Method of detecting bends on a road and system implementing same

• US 7,689,002 B2
• Filed: 12/19/2003
• Issued: 03/30/2010
• Est. Priority Date: 12/20/2002
• Status: Active Grant

• Current Assignee: Valeo Vision (Valeo SE)
• Original Assignee: Valeo Vision (Valeo SE)
• Inventors: Milgram, Maurice, Ivaldi, William, Achard, Catherine, Dumoncel, Didierl
• Primary Examiner(s): Le; Brian Quang
• Assistant Examiner(s): WANG, CLAIRE X
• Application Number:
  US10/743,444
• Publication Number:
  US 20040136568A1
• Time in Patent Office: 2,293 Days
• Field of Search: None
• US Class Current: 382/104
• CPC Class Codes: G06T 7/269 using gradient-based methods
• 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

The invention concerns a method of detecting from a vehicle a bend in a road comprising a surface and road edges, comprising the following operations:

• • taking an image of a road scene unfolding in front of the vehicle and at least partly illuminated by the vehicle,
  • determining, for each pixel in the image a light decrease gradient,
  • analysing these decrease gradients and determining an image of the road edges,
  • mathematically discriminating the gradients from the image of the road edges,
  • analyzing this discrimination and determining edges, and an angle of the bend.

The invention also concerns a system for implementing the method of the invention comprising a camera mounted in the vehicle, an image processing unit and a neural network.

Claims

1. A method of detecting from a vehicle variations in path on a road having a surface and road edges comprising:

at least partly illuminating a road scene near the vehicle;

taking an image of a road scene having a plurality of pixels;

determining, for each of the pixels in the image, a light decrease gradient;

analyzing these light decrease gradients and determining an image of the road edges;

calculating a number of the pixels of the image whose light decrease gradient is oriented from a left towards a center of the image and a number of the pixels of the image whose light decrease gradient is oriented from a right towards the center of the image; and

determining an angle of a bend of the road by comparing a distribution of luminosity of the pixels.

2. A method according to claim 1, wherein the light decrease gradient of an elementary image part corresponds to a decrease vector of light formed between adjacent pixels.

3. A method according to claim 2, wherein the analysis of the light decreases gradients comprises a thresholding of the decrease vectors and an elimination of the decrease vectors outside the threshold.

4. A method according to claim 2, wherein the analysis step comprises counting a number of elementary image parts having a vector oriented in one direction and a number of elementary image parts have decrease vector oriented in an opposite direction.

5. A method according to claim 3, wherein the calculating of the number of the pixels is carried out pixel column by pixel column, or by groups of pixel columns.

6. A method according to claim 1, wherein the analysis is carried out by a neural network.

7. A method according to claim 6, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

8. A system for detecting a bend in a road from a vehicle implementing the method according to claim 1, comprising a camera mounted in the vehicle, an image processing unit and a neural network.

9. A system of detecting a bend according to claim 8, wherein the neural network is integrated in the image processing unit.

10. A system for detecting a bend according to claim 8 that is connected to a vehicle headlight, movable or fixed and modulated for intensity.

11. A method according to claim 3, wherein the analysis step comprises counting a number of elementary image parts having a vector oriented in one direction and a number of elementary image parts have a decrease vector oriented in an opposite direction.

12. A method according to claim 2, wherein the analysis is carried out by a neural network.

13. A method according to claim 3, wherein the analysis is carried out by neural network.

14. A method according to claim 4, wherein the analysis is carried out by neural network.

15. A method according to claim 5, wherein the analysis is carried out by neural network.

16. A method according to claim 1, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

17. A method according to claim 2, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

18. A method according to claim 3, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

19. A method according to claim 4, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

20. A method according to claim 5, wherein the neural network has previously learnt geometries of bends and corresponding distributions of luminosity of the pixels.

21. A system for detecting a bend in a road from a vehicle implementing the method according to claim 2 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

22. A system for detecting a bend in a road from a vehicle implementing the method according to claim 3 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

23. A system for detecting a bend in a road from a vehicle implementing the method according to claim 4 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

24. A system for detecting a bend in a road from a vehicle implementing the method according to claim 5 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

25. A system for detecting a bend in a road from a vehicle implementing the method according to claim 6 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

26. A system for detecting a bend in a road from a vehicle implementing the method according to claim 7 comprising a camera mounted in the vehicle, an image processing unit and a neural network.

27. A system for detecting a bend according to claim 9 that is connected to a vehicle headlight, movable or fixed and modulated for intensity.

28. A method according to claim 1, further comprising controlling lateral orientation of at least one vehicle headlight based on the determination of the angle of the road.

29. A method according to claim 1, wherein the analysis step comprises generating a first curve corresponding to the number of the pixels having a decrease vector oriented from the left of the image towards the center of the image, and generating a second curve corresponding to the number of the pixels having a decrease vector oriented from the left of the image towards the center of the image.

30. A method according to claim 29, wherein the step of analyzing comprises deducing a shape of the road ahead based on a form of the first and second curves.