Method and apparatus for processing an image of an agricultural field
First Claim
1. A method of detecting an edge between a cut crop and an uncut crop in a field, the method comprising:
- processing at least two scanlines of pixel data based on an image of the field;
generating a field boundary approximating the edge between a cut crop and an uncut crop in a field;
calculating a first characteristic of the image after processing each scanline of pixel data;
calculating a second characteristic of the image after processing each scanline of pixel data; and
determining whether to continue processing the pixel data based on the first characteristic, the second characteristic and the location of a particular scanline being processed in relation to the image.
3 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for detecting an edge between a cut crop and an uncut crop in a field is disclosed. At least two scanlines of pixel data are processed based on an image of the field. A characteristic line corresponding to the cut/uncut edge is best fit between center location points of segments from each scanline using sequential regression analysis. A fuzzy logic algorithm is used to determine whether to continue processing each successive scanline of data based on the quality of the characteristic line. If certain fuzzy membership functions determine that the slope and intercept values have sufficiently converged, processing of the scanlines is completed and the agricultural vehicle is automatically guided along the cut/uncut edge corresponding to the characteristic line.
50 Citations
53 Claims
-
1. A method of detecting an edge between a cut crop and an uncut crop in a field, the method comprising:
-
processing at least two scanlines of pixel data based on an image of the field;
generating a field boundary approximating the edge between a cut crop and an uncut crop in a field;
calculating a first characteristic of the image after processing each scanline of pixel data;
calculating a second characteristic of the image after processing each scanline of pixel data; and
determining whether to continue processing the pixel data based on the first characteristic, the second characteristic and the location of a particular scanline being processed in relation to the image. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 24, 25, 26)
-
10. The method according to claim 9, wherein each of pixels are assigned a single bit value.
-
11. The method according to claim 10, wherein each of the pixels in the first class are assigned a value of 1 and each of the pixels in the second class are assigned a value of 0.
-
12. The method according to claim 11, wherein the generating step includes dividing the scanline into a plurality of segments based on calculating a plurality of transition points between the first class and the second class, calculating a length for each of the segments, and calculating a center location for each of the plurality of segments.
-
13. The method according to claim 12, wherein the length of the plurality of segments is calculated based on
-
14. The method according to claim 12, wherein the center location for each of the plurality of segments is calculated based on
-
j 2 ) wherein I is the distance between transitions, x is the column location of the transition, j is the row index and i is an index of transitions within the row.
-
-
15. The method according to claim 12, wherein a characteristic line representing the field boundary is fit to a set of points comprising the center location of the longest segment in each scanline.
-
16. The method of claim 15, wherein the characteristic line is calculated for a first time after a predetermined number of the scanlines is processed.
-
17. The method of claim 16, wherein subsequent to the characteristic line being calculated for the first time, the characteristic line is recalculated after each successive scanline is processed.
-
18. The method according to claim 17, wherein the step of calculating the first characteristic of the image includes determining a slope covariance of the image based on the characteristic line.
-
19. The method according to claim 18, wherein the step of calculating the second characteristic of the image includes determining an intercept covariance of the image based on the characteristic line.
-
20. The method according to claim 19, wherein the slope covariance and the intercept covariance are recalculated after each successive scanline is processed.
-
21. The method according to claim 20, wherein the slope covariance and the intercept covariance are calculated using sequential linear regression analysis.
-
22. The method according to claim 21, wherein the pixel data of a particular scanline is weighted proportional to the location of the particular scanline in relation to the bottom of the image.
-
24. The method according to claim 1, wherein the determining step includes assigning a first membership function to the first characteristic, a second membership function to the second characteristic and a third membership function to the location of the particular scanline.
-
25. The method according to claim 24, wherein the first membership function assigns a first acceptability value to the first characteristic, the second membership function assigns a second acceptability value to the second characteristic, and the third membership function assigns a third acceptability value to the location of the particular scanline.
-
26. The method according to claim 25, wherein the first, second and third acceptability values are combined and compared to a threshold value to determine whether to continue processing the scanlines of the pixel data.
