Movement detection and estimation in wavelet compressed video

• US 20050078873A1
• Filed: 01/29/2004
• Published: 04/14/2005
• Est. Priority Date: 01/31/2003
• Status: Abandoned Application

• Current Assignee: Visioprime Ltd.
• Original Assignee: Visioprime Ltd.
• Inventors: Aksay, Anil, Cetin, Ahmet Enis, Toreyin, Behcet Ugur, Akhan, Mehmet Bilgay
• Application Number:
  US10/768,606
• Publication Number:
  US 20050078873A1
• Time in Patent Office: Days
• US Class Current: 382/236
• CPC Class Codes: G06T 7/262 using transform domain methods, e.g. Fourier domain methods
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

A method and system for moving object and region detection in video compressed using a wavelet transform is disclosed. A plurality of images are inputted to the system in wavelet compressed format in time series. In a first aspect, a method and system determines the motion by comparing the wavelet transform of the current image and the wavelet transform of the previous image of the video. Difference between the wavelet coefficients of the current and previous images indicate motion. Moving regions in the video can be estimated by determining the wavelet coefficients of the current image frame, which are different from the wavelet coefficients of the previous image frame. The method and system does not include performing an inverse wavelet transform on the wavelet transformed image. This leads to a computationally efficient method and a system compared to the existing motion estimation methods.

Claims

1. A method for detecting moving objects and regions in video compressed by a wavelet transform, the method comprising:

comparing the wavelet transform of a current image frame and the wavelet transform of a previous image frame of the video;

wherein a difference between the wavelet coefficients of the current and previous image frames indicate the existence of the motion of the moving objects and regions, wherein an inverse wavelet transformation is not performed.

2. The method of claim 1 wherein locations of moving objects and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of the current image frame differing from the corresponding wavelet coefficients of the previous image.

3. The method of claim 1, wherein the comparing step includes matching a predetermined area in the wavelet transform of one image with a predetermined area in the wavelet transform of the next image by shifting as one unit in a wavelet domain;

calculating a difference of wavelet coefficient values between the predetermined area in the wavelet transform of the one image and each matched area of the wavelet transform of the next image; and

calculating an evaluation value of the difference of the wavelet coefficient values, wherein if this evaluation value is above a threshold then there is motion.

4. The method of claim 1 wherein the locations of moving object and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of a current image frame differing from the wavelet coefficients of previous image frames, wherein given the wavelet coefficients it is possible to determine the location of pixel values on the current image frame producing the wavelet coefficient.

5. A method for estimating the moving objects and regions of a video, the method comprising:

comparing a wavelet transform of a current image of the video with an estimated wavelet transform of a background scene which does not contain moving objects and regions, wherein the motion or the presence or absence of the moving objects and regions in the current image frame of the video is determined without performing an inverse wavelet transformation operation.

6. The method of claim 5, wherein the wavelet transform of the background scene is estimated from the wavelet transforms of past image frames of the video, wherein wavelet coefficients whose value do not change or change below a threshold over time in plurality of images forming the video are classified as wavelet coefficients of the background scene.

7. The method of claim 5, wherein the locations of moving regions on the current image of the video are estimated by determining the indices of the image pixels producing the wavelet coefficients of the current image frame differing from the wavelet coefficients of the estimated background.

8. The method of claim 5, wherein the comparing step comprises:

matching a predetermined area in the wavelet transform of one image with the predetermined area in the estimated wavelet transform of the background image by shifting as one unit in the wavelet domain;

calculating the difference of wavelet coefficient values between the predetermined area in the wavelet transform of the one image and each matched area of the estimated wavelet transform of the background image, and

calculating an evaluation value of the difference of the wavelet coefficient value.

9. The method of claim 6 wherein the threshold for determining the moving wavelet coefficients is estimated in a recursive manner from the threshold value used in previous comparison, and difference of the previous value of the wavelet coefficient and estimated wavelet coefficient of the background.

16. The system of claim 5, wherein the locations of moving regions on the current image of the video are estimated by determining the indices of the image pixels producing the wavelet coefficients of the current image frame differing from the wavelet coefficients of the estimated background.

10. A system for detecting moving objects and regions in video compressed by a wavelet transform, the system comprising:

a comparator mechanism for comparing the wavelet transform of a current image frame and the wavelet transform of a previous image frame of the video;

a mapping mechanism for utilizing a difference between the wavelet coefficient of the current and previous image frames to indicate the motion, wherein an inverse wavelet transform is not performed.

11. The system of claim 10 wherein locations of moving objects and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of the current image frame differing from the corresponding wavelet coefficients of the previous image.

