Method for the detection and tracking of lane markings

• US 9,047,518 B2
• Filed: 10/12/2012
• Issued: 06/02/2015
• Est. Priority Date: 02/15/2012
• Status: Active Grant

• Current Assignee: Aptiv Technologies AG
• Original Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Inventors: Goermer, Steffen, Nunn, Christian, Meuter, Mirko, Mueller, Dennis
• Primary Examiner(s): Thirugnanam, Gandhi
• Application Number:
  US13/650,182
• Publication Number:
  US 20130208945A1
• Time in Patent Office: 963 Days
• Field of Search: 382/100
• US Class Current: 1/1
• CPC Class Codes: G06V 10/62 relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
• CPC Class Codes: G06V 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road

Abstract

In a method for the detection and tracking of lane markings from a motor vehicle, an image of a space located in front of the vehicle is captured by means of an image capture device at regular intervals. The picture elements that meet a predetermined detection criterion are identified as detected lane markings in the captured image. At least one detected lane marking as a lane marking to be tracked is subjected to a tracking process. At least one test zone is defined for each detected lane marking. With the aid of intensity values of the picture elements associated with the test zone, at least one parameter is determined. The detected lane marking is assigned to one of several lane marking categories, depending on the parameter.

Claims

1. A system for the detection and tracking of lane markings proximate to a motor vehicle, said system comprising:

an image capture device mounted on the motor vehicle and configured to capture an image of a space located in front of the vehicle, wherein the image includes a plurality of picture elements, the image is captured by the image capture device at regular intervals,

the picture elements that meet a predetermined detection criterion are identified as detected lane markings in the image,

at least one detected lane marking identified as a lane marking to be tracked is subjected to a tracking process in which the variation of the course of the lane marking in time in the plane of the roadway is tracked,

at least one test zone comprising a plurality of picture elements of the image is defined for each detected lane marking,

with the aid of intensity values of the picture elements associated with the test zone, at least one parameter is determined, and

the detected lane marking is assigned to one of a plurality of lane marking categories, depending on the parameter, wherein from the intensity values of the picture elements associated with the test zone statistical parameters are determined that include an average intensity value, an intensity value total, a maximum intensity value, a minimum intensity value, and a moment of intensity value distribution of the picture elements.

2. The system according to claim 1, wherein each detected lane marking is subjected to the tracking process as a lane marking to be tracked or rejected as an invalid lane marking, depending on the lane marking category to which it is assigned.

3. The system according to claim 1, wherein the position of the test zone within the image is determined by converting the position of a tracked lane marking in the plane of the roadway at a predetermined distance from the vehicle to image coordinates by a projection method.

4. The system according to claim 1, wherein the parameter for a plurality of successive images is determined, and the detected lane marking is assigned to the lane marking category in addition with the aid of the curve of the parameter in time.

5. The system according to claim 4, wherein the current speed of the vehicle is determined and with the aid of the speed the curve of the parameter in time is converted to a curve of the parameter in space in the plane of the roadway, and wherein the detected lane marking is assigned to the lane marking category with the aid of the curve of the parameter in space.

6. The system according to claim 4, wherein for characterization of the curve of the parameter in time, a statistical curve measure is determined, wherein the detected lane marking is assigned to the lane marking category with the aid of the statistical curve measure.

7. The system according to claim 4, wherein the curve of the parameter in time is subjected to a time-frequency transform, wherein the detected lane marking is assigned to the lane marking category with the aid of the transform coefficients determined within the framework of the time-frequency transform.

8. The system according to claim 7, wherein the time-frequency transform is carried out iteratively.

9. The system according to claim 1, wherein for each detected lane marking there is defined a set of a plurality of test zones which in each case correspond to the position of a tracked lane marking in the plane of the roadway at different distances from the vehicle.

10. The system according to claim 9, wherein for each detected lane marking at least five test zones are defined.

11. The system according to claim 1, wherein the detected lane marking is assigned to a lane marking category from a group of lane marking categories that includes the categories of “

invalid image object”

, “

single unbroken line”

, “

double unbroken line”

, “

single broken line”

, “

double broken line”

, “

broken and unbroken line”

, “

wide broken line” and

“

line with surface profile”

.

12. The system according to claim 1, wherein respective parameters of a plurality of test zones of a detected lane marking at different distances from the vehicle and respective parameters of a single test zone in successive images are compared with each other and, with the aid of the result of comparison, combined into a common parameter.

13. The system according to claim 12, wherein with the aid of the result of comparison a degree of statistical confidence is calculated and assigned to the detected lane marking.

14. A method for the detection and tracking of lane markings by an image capture device mounted on a motor vehicle, said method comprising:

capturing an image of a space located in front of the vehicle with the image capture device, wherein said image comprises a plurality of picture elements captured by the image capture device at regular time intervals;

identifying picture elements that meet a predetermined detection criterion as detected lane markings in the captured image by means of an image processing system;

indicating at least one detected lane marking as a lane marking to be tracked according to a tracking process in which the variation of the course of the lane marking over time in the plane of the roadway is tracked by means of a status estimator of the image processing system;

defining for each detected lane marking a test zone comprising a plurality of picture elements;

determining statistical parameters based on an intensity value of the picture elements associated with the test zone, wherein the statistical parameters include an average intensity value, an intensity value total, a maximum intensity value, a minimum intensity value, and a moment of intensity value distribution of the picture elements; and

assigning the detected lane marking to one of a plurality of lane marking categories based on the parameter.

15. The method according to claim 14, wherein each detected lane marking is subjected to the tracking process as a lane marking to be tracked or rejected as an invalid lane marking, depending on the lane marking category to which it is assigned.

16. The method according to claim 14, wherein the parameter for a plurality of successive images is determined, and the detected lane marking is assigned to the lane marking category in addition with the aid of the curve of the parameter in time.

17. The method according to any of claim 16, wherein the curve of the parameter in time is subjected to a time-frequency transform, wherein the detected lane marking is assigned to the lane marking category with the aid of the transform coefficients determined within the framework of the time-frequency transform.

18. The method according to claim 14, wherein for each detected lane marking there is defined a set of a plurality of test zones which in each case correspond to the position of a tracked lane marking in the plane of the roadway at different distances from the vehicle.

19. The method according to claim 14, wherein the detected lane marking is assigned to a lane marking category from a group of lane marking categories that includes the categories of “

invalid image object”

, “

single unbroken line”

, “

double unbroken line”

, “

single broken line”

, “

double broken line”

, “

broken and unbroken line”

, “

wide broken line” and

“

line with surface profile”

.

20. The method according to claim 14, wherein respective parameters of a plurality of test zones of a detected lane marking at different distances from the vehicle and respective parameters of a single test zone in successive images are compared with each other and, with the aid of the result of comparison, combined into a common parameter.