Method and apparatus for determining the position of a vehicle seat

1. A method for preprocessing an image of a vehicle interior to remove a portion of the image corresponding to a seat back comprising:

generating a contour image from the image;

selecting one of a plurality of contour models that most closely matches the generated contour image, a given contour model being associated with one of a plurality of possible seat positions;

determining a plane of separation between a seat back and a seat occupant within the image based on the selected contour model; and

removing a portion of the image corresponding to the seat back based on the determined plane of separation to isolate an image of the seat occupant.

2. The method of claim 1 , wherein each of the contour models represents a three-dimensional contour.

3. The method of claim 1 , further comprising locating a headrest associated with the seat according to the selected contour model.

4. The method of claim 3 , further comprising eliminating a head candidate from an associated head tracking algorithm according to the located headrest.

5. The method of claim 1 , wherein selecting a contour model includes projecting a plurality of representative points associated with each contour model onto a coordinate plane defined by the contour image and determining the sum of the squared distances between the representative points and neighboring points on the contour image.

6. The method of claim 1 , further comprising classifying the image of the seat occupant into one of a plurality of output classes.

7. The method of claim 6 , further comprising providing the selected one of the plurality of output classes to a controller for a vehicle occupant protection system.

8. The method of claim 1 , wherein determining a plane of separation includes determining a rear contour of the seat based on the selected contour model and calculating an associated offset for a plurality of points on the rear contour.

9. The method of claim 1 , wherein generating a contour image includes applying an edge detection algorithm to a grayscale image of the vehicle interior.

10. The method of claim 1 , wherein generating a contour image includes generating a stereo disparity map of the vehicle interior.

11. The method of claim 10 , wherein generating a contour image includes applying an edge detection algorithm to a grayscale image of the vehicle interior and combining the resulting image with the stereo disparity map to produce a three-dimensional contour.

12. The method of claim 1 , wherein selecting one of a plurality of contour models includes determining an associated reference location within the contour image for each of the plurality of contour models.

13. A system for editing an image of a vehicle interior to isolate the occupant of a vehicle seat comprising:

an image generator that generates an image of a vehicle interior;

a contour generator that produces a contour image from the generated image;

a model selector that selects a contour model from a plurality of contour models according to the similarity of the contour model to a portion of the contour image; and

an image editor that removes a portion of the generated image representing at least a portion of the vehicle seat according to the selected contour, model to produce an image of an occupant of the vehicle seat.

14. The system of claim 13 , the contour generator including a Canny edge detector that locates abrupt changes in-intensity in a grayscale image.

15. The system of claim 13 , the image generator comprising a disparity processing component that creates a stereo disparity map from a plurality of images.

16. The system of claim 13 , a given contour model representing the rearmost contour of a vehicle seat when the seat assumes one of a plurality of possible positions.

17. The system of claim 13 , further comprising a pattern recognition classifier that classifies the image of the occupant of the vehicle seat into one of a plurality of output classes.

18. A computer program product, implemented in a computer readable medium and operative in a data processing system, for editing an image of a vehicle interior to isolate the occupant of a vehicle seat comprising:

a contour generator that produces a contour image from one or more generated images;

a model selector that selects a contour model from a plurality of contour models according to the contour image and an associated selection algorithm; and

an image editor that edits the generated image to remove a portion of the image representing the vehicle seat according to the selected contour model.

19. The computer program product of claim 18 , wherein the model selector selects a contour model representing a headrest associated with the vehicle seat and provides the location of the headrest to an associated occupant head tracking algorithm to eliminate a head candidate within the algorithm.

20. The computer program product of claim 18 , a given contour model representing the rearmost contour of the vehicle seat when the seat assumes one of a plurality of possible positions.

21. The computer program product of claim 18 , the model selector being operative to project a plurality of representative points associated with each contour model onto a coordinate plane defined by the contour image and determine the sum of the squared distances between the representative points and neighboring points on the contour image.

22. The computer program product of claim 18 , further comprising a pattern recognition classifier that is operative to classify the edited image into one of a plurality of output classes and provide a classification output to an associated vehicle occupant protection system.