Road lane marker recognition

• US 6,845,172 B2
• Filed: 09/27/2001
• Issued: 01/18/2005
• Est. Priority Date: 09/29/2000
• Status: Expired due to Fees

• Current Assignee: Nissan Motor Co., Ltd.
• Original Assignee: Nissan Motor Co., Ltd.
• Inventors: Furusho, Hiroyuki
• Primary Examiner(s): Ahmed, Samir
• Assistant Examiner(s): BHATNAGAR, ANAND P
• Application Number:
  US09/963,490
• Publication Number:
  US 20020042676A1
• Time in Patent Office: 1,209 Days
• Field of Search: 382/100, 382/104, 382/107, 382/199, 382/202, 340/907, 348/113, 701/23, 701/28, 701/41, 701/200, 701/300, 703/8
• US Class Current: 382/104
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: G06V 10/443 by matching or filtering
• CPC Class Codes: G06V 10/457 by analysing connectivity, e.g. edge linking, connected component analysis or slices
• 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
• CPC Class Codes: G08G 1/167 Driving aids for lane monitoring, lane changing, e.g. blind spot detection

Abstract

A road recognition apparatus estimates road parameters of a road model equation by processing a forward road image of a vehicle. A processor sets a series of windows along a road white line (or lane marker) on the forward road image, and calculates a candidate straight line segment likely to be a white line in each window. The processor further checks an accuracy in position of a candidate white line in each window. The processor estimates the road parameters from the image coordinates of the candidate white line in each or the window higher in the accuracy, and estimates the road parameters exclusive of a vehicle lateral displacement from the gradient of candidate white line in each or the window lower in the accuracy.

Claims

1. A road lane marker recognition apparatus for obtaining a forward road image of a road ahead of a vehicle with an imaging device and estimating road parameters of a road model equation, the recognition apparatus comprising a processor:

to set, in the road image, a plurality of windows along a road lane marker on the road so that the road lane marker is contained in each window;

to calculate a candidate lane marker which is likely to be the road lane marker in each window;

to judge an accuracy in determining the position of the candidate lane marker with respect to the road lane marker in each window; and

to estimate the road parameters of the road model equation representing a road shape of the road and vehicle attitude with respect to the road, the processor being configured to estimate the road parameters inclusive of a vehicle lateral displacement, by using at least a gradient of the candidate lane marker for one window judged to be lower in the accuracy, and by using an image coordinate of the candidate lane marker for one window judged to be higher in the accuracy.

2. The recognition apparatus as claimed in claim 1, wherein the processor is configured to examine each window to determine whether there is a pseudo lane marker approximately equal in gradient to the candidate lane marker, and to judge a window to be lower in the accuracy if there is the pseudo lane marker, and to be higher in the accuracy if there is not the pseudo lane marker.

3. The recognition apparatus as claimed in claim 1, wherein the processor is configured to judge the accuracy of each window in accordance with a distance of a road region in the window from the vehicle, and to judge a window to be lower in the accuracy if the distance is greater, and to be higher in the accuracy if the distance is smaller.

4. The recognition apparatus as claimed in claim 1, wherein the processor is configured to extract candidate lane marker points by processing the image in each window, and to calculate the candidate lane marker from the candidate lane marker points in each window.

5. The recognition apparatus as claimed in claim 1, wherein the processor is configured to calculate the candidate lane marker in each window by straight line approximation.

6. The recognition apparatus as claimed in claim 4, wherein the processor is configured to calculate the candidate lane marker by straight line approximation using Hough transform from the candidate lane marker points in each window.

7. The recognition apparatus as claimed in claim 1, wherein the processor is configured to subject an equation obtained by eliminating the vehicle lateral displacement from the road model equation, to linear approximation about the parameters, for a window judged to be lower in the position accuracy, and to subject the road model equation to linear approximation about the parameters, for a window judged to be higher in the position accuracy, to estimate the road parameters by Kalman filter operation.

8. The recognition apparatus as claimed in claim 1, wherein the road parameters comprise a road curvature, a yaw angle of the vehicle with respect to a road lane, a pitch angle of the vehicle, and a height of the imaging device from a ground level.

9. The recognition apparatus as claimed in claim 1, wherein the processor is configured to classify the windows into a higher accuracy window subset and a lower accuracy window subset in accordance with the position accuracy, and the processor is further configured to estimate the road parameters exclusive of the vehicle lateral displacement, by using the gradient of the candidate lane marker for each window or the window of the lower accuracy window subset, and to estimate the road parameters inclusive of the vehicle lateral displacement for each window or the window of the higher accuracy window subset.

10. A road lane marker recognition process comprising:

obtaining a forward road image of a road ahead of a vehicle;

setting, in the road image, a series of windows arranged along a road lane marker on the road so that the road lane marker is contained in each window;

calculating a candidate lane marker which is likely to be the lane marker in each window;

judging a position accuracy in position of the candidate lane marker with respect to the road lane marker in each window; and

estimating road parameters of a road model equation representing a road shape of the road and vehicle attitude with respect to the road, the road parameters inclusive of a vehicle lateral displacement being estimated by using at least a gradient of the candidate lane marker for one window judged to be lower in the position accuracy, and by using an image coordinate of the candidate lane marker for one window judged to be higher in the position accuracy.

11. A road lane marker recognition apparatus comprising:

means for setting, in a forward road image of a road ahead of a vehicle, a series of windows along a road lane marker on the road so that the road lane marker is contained in each window;

means for calculating a candidate lane marker which is likely to be the road lane marker in each window;

means for judging a position accuracy in position of the candidate lane marker in each window; and

means for estimating road parameters of a road model equation, inclusive of a vehicle lateral displacement, by using at least a gradient of the candidate lane marker for one window judged to be lower in the position accuracy, and by using an image coordinate of the candidate lane marker for one window judged to be higher in the position accuracy.