Method of detecting a structure in a field, a method of steering an agricultural vehicle and an agricultural vehicle

1. A method of steering an agricultural vehicle or implement, comprising the steps of:

providing an image of a field that includes a specific structure;

analysing the image of the field to obtain texture information,

wherein at least during a learning or training step, the image of the field is divided at least in an upper part and a lower part, wherein the upper part is further away from the agricultural vehicle in a direction of travel than the lower part, wherein for the upper part, texture information of respective areas is obtained while the areas are in the upper part of the image and, from a subsequent image captured at a time when the agricultural vehicle or implement has advanced a distance corresponding to the height of the lower half of the image, the areas belonging to the specific structure and the areas belonging to the field beside the specific structure are determined;

assigning to a plurality of areas of the image probability-values reflecting a likelihood that a respective area relates to the specific structure;

assuming at least one geometric property of the specific structure; and

establishing a most possible position parameter of the specific structure taking into account the probability-values and the assumed geometric property.

2. A method of detecting a structure in a field as defined in claim 1 , further comprising selecting the at least one geometric property from the group consisting of a borderline, a centreline, a width, a position and an attitude.

3. A method of detecting a structure in a field as defined in claim 1 , wherein said assigning includes assigning the plurality of areas the probability-values reflecting the likelihood that the respective area relates to a structure different from the specific structure.

4. A method of detecting a structure in a field as defined in claim 1 , further comprising using the probability-values which are relative values.

5. A method of detecting a structure in a field as defined in claim 1 , further comprising establishing the probability-value reflecting the likelihood of the specific structure being present in the image.

6. A method of detecting a structure in a field as defined in claim 1 , further comprising, prior to the analyzing the image to obtain texture information, performing a learning step to obtain information about a texture of the structure in the field and information about a texture of the field beside the structure.

7. A method of detecting a structure in a field as defined in claim 1 , further comprising analyzing an upper part and a lower part of the image individually to obtain the texture information.

8. A method of steering an agricultural vehicle or implement in relation to a structure in a field, comprising the steps of:

defining the structure by providing an image of the field,

analysing the image of the field to obtain texture information,

wherein at least during a learning or training step, the image of the field is divided in an upper part and a lower part that are analyzed separately, wherein the upper part is further away from the agricultural vehicle in a direction of travel than the lower part, wherein for the upper part, texture information of respective areas is obtained while the areas are in the upper part of the image and, from a subsequent image captured at a time when the agricultural vehicle or implement has advanced a distance corresponding to the height of the lower half of the image, the areas belonging to the specific structure and the areas belonging to the field beside the specific structure are determined;

assigning to a plurality of areas of the image probability-values reflecting a likelihood that a respective area relates to the specific structure,

assuming at least one geometric property of the specific structure,

establishing a most possible position parameter of the specific structure taking into account the probability-values and the assumed geometric property; and

providing a steering signal in accordance with the position parameter thus established.

9. A method of steering an agricultural vehicle or implement as defined in claim 8 , further comprising selecting the at least one geometric property from the group consisting of a borderline, a centreline, a width, a position and an attitude.

10. A method of steering an agricultural vehicle or implement as defined in claim 8 , wherein said assigning includes assigning the plurality of areas the probability-values reflecting the likelihood that a respective area relates to a structure different from the specific structure.

11. A method of steering an agricultural vehicle or implement as defined in claim 8 , further comprising using the probability-values which are relative values.

12. A method of steering an agricultural vehicle or implement as defined in claim 8 , further comprising establishing the probability-value reflecting the likelihood of the specific structure being present in the image, and taking the latter probability-value into account when providing the steering signal.

13. An agricultural vehicle, comprising:

a steering system providing steering signals, and

an imaging device for imaging surroundings of the vehicle and an image processing device,

wherein said steering system operates to provide by said imaging device an image of the field that includes a specific structure, to analyze the image of the field to obtain texture information, to assign to a plurality of areas of the image probability-values reflecting a likelihood that a respective area relates to the specific structure, to assume at least one geometric property of the specific structure, and establish a most possible position parameter of the specific structure taking into account the probability-values and the assumed geometric property and to provide a steering signal in accordance with the position parameter thus established,

wherein at least during a learning or training step, the image of the field is divided in an upper part and a lower part are analyzed separately wherein the upper part is further away from the agricultural vehicle in a direction of travel than the lower part, and wherein for the upper part of the image, the texture information of respective areas is obtained while the areas are in the upper part of the image and, from a subsequent image captured at a time when the agricultural vehicle or implement has advanced a distance corresponding to the height of the lower half of the image, the areas belonging to the specific structure and the areas belonging to the field beside the specific structure are determined.

14. An agricultural vehicle as defined in claim 13 , wherein the steering system includes at least one additional navigation sensor and operates taking into account an output of the additional navigation sensor when providing the steering signal.

15. An agricultural vehicle as defined in claim 14 , wherein the at least one additional navigation sensor is a sensor selected from the group consisting of a GPS-sensor, an optical field contour sensor, an inertial measuring unit, and an odometry sensor.