Process and steering system for the automatic steering of an agricultural vehicle

• US 20050102079A1
• Filed: 10/26/2004
• Published: 05/12/2005
• Est. Priority Date: 11/06/2003
• Status: Active Grant

• Current Assignee: Deere & Company
• Original Assignee: Deere & Company
• Inventors: Hofer, Jurgen, Bohner, Manfred, Freichel, Thorsten
• Granted Patent: US 7,400,957 B2
• Time in Patent Office: Days
• US Class Current: 701/41
• CPC Class Codes: A01B 69/001 Steering by means of optical assistance, e.g. television cameras steering devices for road marking vehicles E01C23/163
• CPC Class Codes: A01D 41/1278 for automatic steering

Abstract

An automatic steering of an agricultural tractor on a surface of a field to be processed is accomplished by equipping the tractor with a camera for taking a picture of a region of the field, which is in front of the tractor, including the surface that is to be processed, generating a pixel data file from a picture signal of the camera, generating texture information with respect to the texture of the surroundings of the pixels contained in the data file, classifying the pixels of the pixel data file while considering the texture information, in order to generate a binary information as to whether or not a particular pixel is to be associated with the surface that is to be processed, generating a steering signal based on the results of the classification, and repositioning the steering device of the harvesting machine according to the steering signal so that the machine is steered automatically along the surface that is to be processed.

Claims

1. A process for automatically steering an agricultural machine, equipped with a steering device, on a field along a surface that is to be processed, comprising the following steps:

a) taking a picture of a region of the field lying in front of the machine including the surface to be processed with a picture taking camera;

b) generating a pixel data file from a picture signal from the camera;

c) generating a texture information with respect to the texture of the surroundings of the pixels of the pixel data file;

d) classifying pixels of the pixel data file under consideration of the texture information, in order to generate a binary information, whether or not the particular pixels are to be assigned to the surface that is to be processed;

e) generating a steering signal based on the results of the classifications, and;

f) repositioning the steering device of the agricultural machine according to the steering signal, so that the machine is steered automatically along the surface to be processed.

2. The process, as defined in claim 1, wherein said texture information that is generated in step (c) is generated by a texture analysis that is one of (i) independent of direction, or (ii) dependent upon direction, particularly a grey scale value dependency analysis and/or a multi-spectral analysis.

3. The process, as defined in claim 1, wherein said classifying carried out in step (d) is done using a neuronal network.

4. The process, as defined in claim 1, wherein, prior to step (d) a learning step is performed wherein information relating to a picture taken of a field to be processed is provided showing certain regions that are to be processed and certain regions that are not to be processed.

5. An automatic steering system for automatically steering an agricultural harvesting machine, equipped with a steering device, on a field along a surface containing crop that is to be processed, comprising:

a picture taking camera mounted on said harvesting machine for taking a picture of a region of the field lying in front of the harvesting machine including the surface that is to be processed and for generating a picture signal;

a processing arrangement connected to said camera for receiving said picture signal from said camera and for generating a pixel data file from said picture signal;

said processing arrangement further including structure operating so as to generate texture information relative to the texture of the surroundings of the pixels of the pixel data file, to perform a classification of the pixels of the pixel data file considering the texture information, in order to generate binary information as to whether or not the particular pixel is to be associated with the surface to be processed and, based on the results of the classification, to generate a steering signal that is connected to said steering device so as to be transmitted to said steering device so that the machine can be automatically steered along the surface containing crop to be processed.

6. The steering system, as defined in claim 5, wherein said texture information that is generated is done one of independent of direction, or dependent upon direction, particularly a grey scale value dependency analysis and/or a spectral analysis.

7. The steering system, as defined in claim 5, wherein said classification of whether or not the individual pixels are to be associated with the pixel data file of the harvested crop that in each case is to be taken up is obtained by means of a neuronal network.

8. The steering system, as defined in claim 5, wherein said processing arrangement performing the classification is subjected to a learning process before the beginning of the operation by means of a picture supplied to it and information about regions of the picture that show a surface that is to be processed and regions of the picture that show a surface that is not to be processed.