Determining a geometric parameter from a single image

• US 8,896,686 B2
• Filed: 06/23/2009
• Issued: 11/25/2014
• Est. Priority Date: 06/23/2009
• Status: Active Grant

• Current Assignee: HERE Global B.V. (Nokia Corporation)
• Original Assignee: HERE Global B.V. (Nokia Corporation)
• Inventors: Chen, Xin, Hui, Roger B., Alwar, Narayanan
• Primary Examiner(s): Kim, Hee-Yong
• Application Number:
  US12/489,697
• Publication Number:
  US 20100321490A1
• Time in Patent Office: 1,981 Days
• Field of Search: 348/118
• US Class Current: 348/118
• CPC Class Codes: G01C 11/00 Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
• CPC Class Codes: G01C 21/3811 Point data, e.g. Point of Interest [POI]
• CPC Class Codes: G01C 21/3822 Road feature data, e.g. slope data
• CPC Class Codes: G01C 21/3837 Data obtained from a single source
• CPC Class Codes: G01C 21/3848 Data obtained from both position sensors and additional sensors
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30256 Lane; Road marking
• CPC Class Codes: G06T 7/62 of area, perimeter, diameter or volume

Abstract

Systems, devices, features, and methods for determining a geometric parameter from an image are disclosed. For example, one method for determining the geometric parameter is used to develop a navigation database. The method comprises determining calibration values corresponding to a camera mounted on a vehicle or a pedestrian. A plurality of images of geographic features are captured by the camera. A single image from the plurality of images is identified or selected. A geometric parameter of a region in the single image is determined based on the determined calibration values. For example, the geometric parameter is a real-world distance, such as a real-world length or width.

Claims

1. A method of determining a geometric parameter of a region from an image, wherein the image is collected to develop a navigation database, the method comprising:

determining calibration values corresponding to a camera mounted on a vehicle;

capturing a plurality of images of geographic features, the plurality of images captured by the camera mounted on the vehicle;

identifying a single image from the plurality of images;

calculating a vector from a center of the camera through a point in a single image;

calculating an intersection of a ground plane and the vector from the center of the camera through the point in the single image; and

determining a geometric parameter of a region in the single image based on the determined calibration values and the intersection of the ground plane and the vector from the center of the camera through the point in the single image, wherein the geometric parameter is a real-world distance.

2. The method of claim 1, wherein the real-world distance comprises a ground distance.

3. The method of claim 2, wherein the ground distance comprises a width or length of an object in the region represented in the single image.

4. The method of claim 3, wherein the object comprises a road lane or a lane marking.

5. The method of claim 1, wherein the real-world distance comprises a height.

6. The method of claim 5, wherein the height comprises a height of a road sign.

7. The method of claim 1, wherein the calibration values include a focal length of the camera and a camera roll angle.

8. The method of claim 1, further comprising:

storing the determined geometric parameter in the navigation database.

9. The method of claim 1, wherein determining the calibration values comprises projecting a three-dimensional ground coordinate system from an image plane.

10. The method of claim 9, wherein real-world ground points represented in the three-dimensional ground coordinate system are modeled as a function of an intrinsic value of the camera and an extrinsic value of the camera.

11. The method of claim 10, wherein the calibration values are determined based on comparing a real-world measurement between the real-world ground points and a computed measurement between the real-world ground points.

12. A method of determining a real-world parameter of an object in an image, the method comprising:

determining calibration values relating to a camera, wherein the camera captures an image at the determined calibration values;

identifying a first point in the image;

identifying a second point in the image, wherein the first point and the second point correspond to endpoints of an object in the image;

identifying a ray from a camera center to the image;

determining a distance between the first point and the second point in the image based on the determined calibration values and the ray from the camera, the distance corresponding to a real-world length or a real-world width of the object in the image.

13. The method of claim 12, wherein the object comprises a road lane, a road marking, a road sign, or portions thereof.

14. The method of claim 12, wherein the calibration values include a focal length of the camera and a camera roll angle.

15. The method of claim 12, wherein determining the distance comprises entering the calibration values into mathematical models representing the first point and the second point in a three-dimensional ground coordinate system, the three-dimensional ground coordinate system generated as a function of an image plane.

16. A non-transitory computer-readable medium configured to store a computer program that performs a method of determining a real-world ground parameter from a single image, the method comprising:

identifying an image captured by a camera;

identifying calibration values associated with the camera, the identified calibration values including a determined extrinsic value and a determined intrinsic value of the camera at the time the single image was captured;

calculating a vector through a vanishing point in the single image; and

determining a real-world ground parameter of a region in the single image based on the identified calibration values and the vector through the vanishing point in the single image, the real-world ground parameter including a ground distance value.

17. The non-transitory computer readable medium of claim 16, wherein the determined extrinsic value includes a camera roll angle and the determined intrinsic value includes a focal length of the camera.

18. The non-transitory computer-readable medium of claim 16, wherein the ground distance value comprises a length or width of a feature on a road.

19. A device for determining a real-world geometric parameter from a single image, the device comprising:

a user interface configured to display an image captured by a camera;

a processor in communication with the user interface and configured to calculate a vector from the camera through a point in the image;

a memory in communication with the processor, the memory configured to store data of the image and data of a determined calibration value of the camera,

wherein an image is displayed on the user interface, and

wherein the processor is configured to determine a geometric parameter of a region represented in the displayed image based on the determined calibration value and the vector from the camera through the point in the image, the geometric parameter being a value representing a real-world distance between selected points in the displayed image.