-
-
23. A method of detecting an edge between a cut crop and an uncut crop in a field, the method comprising:
-
processing at least two scanlines of pixel data based on an image of the field;
generating a field boundary;
calculating a first characteristic of the image after processing each scanline of pixel data;
calculating a second characteristic of the image after processing each scanline of pixel data; and
determining whether to continue processing the pixel data based on the first characteristic, the second characteristic and the location of a particular scanline being processed in relation to the image, wherein the scanline of pixel data that is processed first is obtained from a bottom of the image, wherein the processing step further includes classifying the pixel data into a first class of data and a second class of data, wherein the first class of data represents the cut crop and the second class of data represents the uncut crop, wherein the pixel data is processed based on an adaptive clustering algorithm, wherein the pixel data includes a plurality of pixels and each of the pixels include a red component, a green component and a blue component, wherein each of the pixels are assigned to the first class or the second class based on whether the red, green and blue components of each of the pixels are closest to the first class or the second class, wherein each of the pixels are determined to be closest to the first class or the second class by calculating a first minimum RGB distance between each of the pixels and the first class, calculating a second minimum RGB distance between each of the pixels and the second class, and assigning each of the pixels to the first class if the first minimum RGB distance is smaller than the second minimum RGB distance, wherein the first and second minimum RGB distances are calculated based on
-
-
27. An agricultural vehicle configured to be guided through a field of crops by an automatic guidance system, the agricultural vehicle comprising:
-
at least one camera mounted on the agricultural vehicle;
an image processor configured to process at least two scanlines of pixel data based on an image of the field; and
a central processing unit configured to generate a field boundary approximating an edge between a cut crop and an uncut crop in a field, calculate a first characteristic of the image, calculate a second characteristic of the image, and determine whether to continue processing the pixel data based on the first characteristic, the second characteristic and the location of a particular scanline being processed in relation to the image. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47)
-
36. The agricultural vehicle according to claim 35, wherein each of pixels are assigned a single bit value.
-
37. The agricultural vehicle according to claim 36, wherein each of the pixels in the first class are assigned a value of 1 and each of the pixels in the second class are assigned a value of 0.
-
38. The agricultural vehicle according to claim 37, wherein a particular scanline being processed is divided into a plurality of segments by calculating a plurality of transition points between the first class and the second class, calculating a length for each of the segments, and calculating a center location for each of the plurality of segments.
-
39. The agricultural vehicle according to claim 38, wherein the length of the plurality of segments is calculated based on
-
40. The agricultural vehicle according to claim 39, wherein the center location for each of the plurality of segments is calculated based on
-
j 2 ) wherein i is the distance between transitions, x is the column location of the transition, j is a row index and I is an index of transitions within the row.
-
-
41. The agricultural vehicle according to claim 40, wherein a characteristic line representing the field boundary is fit to a set of points comprising the center location of the longest segment in each scanline.
-
42. The agricultural vehicle according to claim 41, wherein the characteristic line is calculated for a first time after a predetermined number of the scanlines is processed.
-
43. The agricultural vehicle according to claim 42, wherein subsequent to the characteristic line being calculated for the first time, the characteristic line is recalculated after each successive scanline is processed.
-
44. The agricultural vehicle according to claim 43, wherein the first characteristic of the image is calculated by determining a slope covariance of the image based on the characteristic line.
-
45. The agricultural vehicle according to claim 44, wherein the second characteristic of the image is calculated by determining an intercept covariance of the image based on the characteristic line.
-
46. The agricultural vehicle according to claim 45, wherein the slope covariance and the intercept covariance are recalculated after each successive scanline is processed.
-
47. The agricultural vehicle according to claim 46, wherein the slope covariance and the intercept covariance are calculated using sequential linear regression analysis.
-
-
48. A method of detecting an edge between a cut crop and an uncut crop in a field, the method comprising:
-
processing at least two scanlines of pixel data based on an image of the field;
generating a field boundary approximating the edge between the cut crop and the uncut crop in a field and dividing the cut crop from the uncut crop;
calculating a first characteristic of the image;
calculating a second characteristic of the image; and
determining whether to continue processing the pixel data based on a convergence of the first and second characteristics to a first value and a second value, respectively, wherein the first and second values are compared to a predetermined covariance threshold. - View Dependent Claims (49, 50, 51, 52, 53)
-
Specification