Video segmentation using statistical pixel modeling
View Patent ↗A method for segmenting video data into foreground and background ( 324 ) portions utilizes statistical modeling of the pixels Λ statistical model of the background is built for each pixel, and each pixel in an incoming video frame is compared ( 326 ) 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 video segmentation, for differentiating between foreground and background portions of video, comprising the steps of:
obtaining a frame sequence from a real-time video stream; and
for two or more frames in the frame sequence, performing the following steps:
labeling pixels in the frames;
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.
2. A method of video segmentation, for differentiating between foreground and background portions of video, comprising the steps of:
obtaining a frame sequence from a real-time video stream; and
for two or more frames in the frame sequence, performing the following steps:
labeling at least a portion of pixels in the frames;
performing spatial or temporal filtering;
for at least one labeled pixel, determining if the labeled pixel matches a first background statistical model;
updating the first background statistical model if the labeled pixel matches the first background statistical model; and
updating a secondary background statistical model if the labeled pixel does not match the first background statistical model.
3. A method according to claim 2 , wherein the secondary background statistical model represents a potential foreground pixel.
4. A method according to claim 3 , further comprising:
replacing the first background statistical model with the second background statistical model when the second background statistical model is stable.
5. A method of video segmentation, comprising the steps of:
obtaining a plurality of frames from a video stream; and
for at least two of the obtained frames, performing the following steps:
labeling a portion of the pixels in the frames;
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.
6. A method of video segmentation, comprising the steps of:
performing the following with at least one of a computer or circuitry for two or more frames obtained from a video stream:
labeling at least a portion of pixels in the frames;
performing spatial or temporal filtering;
for at least one labeled pixel, determining if the labeled pixel matches a first background statistical model;
updating the first background statistical model if the labeled pixel matches the first background statistical model; and
updating a secondary background statistical model if the labeled pixel does not match the first background statistical model.
7. A method according to claim 6 , wherein the secondary background statistical model represents a potential foreground pixel.
8. A method according to claim 7 , further comprising:
replacing the first background statistical model with the second background statistical model when the second background statistical model is stable.