Apparatus and method for providing vehicle camera calibration

• US 9,275,458 B2
• Filed: 06/10/2013
• Issued: 03/01/2016
• Est. Priority Date: 11/19/2012
• Status: Expired due to Fees

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Oh, Jaehong, Kim, Kyong Ho, Yoon, Chang Rak, Cho, Seong Yun, Park, Hye Sun
• Primary Examiner(s): Ustaris, Joseph
• Assistant Examiner(s): NAWAZ, TALHA M
• Application Number:
  US13/914,257
• Publication Number:
  US 20140139671A1
• Time in Patent Office: 995 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: B60R 2300/402 Image calibration
• CPC Class Codes: G06T 7/80 Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration

Abstract

A method and apparatus for providing a vehicle camera calibration includes choosing overlap images from among the image information captured by a camera mounted on a vehicle, and filtering images having the same vehicle attitude from among the chosen overlap images on a basis of a map coordinate system and status information of the vehicle, picking out objects in the filtered images, separating stationary objects from the picked out objects, and predicting an angular misalignment with respect to the separated stationary objects.

Claims

1. An apparatus for providing calibration of a vehicle camera mounted on a vehicle, the apparatus comprising:

an overlap image selection unit that determines a predetermined time period and chooses images from images captured by the vehicle camera during the predetermined time period, the chosen images including at least a same object;

a filtering unit that receives the chosen images from the overlap image selection unit and extracts images captured at a substantially identical vehicle attitude of the vehicle from the chosen images on a basis of a map coordinate system and status information of the vehicle, the status information including at least one of location and attitude of the vehicle;

an object pick-out unit that receives the extracted images from the filtering unit and picks out objects commonly existing in the extracted images;

an object separation unit that receives the picked out objects from the object pick-out unit and separates stationary objects from the picked out objects; and

an angular misalignment prediction unit that receives the stationary objects from the object separation unit to predict an angular misalignment of the vehicle camera based on the stationary objects separated by the object separation unit,

wherein a three-dimensional coordinate of the vehicle from a first time (t) to a second time (t+1) is a summation of a reference coordinate of a vehicle coordinate system of the stationary objects and a distance that the vehicle moves on a basis of the vehicle coordinate system at the first time (t) for the first time (t) to the second time (t+1).

2. The apparatus of claim 1, wherein the object pick-out unit is configured to pick out the stationary objects using image information on road facilities.

3. The apparatus of claim 1, wherein the object pick-out unit is configured to utilize locations of the objects predicted at a tracking time when there exist objects being tracked.

4. The apparatus of claim 3, wherein the stationary objects separated by the object separation unit are stored or updated in an object list.

5. The apparatus of claim 1, wherein the angular misalignment prediction unit is configured to eliminate a camera distortion at an arbitrary location of the stationary objects separated by the object separation unit.

6. The apparatus of claim 5, wherein the angular misalignment prediction unit is configured to predict the angular misalignment with respect to the stationary objects using the Method of Least Square.

7. The apparatus of claim 1, wherein the reference coordinate of the vehicle coordinate system of the stationary objects is a summation of a relative coordinate of the stationary objects based on a camera coordinate system at an arbitrary time and a distance from an origin of the vehicle coordinate system to an origin of the camera coordinate system.

8. The apparatus of claim 1, wherein the three-dimensional coordinate of the vehicle from the first time (t) to the second time (t+1) is a coordinate when having the same vehicle attitudes on the basis of the map coordinate system.

9. The apparatus of claim 1, wherein the images captured by the vehicle camera and the status information on the vehicle are temporally stored for the predetermined time period in the apparatus for providing a vehicle camera calibration.

10. A method for providing calibration of a vehicle camera mounted on a vehicle in an apparatus for providing calibration of the video camera, the method comprising:

determining a predetermined time period;

choosing images from images captured by the vehicle camera during the predetermined time period, the chosen images including at least a same object;

extracting images captured at a substantially identical vehicle attitude from the chosen images on a basis of a map coordinate system and status information of the vehicle, the status information including at least one of location and attitude of the vehicle;

picking out objects in the extracted images;

separating stationary objects from the picked out objects; and

predicting an angular misalignment of the vehicle camera based on the separated stationary objects, wherein said predicting the angular misalignment comprises;

calculating a three-dimensional coordinate of the vehicle from a first time (t) to a second time (t+1) by summing a reference coordinate of a vehicle coordinate system of the stationary objects and a distance that the vehicle moves on a basis of the vehicle coordinate system at the first time (t) for the first time (t) to the second time (t+1).

11. The method of claim 10, wherein said picking out objects comprises:

picking out the stationary objects using image information on road facilities.

12. The method of claim 10, wherein said picking out objects comprises:

utilizing locations of the objects predicted at a tracking time when there exist objects being tracked.

13. The method of claim 12, further comprising:

storing the stationary objects in a list of the objects being tracked; and

updating the stationary objects in the list of the objects being tracked.

14. The method of claim 10, wherein said predicting the angular misalignment comprises:

eliminating a camera distortion at an arbitrary location of the separated stationary objects.

15. The method of claim 14, wherein said predicting the angular misalignment comprises:

predicting the angular misalignment with respect to the stationary objects using the Method of Least Square.

16. The method of claim 10, wherein said predicting the angular misalignment comprises:

calculating the reference coordinate of the vehicle coordinate system of the stationary objects by summing a relative coordinate of the stationary objects based on a camera coordinate system at an arbitrary time and a distance from an origin of the vehicle coordinate system to an origin of the camera coordinate system.

17. The method of claim 10, wherein the three-dimensional coordinate of the vehicle from the first time (t) to the second time (t+1) is a coordinate when having the same vehicle attitudes on the basis of the map coordinate system.

18. The method of claim 10, wherein the images captured by the vehicle camera and the status information on the vehicle are temporally stored for the predetermined time period in the apparatus for providing a vehicle camera calibration.