Video segmentation using statistical pixel modeling
View Patent ↗A method for segmenting video data into foreground and background portions utilizes statistical modeling of the pixels. A statistical model of the background is built for each pixel, and each pixel in an incoming video frame is compared with the background statistical model for that pixel. Pixels are determined to be foreground or background based on the comparisons. The method for segmenting video data may be further incorporated into a method for implementing an intelligent video surveillance system. The method for segmenting video data may be implemented in hardware.
1. Circuitry adapted to perform a one-pass video segmentation for differentiating between foreground and background portions of video, the circuitry comprising:
circuitry to obtain a frame sequence from a video stream; and
for each frame in the frame sequence:
circuitry to align the frame with a scene model;
circuitry to build a background statistical model, the background statistical model comprising values corresponding to regions of frames of the frame sequence and variances for the regions;
circuitry to label the regions of the frame; and
circuitry to perform spatial or temporal filtering.
2. A one-pass method of video segmentation, for differentiating between foreground and background portions of video, comprising the steps of:
obtaining a real-time video stream; and
for each frame in the real-time frame stream, performing the following steps:
labeling pixels in the frame;
performing spatial or temporal filtering;
updating a background statistical model, after the pixels are labeled; and
at least one of building or updating at least one foreground statistical model, after the pixels are labeled.
3. A method according to claim 2 , wherein for each frame in the real-time frame stream, further performing the following step:
inserting an object into the background statistical model.
4. A method according to claim 2 , wherein labeling pixels in the frame comprises:
compensating for global lighting change to the frame.
5. A method according to claim 2 , wherein updating the background statistical model and the foreground statistical model comprises:
compensating for global lighting change to the frame.
6. A method according to claim 2 , wherein pixels are labeled in the frame for an area of interest in the frame.
7. A method according to claim 2 , further comprising:
at least one of building or updating the background statistical model and the foreground statistical model pixels for an area of interest in the frame.
8. A method according to claim 2 , further comprising:
updating the background statistical model based on an infinite impulse response filter applied to previous statistical values of the background statistical model.
9. A method according to claim 2 , further comprising:
updating the foreground statistical model based on an infinite impulse response filter applied to previous statistical values of the foreground statistical model.
10. Circuitry adapted to perform a one-pass video segmentation for differentiating between foreground and background portions of video, comprising:
circuitry to obtain a real-time video stream; and
for each frame in the real-time frame stream:
circuitry to label pixels in the frame;
circuitry to perform spatial or temporal filtering;
circuitry to update a background statistical model, after the pixels are labeled; and
circuitry to at least one of build or update at least one foreground statistical model, after the pixels are labeled.
11. A circuitry according to claim 10 , further comprising:
for each frame in the real-time frame stream: circuitry to insert an object into the background statistical model.
12. A circuitry according to claim 10 , wherein the circuitry to label pixels in the frame comprises:
circuitry to compensate for global lighting change to the frame.
13. A circuitry according to claim 10 , wherein at least one of the circuitries to update the background statistical model and the foreground statistical model comprises:
circuitry to compensate for global lighting change to the frame.
14. A circuitry according to claim 10 , wherein pixels are labeled in the frame for an area of interest in the frame.
15. A circuitry according to claim 10 , wherein the background statistical model and the foreground statistical model pixels are built or updated for an area of interest in the frame.
16. A circuitry according to claim 10 , further comprising:
circuitry to update the background statistical model based on an infinite impulse response filter applied to previous statistical values of the background statistical model.
17. A circuitry according to claim 10 , further comprising:
circuitry to update the foreground statistical model based on an infinite impulse response filter applied to previous statistical values of the foreground statistical model.
18. A computer-readable medium comprising software for differentiating between foreground and background portions of video, the medium comprising:
instructions for obtaining a real-time video stream; and
for each frame in the real-time frame stream, instructions for:
labeling pixels in the frame;
performing spatial or temporal filtering;
updating a background statistical model, after the pixels are labeled; and
at least one of building or updating a foreground statistical model, after the pixels are labeled.
19. A computer-readable medium comprising software for differentiating between foreground and background portions of video, the medium comprising:
instructions for obtaining a frame sequence from a video stream; and
for each frame in the frame sequence, instructions for:
aligning the frame with a scene model,
building a background statistical model, the background statistical model comprising values corresponding to regions of frames of the frame sequence and variances for the regions,
labeling the regions of the frame, and
performing spatial or temporal filtering.