12. The system of claim 10 wherein the comparator mechanism comprises:

means for matching a predetermined area in the wavelet transform of one image with a predetermined area in the wavelet transform of a next image by shifting as one unit in a wavelet domain;

means for calculating a difference of wavelet coefficient values between the predetermined area in a wavelet transform of the one image and each matched area of a wavelet transform of the next image; and

means for calculating an evaluation value of the difference of the wavelet coefficient values wherein if the evaluation value is above a threshold then there is motion.

13. The system of claim 10 wherein the locations of moving object and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of a current image frame differing from the wavelet coefficients of previous image frames, wherein given the wavelet coefficients it is possible to determine the location of pixel values on the current image frame producing the wavelet coefficient.

14. A system for estimating the moving objects and regions of a video;

the system comprising;

a comparator for comparing a wavelet transform of a current image frame of the video with an estimated wavelet transform of a background scene which does not contain moving objects and regions; and

means for determining whether a motion or the presence or absence of the moving object and regions is within the current image frame without performing an inverse wavelet transformation operation.

15. The system of claim 14, wherein the wavelet transform of the background scene is estimated from the wavelet transforms of past image frames of the video, wherein wavelet coefficients whose value do not change or change below a threshold over time in plurality of images forming the video are classified as wavelet coefficients of the background scene.

17. The system of claim 14 wherein the comparator comprises:

means for matching a predetermined area in the wavelet transform of one image with a predetermined area in the estimated wavelet transform of the background image by shifting as one unit in the wavelet domain;

means for calculating the difference of wavelet coefficient values between a predetermined error in the wavelet transformation of the one image and each matched area of the estimated wavelet transform of the background image; and

means for calculating an evaluation value of the difference of the wavelet coefficient values.

18. The system of claim 15 wherein the threshold for determining the moving wavelet coefficients is estimated in a recursive manner from the threshold value used in previous comparison, and difference of the previous value of the wavelet coefficient and estimated wavelet coefficient of the background.

19. A computer readable medium containing program instructions for detecting moving objects and regions in video compressed by a wavelet transform, the program instructions for:

comparing the wavelet transform of a current image frame and the wavelet transform of a previous image frame of the video;

wherein a difference between the wavelet coefficients of the current and previous image frames indicate the existence of the motion of the moving objects and regions, wherein an inverse wavelet transformation is not performed.

20. The computer readable medium of claim 19 wherein locations of moving objects and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of the current image frame differing from the corresponding wavelet coefficients of the previous image.

21. The computer readable medium of claim 19, wherein the comparing step includes matching a predetermined area in the wavelet transform of one image with a predetermined area in the wavelet transform of the next image by shifting as one unit in a wavelet domain;

calculating a difference of wavelet coefficient values between the predetermined area in the wavelet transform of the one image and each matched area of the wavelet transform of the next image; and

calculating an evaluation value of the difference of the wavelet coefficient values, wherein if this evaluation value is above a threshold then there is motion.

22. The computer readable medium of claim 19 wherein the locations of moving object and regions on the current image of the video are estimated by determining the indices of image pixels of the video producing the wavelet coefficients of a current image frame differing from the wavelet coefficients of previous image frames, wherein given the wavelet coefficients it is possible to determine the location of pixel values on the current image frame producing the wavelet coefficient.

23. A computer readable medium containing program instructions for estimating the moving objects and regions of a video, the program instructions for:

comparing a wavelet transform of a current image of the video with an estimated wavelet transform of a background scene which does not contain moving objects and regions, wherein the motion or the presence or absence of the moving objects and regions in the current image frame of the video is determined without performing an inverse wavelet transformation operation.

24. The computer readable medium of claim 23, wherein the wavelet transform of the background scene is estimated from the wavelet transforms of past image frames of the video, wherein wavelet coefficients whose value do not change or change below a threshold over time in plurality of images forming the video are classified as wavelet coefficients of the background scene.

25. The computer readable medium of claim 23, wherein the locations of moving regions on the current image of the video are estimated by determining the indices of the image pixels producing the wavelet coefficients of the current image frame differing from the wavelet coefficients of the estimated background.

26. The computer readable medium of claim 23, wherein the comparing step comprises:

matching a predetermined area in the wavelet transform of one image with the predetermined area in the estimated wavelet transform of the background image by shifting as one unit in the wavelet domain;

calculating the difference of wavelet coefficient values between the predetermined area in the wavelet transform of the one image and each matched area of the estimated wavelet transform of the background image, and

calculating an evaluation value of the difference of the wavelet coefficient value.

27. The computer readable medium of claim 24 wherein the threshold for determining the moving wavelet coefficients is estimated in a recursive manner from the threshold value used in previous comparison, and difference of the previous value of the wavelet coefficient and estimated wavelet coefficient of the background.