Object tracking systems and methods utilizing compressed-domain motion-based segmentation

• US 7,835,542 B2
• Filed: 05/22/2006
• Issued: 11/16/2010
• Est. Priority Date: 12/29/2005
• Status: Active Grant

• Current Assignee: Industrial Technology Research Institute
• Original Assignee: Industrial Technology Research Institute
• Inventors: Lee, Kual-Zheng, Lin, Chia-Wen, Ling, Zhi-Hong
• Primary Examiner(s): Ahmed; Samir A.
• Assistant Examiner(s): Fitzpatrick; Atiba O
• Application Number:
  US11/419,600
• Publication Number:
  US 20070154066A1
• Time in Patent Office: 1,639 Days
• Field of Search: 382/101, 382/103, 348169-180, 348154-155, 356 27- 29
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/30241 Trajectory
• CPC Class Codes: G06T 7/262 using transform domain methods, e.g. Fourier domain methods
• CPC Class Codes: G06T 7/277 involving stochastic approaches, e.g. using Kalman filters
• CPC Class Codes: G06V 10/24 Aligning, centring, orientation detection or correction of the image

Abstract

An object tracking method may include: receiving frames of data containing image information of an object; performing an object segmentation to obtain an object motion result; and using the object motion result to conduct an object tracking. In particular, the object segmentation may include: extracting motion vectors from the frames of data; estimating a global motion using the motion vectors; and subtracting the global motion from the motion vectors to generate an object motion result.

Claims

1. An object tracking method comprising:

receiving frames of data containing compressed-domain image information of an object;

performing an object segmentation using the compressed-domain image information, the object segmentation comprising;

extracting motion vectors from the frames of data;

estimating a global motion using the motion vectors; and

subtracting the global motion from the motion vectors to generate an object motion result; and

using the object motion result to conduct an object tracking, wherein using the object motion result comprises conducting a similarity analysis between sampling points identified based on the object motion result and an initial object model of the frames of data.

2. The method of claim 1, wherein estimating the global motion comprises distinguishing local object motions from camera motions to derive a rough object mask.

3. The method of claim 1, wherein estimating the global motion comprises applying a simplified affine motion model to process the motion vectors.

4. The method of claim 1, further comprising performing a label set process after generating the object motion result, wherein performing the label set process comprises distinguishing the object from a second object for conducting multiple-object tracking.

5. The method of claim 1, further comprising performing a label set process after generating the object motion result, wherein performing the label set process comprises clustering the object motion result.

6. The method of claim 1, further comprising refining the object motion result by applying particle filtering.

7. The method of claim 1, wherein the frames of data belong to a portion of a compressed video stream and performing the object segmentation comprises performing the object segmentation in a compressed domain.

8. An apparatus comprising a processor configured to perform or cause the apparatus to at least perform the following:

receiving frames of data in a coded format containing image information of an object;

performing an object segmentation, the object segmentation comprising;

extracting motion vectors from the frames of data;

estimating a global motion using the motion vectors; and

subtracting the global motion from the motion vectors to generate an object motion result; and

conducting an object tracking based on the object motion result, wherein conducting the object tracking comprises performing a similarity analysis between sampling points identified based on the object motion result and an initial object model of the frames of data.

9. The apparatus of claim 8, wherein estimating the global motion comprises distinguishing local object motions from camera motions to derive a rough object mask.

10. The apparatus of claim 8, wherein estimating the global motion comprises applying a simplified affine motion model to process the motion vectors.

11. The apparatus of claim 8, wherein the processor is further configured to perform or cause the apparatus to further perform a label set process after generating the object motion result, wherein performing the label set process comprises distinguishing the object from a second object for conducting multiple-object tracking.

12. The apparatus of claim 8, wherein conducting the object tracking comprises refining the object motion result by applying particle filtering.

13. The apparatus of claim 8, wherein the frames of data belong to a portion of a compressed video stream in the coded format and the object segmentation processor performs the object segmentation in a compressed domain.

14. The apparatus of claim 8 further comprising a memory storing executable instructions that in response to execution by the processor cause the processor or apparatus to perform the object segmentation and the object tracking.

15. A non-transitory computer-readable storage medium having computer-readable program code portions stored therein that in response to execution by a processor cause an apparatus to at least perform the following:

receiving frames of data containing compressed-domain image information of an object;

performing an object segmentation using the compressed-domain image information, the object segmentation comprising;

extracting motion vectors from the frames of data;

estimating a global motion using the motion vectors; and

subtracting the global motion from the motion vectors to generate an object motion result; and

using the object motion result to conduct an object tracking, wherein using the object motion result comprises conducting a similarity analysis between sampling points identified based on the object motion result and an initial object model of the frames of data.

16. The non-transitory computer-readable storage medium of claim 15, wherein estimating the global motion comprises distinguishing local object motions from camera motions to derive a rough object mask, or applying a simplified affine motion model to process the motion vectors.

17. The non-transitory computer-readable storage medium of claim 15, wherein execution of the computer-readable program code portions by the processor further causes the apparatus to perform a label set process after generating the object motion result, wherein performing the label set process comprises distinguishing the object from a second object for conducting multiple-object tracking, or clustering the object motion result.

18. The non-transitory computer-readable storage medium of claim 15, further comprising refining the object motion result by applying particle filtering.

19. The non-transitory computer-readable storage medium of claim 15, wherein the frames of data belong to a portion of a compressed video stream and performing the object segmentation comprises performing the object segmentation in a compressed domain.