CLEAR PATH DETECTION USING PATCH APPROACH

• US 20100104137A1
• Filed: 10/19/2009
• Published: 04/29/2010
• Est. Priority Date: 04/24/2008
• Status: Active Grant

• Current Assignee: Carnegie Mellon University, GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: Carnegie Mellon University, GM Global Technology Operations Incorporated (Motors Liquidation Co. GUC Trust)
• Inventors: Wu, Qi, Zhang, Wende, Chen, Tsuhan
• Granted Patent: US 8,611,585 B2
• Time in Patent Office: Days
• US Class Current: 382/104
• CPC Class Codes: B60W 2420/403 Image sensing, e.g. optical camera
• CPC Class Codes: G06V 10/443 by matching or filtering
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• 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

A method for detecting a clear path of travel for a vehicle utilizing an image generated by a camera device located upon the vehicle includes monitoring the image, and generating a plurality of patches onto the image, wherein the patches are collectively arranged to substantially include the clear path of travel. Feature extraction is utilized to analyze the patches and includes convolving each of the patches with a feature detection filter generating a feature-based filter response, extracting features based upon the feature-based filter response, and determining each of the patches to represent the clear path of travel or to not represent the clear path of travel based upon the extracted features. The clear path of travel is determined based upon the plurality of patches, and is utilized to navigate the vehicle.

Claims

1. Method for detecting a clear path of travel for a vehicle utilizing an image generated by a camera device located upon the vehicle, the method comprising:

monitoring the image;

generating a plurality of patches onto the image, wherein the patches are collectively arranged to substantially include the clear path of travel;

utilizing feature extraction to analyze the patches comprising

convolving each of the patches with a feature detection filter generating a feature-based filter response,

extracting features based upon the feature-based filter response, and

determining each of the patches to represent the clear path of travel or to not represent the clear path of travel based upon the extracted features;

determining the clear path of travel based upon the plurality of patches; and

utilizing the clear path of travel to navigate the vehicle.

2. The method of claim 1, wherein generating a plurality of patches onto the image comprises:

generating fixed-grid patches onto the image, wherein each patch is of equal size according to image coordinates.

3. The method of claim 1, further comprising estimating a horizon line for the image;

and

wherein generating a plurality of patches onto the image comprises generating the plurality of patches substantially below the horizon line.

4. The method of claim 1, wherein generating a plurality of patches onto the image comprises:

monitoring a desired analysis grid upon three-dimensional world coordinates; and

determining the patches for the image based upon the desired analysis grid.

5. The method of claim 4, further comprising determining an estimated road shape based upon the image;

and

wherein the desired analysis grid is selected based upon the estimated road shape.

6. The method of claim 4, wherein the desired analysis grid is selected to generate the patches comprising larger patches at a bottom of the image and increasingly smaller patches away from the bottom of the image.

7. The method of claim 4, wherein determining the patches for the image based upon the desired analysis grid comprises projecting the patches upon image coordinates of the image.

8. The method of claim 7, wherein the patches upon the image coordinates are quadrilaterals.

9. The method of claim 7, further comprising projecting the patches upon the image coordinates to generate perspective patches.

10. The method of claim 9, wherein the perspective patches are rectangular.

11. The method of claim 1, wherein convolving each of the patches with the feature detection filter generating the feature-based filter response comprises utilizing a Leung-Malik filter bank.

12. The method of claim 1, wherein convolving each of the patches with the feature detection filter generating the feature-based filter response comprises utilizing a Gabor filter bank.

13. The method of claim 1, wherein extracting features based upon the feature-based filter response comprises features including vehicles, pedestrians, lane-markers, road-edges, and lanes of travel.

14. Method for detecting a clear path of travel for a vehicle utilizing an image generated by a camera device located upon the vehicle, the method comprising:

monitoring the image, the image including a view of a road scene;

generating a plurality of patches onto the image, wherein the patches are collectively arranged to substantially include the clear path of travel;

analyzing the patches comprising

within each patch, discriminating image properties indicative of obstacles based upon trained classifiers, and

determining each of the patches to represent the clear path of travel or to not represent the clear path of travel based upon the image properties;

determining the clear path of travel based upon the plurality of patches; and

utilizing the clear path of travel to navigate the vehicle.

15. The method of claim 14, wherein discriminating image properties indicative of obstacles comprises monitoring color in the patches.

16. The method of claim 14, wherein discriminating image properties indicative of obstacles comprises monitoring texture in the patches.

17. The method of claim 14, wherein discriminating image properties indicative of obstacles comprises monitoring recognizable feature patterns in the patches.

18. The method of claim 14, wherein discriminating image properties comprises extracting features from the image;

and

the method further comprising classifying the extracted features, wherein the classifications are based upon a predetermined threshold response.

19. A system for detecting a clear path of travel for a vehicle utilizing an image, the method comprising:

a camera device located upon the vehicle and configured to generate an image; and

a processing module monitoring the image, generating a plurality of patches onto the image, and analyzing the patches utilizing feature extraction, wherein analyzing the patches comprises

convolving each of the patches with a feature detection filter generating a feature-based filter response,

extracting features based upon the feature-based filter response,

determining each of the patches to represent the clear path of travel or to not represent the clear path of travel based upon the extracted features,

determining the clear path of travel based upon the plurality of patches, and

utilizing the clear path of travel to navigate the vehicle;

wherein the patches are collectively arranged to substantially include the clear path of travel.

20. The system of claim 19, further comprising an autonomous steering system and a speed control device, wherein the autonomous steering system and the speed control device are operative based upon the analyzing the patches on the image.