Systems and methods for tracking points within an encasement

• US 9,286,695 B2
• Filed: 03/13/2014
• Issued: 03/15/2016
• Est. Priority Date: 03/13/2014
• Status: Active Grant

• Current Assignee: RM Acquisition LLC (Rand McNally & Company)
• Original Assignee: Bendix Commercial Vehicle Systems LLC (Estate of Heinz Hermann Thiele)
• Inventors: Kuehnle, Andreas U., Boon, Cathy L., Molin, Hans M.
• Primary Examiner(s): Abdi, Amara
• Application Number:
  US14/209,227
• Publication Number:
  US 20150262378A1
• Time in Patent Office: 733 Days
• US Class Current: 1/1
• CPC Class Codes: G06T 11/001 Texturing; Colouring; Generation of texture or colour inpainting G06T5/77
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20164 Salient point detection; Corner detection
• CPC Class Codes: G06T 2207/30241 Trajectory
• CPC Class Codes: G06T 2207/30244 Camera pose
• CPC Class Codes: G06T 2207/30261 Obstacle
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments

Abstract

Systems and methods for tracking points within an encasement are provided. According to an aspect of the invention, a processor designates an encasement at a first location within a first image acquired at a first time; identifies points to track within the encasement; determines characteristics of the points to track; tracks the points over time based on the characteristics; and determines a second location of the encasement within a second image acquired at a second time based on positions of the tracked points at the second time. Identifying the points to track may include identifying points within the encasement that are significant and persistent.

Claims

1. A tracking method comprising:

designating, by a processor, an encasement at a first location within a first image acquired at a first time;

identifying, by the processor, points to track within the encasement;

determining, by the processor, characteristics of the points to track;

tracking, by the processor, the points over time based on the characteristics; and

determining, by the processor, a second location of the encasement within a second image acquired at a second time based on positions of the tracked points at the second time,

wherein the determining of the second location of the encasement comprises;

selecting pairs of the tracked points;

for each of the pairs, determining a length of the encasement based on a position of each of the points with respect to the encasement and a distance between the points;

for each of the pairs, determining the second location of the encasement within the second image based on the length of the encasement, the position of each of the points with respect to the encasement, and a position of each of the points at the second time and

selecting an optimal second location of the encasement from the second locations of the encasement determined for each of the pairs.

2. The method according to claim 1, wherein identifying the points to track comprises:

identifying points within the encasement that are significant; and

identifying significant points that are persistent.

3. The method according to claim 2, wherein identifying the significant points comprises, for each of a plurality of points within the encasement:

determining a local gradient of the point along each of a plurality of directions;

comparing a magnitude of the local gradient to a threshold for each of the plurality of directions; and

identifying the point as significant if the magnitude of the local gradient is not less than the threshold for any one of the plurality of directions.

4. The method according to claim 2, wherein identifying the significant points comprises, for each of a plurality of points within the encasement:

determining a local gradient of the point along each of a plurality of directions;

comparing a magnitude of the local gradient to a threshold for each of the plurality of directions; and

identifying the point as significant if the magnitude of the local gradient is not less than the threshold for each of the plurality of directions, and a sign of the local gradient is the same for each of the plurality of directions.

5. The method according to claim 3, further comprising assigning a type to each of the significant points based on which of the plurality of directions of the local gradient is not less than the threshold.

6. The method according to claim 3, wherein identifying the significant points further comprises, for each of the plurality of points within the encasement:

determining a desirability of the point as a product of the magnitudes of the local gradients for each of the plurality of directions; and

if a plurality of adjacent points are identified as significant, removing significant points whose desirability is not greater than a desirability of all immediately neighboring points.

