System and method for detecting settle down time using computer vision techniques

1. A method for detecting settle-down time in a space, the method comprising:

acquiring a sequence of frames capturing a select space from a first camera;

determining an initial time for computing a duration it takes for an associated occupant to settle into a seat in the select space;

determining one or more candidate frames from the sequence of frames where one or both of a sitting behavior and seat occupancy is observed at the seat;

determining a final frame and a final time associated with the final frame from the one or more candidate frames; and

computing the settle-down time using the initial and the final times.

2. The method of claim 1 , wherein determining a candidate frame includes:

obtaining a score for each frame of the sequence;

mapping scores for the sequence of frames onto a score map;

identifying the maximum score;

computing a difference between a score associated with the candidate frame and the maximum score;

comparing the difference to a predetermined threshold range; and

in response to the difference falling below the predetermined threshold range, associating the candidate frame as the final frame.

3. The method of claim 2 , wherein the predetermined threshold range includes a predetermined number of frames.

4. The method of claim 2 , further comprising:

in response to the difference falling outside the predetermined range threshold, associating the candidate frame as including false sitting behavior;

acquiring a second candidate frame number from a second camera; and,

repeating the thresholding for the second candidate frame.

5. The method of claim 2 , wherein the obtaining the score includes:

determining a region of interest (“ROI”) where sitting behavior is expected within the select space;

extracting activity features from the ROI and generating an activity feature vector using the activity features; and,

applying the activity features to a previously trained classifier to obtain the score.

6. The method of claim 1 , wherein the determining the initial time includes determining when the associated occupant enters the select space.

7. The method of claim 1 , wherein a first camera is positioned to capture the select space from an oblique angle and a second camera is positioned to capture the select space from an overhead view.

8. The method of claim 1 , wherein the determining the initial frame when the associated occupant enters the select space includes:

searching for a moving object in a foreground of a given frame;

associating a detected foreground object as being the associated occupant that entered the select space; and

identifying a number of a frame where the moving object is detected.

9. The method of claim 1 further comprising:

using the settle-down time for each seat in the select space, aggregating an overall time for all occupied seats in the select space.

10. The method of claim 1 further comprising:

providing a trained classifier operative to classify between an occupied seat and an unoccupied.

11. A system for detecting settle-down time in a space, the system comprising a computer including a memory and a processor in communication with the processor configured to:

acquire a sequence of frames capturing a select space from a first camera;

determine an initial time for computing a duration it takes for an associated occupant to settle into a seat in the select space;

determine one or more candidate frames from the sequence of frames where one or both of a sitting behavior and seat occupancy is observed at the seat;

determine a final frame and a final time associated with the final frame from the one or more candidate frames; and

compute the settle-down time using the initial and the final times.

12. The system of claim 11 , wherein the processor is further configured to:

obtain a score for each frame of the sequence;

map scores for the sequence of frames onto a score map;

identify the maximum score;

compute a difference between a score associated with the candidate frame and the maximum score;

compare the difference to a predetermined threshold range; and,

in response to the difference falling below the predetermined threshold range, associating the candidate frame as the final frame.

13. The system of claim 12 , wherein the predetermined threshold range includes a predetermined number of frames.

14. The system of claim 12 , wherein the processor is further configured to:

in response to the difference falling outside the predetermined range threshold, associate the candidate frame as including false sitting behavior;

acquire a second candidate frame number from a second camera; and,

repeat the thresholding for the second candidate frame.

15. The system of claim 12 , wherein the processor is further configured to:

determine a region of interest (“ROI”) where sitting behavior is expected within the select space;

extract activity features from the ROI and generating an activity feature vector using the activity features; and,

apply the activity features to a previously trained classifier to obtain the score.

16. The system of claim 11 , wherein the processor is further configured to:

determine the initial time when the associated occupant enters the select space.

17. The system of claim 11 , wherein a first camera is positioned to capture the select space from an oblique angle and a second camera is positioned to capture the select space from an overhead view.

18. The system of claim 11 , wherein the processor is further configured to:

search for a moving object in a foreground of a given frame;

associate a detected foreground object as being the associated occupant that entered the select space;

identify a number of a frame where the moving object is detected; and,

associate the number of the frame as the initial frame when the associated occupant enters the select space.

19. The system of claim 11 , wherein the processor is further configured to:

using the settle-down time for each seat in the select space, aggregate an overall time for all occupied seats in the select space.

20. The system of claim 11 , wherein the processor is further configured to:

provide a trained classifier operative to classify between an occupied seat and an unoccupied.