Locally invariant global hypothesis tracking

• US 9,441,986 B2
• Filed: 04/12/2013
• Issued: 09/13/2016
• Est. Priority Date: 04/17/2012
• Status: Active Grant

• Current Assignee: Raytheon Company (Rtx Corporation)
• Original Assignee: Raytheon Company (Rtx Corporation)
• Inventors: Cronin, Brian A., Ploplys, Nicholas J., Huffaker, Ray B., Sun, Nicholas, Boardman, Michael
• Primary Examiner(s): Badii, Behrang
• Assistant Examiner(s): GREENE, DANIEL LAWSON
• Application Number:
  US13/862,075
• Publication Number:
  US 20130275034A1
• Time in Patent Office: 1,250 Days
• Field of Search: 701/300
• US Class Current: 1/1
• CPC Class Codes: G01C 23/00 Combined instruments indicating more than one navigational value, e.g. for aircraft; Combined measuring devices for measuring two or more variables of movement, e.g. distance, speed or acceleration
• CPC Class Codes: G01S 13/726 Multiple target tracking

Abstract

The technology described herein includes a system and/or a method for global hypothesis tracking. In some examples, a method generates one or more paired segments based on track data representing kinematic data of target objects. Each paired segment includes a list of tracks incompatible with the paired segment, which are tracks sharing common track data, and a likelihood score. The method generates a transition probability between each pair of the paired segments based on the list of tracks incompatible with the paired segment and the likelihood score associated with each paired segment. The method further generates one or more multi-segment tracks based on the one or more paired segments also based on the transition probability between each pair of the paired segments.

Claims

1. A system for generating global hypothesis tracks associated with one or more target objects, the system comprising:

one or more processors and a memory coupled to the one or more processors for storing computer program instructions, wherein the computer program instructions are configured to cause, when executed by the one or more processors, the one or more processor to;

store track data, received from a sensor observing one or more target objects, in the memory, the track data comprising kinematic information associated with the one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generate one or more paired segments based on the track data loaded from the memory, each of the one or more paired segments comprising a list of one or more tracks incompatible with the paired segment, the paired segment having two segments, and a likelihood score, the tracks incompatible with the paired segment being tracks sharing common track data; and

generate a transition probability between each pair of the one or more paired segments based on the list of one or more tracks incompatible with the paired segment and the likelihood score associated with each paired segment of the pair of the one or more paired segments, and

generate one or more multi-segment tracks associated with the one or more observed target objects and based on the one or more paired segments and the generated transition probability between each pair of the one or more paired segments, the generated one or more multi-segment tracks having a transition probability above a threshold.

2. The system of claim 1, wherein the one or more processors are further caused to:

receive second track data, the second track data comprises kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generate one or more second paired segments based on the second track data, each of the one or more second paired segments comprises a list of one or more tracks incompatible with the paired segment and a likelihood score; and

generate one or more second multi-segment tracks based on the one or more paired segments and the one or more second paired segments, and

generate a second transition probability between each pair of the one or more second paired segments based on the list of one or more incompatible tracks and the likelihood score associated with each paired segment of the pair of the one or more second paired segments.

3. The system of claim 2, wherein the track data is associated with a time period, the second track data is associated with a second time period, and the second time period occurs in time before the time period.

4. The system of claim 1, wherein the track data is associated with a time period and the one or more processors are further caused to:

receive second track data, the second track data associated with a second time period and the second time period occurs in time after the time period and the second track data comprises kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generate one or more second paired segments based on the second track data, each of the one or more second paired segments comprises a list of one or more tracks incompatible with the paired segment and a likelihood score;

generate one or more second multi-segment tracks based on the one or more paired segments and the one or more second paired segments,

generate a second transition probability between each pair of the one or more second paired segments based on the list of one or more incompatible tracks and the likelihood score associated with each paired segment of the pair of the one or more second paired segments,

associate the one or more second multi-segment tracks with the one or more multi-segment tracks based on a transition probability between intersecting paired segments, and

generate one or more third multi-segment tracks based on the one or more multi-segment tracks, the one or more second multi-segment tracks, and the associations between the one or more second multi-segment tracks with the one or more multi-segment tracks.

5. The system of claim 1, wherein the one or more processors are further caused to:

determine the feature likelihood information based on a target object in the track data, and

generate the likelihood score for each of the one or more paired segments based on the kinematic information and the feature likelihood information.

6. The system of claim 1, wherein the one or more processors are further caused to:

determine the feature likelihood information based on a target object shape in the track data, an image feature in the track data, a radar related feature in the track data, a spectral feature in the track data, measurement and signature intelligence data in the track data, intelligence information associated with the track data, or any combination thereof, and

generate the likelihood score for each of the one or more paired segments based on the kinematic information and the feature likelihood information.

7. The system of claim 1, wherein the one or more processors are further caused to generate the list of the one or more tracks incompatible with the paired segment for each of the one or more paired segments based on the kinematic information.