7. The method according to claim 2, wherein identifying the significant points that are persistent comprises:

scaling and shifting locations of the significant points with respect to locations of corresponding points within a third image acquired at a third time, such that an overlap between the significant points and the corresponding points is maximized;

repeating the scaling and shifting with respect to locations of corresponding points within a plurality of additional images acquired at a plurality of additional times until enough information to analyze the significant points is available; and

for each of the significant points, determining a percentage overlap between the significant point and the corresponding points within the third image and the plurality of additional images, and identifying the significant point as persistent if the percentage overlap over time is not less than a minimum percentage.

8. The method according to claim 1, wherein the characteristics of the points to be tracked comprise, for each of the points to be tracked, a local gradient along each of a plurality of directions and a ruler location with respect to the encasement.

9. The method according to claim 1, wherein the characteristics of the points to be tracked comprise, for each of the points to be tracked, a grayscale or a color of the points.

10. The method according to claim 1, wherein tracking the points over time comprises, for each of the points:

determining whether there is a second point of a same type as the point and near the point within a third image acquired at a third time; and

if it is determined that there is only one second point of the same type near the point within the third image, updating a position of the point as a position of the second point.

11. The method according to claim 10, wherein the type is a corner type or an edge type.

12. The method according to claim 10, wherein if it is determined that there is only one second point of the same type near the point within the third image, a local gradient of the point is updated as an average of the local gradient of the point and a local gradient of the second point.

13. The method according to claim 10, wherein if it is determined that there is more than one second point of the same type near the point within the third image, the position of the point is updated as the position of the one of the second points whose neighborhood is most similar to a neighborhood of the point.

14. The method according to claim 13, wherein if it is determined that there is more than one second point of the same type near the point within the third image, a local gradient of the point is updated as an average of the local gradient of the point and a local gradient of the one of the second points whose neighborhood is most similar to a neighborhood of the point.

15. The method according to claim 10, wherein if it is determined that there is no second point of the same type near the point within the third image, a hold counter is determined, and the point is removed from the points to track if a maximum value of the hold counter is reached.

16. The method according to claim 1, wherein the optimal second location is the median or the average of the second locations of the encasement determined for each of the pairs.

17. The method according to claim 1, wherein the determining of the second location of the encasement comprises:

scaling and shifting locations of the tracked points within the first image with respect to locations of corresponding points within the second image according to a transformation in which an overlap between the tracked points and the corresponding points is maximized; and

scaling and shifting the location of the encasement according to the transformation.

18. A tracking system comprising:

a memory; and

a processor coupled to the memory, the processor comprising;

encasement designation logic that designates an encasement at a first location within a first image acquired at a first time;

point identification logic that identifies points to track within the encasement;

point characterization logic that determines characteristics of the points to track;

point tracking logic that tracks the points over time based on the characteristics; and

location determination logic that determines a second location of the encasement within a second image acquired at a second time based on positions of the tracked points at the second time,

wherein the location determination logic determines the second location by;

selecting pairs of the tracked points;

for each of the pairs, determining a length of the encasement based on a position of each of the points with respect to the encasement and a distance between the points;

for each of the pairs, determining the second location of the encasement within the second image based on the length of the encasement, the position of each of the points with respect to the encasement, and a position of each of the points at the second time; and

selecting an optimal second location of the encasement from the second locations of the encasement determined for each of the pairs.

19. A non-transitory computer-readable medium comprising computer instructions executable by a processor to cause the processor to perform a tracking method comprising:

designating an encasement at a first location within a first image acquired at a first time;

identifying points to track within the encasement;

determining characteristics of the points to track;

tracking the points over time based on the characteristics; and

determining a second location of the encasement within a second image acquired at a second time based on positions of the tracked points at the second time,

wherein the determining of the second location of the encasement comprises;

selecting pairs of the tracked points;

for each of the pairs, determining a length of the encasement based on a position of each of the points with respect to the encasement and a distance between the points;

for each of the pairs, determining the second location of the encasement within the second image based on the length of the encasement, the position of each of the points with respect to the encasement, and a position of each of the points at the second time; and

selecting an optimal second location of the encasement from the second locations of the encasement determined for each of the pairs.