame data is key frame data of the previous frame; and
matching, by the processor, the estimated pose of the laser data of the current frame with the preset probability grid map to obtain a pose of the laser data of the current frame, in response to the laser data of the current frame being different from the reference key frame data.

3. The method of claim 2 , wherein determining, by the processor, whether the laser data of the current frame is different from the reference key frame data based on the estimated pose of the laser data of the current frame and the pose of the reference key frame data comprises:
calculating, by the processor, a distance difference and an angle difference between the estimated pose of the laser data of the current frame and the pose of the reference key frame data; and
determining, by the processor, the laser data of the current frame is different from the reference key frame data, in response to the distance difference being greater than a preset distance threshold or the angle difference being greater than a preset angle threshold.

4. The method of claim 2 , wherein matching, by the processor, the estimated pose of the laser data of the current frame with the preset probability grid map to obtain the pose of the laser data of the current frame comprises:
using, by the processor, the estimated pose of the laser data of the current frame as an initial coordinate to perform an optimization matching with respect to the preset probability grid map so as to obtain the pose of the laser data of the current frame, in response to an accuracy of an odometer of the robot being greater than a preset accuracy threshold.

5. The method of claim 2 , wherein matching, by the processor, the estimated pose of the laser data of the current frame with the preset probability grid map to obtain the pose of the laser data of the current frame comprises:
using, by the processor, the estimated pose of the laser data of the current frame as the initial coordinate to perform one of an iterative closest point (ICP) matching and a template matching with respect to the preset probability grid map so as to obtain a first coordinate of the laser data of the current frame; and
using, by the processor, the first coordinate of the laser data of the current frame as the initial coordinate to perform an optimization matching with respect to the preset probability grid map so as to obtain the pose of the laser data of the current frame.

6. The method of claim 2 , wherein the odometer value of the laser data of the current frame and the odometer value of laser data of the previous frame of the current frame are obtained based on data of an odometer composed of a chassis disposed on the robot or based on integral data of an odometer composed of an inertial measurement unit (IMU) disposed on the robot.

7. The method of claim 2 , wherein the odometer value of the laser data of the current frame and the odometer value of laser data of the previous frame of the current frame are obtained by performing a fusion computing on data of an odometer composed of a chassis disposed on the robot and integral data of an odometer composed of an inertial measurement unit (IMU) disposed on the robot.

8. The method of claim 1 , wherein after covering, by the processor, the key frame data of each frame on the designated path onto the first navigation map to obtain the second navigation map, further comprises:
covering, by the processor, pixels outside a laser beam as gray pixels in response to the pixels outside the laser beam being not the gray pixels, wherein the gray pixels are configured for representing an unknown area.

9. The method of claim 1 , wherein a resolution conversion is performed during copying the information in the existing navigation map into the created navigation map.

10. The method of claim 1 , wherein denoising, by the processor, the second navigation map based on the second probability grid map to obtain the third navigation map comprises:
identifying, by the processor, a black area in the second navigation map that has an area smaller than a preset area threshold; and
in response to a probability of black points in the identified area in the second probability grid map being less than a preset probability threshold, modifying, by the processor, the black points in the identified area to white.

11. The method of claim 1 , wherein denoising, by the processor, the second navigation map based on the second probability grid map to obtain the third navigation map comprises:
identifying, by the processor, an area in the second navigation map that has a total number of pixels of a black area smaller than a preset number threshold; and
modifying, by the processor, points in the identified area to white.

12. The method of claim 1 , wherein denoising, by the processor, the second navigation map based on the second probability grid map to obtain a third navigation map comprises:
identifying, by the processor, a gray area in the second navigation map; and
in response to an area of the gray area being smaller than a preset area threshold and a color of surrounding area of the gray area being white, filling, by the processor, the gray area with white.

13. A robot comprising:
a non-transitory memory;
a sensor; and
a process

Assignees

• Ubtech Robotics Corp Ltd

Inventors

• Zhao Yongsheng
• Liu Zhichao
• Pang Jianxin
• Xiong Youjun

Classifications

• G01S17/875

  for determining attitude · CPC title

• G01S17/88

  Lidar systems specially adapted for specific applications · CPC title

• G01C21/32 Primary

  Structuring or formatting of map data · CPC title

• G01C21/3848

  Data obtained from both position sensors and additional sensors · CPC title

• G05D1/0272

  comprising means for registering the travel distance, e.g. revolutions of wheels (measuring distance traversed on the ground by vehicles, e.g. using odometers G01C22/00) · CPC title