REFINED ROW GUIDANCE PARAMETERIZATION WITH HOUGH TRANSFORM

• US 20150321694A1
• Filed: 05/07/2015
• Published: 11/12/2015
• Est. Priority Date: 05/09/2014
• Status: Active Grant

• Current Assignee: Raven Industries Incorporated (CNH Industrial N.V.)
• Original Assignee: Raven Industries Incorporated (CNH Industrial N.V.)
• Inventors: Nelson, Robert Leonard Jr., Wilson, Edwin Ernest
• Granted Patent: US 9,446,791 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: A01B 69/008 automatic
• CPC Class Codes: B62D 6/00 Arrangements for automatically controlling steering depending on driving conditions sensed and responded to, e.g. control circuits means for initiating a change in direction B62D1/00; steering valves B62D5/06; combined with means for inclining the vehicle body or wheels on bends B62D9/00
• CPC Class Codes: B62D 6/005 treating sensor outputs to obtain the actual yaw rate
• CPC Class Codes: G05D 1/021 specially adapted to land vehicles
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06K 2207/1016 Motor control or optical moving unit control
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/20061 Hough transform
• CPC Class Codes: G06T 2207/30188 Vegetation; Agriculture
• CPC Class Codes: G06T 2207/30241 Trajectory
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• 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: G06V 20/68 Food, e.g. fruit or vegetables

Abstract

Systems and techniques for refined row guidance parameterization with Hough transform are described herein. An electronic representation of a field (ERF) can be received. A steering variable calculation can be performed by executing a cascaded Hough transform on the ERF. An intermediate value used in calculating the steering variable can be selected. A refined intermediate value can be identified from the intermediate value by measuring a difference between the intermediate value and an anchor value. The steering variable calculation can be adjusted using the refined intermediate value

Claims

1. A machine-readable medium that is not a transitory propagating signal, the machine-readable medium including instructions that, when executed by a machine, cause the machine to perform operations comprising:

receiving an electronic representation of a field (ERF);

performing a steering variable calculation by executing a cascaded Hough transform on the ERF;

selecting an intermediate value used in calculating the steering variable;

identifying a refined intermediate value from the intermediate value by measuring a difference between the intermediate value and an anchor value; and

adjusting the steering variable calculation using the refined intermediate value.

2. The machine-readable medium of claim 1, wherein the anchor value is a combination of one or more previously calculated intermediate values;

wherein measuring the intermediate value to the anchor value includes;

determining a difference between the intermediate value and the anchor value; and

comparing the difference to a threshold;

wherein identifying the refined intermediate value includes rejecting the intermediate value when the difference is beyond the threshold; and

wherein adjusting the steering variable calculation using the refined intermediate value in place of the intermediate value includes abandoning the steering variable calculation in response to rejecting the intermediate value.

3. The machine-readable medium of claim 2, wherein the combination of one or more previously calculated intermediate values is an average of the one or more previously calculated intermediate values.

4. The machine-readable medium of claim 2, wherein the intermediate value is a current vanishing point, wherein the anchor value is a combination of previous vanishing points, and wherein the difference between the intermediate value and the anchor value is a distance between the current vanishing point and the combination of previous vanishing points.

5. The machine-readable medium of claim 4, wherein the distance is determined by at least one of sum of squares, root sum of squares, or sum of absolute values.

6. The machine-readable medium of claim 4, wherein the threshold is calibrated to a mechanism used to determine the distance.

7. The machine-readable medium of claim 1, wherein the intermediate value is a mousetail magnitude from a plurality of moustail magnitudes computed for the steering variable calculation, each of the plurality of mousetail magnitudes being a magnitude of a sum of vectors, the vectors each having an angle representing a distance of a corresponding line intercept from a camera intercept represented in a first parameter space of the cascaded Hough transform;

wherein the anchor value is a greatest mousetail magnitude of the plurality of mousetail magnitudes within the steering variable calculation; and

wherein identifying the refined intermediate value includes modifying a component that contributes to a vector of the vectors to produce the refined intermediate value, the refined intermediate value being a mousetail magnitude not less than the anchor and greater than the intermediate value.

8. The machine-readable medium of claim 7, wherein the component is a threshold used to classify pixels in the ERF.

9. The machine-readable medium of claim 8, wherein the threshold is a color threshold used to distinguish between crop pixels and furrow pixels.

10. A machine-implemented method comprising:

receiving an electronic representation of a field (ERF);

performing a steering variable calculation by executing a cascaded Hough transform on the ERF;

selecting an intermediate value used in calculating the steering variable;

identifying a refined intermediate value from the intermediate value by measuring a difference between the intermediate value and an anchor value; and

adjusting the steering variable calculation using the refined intermediate value.

