Apparatus and method for extracting feature information of object and apparatus and method for creating feature map

• US 20100114374A1
• Filed: 06/03/2009
• Published: 05/06/2010
• Est. Priority Date: 11/03/2008
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Yeon-Ho, Cho, Joon-Kee, Park, Dong-Ryeol
• Granted Patent: US 8,352,075 B2
• Time in Patent Office: Days
• US Class Current: 700/259
• CPC Class Codes: G06T 7/514 from specularities
• CPC Class Codes: G06V 10/44 Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
• 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
• CPC Class Codes: G06V 20/64 Three-dimensional objects

Abstract

Technology for creating a feature map for localizing a mobile robot and extracting feature information of surroundings is provided. According to one aspect, feature information including a reflection function is extracted from information acquired using a 3D distance sensor and used as a basis for creating a feature map. Thus, a feature map that is less sensitive to change in the surrounding environment can be created, and a success rate of feature matching can be increased.

Claims

1. An apparatus for extracting feature information of environmental objects, comprising:

a sensor to detect light reflected from the environmental objects to acquire distance information and remission information regarding the environmental objects;

a corner detector to detect corner information regarding the environmental objects using the remission information;

a plane detector to detect planar patch information using the corner information and distance information of surroundings corresponding to a corner; and

a reflection function detector to detect a reflection function of the planar patch using the remission information depending on angle of light incident on the planar patch.

2. The apparatus according to claim 1, wherein at least one of the corner information, the planar patch information, and the reflection function is used as the feature information of the environmental objects.

3. The apparatus according to claim 2, wherein the corner information includes location coordinates of the corresponding corner, the planar patch information includes plane equation parameters of the planar patch, and the reflection function includes reflection function parameters of the planar patch.

4. The apparatus according to claim 1, wherein the corner information includes location coordinates of corners of the environmental objects, and the plane detector establishes a cube centered on the corner location coordinates and detects at least one piece of planar patch information from distance information of the cube.

5. The apparatus according to claim 1, wherein the reflection function detector computes the reflection function by curve fitting a Gaussian function of the remission information.

6. An apparatus for creating a feature map for localization of a mobile robot, comprising:

a sensor to detect light reflected from environmental objects to acquire distance information and remission information regarding the environmental objects; and

a data processor to extract feature information which includes at least one of corner information, planer patch information and a reflection function of the planer patch using the distance information and the remission information, and to store the extracted feature information to create the feature map.

7. The apparatus according to claim 6, wherein the corner information includes location coordinates of a corresponding corner, the planar patch information includes plane equation parameters of the planar patch, and the reflection function includes reflection function parameters of the planar patch.

8. The apparatus according to claim 6, wherein the data processor comprises:

a corner detector to detect the corner information regarding the environmental objects using the remission information;

a plane detector detects the planar patch information using the corner information and distance information of surroundings corresponding to a corner;

a reflection function detector to detect the reflection function of the planar patch using the remission information which depends on angle of light incident on the planar patch; and

a feature map storage to store at least one of the corner information, the planar patch information, or the reflection function of the planar patch.

9. The apparatus according to claim 8, wherein the data processor further comprises:

an update unit to match current feature information with stored feature information to correct the location of the robot and to update the stored feature information.

10. The apparatus according to claim 8, wherein the corner information includes location coordinates of corners of the environmental objects, and the plane detector establishes a cube centered on the corner location coordinates and detects at least one piece of planar patch information from distance information of the cube.

11. The apparatus according to claim 8, wherein the reflection function detector computes the reflection function by curve fitting a Gaussian function of the remission information.

12. A feature information extraction method for creation of a feature map, comprising:

detecting light reflected from environmental objects to acquire distance information and remission information regarding the environmental objects;

detecting corner location coordinates regarding the environmental objects as corner information using the remission information;

establishing a cube centered on the corner location coordinates using the detected corner location coordinates and distance information of surroundings corresponding to a corner, and detecting planar patch information from distance information of the cube;

detecting a reflection function of the planar patch using the remission information which depends on angle of light incident on the planar patch; and

extracting and storing at least one of the corner information, the planar patch information, and the reflection function as the feature information.

13. The method according to claim 12, wherein the remission information is an amount of light reflected from the environmental objects or a plane image created based on the amount of light reflected from the environmental objects.

14. The method according to claim 12, wherein the reflection function is computed by curve fitting a Gaussian function of the remission information.

18. At least one computer readable medium storing computer readable instructions that control at least one processor to implement the method of claim 12.

15. A method for creating a feature map for localization of a mobile robot, comprising:

detecting light reflected from environmental objects to acquire distance information and remission information regarding the environmental objects;

detecting corner location coordinates regarding the environmental objects as corner information based on the remission information;

establishing a cube centered on the corner location coordinates using the detected corner location coordinates and distance information of surroundings corresponding to a corner, and detecting planer patch information from distance information of the cube;

detecting a reflection function of the planar patch using remission information which depends on angle of light incident on the planar patch;

extracting and storing at least one of the corner information, the planar patch information, and the reflection function as feature information; and

matching current feature information with stored feature information to correct a position of the robot and updating the stored feature information.

16. The method according to claim 15, wherein the remission information is an amount of light reflected from the environmental objects or a plane image created based on the amount of light reflected from the environmental objects.

17. The method according to claim 15, wherein the reflection function is computed by curve fitting a Gaussian function of the remission information.

19. At least one computer readable medium storing computer readable instructions that control at least one processor to implement the method of claim 15.