Dynamic radial contour extraction by splitting homogeneous areas
View Patent ↗Systems and methods for extracting a radial contour around a given point in an image includes providing an image including a point about which a radial contour is to be extracted around. A plurality of directions around the point and a plurality of radius lengths for each direction are provided. Local costs are determined for all radius lengths for each direction by comparing texture variances at each radius length with the texture variance at a further radius length. A radius length is determined, using a processor, for each direction based on the accumulated value of the local costs to provide a radial contour.
1. A method for extracting a radial contour around a given point in an image, comprising:
providing an image including a point about which a radial contour is to be extracted around;
providing a plurality of directions around the point;
for each direction, providing a plurality of radius lengths;
determining local costs for all radius lengths for each direction by comparing texture variances at each radius length with the texture variance at a further radius length; and
determining a radius length, using a processor, for each direction based on the accumulated value of the local costs to provide a radial contour.
2. The method as recited in claim 1 , wherein determining local costs include subtracting the texture variance at each radius length from the texture variance at a further radius length and adding a constant.
3. The method as recited in claim 2 , wherein the constant ensures the local costs are nonnegative.
4. The method as recited in claim 1 , wherein determining a radius length includes applying a dynamic programming method to determine a plurality of sets, each set including a radius length for each direction, wherein a set is selected as the radial contour such that the sum of the local costs for the selected set is minimal compared to the other sets.
5. The method as recited in claim 4 , further comprising rejecting a set where the first radius length does not equal the last radius length.
6. The method as recited in claim 1 , wherein a penalty is imposed for changes in radius lengths at two consecutive directions that are larger than a length constraint value.
7. The method as recited in claim 6 , wherein the penalty is an infinite penalty.
8. The method as recited in claim 1 , wherein the image is an image of tissue.
9. A computer readable storage medium comprising a computer readable program, wherein the computer readable program when executed on a computer causes the computer to execute the method as recited in claim 1 .
10. A method for extracting a radial contour around a given point in an image, comprising:
providing an image including a point about which a radial contour is to be extracted around;
providing a plurality of directions around the point;
for each direction, providing a plurality of radius lengths;
determining local costs for all radius lengths for each direction; and
applying a dynamic programming method, using a processor, to determine a plurality of sets, each set including a radius length for each direction, wherein a set is rejected where the first radius length does not equal the last radius length, and further wherein a set is selected as the radial contour such that the sum of the local costs for the selected set is minimal compared to the other non-rejected sets.
11. The method as recited in claim 10 , wherein determining local costs include subtracting the texture variance at each radius length from the texture variance at a further radius length and adding a constant.
12. The method as recited in claim 11 , wherein the constant ensures that the local costs are nonnegative.
13. The method as recited in claim 10 , wherein an infinite penalty is imposed for changes in radius lengths at two consecutive directions that are larger than a length constraint value.
14. A system for extracting a radial contour around a given point in an image, comprising:
providing an image including a point about which a radial contour is to be extracted around;
providing a plurality of directions around the point;
for each direction, providing a plurality of radius lengths;
a feature measurement module configured to determine texture variances for each radius length at each direction;
a cost determination module configured to determine local costs for all radius lengths for each direction by comparing texture variances at each radius length with the texture variance at a further radius length; and
a minimization module configured to determine a radius length for each direction based on the accumulated value of the local costs to provide a radial contour.
15. The system as recited in claim 14 , wherein determining local costs include subtracting the texture variance at each radius length from the texture variance at a further radius length and adding a constant.
16. The system as recited in claim 15 , wherein the constant ensures the local costs are nonnegative.
17. The system as recited in claim 14 , wherein determining a radius length for each direction includes applying a dynamic programming method to determine a plurality of sets, each set including a radius length for each direction, wherein a set is selected as the radial contour such that the sum of the local costs for the selected set is minimal compared to the other sets.
18. The system as recited in claim 17 , further comprising rejecting a set where the first radius length does not equal the last radius length.
19. The system as recited in claim 14 , wherein a penalty is imposed for changes in radius lengths at two consecutive directions that are larger than a length constraint value.
20. The system as recited in claim 19 , wherein the penalty is an infinite penalty.