11. The method of claim 10, wherein the anchor value is a combination of one or more previously calculated intermediate values;

wherein measuring the intermediate value to the anchor value includes;

determining a difference between the intermediate value and the anchor value; and

comparing the difference to a threshold;

wherein identifying the refined intermediate value includes rejecting the intermediate value when the difference is beyond the threshold; and

wherein adjusting the steering variable calculation using the refined intermediate value in place of the intermediate value includes abandoning the steering variable calculation in response to rejecting the intermediate value.

12. The method of claim 11, wherein the combination of one or more previously calculated intermediate values is an average of the one or more previously calculated intermediate values.

13. The method of claim 11, wherein the intermediate value is a current vanishing point, wherein the anchor value is a combination of previous vanishing points, and wherein the difference between the intermediate value and the anchor value is a distance between the current vanishing point and the combination of previous vanishing points.

14. The method of claim 13, wherein the distance is determined by at least one of sum of squares, root sum of squares, or sum of absolute values.

15. The method of claim 13, wherein the threshold is calibrated to a mechanism used to determine the distance.

16. The method of claim 10, wherein the intermediate value is a mousetail magnitude from a plurality of moustail magnitudes computed for the steering variable calculation, each of the plurality of mousetail magnitudes being a magnitude of a sum of vectors, the vectors each having an angle representing a distance of a corresponding line intercept from a camera intercept represented in a first parameter space of the cascaded Hough transform;

wherein the anchor value is a greatest mousetail magnitude of the plurality of mousetail magnitudes within the steering variable calculation; and

wherein identifying the refined intermediate value includes modifying a component that contributes to a vector of the vectors to produce the refined intermediate value, the refined intermediate value being a mousetail magnitude not less than the anchor and greater than the intermediate value.

17. The method of claim 16, wherein the component is a threshold used to classify pixels in the ERF.

18. The method of claim 17, wherein the threshold is a color threshold used to distinguish between crop pixels and furrow pixels.

19. A system comprising:

a scene module arranged to receive an electronic representation of a field (ERF);

a transform module arranged to perform a cascaded Hough transform on the ERF; and

a steering parameter calculation module arranged to;

perform a steering variable calculation by executing a cascaded Hough transform on the ERF;

select an intermediate value used in calculating the steering variable;

identify a refined intermediate value from the intermediate value by measuring a difference between the intermediate value and an anchor value; and

adjust the steering variable calculation using the refined intermediate value.

20. The system of claim 19, wherein the anchor value is a combination of one or more previously calculated intermediate values;

wherein to measure the intermediate value to the anchor value includes the steering parameter calculation module arranged to;

determine a difference between the intermediate value and the anchor value; and

compare the difference to a threshold;

wherein to identify the refined intermediate value includes the steering parameter calculation module arranged to reject the intermediate value when the difference is beyond the threshold; and

wherein to adjust the steering variable calculation using the refined intermediate value in place of the intermediate value includes the steering parameter calculation module arranged to abandon the steering variable calculation in response to rejecting the intermediate value.

21. The system of claim 20, wherein the combination of one or more previously calculated intermediate values is an average of the one or more previously calculated intermediate values.

22. The system of claim 20, wherein the intermediate value is a current vanishing point, wherein the anchor value is a combination of previous vanishing points, and wherein the difference between the intermediate value and the anchor value is a distance between the current vanishing point and the combination of previous vanishing points.

23. The system of claim 22, wherein the distance is determined by at least one of sum of squares, root sum of squares, or sum of absolute values.

24. The system of claim 22, wherein the threshold is calibrated to a mechanism used to determine the distance.

25. The system of claim 19, wherein the intermediate value is a mousetail magnitude from a plurality of moustail magnitudes computed for the steering variable calculation, each of the plurality of mousetail magnitudes being a magnitude of a sum of vectors, the vectors each having an angle representing a distance of a corresponding line intercept from a camera intercept represented in a first parameter space of the cascaded Hough transform;

wherein the anchor value is a greatest mousetail magnitude of the plurality of mousetail magnitudes within the steering variable calculation; and

wherein to identify the refined intermediate value includes the steering parameter calculation module arranged to modify a component that contributes to a vector of the vectors to produce the refined intermediate value, the refined intermediate value being a mousetail magnitude not less than the anchor and greater than the intermediate value.

26. The system of claim 25, wherein the component is a threshold used to classify pixels in the ERF.

27. The system of claim 26, wherein the threshold is a color threshold used to distinguish between crop pixels and furrow pixels.