8. A computer-implemented method for generating global hypothesis tracks associated with one or more target objects, the method comprising:

receiving at one or more processors track data from a sensor observing one or more target objects, the track data comprising kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generating using the one or more processors one or more paired segments based on the track data, each of the one or more paired segments comprising a list of one or more tracks incompatible with the paired segment, the paired segment having two segments, and a likelihood score, the tracks incompatible with the paired segment tracks sharing common track data;

generating using the one or more processors a transition probability between each pair of the one or more paired segments based on the list of one or more incompatible tracks and the likelihood score associated with each paired segment of the pair of the one or more paired segments; and

generating using the one or more processors one or more multi-segment tracks based on the one or more paired segments and the generated transition probability between each pair of the one or more paired segments, the generated one or more multi-segment tracks having a transition probability above a threshold.

9. The method of claim 8, further comprising:

receiving at the one or more processors second track data, the second track data comprises kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generating using the one or more processors one or more second paired segments based on the second track data, each of the one or more second paired segments comprises a list of one or more tracks incompatible with the paired segment and a likelihood score;

generating using the one or more processors a transition probability between each pair of the one or more second paired segments based on the list of one or more incompatible tracks and the likelihood score associated with each paired segment of the pair of the one or more second paired segments; and

generating using the one or more processors one or more second multi-segment tracks based on the one or more paired segments, the one or more second paired segments, and the transition probability between each of the one or more second paired segments.

10. The method of claim 8, wherein the track data is associated with a time period and the method further comprising:

receiving at the one or more processors second track data, the second track data associated with a second time period and the second time period occurs in time after the time period and the second track data comprises kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generating using the one or more processors one or more second paired segments based on the second track data, each of the one or more second paired segments comprises a list of one or more tracks incompatible with the paired segment and a likelihood score;

generating using the one or more processors a second transition probability between each pair of the one or more second paired segments based on the list of one or more tracks incompatible with the paired segment and the likelihood score associated with each paired segment of the pair of the one or more second paired segments;

generating using the one or more processors one or more second multi-segment tracks based on the one or more paired segments, the one or more second paired segments, and the transition probability between each of the one or more second paired segments;

associating using the one or more processors the one or more second multi-segment tracks with the one or more multi-segment tracks based on a transition probability between intersecting paired segments; and

generating using the one or more processors one or more third multi-segment tracks based on the one or more multi-segment tracks, the one or more second multi-segment tracks, and the associations between the one or more second multi-segment tracks with the one or more multi-segment tracks.

11. The method of claim 8, further comprising:

determining using the one or more processors the feature likelihood information based on a target object shape in the track data, an image intensity frequency in the track data, or any combination thereof, and

generating using the one or more processors the likelihood score for each of the one or more paired segments based on the kinematic information and the feature likelihood information.

12. The method of claim 8, further comprising:

receiving at the one or more processors second track data, the second track data comprises feature likelihood information associated with at least one of the one or more target objects;

associating using the one or more processors the second track data with the at least one of the one or more target objects;

generating using the one or more processors a second transition probability between each pair of the one or more second paired segments based on the list of one or more incompatible tracks, the likelihood score associated with each paired segment of the pair of the one or more second paired segments, and the second track data associated the at least one of the one or more target objects; and

generating using the one or more processors one or more second multi-segment tracks based on the one or more paired segments, the one or more second paired segments, and the transition probability between each of the one or more second paired segments.

13. The method of claim 12, wherein the second track data further comprises kinematic information that is not assigned to the one or more target objects.

14. The method of claim 8, further comprising:

generating using the one or more processors a plurality of target tracks based on the one or more multi-segment tracks; and

selecting using the one or more processors a long-term track for each of the one or more target objects based on the plurality of target tracks and the transition probability between each part of the one or more paired segments.

15. The method of claim 8, further comprising generating using the one or more processors the list of the one or more tracks for each of the one or more paired segments based on the kinematic information.

16. The method of claim 8, wherein the likelihood score for each paired segment of the one or more paired segments is indicative of a probability of a transition in a track between the paired segment.

17. The method of claim 8, wherein the feature likelihood information is indicative of a probability of a feature type in the track data.

18. The method of claim 8, wherein each of the one or more paired segments are indicative of a local cluster hypothesis of a movement of a target object along a track.

19. The method of claim 8, wherein the one or more multi-segment tracks are indicative of a global hypothesis of a movement of a target object along a track.

20. The method of claim 8, wherein the track data comprises measurement and signature intelligence data, movement intelligence data, signal intelligence data, imagery intelligence data, human intelligence data, or any combination thereof.

21. A non-transitory computer readable program product, tangibly embodied in an information carrier, the computer program product including instructions that when executed by a data processing apparatus cause the data processing apparatus to:

receive track data from a sensor observing one or more target objects, the track data comprising kinematic information associated with one or more target objects, feature likelihood information associated with the one or more target objects, or any combination thereof;

generate one or more paired segments based on the track data, each of the one or more paired segments comprising a list of one or more tracks incompatible with the paired segment, the paired segment having two segments, and a likelihood score, the tracks incompatible with the paired segment tracks sharing common track data;

generate a transition probability between each pair of the one or more paired segments based on the list of one or more tracks incompatible with the paired segment and the likelihood score associated with each paired segment of the pair of the one or more paired segments; and

generate one or more multi-segment tracks based on the one or more paired segments and the transition probability between each of the one or more paired segments, the generated one or more multi-segment tracks having a transition probability above a threshold.