Granted Patent B2

US 8,194,927 · App. 12/175,622 · Granted Jun 5, 2012

Abstract

Claims (178)

1. A method of detecting road lane markers using a light-based sensing technology, the method comprising the steps of:

capturing reflectivity data using a light-based sensing device;

generating a light intensity signal based on the captured reflectivity data received by the light-based sensing device;

convolving the light intensity signal with a differential filter for generating a filter response that identifies a candidate lane marker region and ground segment regions juxtaposed on each side of the candidate lane marker region;

calculating a weighted standard deviation of light intensity-based data points within the identified candidate lane marker region;

calculating a standard deviation of light intensity-based data points within the ground segment regions juxtaposed to the candidate lane marker region;

determining an objective value for the identified candidate lane marker region as a function of the weighted standard deviation of light intensity-based data points within the candidate lane marker region, the weighted standard deviation of light intensity-based data points within the ground segment regions juxtaposed to the identified candidate lane marker region, and a number of data points contained within the identified candidate lane marker region; and

comparing the objective value to a respective threshold for determining whether the identified candidate lane marker region is a lane marker.

2. The method of claim 1 further comprising the step of applying a false alarm mitigation analysis for verifying whether the identified candidate lane marker region is a lane marker.

3. The method of claim 2 wherein the false alarm mitigation analysis further comprises the steps of:

determining a width of the identified candidate lane marker region;

comparing the width to a predetermined width; and

determining that the identified candidate lane marker region is the lane marker in response to the width being less than the predetermined width, otherwise determining that the identified candidate lane marker region is not the lane marker.

4. The method of claim 2 wherein the false alarm mitigation analysis further comprises the steps of:

calculating a weighted mean of the light intensity-based data points within the identified candidate lane marker region;

calculating a weighted mean of the light intensity-based data points within the ground segment regions;

comparing the weighting mean of the identified candidate lane marker region to the weighted mean of the ground segment regions; and

determining that the identified candidate lane marker region is the lane marker in response to the weighting mean of the identified candidate lane marker region being greater than the weighted mean of the ground segment regions.

5. The method of claim 1 wherein a first edge of the identified candidate lane marker region is identified by a negative peak in the filter response and a second response is identified by a positive peak in the filter response.

6. The method of claim 5 wherein the negative peak is identified at an abrupt shifting downward of data points in the filtered response in comparison to substantially horizontal adjacent data points of the ground segment regions.

7. The method of claim 4 wherein the positive peak is identified at an abrupt shifting upward of data points in the filtered response in comparison to substantially horizontal adjacent data points of the ground segment regions.

8. The method of claim 1 wherein a determined weight applied for determining the weighted standard deviation is represented by the equation:

w

⁡

(

i

)

=

{

2

*

arc

⁢

⁢

sin

(

i

*

π

/

(

N

L

-

1

)

,

2

*

arc

⁢

⁢

sin

⁡

(

i

*

π

/

(

N

L

-

1

)

)

≤

1

}

,

{

1

,

otherwise

}

i

=

0

,

1

,

2

,

…

⁢

,

N

L

-

1

,

where N L is the number of data points within the identified candidate lane marker region.

9. The method of claim 1 wherein the step of determining a weighted standard deviation of the light intensity-based data points within the ground segment regions include calculating a weighted standard deviation of the light intensity ground segment regions.

10. The method of claim 1 wherein a determined weight applied for determining the weighted standard deviation is represented by the equation:

w

⁡

(

i

)

=

{

2

*

arc

⁢

⁢

sin

(

i

*

π

/

(

N

G

-

1

)

,

2

*

arc

⁢

⁢

sin

⁡

(

i

*

π

/

(

N

G

-

1

)

)

≤

1

}

,

{

1

,

otherwise

}

i

=

0

,

1

,

2

,

…

⁢

,

N

G

-

1

,

where N G is the number of data points within the identified ground segment region juxtaposed on both sides of the identified candidate lane marker region.

11. The method of claim 1 wherein determining the objective value (f) is derived from the following formula:

f=α*σ L +β*σ G +γ/( N*N )

where σ L is the weighted standard deviation of the light intensity-based data points within the identified candidate lane marker region, α is a balancing parameter applied to the weighted standard deviation of the identified candidate lane marker region, σ G is the weighted standard deviation of the light intensity-based data points within the ground segment regions, β is a balancing parameter applied to the weighted standard deviation of the ground segment regions, N L is the number of data points within the identified candidate lane marker region, and γ is a balancing parameter applied to the number of data points within the identified candidate lane marker region.

12. The method of claim 11 wherein the respective balancing parameters are determined by a classifier.

13. The method of claim 11 wherein the respective balancing parameters are determined by a support vector machine classifier.

14. The method of claim 11 wherein the respective balancing parameters are determined by a neural network-base training program.

15. The method of claim 11 wherein the respective balancing parameters are selected to provide a balance between each of weighted standard deviation of the identified candidate lane marker region, the standard deviation of the ground segment regions, and the number of data points contained in the identified candidate lane marker region.

16. A lane marker detection system comprising:

a light-based sensing device for capturing reflectivity data;

a processor for receiving the captured reflectivity data received by the light-based sensing system, the processor generating a light intensity signal based on the captured reflectivity data, the processor convolving the light intensity signal with a filter for generating a filter response for identifying a candidate lane marker region and adjacent ground segment regions, the processor determining a weighted standard deviation of light intensity-based data points within the candidate lane marker region and a weighted standard deviation of light intensity-based data points within the adjacent ground segments, the processor calculating an objective value for the identified candidate lane marker region as a function of the respective weighted standard deviations and a number of data points within the identified candidate lane marker region, the processor comparing the objective value to a threshold for determining whether the candidate lane marker region is a lane marker; and

an output device identifying a location of each of the lane markers.

17. The lane marker detection system of claim 16 further comprising a false alarm mitigation module for verifying whether the identified candidate lane marker region is the lane marker.

18. The lane marker detection system of claim 16 further comprising a classifier for generating the weighting parameters.

19. The lane marker detection system of claim 16 wherein the classifier is a support vector machine classifier.

Assignments (12)