Method and system for tracking multiple regional objects by multi-dimensional relaxation
First Claim
1. A method for tracking a plurality of objects comprising:
- using at least one sensor to repeatedly scan a region containing at least one object and generating a variable number of N frames of said region that provide a plurality of observations associated with said N frames that include information for said at least one object in said N frames;
determining a most probable partition of said observations into tracks and false observations by computing likelihood functions to generate a score to assign a sequence of said observations to a track and by determining a correct combination of said tracks by transforming a binary partitioning problem to a continuous partitioning problem that is solved and from which a binary solution is recovered;
using a sliding window having a length N that encompasses said N frames of said region to allow both provisional and irrevocable data association decisions to be made to assign said observations to said tracks within said sliding window by determining said most probable partition of said observations into said tracks.
0 Assignments
0 Petitions
Accused Products
Abstract
A method and system for real-time tracking of objects are disclosed. A region is repeatedly scanned providing a plurality of images or data sets having points corresponding to objects in the region to be tracked. Given a previously determined track for each object in the region, an M-dimensional combinatorial optimization assignment problem is formulated using the points from M−1 of the images or data sets, wherein each point is preferably used in extending at most one track. The M-dimensional problem is subsequently solved for an optimal or near-optimal assignment of the points to the tracks, extending the tracking of the objects so that a response to each object can be initiated by the system in real-time. Speed and accuracy is provided by an iterative Lagrangian Relaxation technique wherein a plurality of constraint dimensions are relaxed simultaneously to yield a reduced dimensional optimization problem whose solution is used to formulate an assignment problem of dimensionality less than M. The iterative reducing of dimensions terminates when exact solutions are determined for two-dimensional cases. A recovery procedure is used for determining a higher dimensional assignment problem solution from a problem having one less dimension. The procedure is useful when the reduced dimensional optimizational problem has two constraint dimensions.
59 Citations
20 Claims
-
1. A method for tracking a plurality of objects comprising:
-
using at least one sensor to repeatedly scan a region containing at least one object and generating a variable number of N frames of said region that provide a plurality of observations associated with said N frames that include information for said at least one object in said N frames;
determining a most probable partition of said observations into tracks and false observations by computing likelihood functions to generate a score to assign a sequence of said observations to a track and by determining a correct combination of said tracks by transforming a binary partitioning problem to a continuous partitioning problem that is solved and from which a binary solution is recovered;
using a sliding window having a length N that encompasses said N frames of said region to allow both provisional and irrevocable data association decisions to be made to assign said observations to said tracks within said sliding window by determining said most probable partition of said observations into said tracks. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
wherein each cl 1 . . . iN is included in said plurality of costs, each M1, i=1, . . . ,N, being one of;
(a) a number of observations in an ith frame of said N frames;
(b) a sum of a number of tracks in said plurality of tracks and a number of said observations in the ith frame not assigned to one of said tracks;
(c) a number of tracks in said plurality of tracks; and
(d) a number of said observations in the if frame not assigned to one of said tracks.
-
-
3. The method of claim 2 wherein said step of determining a most probable partition further comprises the steps of:
-
solving said continuous partitioning problem either approximately or optimally for values of for each zi 1 . . . iN for each i1, . . . , iN using a dual ascent method;
determining a binary value of zi 1 . . . iN in {0,1} for each i1 . . . iN corresponding to each zi1 . . . iN ,wherein said values zi 1 . . . iN provide an optimal or near optimal solution to said continuous partitioning problem.
-
-
4. The method of claim 2 wherein said step of determining a most probable partition further comprises the steps of:
-
solving said continuous partitioning problem either approximately or optimally for values of zi 1 . . . iN for each i1, . . . , iN using a linear programming algorithm that employs the primal simplex method or interior point method;
determining a binary value of zl 1 . . . iN in {0,1} for each i1 . . . iN corresponding to each zi1 . . . iN wherein said values zi 1 . . . iN provide an optimal or near optimal solution to said continuous partitioning problem.
-
-
5. The method of claim 2 wherein said step of determining a most probable partition further comprises the steps of:
-
solving said continuous partitioning problem either approximately or optimally for values of zl 1 . . . iN for each i1, . . . ,iN using a Lagrangian relaxation and a non-smooth optimization technique, a dual simplex method, or dual interior point method to determine the Lagrange multipliers of the dual problem and then recovering a solution to the continuous partitioning problem using duality;
determining a binary value of zi 1 . . . iN in {0,1} for each i1 . . . iN corresponding to each zi1 . . . iN ,wherein said values zi 1 . . . iN provide an optimal or near optimal solution to said continuous partitioning problem.
-
-
6. The method of claim 2 wherein said step of determining a most probable partition further comprises The steps of:
-
solving said continuous partitioning problem either approximately or optimally for values of zi 1 . . . iN or each i1, . . . , iN using a primal-dual simplex method;
determining a binary value of zi 1 . . . iN in {0,1} for each i1 . . . iN corresponding to each zi1 . . . iN ,wherein said values zi 1 . . . iN provide an optimal or near optimal solution to said continuous partitioning problem.
-
-
7. The method of claim 1 further comprising the step of sending a warning to aircraft or ground or sea facility based on said data association decisions.
-
8. The method of claim 1 further comprising the step of controlling air traffic based on said data association decisions.
-
9. The method of claim 1 further comprising the step of controlling anti-aircraft or anti-missile equipment based on said data association decisions.
-
10. The method of claim 1 further comprising the step of taking evasive action based on said data association decisions.
-
11. The method of claim 1 further comprising the step of working on one of said one or more objects based on said data association decisions.
-
12. The method of claim 1 further comprising the step of surveilling one of said one or more objects based on said data association decisions.
-
13. The method of claim 1 wherein said step of using a sliding window is subject to any one track of said tracks being assigned to a maximum of one observation in a single frame and subject to a maximum of a single track being assigned to each observation in said single frame.
-
14. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include kinematic information.
-
15. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include feature information.
-
16. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include attribute information.
-
17. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include kinematic and feature information.
-
18. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include kinematic and attribute information.
-
19. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include feature and attribute information.
-
20. The method of claim 1 wherein said step of using at least one sensor to repeatedly scan a region further comprises providing a plurality of observations that include kinematic and target identification information.
Specification