Device and method for detecting plantation rows

1. A method comprising:

producing, by a processor, a frequency domain representation of a digital image comprising plants and identifying at least one desired patch in said frequency domain representation, according to predefined frequency characteristics of said desired patch;

removing, by the processor, all frequencies outside said desired patch to obtain a masked frequency domain; and

reverse transforming, by the processor, said masked frequency domain to obtain a reverse transformed image comprising spatial elements representing plant rows corresponding to said desired patch;

wherein said predefined frequency characteristics comprise:

at least one dominant frequency value;

at least one predefined frequency characteristic corresponding to at least one predefined spatial characteristic relating to at least one attribute of said plants; or

any combination thereof,

and wherein the attribute includes one of: planted row spacing, spacing between plants on the same row and direction of plant rows.

2. The method according to claim 1 , further comprising:

stretching said reverse transformed image to obtained a stretched image;

converting said stretched image to a single band image comprising black connected pixel components; and

transforming said black connected pixel components to a pattern of sub-area polylines representing plant rows.

3. A device comprising:

a memory to store a representation of plant rows in a predefined area; and

a processor configured to:

produce a frequency domain representation of the representation and identifying at least one desired patch in said frequency domain representation, according to predefined frequency characteristics of said desired patch;

remove all frequencies outside said desired patch to obtain a masked frequency domain; and

reverse transform, by a computing device, said masked frequency domain to obtain a reverse transformed image comprising spatial elements representing plant rows corresponding to said desired patch;

wherein said predefined frequency characteristics comprise:

at least one dominant frequency value;

at least one predefined frequency characteristic corresponding to at least one predefined spatial characteristic relating to at least one attribute of said plants; or

any combination thereof,

and wherein the attribute of said plant is planted row spacing, spacing between plants on the same row, direction of plant rows, typical row samples manually marked on the aerial image providing any of the above attributes, or any combination thereof.

4. The device according to claim 3 , further comprising causing the processor to:

stretch said reverse transformed image to obtained a stretched image;

convert said stretched image to a single band image comprising black connected pixel components; and

transform said black connected pixel components to a pattern of sub-area polylines representing plant rows.

5. A method comprising:

receiving, by a processor, a representation of plants in a predefined area;

assigning, by the processor, a vector to each plant in the representation according to its position, the position of neighbor plants, the distance between said plant and said neighbor plants, and the direction of a line formed by said plants and said neighbor plants;

associating, by the processor, the plants with curved lines, according to said vectors, to obtain estimated rows;

extending, by the processor, the estimated rows in the direction defined by the curved lines, to obtain extended rows;

defining, by the processor, a line connecting a first and second extended row having a common orientation, when the plants positioned on said line have a predefined density to obtain intermediate rows having a position and an orientation;

predicating, by the processor, rows parallel to any one of the intermediate rows, according to said position and orientation of intermediate rows, to obtain predicated parallel plant rows;

grouping, by the processor, plant rows, comprising intermediate plant rows and predicated parallel plant rows, into groups; and

removing, by the processor, intersecting groups, thereby obtaining plant rows.

6. The method according to claim 5 , further comprising

building a local directions map by dividing the predefined area into cells;

determining the dominant direction of the plants in each cell, thereby obtaining plant directions;

reiterating the method of claim 5 , wherein the vectors are reevaluated according to said plant directions.

7. A device comprising:

a memory to a representation of plants in a predefined area; and

a processor configured to:

assign a vector to each plant in the representation according to its position, the position of neighbor plants, the distance between said plant and said neighbor plants, and the direction of a line formed by said plants and said neighbor plants;

associate the plants with curved lines, according to said vectors, to obtain estimated rows;

extend the estimated rows in the direction defined by the curved lines, to obtain extended rows;

define a line connecting a first and second extended row having a common orientation, when the plants positioned on said line have a predefined density to obtain intermediate rows having a position and an orientation;

predicate rows parallel to any one of the intermediate rows, according to said position and orientation of intermediate rows, to obtain predicated parallel plant rows;

group plant rows, comprising intermediate plant rows and predicated parallel plant rows, into groups; and

remove intersecting groups, thereby obtaining plant rows.

8. The device according to claim 7 , further comprising causing the processor to:

build a local directions map by dividing the predefined area into cells;

determine the dominant direction of the plants in each cell, thereby obtaining plant directions;

reiterating the steps of claim 7 , wherein the vectors are reevaluated according to said plant directions.

9. A method comprising

storing in a memory a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves;

obtaining, by an image acquisition system, an image of said pre-defined area;

dividing, by a processor, said spatial elements, polylines or curves into segments according to pre-defined classifications relating to attributes of said image.

10. The method according to claim 9 , wherein the attributes are color levels or gray levels.

11. A device comprising:

a memory having stored thereon:

a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves; and

an image of said predefined area;

and

a processor configured to:

divide said spatial elements, polylines or curves into segments according to pre-defined classifications relating to attributes of said image.

12. The device according to claim 11 , wherein the attributes are color levels or gray levels.

13. A method comprising

storing in a memory a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves;

obtaining, by an image acquisition system, an image of said pre-defined area; and

distinguishing, by a processor, between spatial elements, polylines or curves representing planted and non-planted rows according to height attributes derived from said image, color of areas in said image, pattern of areas in said image or any combination thereof.

14. The method according to claim 13 , wherein the height attributes include linear segments representing ground altitude rows.

15. A device comprising:

a memory having stored thereon:

a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves; and

an image of said predefined area;

and

a processor configured to:

distinguish between spatial elements, polylines or curves representing planted and non-planted rows according to height attributes derived from said image, color of areas in said image, pattern of areas in said image or any combination thereof.

16. The device according to claim 15 , wherein the height attributes include linear segments representing ground altitude rows.

17. A method comprising

storing in a memory a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves;

obtaining attributes of said predefined area comprising height, surface color, vegetation class, soil data or any combination thereof; and

distinguishing, by a processor, between spatial elements, polylines or curves representing planted and non-planted rows according to said attributes;

wherein the attributes are not derived from an image of said predefined area.

18. The method according to claim 17 , wherein said attributes are obtained from a surface model containing height data.

19. A device comprising:

a memory having stored thereon:

a representation of plant rows in a predefined area, wherein the plant rows are represented by spatial elements corresponding a at least one frequency patch, polylines or curves; and

attributes of said predefined area comprising height, surface color, vegetation class, soil data or any combination thereof;

and

a processor configured to:

distinguish between spatial elements, polylines or curves representing planted and non-planted rows according to said attributes, wherein the attributes are not derived from an image of said predefined area.

20. The device according to claim 19 , wherein said attributes are obtained from a surface model containing height data.