Let m 1 denotes the number of ξ pixels, and m 2 denotes the number of ( overset{sim }{phi_{EM}} ) pixels, where ξ is the uniform segmentation frame and ( overset{sim }{phi_{EM}} ) is the EM segmented frame.
Based on the observation that forgeries break the original frame segmentation, frame offsets are used to locate forgeries automatically, allowing to detect most common forgeries, such as deletion, insertion, substitution, and splicing.
After foreground segmentation, both segmented frames are fused to get a new foreground which embeds more information compared to a single segmented frame.
The motion information is segmented by uniform segmentation and EM segmentation [45] and then fuse both segmented frames by implementing multiplication law of probability.
To optimize segmentation, binary frames were sometimes manually corrected with appropriate tools built into the software.
Ikizler-Cinbis and Sclaroff [31] used the segmentation of frame images for separation of foreground objects and background planes, and then, camera motions are estimated from background planes.
The segmentation was done frame-by-frame in 2D.
In the first frame, markers are extracted which provide reliable seed regions for segmentation in subsequent frames.
The segmentation in both frames is not consistent.
