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. A method of one-pass video segmentation for differentiating between portions of video, the method comprising:
obtaining, by a processor, two or more frames from a real-time video stream;
building, by the processor, a background model using the two or more frames;
labeling, by the processor, at least a portion of pixels in the two or more frames; and
performing spatial or temporal filtering on the two or more frames by the processor.
2. The method according to claim 1 , further comprising:
aligning the two or more frames with a scene model.
3. The method according to claim 1 , further comprising:
updating the background model if at least one of the labeled pixels matches the background model.
4. The method according to claim 3 , further comprising:
building or updating a second background model if at least one of the labeled pixels do not match the background model.
5. The method according to claim 4 , further comprising:
replacing the background model with the second background model when the second background model is stable.
6. The method according to claim 1 , further comprising:
for at least one labeled pixel, determining if at least one of the labeled pixels match the background model.
7. The method according to claim 6 , further comprising:
updating the background model if at least one of the labeled pixels match the background model.
8. The method according to claim 7 , further comprising:
updating a second background model if at least one of the labeled pixels do not match the background model.
9. The method according to claim 8 , further comprising:
replacing the background model with the second background model when the second background model is stable.
10. A method of one-pass video segmentation for differentiating between portions of video, the method comprising:
obtaining, by a processor, two or more frames from a real-time video stream;
labeling, by the processor, at least a portion of pixels in the two or more frames;
performing spatial or temporal filtering on the two or more frames by the processor; and
updating, by the processor, a statistical model of the two or more frames.
11. The method according to claim 10 , further comprising:
building or updating a second statistical model of the two or more frames.
12. The method according to claim 11 , further comprising:
replacing the statistical model with the second statistical model when the second statistical model is stable.
13. A method of one-pass video segmentation, comprising:
receiving, by a processor, two or more frames from a real-time video stream;
comparing, by the processor, the received frames to a background model;
labeling, by the processor, at least a portion of pixels in the received frames;
performing spatial or temporal filtering on the two or more frames by the processor; and
building or updating, by the processor, the background model of the received frames.
14. The method according to claim 13 , further comprising:
updating the background model if at least one of the labeled pixels match the background model.
15. The method according to claim 14 , further comprising:
building or updating a second background model if at least one of the labeled pixels do not match the background model.
16. The method according to claim 15 , further comprising:
replacing the background model with the second background model when the second background model is stable.
17. The method according to claim 13 , further comprising:
updating the background model if at least one of the labeled pixels match the background model.
18. The method according to claim 17 , further comprising:
updating a second background model if at least one of the labeled pixels do not match the background model.
19. The method according to claim 18 , further comprising:
replacing the background model with the second background model when the second background model is stable.
20. A non-transitory computer-readable medium comprising instructions executable by one or more processors to perform a one-pass video segmentation for differentiating between portions of video, the computer-readable medium comprising one or more instructions for:
obtaining, by the one or more processors, two or more frames from a real-time video stream;
building, by the one or more processors, a background model using the two or more frames;
labeling, by the one or more processors, at least a portion of pixels in the two or more frames; and
performing, by the one or more processors, spatial or temporal filtering on the two or more frames.
21. A non-transitory computer-readable medium comprising instructions executable by one or more processors to perform a one-pass video segmentation for differentiating between portions of video, the computer-readable medium comprising one or more instructions for:
obtaining, by the one or more processors, two or more frames from a real-time video stream;
labeling, by the one or more processors, at least a portion of pixels in the two or more frames;
performing, by the one or more processors, spatial or temporal filtering on the two or more frames; and
updating, by the one or more processors, a statistical model of the two or more frames.
22. A non-transitory computer-readable medium comprising instructions executable by one or more processors for one-pass video segmentation, the computer-readable medium comprising one or more instructions for:
receiving, by the one or more processors, two or more frames from a real-time video stream;
comparing, by the one or more processors, the received frames to a background model;
labeling, by the one or more processors, at least a portion of pixels in the received frames;
performing, by the one or more processors, spatial or temporal filtering on the two or more frames; and
building or updating, by the one or more processors, the background model of the received frames.
23. An apparatus for one-pass video segmentation for differentiating between portions of video, the apparatus comprising:
at least one processor;
a memory coupled to the at least one processor; and
a computer program residing in memory and being executed by the at least one processor, wherein the computer program obtains two or more frames from a real-time video stream, builds a background model using the two or more frames, labels at least a portion of pixels in the two or more frames, and performs spatial or temporal filtering on the two or more frames.
24. An apparatus for one-pass video segmentation for differentiating between portions of video, the apparatus comprising:
at least one processor;
a memory coupled to the at least one processor; and
a computer program residing in memory and being executed by the at least one processor, wherein the computer program obtains two or more frames from a real-time video stream, labels at least a portion of pixels in the two or more frames, performs spatial or temporal filtering on the two or more frames, and updates a statistical model of the two or more frames.
25. An apparatus for one-pass video segmentation, the apparatus comprising:
at least one processor;
a memory coupled to the at least one processor; and
a computer program residing in memory and being executed by the at least one processor, wherein the computer program receives two or more frames from a real-time video stream, compares the received frames to a background model, labels at least a portion of pixels in the received frames, performs spatial or temporal filtering on the two or more frames, and builds or updates the background model of the received frames.