Apparatus and method with composite sensor calibration

• US 20110182476A1
• Filed: 12/07/2010
• Published: 07/28/2011
• Est. Priority Date: 01/27/2010
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Park, Dong-Ryeol, Kim, Dong-Jo
• Granted Patent: US 8,639,021 B2
• Time in Patent Office: Days
• US Class Current: 382/106
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 7/85 Stereo camera calibration

Abstract

An apparatus and method capable of calculating a coordinate transformation parameter without having to utilize a rig are provided. The apparatus and method extract a first feature point based on a plane of first data, project the first feature point onto second data and then extract a second feature point from a part of the second data onto which the first feature point is projected. Then, calibration is performed based on the extracted feature points. Therefore, it is possible to perform calibration immediately as necessary without having to utilize a separate device such as a rig.

Claims

1. An apparatus with composite sensor calibration, comprising:

a plane extractor to extract at least one plane from first data obtained by a first sensor;

a first feature point extractor to extract a first feature point from the first data based on the at least one plane;

a second feature point extractor to project the first feature point onto second data obtained by a second sensor and to extract a second feature point corresponding to the first feature point from a part of the second data onto which the first feature point is projected; and

a transformation parameter calculator to calculate a coordinate transformation parameter between the first sensor and the second sensor using the first feature point and the second feature point.

2. The calibration apparatus of claim 1, wherein the first sensor is a 3-dimensional (3D) distance sensor and the first data is 3D range data obtained from the 3D distance sensor.

3. The calibration apparatus of claim 1, wherein the second sensor is an image sensor and the second data is image data obtained from the image sensor.

4. The calibration apparatus of claim 1, wherein the first feature point extractor extracts a part of the first data at which at least three planes cross each other as the first feature point.

5. The calibration apparatus of claim 1, wherein the transformation parameter calculator registers the first feature point and the second feature point when a difference between coordinate values of the first and second feature points is equal to or less than a first threshold value, and calculates, when the number of the registered first and second feature points is equal to or greater than a second threshold value, the coordinate transformation parameter using the registered first and second feature points.

6. An apparatus with composite sensor calibration, comprising:

a plane extractor to extract at least one plane from first data obtained by a first sensor;

a first feature point extractor to extract a first feature point from the first data based on the at least one plane;

a first line extractor to extract a first line from the first data based on the at least one plane;

a second line extractor to project the first line onto second data obtained by a second sensor and to extract a second line corresponding to the first line from a part of the second data onto which the first line is projected;

a second feature point extractor to extract a second feature point corresponding to the first feature point based on the second line; and

a transformation parameter calculator to calculate a coordinate transformation parameter between the first sensor and the second sensor using the first feature point and the second feature point.

7. The calibration apparatus of claim 6, wherein the first sensor is a 3-dimensional (3D) distance sensor and the first data is 3D range data obtained from the 3D distance sensor.

8. The calibration apparatus of claim 6, wherein the second sensor is an image sensor and the second data is image data obtained from the image sensor.

9. The calibration apparatus of claim 6, wherein the first feature point extractor extracts a part of the first data at which at least three planes cross each other as the first feature point.

10. The calibration apparatus of claim 6, wherein the first line extractor extracts a part of the first data at which at least two planes cross each other as the first feature point.

11. The calibration apparatus of claim 6, wherein the transformation parameter calculator registers the first feature point and the second feature point when a difference between coordinate values of the first and second feature points is equal to or less than a first threshold value, and calculates, when the number of the registered first and second feature points is equal to or greater than a second threshold value, the coordinate transformation parameter using the registered first and second feature points.

12. A method with composite sensor calibration, comprising:

extracting at least one plane from first data obtained by a first sensor;

extracting a first feature point from the first data based on the at least one plane;

projecting the first feature point to second data obtained by the second sensor, and extracting a second feature point corresponding to the first feature point from a part of the second data to which the first feature point is projected; and

calculating a coordinate transformation parameter between the first sensor and the second sensor using the first feature point and the second feature point.

13. The calibration method of claim 12, wherein the extracting of the first feature point comprises extracting a part of the first data at which at least three planes cross each other as the first feature point.

14. The calibration method of claim 12, wherein the calculating of the coordinate transformation parameter comprises:

registering, when a difference between coordinate values of the first and second feature points is equal to or less than a first threshold value, the first feature point and the second feature point; and

calculating, when the number of the registered first and second feature points is equal to or greater than a second threshold value, the coordinate transformation parameter using the registered first feature point and the second feature point.

15. A method with composite sensor calibration, comprising:

extracting at least one plane from first data obtained by a first sensor;

extracting a first feature point from the first data based on the at least one plane;

extracting a first line from the first data based on the at least one plane;

projecting the first line to second data obtained by the second sensor, and extracting a second line corresponding to the first line from a part of the second data to which the first line is projected;

extracting a second feature point corresponding to the first feature point based on the second line; and

calculating a coordinate transformation parameter between the first sensor and the second sensor using the first feature point and the second feature point.

16. The calibration method of claim 15, wherein the extracting of the first feature point comprises extracting a part of the first data at which at least three planes cross each other as the first feature point.

17. The calibration method of claim 15, wherein the extracting of the first line comprises extracting a part of the first data along which at least two planes cross each other as the first line.

18. The calibration method of claim 15, wherein the calculating of the coordinate transformation parameter comprises:

registering the first feature point and the second feature point when a difference between coordinate values of the first feature points is equal to or less than a first threshold value; and

calculating the coordinate transformation parameter using the registered first and second feature points when the number of the registered first and second feature points is equal to or greater than a second threshold value.